Professional Documents
Culture Documents
Falls Guidance
Falls Guidance
NICE clinical guideline 161 Falls: Assessment and prevention of falls in older people You can download the following documents from www.nice.org.uk/guidance/CG161 The NICE guideline all the recommendations. The NICE pathway a set of online diagrams that brings together all NICE guidance and support tools. Information for the public a summary for patients and carers. The full guideline (this document) all the recommendations, details of how they were developed, and reviews of the evidence they were based on. NICE clinical guidelines are recommendations about the treatment and care of people with specific diseases and conditions in the NHS in England and Wales. This guidance represents the view of NICE, which was arrived at after careful consideration of the evidence available. Healthcare professionals are expected to take it fully into account when exercising their clinical judgement. However, the guidance does not override the individual responsibility of healthcare professionals to make decisions appropriate to the circumstances of the individual patient, in consultation with the patient and/or guardian or carer, and informed by the summary of product characteristics of any drugs. Implementation of this guidance is the responsibility of local commissioners and/or providers. Commissioners and providers are reminded that it is their responsibility to implement the guidance, in their local context, in light of their duties to have due regard to the need to eliminate unlawful discrimination, advance equality of opportunity and foster good relations. Nothing in this guidance should be interpreted in a way that would be inconsistent with compliance with those duties. National Institute for Health and Care Excellence Level 1A, City Tower Piccadilly Plaza Manchester M1 4BT www.nice.org.uk
National Institute for Health and Care Excellence, 2013. All rights reserved. This material may be freely reproduced for educational and not-for-profit purposes. No reproduction by or for commercial organisations, or for commercial purposes, is allowed without the express written permission of NICE.
Page 2 of 315
This clinical guideline provides evidence and recommendations on the assessment and prevention of falls in older people. It extends and replaces Falls: assessment and prevention of falls in older people (NICE clinical guideline 21; 2004), by including additional recommendations about preventing falls in people admitted to hospital (inpatients). This document includes all the recommendations, details of how they were developed and summaries of the evidence they were based on. This guideline has been developed following the methods and processes outlined in the NICE guidelines manual (2009).
Page 3 of 315
Contents
Introduction ...................................................................................................... 5 Methods used to develop the guideline ........................................................ 5 Who this guideline is for ............................................................................... 6 Populations covered by this guideline .......................................................... 6 Patient-centred care......................................................................................... 7 Strength of recommendations .......................................................................... 8 Interventions that must (or must not) be used .............................................. 8 Interventions that should (or should not) be used a strong recommendation........................................................................................... 8 Interventions that could be used................................................................... 8 Wording of 2004 recommendations .............................................................. 9 Labelling of recommendations ....................................................................... 10 Key priorities for implementation .................................................................... 11 1 Recommendations .................................................................................. 13 1.1 Preventing falls in older people ......................................................... 13 1.2 Preventing falls in older people during a hospital stay ...................... 19 2 Care pathway.......................................................................................... 22 3 The assessment and prevention of falls in older people ......................... 23 Disclaimer................................................................................................... 23 3.1 Executive summary .......................................................................... 23 3.2 Aims .................................................................................................. 29 3.3 Methods ............................................................................................ 32 3.4 Guideline recommendations with supporting evidence reviews ...... 139 3.5 Recommendations for research ...................................................... 186 3.6 Validation ........................................................................................ 187 4 Assessment and prevention of falls in older people during a hospital stay 188 4.1 Background..................................................................................... 188 4.2 Methods used to develop this part of guideline ............................... 188 4.3 Inpatient risk prediction : evidence review and recommendations .. 189 4.4 Inpatient falls prevention interventions: evidence review and recommendations ..................................................................................... 198 4.5 Inpatient information: evidence review and recommendations ....... 245 4.6 List of research recommendations .................................................. 252 Environmental adaptions .......................................................................... 252 Prevalence of risk factors ......................................................................... 252 Unwitnessed falls ..................................................................................... 253 Inpatient falls prevention .......................................................................... 253 5 Further information ............................................................................... 254 5.1 Guideline development group information ...................................... 254 5.2 Notes on the scope of the guideline ................................................ 260 5.3 Implementation ............................................................................... 260 5.4 Other versions of this guideline ....................................................... 260 5.5 Related NICE guidance .................................................................. 261 5.6 References ..................................................................................... 262 5.7 Glossary and abbreviations ............................................................ 312 All appendices are in separate files.
Page 4 of 315
Introduction
Falls and fall-related injuries are a common and serious problem for older people. People aged 65 and older have the highest risk of falling, with 30% of people older than 65 and 50% of people older than 80 falling at least once a year. The human cost of falling includes distress, pain, injury, loss of confidence, loss of independence and mortality. Falling also affects the family members and carers of people who fall. Falls are estimated to cost the NHS more than 2.3 billion per year (College of Optometrists/British Geriatrics Society, 2011). Therefore falling has an impact on quality of life, health and healthcare costs. This guideline provides recommendations for the assessment and prevention of falls in older people. It is an extension to the remit of NICE clinical guideline 21 (published November 2004) to include assessing and preventing falls in older people during a hospital stay (inpatients). The 2 main parts of the guideline are as follows: Recommendations for all older people (sections 1.1 and 3). Evidence on the assessment and prevention of falls in older people has not been updated and the original 2004 recommendations remain unchanged (except for some minor wording changes for the purposes of clarification only). This part was originally developed by the National Collaborating Centre for Nursing and Supportive Care (now part of the National Clinical Guideline Centre) and published by the Royal College of Nursing. Additional recommendations for older people who are admitted to hospital (sections 1.2 and 4). New evidence has been reviewed and new recommendations have been made for assessing and preventing falls in older people during a hospital stay. This part was developed by the Internal Clinical Guidelines Programme in the Centre for Clinical Practice at NICE.
Page 5 of 315
The section for all older people (section 3) comprises the previous 2004 guideline (CG21), which has been maintained as far as possible in its original structure and style (although it has been renumbered to maintain consistency throughout the entire guideline). The additional section for older people who are admitted to hospital (section 4) contains new evidence and recommendations that were developed using The guidelines manual (2009). This has inevitably led to inconsistencies in style. For example, GRADE methodology is used in section 4 to assess the quality and strength of the evidence and recommendations, whereas in section 3 the evidence and recommendations are graded using the old evidence hierarchy. In addition, in section 3 (developed in 2004) older people are defined as those aged 65 and older, but more recently the Department of Health has recognised that there are different interpretations of older people, some of which include people from the age of 50. Section 4 of the guideline reflects this, and the new recommendations also cover people aged between 50 and 64. The different sections of the guideline contain details of the methodology used at the time of development (see appendix J [2004] and appendix J [2013]). It is important to emphasise that although guideline methodology has changed over time, all of the 2004 recommendations are just as relevant and important now as they were when they were originally published.
Page 6 of 315
Patient-centred care
This guideline offers best practice advice on the care of older people who are at risk of falling. Patients and healthcare professionals have rights and responsibilities as set out in the NHS Constitution for England all NICE guidance is written to reflect these. Treatment and care should take into account individual needs and preferences. Patients should have the opportunity to make informed decisions about their care and treatment, in partnership with their healthcare professionals. Healthcare professionals should follow the Department of Healths advice on consent. If someone does not have the capacity to make decisions, healthcare professionals should follow the code of practice that accompanies the Mental Capacity Act and the supplementary code of practice on deprivation of liberty safeguards. In Wales, healthcare professionals should follow advice on consent from the Welsh Government. NICE has produced guidance on the components of good patient experience in adult NHS services. All healthcare professionals should follow the recommendations in Patient experience in adult NHS services.
Page 7 of 315
Strength of recommendations
Some recommendations can be made with more certainty than others. The Guideline Development Group makes a recommendation based on the tradeoff between the benefits and harms of an intervention, taking into account the quality of the underpinning evidence. For some interventions, the Guideline Development Group is confident that, given the information it has looked at, most patients would choose the intervention. The wording used in the recommendations labelled [new 2013] in this guideline denotes the certainty with which the recommendation is made (the strength of the recommendation). For all recommendations, NICE expects that there is discussion with the patient about the risks and benefits of the interventions, and their values and preferences. This discussion aims to help them to reach a fully informed decision (see also Patient-centred care).
have the intervention at all, is more likely to depend on the patients values and preferences than for a strong recommendation, and so the healthcare professional should spend more time considering and discussing the options with the patient.
Page 9 of 315
Labelling of recommendations
This guideline extends and replaces NICE clinical guideline 21 (published November 2004). New recommendations have been added about preventing falls in older people during a hospital stay (labelled [2013]). The original recommendations from NICE clinical guideline 21 are incorporated unchanged (except for minor wording changes for the purposes of clarification only). These are labelled [2004] or [2004, amended 2013].
Page 10 of 315
footwear that is unsuitable or missing health problems that may increase their risk of falling medication postural instability, mobility problems and/or balance problems syncope syndrome visual impairment. [new 2013]
Page 12 of 315
1
1.1
1.1.1
1.1.1.1
Recommendations
Preventing falls in older people
Case/risk identification
Older people in contact with healthcare professionals should be asked routinely whether they have fallen in the past year and asked about the frequency, context and characteristics of the fall/s. [2004]
1.1.1.2
Older people reporting a fall or considered at risk of falling should be observed for balance and gait deficits and considered for their ability to benefit from interventions to improve strength and balance. (Tests of balance and gait commonly used in the UK are detailed in section 3.3.) [2004]
1.1.2
1.1.2.1
1.1.2.2
Multifactorial assessment may include the following: identification of falls history assessment of gait, balance and mobility, and muscle weakness assessment of osteoporosis risk assessment of the older persons perceived functional ability and fear relating to falling assessment of visual impairment assessment of cognitive impairment and neurological examination
Page 13 of 315
assessment of urinary incontinence assessment of home hazards cardiovascular examination and medication review. [2004]
1.1.3
1.1.3.1
Multifactorial interventions
All older people with recurrent falls or assessed as being at increased risk of falling should be considered for an individualised multifactorial intervention. [2004] In successful multifactorial intervention programmes the following specific components are common (against a background of the general diagnosis and management of causes and recognised risk factors): strength and balance training home hazard assessment and intervention vision assessment and referral medication review with modification/withdrawal. [2004]
1.1.3.2
Following treatment for an injurious fall, older people should be offered a multidisciplinary assessment to identify and address future risk and individualised intervention aimed at promoting independence and improving physical and psychological function. [2004]
1.1.4
1.1.4.1
Page 14 of 315
1.1.5
1.1.5.1
1.1.6
1.1.6.1
1.1.6.2
Home hazard assessment is shown to be effective only in conjunction with follow-up and intervention, not in isolation. [2004]
1.1.7
1.1.7.1
Psychotropic medications
Older people on psychotropic medications should have their medication reviewed, with specialist input if appropriate, and discontinued if possible to reduce their risk of falling. [2004]
1.1.8
1.1.8.1
Cardiac pacing
Cardiac pacing should be considered for older people with cardioinhibitory carotid sinus hypersensitivity who have experienced unexplained falls. [2004]
1.1.9
1.1.9.1
To promote the participation of older people in falls prevention programmes the following should be considered. Healthcare professionals involved in the assessment and prevention of falls should discuss what changes a person is willing to make to prevent falls.
Page 15 of 315
Information should be relevant and available in languages other than English. Falls prevention programmes should also address potential barriers such as low self-efficacy and fear of falling, and encourage activity change as negotiated with the participant. [2004] 1.1.9.2 Practitioners who are involved in developing falls prevention programmes should ensure that such programmes are flexible enough to accommodate participants different needs and preferences and should promote the social value of such programmes. [2004]
1.1.10
1.1.10.1
1.1.10.2
Individuals at risk of falling, and their carers, should be offered information orally and in writing about: what measures they can take to prevent further falls how to stay motivated if referred for falls prevention strategies that include exercise or strength and balancing components the preventable nature of some falls the physical and psychological benefits of modifying falls risk where they can seek further advice and assistance how to cope if they have a fall, including how to summon help and how to avoid a long lie. [2004]
Page 16 of 315
1.1.11
1.1.11.1
1.1.12
We do not recommend implementation of the following interventions at present. This is not because there is strong evidence against them, but because there is insufficient or conflicting evidence supporting them 1. [2004] 1.1.12.1 Low intensity exercise combined with incontinence programmes. There is no evidence1 that low intensity exercise interventions combined with continence promotion programmes reduce the incidence of falls in older people in extended care settings. [2004] 1.1.12.2 Group exercise (untargeted). Exercise in groups should not be discouraged as a means of health promotion, but there is little evidence1 that exercise interventions that were not individually prescribed for older people living in the community are effective in falls prevention. [2004] 1.1.12.3 Cognitive/behavioural interventions. There is no evidence1 that cognitive/behavioural interventions alone reduce the incidence of falls in older people living in the community who are of unknown risk status. Such interventions included risk assessment with feedback and counselling and individual education discussions. There is no evidence1 that complex interventions in which group activities included education, a behaviour modification programme
1
Page 17 of 315
aimed at moderating risk, advice and exercise interventions are effective in falls prevention with older people living in the community. [2004] 1.1.12.4 Referral for correction of visual impairment. There is no evidence2 that referral for correction of vision as a single intervention for older people living in the community is effective in reducing the number of people falling. However, vision assessment and referral has been a component of successful multifactorial falls prevention programmes. [2004] 1.1.12.5 Vitamin D. There is evidence2 that vitamin D deficiency and insufficiency are common among older people and that, when present, they impair muscle strength and possibly neuromuscular function, via CNS-mediated pathways. In addition, the use of combined calcium and vitamin D3 supplementation has been found to reduce fracture rates in older people in residential/nursing homes and sheltered accommodation. Although there is emerging evidence2 that correction of vitamin D deficiency or insufficiency may reduce the propensity for falling, there is uncertainty about the relative contribution to fracture reduction via this mechanism (as opposed to bone mass) and about the dose and route of administration required. No firm recommendation can therefore currently be made on its use for this indication.3 [2004, amended 2013] 1.1.12.6 Hip protectors. Reported trials that have used individual patient randomisation have provided no evidence2 for the effectiveness of hip protectors to prevent fractures when offered to older people living in extended care settings or in their own homes. Data from cluster randomised trials provide some evidence2 that hip
2 3
This refers to evidence reviewed in 2004. The following text has been deleted from the 2004 recommendation: Guidance on the use of vitamin D for fracture prevention will be contained in the forthcoming NICE clinical practice guideline on osteoporosis, which is currently under development. As yet there is no NICE guidance on the use of vitamin D for fracture prevention.
Page 18 of 315
protectors are effective in the prevention of hip fractures in older people living in extended care settings who are considered at high risk. [2004]
1.2
1.2.1
1.2.1.1
1.2.1.2
Regard the following groups of inpatients as being at risk of falling in hospital and manage their care according to recommendations 1.2.2.1 to 1.2.3.2: all patients aged 65 years or older patients aged 50 to 64 years who are judged by a clinician to be at higher risk of falling because of an underlying condition. [new 2013]
1.2.2
1.2.2.1
1.2.2.2
For patients at risk of falling in hospital (see recommendation 1.2.1.2), consider a multifactorial assessment and a multifactorial intervention. [new 2013]
1.2.2.3
Ensure that any multifactorial assessment identifies the patients individual risk factors for falling in hospital that can be treated, improved or managed during their expected stay. These may include: cognitive impairment continence problems
Page 19 of 315
falls history, including causes and consequences (such as injury and fear of falling) footwear that is unsuitable or missing health problems that may increase their risk of falling medication postural instability, mobility problems and/or balance problems syncope syndrome visual impairment. [new 2013] 1.2.2.4 Ensure that any multifactorial intervention: promptly addresses the patients identified individual risk factors for falling in hospital and takes into account whether the risk factors can be treated, improved or managed during the patients expected stay. [new 2013] 1.2.2.5 Do not offer falls prevention interventions that are not tailored to address the patients individual risk factors for falling. [new 2013]
1.2.3
1.2.3.1
Page 20 of 315
helping the patient to engage in any multifactorial intervention aimed at addressing their individual risk factors. [new 2013] 1.2.3.2 Ensure that relevant information is shared across services. Apply the principles in Patient experience in adult NHS services (NICE clinical guideline 138) in relation to continuity of care. [new 2013]
Page 21 of 315
Care pathway
Older Older person person aged aged 65 65 and and older older
Healthcare professionals should routinely ask older people (aged 65 and older) about falls
Older person aged 65 or older Identified as being at risk* of falling; not admitted to hospital
*presents for medical attention with a fall, reports recurrent falls and/or has abnormalities of gait or balance
Older person aged 65 or older Admitted to hospital with a fall, or admitted to hospital for any reason and has a history of falls
Older person aged 65 or older Admitted to hospital for any reason Has not fallen and has no history of falls
Older person aged 50 to 64 Admitted to hospital for any reason Identified by a clinician as being at higher risk of falling
Multifactorial risk assessment to identify any risk factors for falling and specific risk factors for falling in hospital
NB the assessments may be done together or separately
Multifactorial risk assessment to identify specific risk factors for falling in hospital
Multifactorial intervention to address identified risk factors for falling and specific risk factors for falling in hospital
Multifactorial intervention to address identified specific risk factors for falling in hospital
Encourage the participation of older people in falls prevention programmes Provide information and education
Page 22 of 315
This work was undertaken by the National Collaborating Centre for Nursing and Supportive Care (NCC-NSC) and the Guideline Development Group (GDG) formed to develop this guideline. Funding was received from the National Institute for Clinical Excellence (NICE). The NCC-NSC consists of a partnership between: Centre for Evidence-Based Nursing; Centre for Statistics in Medicine; Clinical Effectiveness Forum for Allied Health Professionals, College of Health; Health Care Libraries (University of Oxford); Health Economics Research Centre, Royal College of Nursing and UK Cochrane Centre.
Disclaimer
As with any clinical guideline, recommendations may not be appropriate for use in all circumstances. A limitation of a guideline is that it simplifies clinical decision-making (Shiffman 1997). Decisions to adopt any particular recommendations must be made by the practitioners in the light of: available resources local services, policies and protocols the patients circumstances and wishes available personnel and devices clinical experience of the practitioner knowledge of more recent research findings.
3.1
Executive summary
The National Institute for Clinical Excellence (NICE) commissioned the National Collaborating Centre for Nursing and Supportive Care (NCC-NSC) to develop guidelines on the assessment and prevention of falls in older people. This follows referral of the topic by the Department of Health and Welsh Assembly Government. This document describes the methods for developing
Page 23 of 315
the guidelines and presents the resulting recommendations. It is the source document for the NICE (abbreviated version for health professionals) and Information for the public (patient) versions of the guidelines that are published by NICE. A multidisciplinary Guideline Development Group produced the guidelines and the development process was undertaken by the NCC-NSC. The main areas examined by the guideline were: The evidence for factors that increase the risk of falling. The most effective methods of assessment and identification of older people at risk of falling. The most clinically and cost effective interventions and preventative strategies for the prevention of falls. The clinical effectiveness of hip protectors for the prevention of hip fracture. The most clinically and cost effective interventions and rehabilitation programmes for the prevention of further falls. Older peoples views and experiences of falls prevention strategies and programmes. Recommendations for good practice based on the best available evidence of clinical and cost effectiveness are presented. Evidence published after October 2003 was not considered. Health care professionals should use their clinical judgement and consult with patients when applying the recommendations, which aim to reduce the negative physical, social and financial impact of falling.
Page 24 of 315
Background In March 2002, the National Collaborating Centre for Nursing and Supportive Care (NCC-NSC) was commissioned by NICE to develop clinical guideline on the assessment and prevention of falls in older people for use in the NHS in England and Wales. The remit from the DH and Welsh Assembly Government was as follows: To prepare clinical guidelines for the NHS in England and Wales for the assessment and prevention of falls, including recurrent falls in older people; with an associated clinical audit system. Clinical need Falls are a major cause of disability and the leading cause of mortality resulting from injury in people aged above 75 in the UK (Scuffham & Chaplin 2002). Furthermore, more than 400,000 older people in England attend accident and emergency departments following an accident, while up to 14,000 people die annually in the UK as a result of an osteoporotic hip fracture (National Service Framework for Older People 2001). Its clear that falling has an impact on quality of life, health and health care costs. Falls are not an inevitable result of ageing, but they do pose a serious concern to many older people and to the health system. Older people have a higher risk of accidental injury that results in hospitalisation or death than any other age group (Cryer 2001). The Royal Society for the Prevention of Accidents (ROSPA) estimates that one in three people aged 65 years and over experience a fall at least once a year rising to one in two among 80 yearolds and older. Although most falls result in no serious injury, approximately 5 per cent of older people in community-dwelling settings who fall in a given year experience a fracture or require hospitalisation (Rubenstein et al. 2001). Incidence rates for falls in nursing homes and hospitals are two to three times greater than in the community and complication rates are also considerably higher. Ten to 25 per cent of institutional falls result in fracture, laceration or need for hospital care (Rubenstein 2001).
Page 25 of 315
The key issue of concern is not simply the high incidence of falls in older people since children and athletes have a very high incidence of falls but rather the combination of a high incidence and a high susceptibility to injury (Rubenstein 2001). In 1999, there were 647,721 A&E attendances and 204,424 admissions to hospital for fall-related injuries in the UK population aged 60 years or over (Scuffham and Chaplin 2002). The associated cost of these falls to the NHS and PSS was 908.9 million and 63 per cent of these costs were incurred from falls in those aged 75 years and over (Scuffham and Chaplin 2002). In addition, 86, 000 hip fractures occur annually in the UK (Torgerson 2001) and 95 per cent of hip fractures are the result of a fall (Youm 1999). Although only 5 per cent of falls result in fracture (Tinetti 1988), the total annual cost of these fractures to the NHS has been calculated as 1.7 billion (Torgerson 2001) with many individuals losing independence and quality of life (Cooper 1993). Some older people have stated that they would rather die than fracture their hip and have to live in a nursing home (Salkeld 2000). Although most falls do not result in serious injury, the consequences for an individual of falling or of not being able to get up after a fall can include: psychological problems, for example, a fear of falling and loss of confidence in being able to move about safely loss of mobility, leading to social isolation and depression increase in dependency and disability hypothermia pressure-related injury infection. Falls have a multifactorial aetiology, with more than 400 separate risk factors described (Oliver 2000). The major risk factors for falling are diverse, and many of them such as balance impairment, muscle weakness, polypharmacy and environmental hazards are potentially modifiable. Since the risk of falling appears to increase with the number of risk factors, multifactorial interventions have been suggested as the most effective
Page 26 of 315
strategy to reduce declines in function and independence and also to prevent the associated costs of complications (Gillespie et al. 2001). Preventive programmes based on risk factors for falling include exercise programmes, education programmes, medication review, environmental modification in homes or institutions and nutritional or hormonal supplementation (Cummings et al. 2001). Interventions need to target extrinsic factors such as hazards within the home environment and intrinsic risk factors, such as mobility, strength, gait, medicine use and sensory impairment (HDA 2002). Numerous interventions have been studied in the prevention of falls. Few trials have been carried out in the UK. The prevention and management of falls in older people is a key Government target in reducing morbidity and mortality. This is outlined in the National Service Framework (NSF) for England, standard six for older people, which covers falls and specifically aims to: reduce the number of falls which result in serious injury and ensu re effective treatment and rehabilitation for those who have fallen (NSF 2001). The NSF also outlines key changes needed to reduce the number of falls and their impact by: a) prevention including the prevention and treatment of osteoporosis b) improving the diagnosis, care and treatment of those who have fallen c) rehabilitation and long-term support d) ensuring that older people who have fallen receive effective treatment and rehabilitation e) ensuring that patients and their carers receive advice on prevention, through a specialised falls service.
Page 27 of 315
In the light of the serious and costly impact of falls in the community and longterm care setting among older people, plus the potential of interventions to positively influence this problem, risk assessment and preventative interventions were selected as the focus for this NICE guideline. These guidelines will support the implementation of standards two and six of the National Service Framework for Older People in England (2001).
Page 28 of 315
3.2
Aims
To evaluate and summarise the evidence for assessing and preventing falls in older people. To highlight gaps in the research evidence. To formulate evidence-based and, where possible, clinical practice recommendations on the assessment of older people and prevention of falls in older people based on the best evidence available to the GDG. To provide audit criteria to assist with the implementation of the recommendations.
3.2.1
As detailed in the guideline scope, the guideline is of relevance to: those older people aged 65 and above who are vulnerable to or at risk of falling families and carers health care professionals who share in caring for those who are vulnerable or at risk of falling those responsible for service delivery.
3.2.2
The recommendations made in the guideline cover the care of older people: a) in the community or extended care, who are at risk of falling or who have fallen b) who attend primary or secondary care settings, following a fall.
3.2.3
Page 29 of 315
a) hospitalised patients who sustain a fall while in hospital or who may be at risk of falling during hospitalisation4 b) people who are confined to bed for the long-term.
3.2.4
This guideline makes recommendations on the care given by health care professionals who have direct contact with and make decisions concerning the care of older people who have fallen or are at risk of falling. It also makes recommendations on the care given by health care professionals or carers where applicable, involved in the care of older people who have been taken to hospital following a fall. This is an NHS guideline, but also addresses the interface with other services, such as those provided by social services, secure settings, care homes and the voluntary sector. It does not include services exclusive to these sectors.
3.2.5
Interventions covered
The following interventions are covered: exercise, including balance training multifactorial interventions packages of care, for example, exercise, education and home modifications vision assessment and correction of impaired vision home hazard assessment and modification patient and staff education medication review hip protectors rehabilitation strategies. Podiatric interventions were in the scope of the guideline, however no controlled trials were identified with falls as an outcome.
4
Hospitalised patients who sustain a fall while in hospital or who may be at risk of falling during hospitalisation are included in section 4 of this guideline.
Page 30 of 315
Recommendations also take account of the psychosocial aspects of falling, including fear of falling and loss of confidence resulting from a fall.
3.2.6
The prevention and treatment of osteoporosis (currently guidelines on this area are being developed by NICE). The management of hip and other fractures. The prevention of falls in acute settings.
3.2.7
The guideline recommendations were developed by a multidisciplinary and lay GDG convened by the NICE-funded NCC-NSC, with membership approved by NICE. Members include representatives from: nursing general practice allied health NSF working party falls researchers falls clinicians patient groups. A list of GDG members is attached (Appendix A, 2004). The GDG met eight times between September 2002 and December 2003. All members of the GDG were required to make formal declarations of interest at the outset, which were recorded. GDG members were also asked to declare interests at the beginning of each GDG meeting. This information is recorded in the meeting minutes and kept on file at the NCC-NSC.
Page 31 of 315
3.3
Methods
This section describes the systematic review methods used to inform the clinical questions. Results are presented that provided the basis for the evidence statements and recommendations, which are reported in Section 3.4.
3.3.1
The methods used to develop this guideline are based on those outlined by Eccles and Mason (2001) and in the draft NICE technical manual. The structure of the recommendations section (Section 3.4) that is recommendations; evidence statements, evidence narrative and GDG commentary came from McIntosh et al. (2001). The following sources of evidence were used to inform the guideline: The Cochrane reviews: a) Interventions for the prevention of falls in older people (Gillespie et al. 2003) and b) Hip protectors for the prevention of hip fractures (Parker et al. 2003). American Geriatric Society/British Geriatric Society (2001) clinical guidelines that were based on the systematic review Falls prevention interventions in the Medicare population (Shekelle et al. 2002). Analysis of epidemiological data relating to risk factors (NCC-NSC). Reviews of assessment processes, tools, tests and instruments for identifying those at risk (NCC-NSC). Review of studies examining patients views and experiences of falls prevention programmes and methods to maximise participation (NCC-NSC). Reviews of studies on fear of falling and interventions to reduce the psychosocial consequences of falling (NCCNSC). Reviews of the evidence on costs and economic evaluations (SCHARR). Reviews of rehabilitation strategies (NCC-NSC).
Page 32 of 315
The stages used to develop this guideline were as follows: develop scope of guideline convene multidisciplinary GDG review questions set identify sources of evidence retrieve potential evidence evaluate potential evidence utilise the updated Cochrane reviews Interventions for preventing falls in older people (2003) and Hip protectors (2003) utilise the AGS/BGS clinical guidelines and Shekelle systematic review (2002) undertake systematic review on guideline areas not covered by either the Cochrane review, AGS/BGS guidelines and Shekelle review extract relevant data from studies meeting methodological and clinical criteria interpret each paper, taking into account the results including, where reported, the beneficial and adverse effects of the interventions; cost; acceptability to patients; level of evidence; quality of studies; size and precision of effect;and relevance and generalisability of included studies to the scope of the guideline prepare evidence reviews and tables that summarise and grade the body of evidence formulate conclusions about the body of available evidence, based on the evidence reviews, by taking into account the factors above agree final recommendations and apply recommendation gradings submit first drafts short and full versions of guidelines for feedback from NICE registered stakeholders GDC to consider stakeholders comments, following first stage consultation submit final drafts of all guideline versions including Information for the public version and algorithm to NICE for second stage of consultation GDG to consider stakeholders comments final copy submitted to NICE.
Page 33 of 315
Questions addressed by the evidence reviews included: What is the best method of identifying those at highest risk of a first or subsequent fall? (Source of evidence: risk factor evidence review) What assessment tool or process should be used to identify modifiable risk factors for falling? (Source of evidence: assessment evidence review) What are the most clinically effective and cost effective methods for falls prevention? (Source of evidence: clinical and cost effectiveness reviews) What interventions are there to reduce the psychosocial consequences of falling? (Source of evidence: Cochrane review) What is the evidence for the effectiveness of hip protectors? (Cochrane review) What is the best method for maximising participation and compliance in falls prevention programmes and modification of specific risk factors, for example, medication withdrawal/review? (Source of evidence: patients views and experiences) Are falls prevention programmes acceptable to patients? (Source of evidence: patients views and experiences review) What is the best method of rehabilitation/intervention/process of care following a fall requiring treatment? (Source of evidence: rehabilitation review, hip protector review and Cochrane falls prevention review) The methods and the main results for each review are reported in Sections 3.3.1 to 3.3.14. The detailed evidence summaries including economic evidence, where relevant evidence statements, GDG considerations and recommendations are in Section 3.4.
Page 34 of 315
3.3.2
3.3.2.1
To identify those at risk of falling, it is necessary to review the evidence base for risk factors, looking at older people in both community dwelling and residential/extended care settings. Although some risk factors are intuitive, an examination of the empirical evidence provides a comprehensive and thorough overview, with information on the risk factors that should be considered for inclusion in screening/assessment tools and protocols. Because the literature in this area is vast, the evidence statements and recommendations presented in the American and British Geriatric Society (AGS/BGS) 2001 guidelines, and an analytic review by Perell et al. (2001) formed the foundation for the current review. The Perell review provided information on the assessment of older people at risk and a summary of the risk factors predictive of falling. This section reports the findings of these key documents and the review of evidence undertaken to update these documents. Although risk factors for subsequent falls have face validity (Colon-Emeric & Laing 2002), interpretation of the evidence base is often problematic. A variety of study designs have been employed to study this topic, with resulting issues of bias and confounding. This means that summarising such studies is challenging. Furthermore, there is no formal guidance on how best to review the risk factor evidence base. The gold standard approach for researching risk factors is to carry out a prospective cohort study, in which predictors or risk factors are recorded at baseline, and participants are followed-up, with falls outcomes measured. Often study designs, such as case-control and cross-sectional, are used but these are more susceptible to confounding and other biases (Eggar et al. 2001). Therefore, to build on the existing evidence base (provided by the AGS/BGS guidelines and the Perell review), we restricted the review to evidence from
Page 35 of 315
prospective cohort studies. This decision was made following initial screening of search results, which indicated that many different study designs have been used to attempt to identify risk factors, and after consultation with methodological experts. The time and resources available to undertake an evidence review on this complex topic also provided further justification for restricting the study design criteria. 3.3.2.2 Objectives
The review sought to answer the following question: What are the key risk factors that should be used to identify those at highest risk of a first or subsequent fall? 3.3.2.3 Selection criteria
Types of studies Reviews of risk factors with preference given to systematic reviews. Prospective cohort studies of risk factors of falls in older people who are either community-dwelling or living in extended care settings. Types of participants Older people aged 65 and over. Types of outcome Those studies that report falls as an outcome. Risk factors that were conceptually relevant. Explicit details of how risk factors were measured. 3.3.2.4 Search strategy
Twelve electronic databases were searched between 1998 and December 2002, using a sensitive search strategy used for both the risk factor and risk assessment review questions. The bibliographies of all retrieved and relevant publications were searched for further studies. Following guidance from NICE, we searched from the present, looking back over a five-year period, to assess the likely volume of papers that would
Page 36 of 315
require eligibility assessment and critical appraisal. The volume of papers requiring screening and appraisal was considerable. As we were contributing to existing evidence bases (Perell 2001; AGS/BGS 2001), which would have captured the key studies prior to 1998, no further searching was carried out. Hand searching was not undertaken following NICE advice that exhaustive searching on every guideline review topic is not practical and efficient (Mason et al. 2002). (Note: this applies to all reviews reported here, except for the Cochrane reviews summarised here). Reference lists of articles were checked for articles of potential relevance (Note: this was done for all reviews reported in this guideline and will not be repeated in other methods sections). The search strategies and the databases searched are presented in Appendix B 2004. All searches were comprehensive and included a large number of databases. 3.3.2.5 Sifting process
Once articles were retrieved the following sifting process took place: First sift: for material that potentially meets eligibility criteria on basis of title/abstract by one reviewer. Second sift: full papers ordered that appear relevant and eligible and where relevance/eligibility not clear from the abstract. Third sift: one reviewer appraised full articles that met eligibility criteria. Time did not allow for an independent reviewer to identify and appraise studies. (Note: this sifting process applies to all of the non-Cochrane reviews reported in this document and will not be repeated). 3.3.2.6 Data abstraction
Papers were screened for relevance and prospective cohort studies identified. Methodological quality was assessed using pre-defined principles as outlined
Page 37 of 315
in 4.5.2.7 and epidemiological appraisal criteria, which were adapted for this review. Data were extracted by a single reviewer and evidence tables compiled. The following information was extracted: Author, setting,number of participants at baseline and follow-up, methods and details of baseline and outcome measurement, results including summary statistics and 95 per cent confidence intervals, and comments made on the methodological quality. Masked assessment whereby data extractors are blind to the details of journal, authors etc was not undertaken because there is no evidence to support the claim that this minimises bias. 3.3.2.7 Appraisal of methodological quality
Each study was assessed against the following quality criteria: Selection Cohort of eligible older people with well defined demographic information. High recruitment rate of participants equal to or greater than 80 per cent of those approached. Identification of risk factors Risk factors conceptually relevant. Explicit details of how risk factor information is measured. Confounding Statistical adjustment carried out/ sensitivity analysis. Analytic methods described. Follow-up/outcomes Method of measurement of outcome given.
Page 38 of 315
Where quality was low, this is indicated in the evidence tables (Evidence table 1). 3.3.2.8 Data synthesis
No quantitative analysis was carried out for this review. Summary statistics and vote counting of statistical significance for each risk factor were reported in the evidence tables. 3.3.2.9 Details of studies included in the review
Results of the search and sift are shown in Table 1 below. TABLE 1: SIFTING RESULTS FOR RISK FACTOR REVIEW
Initial search results N screened for relevance following sift N identified as relevant N included N excluded 1396 223 37 28 9
Participants and settings Most studies reported findings from community-dwelling participants with varying sample sizes, method of recruitment, participation and follow-up rates. Three studies were conducted in an extended care setting. Baseline data collected ranged from detailed socio-demographic characteristics and full examination of health and functioning. Methodological quality of studies The quality of the identified studies that met the inclusion criteria was variable. Shortcomings included: self-reported data, low participation and follow-up rates; no details of how outcomes were ascertained; small sample sizes; no information on reliability and validity of outcome ascertainment. Often no justification was given for the selection of risk factors to study. Outcome measurement Methods of data collection included self-completed questionnaires, face-toface interview and full medical examination. Measurement of baseline data
Page 39 of 315
included self-report of falls history as a predictor, relying on the participants recall of events. Other measurements, such as participants perception of health status and functioning, were often recorded using self-reported rating scales, which are subjective and prone to bias. Outcome measurement also differed between studies and included: a final interview with a self-reported fall record during the follow-up period; falls diaries completed weekly by participants and posted monthly to researchers; and examination of medical and hospital admission records of fall events of the participants. Statistical adjustment for confounding and/or sensitivity analysis was carried out in most of the studies and analytical methods described. Characteristics of excluded studies are shown in Appendix G (2004). Table 2: STATISTICAL SUMMARIES OF RISK FACTORS FOR FALLS FROM PERELL (2001)
Risk factor Muscle weakness History of falls Gait deficit Balance deficit Use of assist devices Visual deficit Arthritis Impaired activities of daily living Depression Cog impairment Age 80 Mean RR/ OR (Range) 4.4 (1.5-10.3) 3.0 (1.7-7.0) 2.9 (1.3-5.6 2.9 (1.6-5.4) 2.6 (1.2-4.6) 2.5 (1.6-3.5) 2.4 (1.9-2.9) 2.3 (1.5-3.1) 2.2 (1.7-2.5) 1.8 (1.0-2.3) 1.7 (1.1-2.5
3.3.2.10
A review of the empirical evidence relating to risk factors is provided by Perell et al. (2001). This review reported the mean relative risk (RR) or odds ratio (OR) and rank for each factor. However, no details were given of the study design of the included studies. These statistical summaries are reproduced in Table 2. The included studies from the evidence update are presented in Evidence table 1 (Appendix E, 2004). Results of the studies are presented as either
Page 40 of 315
relative risk or odds ratios. The risk factors reported in the evidence table of included studies are those that were reported as statistically significant. Individual risk factors from the evidence update are summarised below. Table 3, column 3 reports the frequency that the risk factor was reported in the included studies. Heterogeneity between studies prohibited aggregation of results. TABLE 3: FREQUENCY OF REPORTING OF RISK FACTOR IN INCLUDED STUDIES
Risk factor Falls history Mobility impairment Visual impairment Balance deficit Gait deficit Mental status Functional dependence Fear Low body mass Depression Diabetes Environmental hazards Incontinence Multiple medications Anti-arrhythmic Psychotropic drugs RR/OR Range OR = 2.4-2.6 RR = 1.9-2.4 OR = 2.0-3.0 OR = 2.6-5.8 RR = 1.6 OR =1.8-3.9 RR = 1.7 OR = 1.8-2.2 RR = 2.2 OR = 2.2-6.7 RR = 6.2 OR = 1.7 RR = 5.6 OR = 1.7-2.8 OR = 1.8-4.1 OR = 1.5-2.2 OR = 3.8-4.1 OR = 2.3-2.5 OR = 1.8-2.3 OR = 2.02-3.16 OR = 1.59 OR = 1.66 (1.40-1.97) Mean RR/OR (Range) 11 8 5 5 4 4 4 3 3 3 2 2 2 Meta-analysis: n=14 studies Meta-analysis: n=10 studies Meta-analysis: n=11 studies
In addition to those risk factors shown in Table 3, other risk factors were reported as significant in single studies that is those studies reporting on one risk factor as follows: generalised pain
Falls: NICE clinical guideline 161 (June 2013) Page 41 of 315
reduced activity high alcohol consumption parkinsons disease arthritis diabetes stroke low body mass. Whilst identification of single risk factors is informative, especially when planning interventions for prevention, it is also the interaction between multiple risk factors that needs to be considered (AGS/BGS 2001). Furthermore, within study analysis demonstrates association of different factors. Further details are reported in Evidence table 1 but a brief summary of such studies is presented below. Covinsky et al. (2001) carried out regression analysis with significant risk factors and a final model (model 3) suggested that abnormal mobility, balance deficit and previous falls history were predictive of further falls. Stalenhoef et al. (2002) developed a risk model with postural sway, falls history, reduced grip strength and depression as significant predictors. Cwikel et al. (1998) developed a risk model (elderly falls screening test), which included: fall in last year, injurious fall in last year, frequent falls, slow walking speed, and unsteady gait. It is clear from the evidence that a previous fall and/or gait and balance disorders may be predictive of those at highest risk, but the presence of other less obvious factors should be considered in combination. The results described above were obtained mainly from community-dwelling participants. The results from studies conducted with extended care participants were similar, in that a previous fall was predictive of a further fall. Medications also featured as important risk factors for both those in community and extended care settings for example, benzodiazepines, antidepressants, neuroleptics and cardiotonic glycosides as single predictors, but also the use of multiple medications (Leipzig et al. 1999).
Page 42 of 315
Analysis of multivariate studies of risk factors for falling of the included studies displayed in Evidence table 1, some reported adjusted summary statistics in which multivariate analysis had been carried out. Others had conducted bivariate analysis, with the reporting of unadjusted significant factors. Therefore, to assist with clarification of the risk factor evidence, the multivariate studies were analysed in depth. This section reports on: a detailed examination of studies in which multivariate analysis had been carried out further detailed examination of the quality of each multivariate study the results for each risk factor. Methods Multivariate analysis allows for the efficient estimate of measures of association, while controlling for a number of confounding factors simultaneously. Mathematical multivariate regression models include: linear regression when the dependant outcome variable is continuous data logistical regression for binary data. While this information can be obtained from the studies included in our evidence review, there were several associated methodological issues that made data extraction and synthesis of the multivariate studies difficult. These included: a) different methods of analysis are employed within each study b) methods of conducting systematic reviews of prognostic studies are unclear. The clinical interpretability of information from each study and risk factors is both complex and challenging due to the heterogeneity of the studies. Methodological advice was sought on how to best appraise the studies and how to illustrate the results in a rigorous, but clinically relevant and meaningful way. We were advised to extract adjusted summary statistics and report
Page 43 of 315
details of both the statistical methods and adjusted variables within each study. To aid interpretation, these results were presented in an evidence table (Evidence table 2, Appendix E) and a narrative summary was produced. Study design inclusion criteria Prospective cohort studies with multivariate statistical analysis, including those studies reporting statistical significance for the specified risk factor. Also included are studies reporting statistically non-significant results. This avoids introducing reporting bias. Detailed quality assessment of risk factor studies Studies were quality assessed using the following criteria. All studies had to fulfil the following criteria for inclusion: eligible cohort of participants high participation at baseline and follow-up > 70 per cent risk factors conceptually relevant baseline measurement of risk factors reporting of methods, explicit inclusion criteria and demographic information adequate length of follow-up > six months measurement of falls as outcome statistical methods detailed. Adequate reporting for data extraction. For methods of adjustment for confounding reported, see below. Quality was then classified as follows: High quality large sample >200 high participation at baseline and follow-up > 80 per cent baseline measurement of risk factors: clear methods of measurement given. Balance between clinical tests and subjective measurement methods of outcome measurement clear. Falls diaries with frequent researcher follow-up. Minimal reliance on recall of fall events
Page 44 of 315
methods of adjustment: all factors adjusted and reported. Medium quality large sample >200 participation at baseline and follow-up 70-80 per cent baseline measurement of risk factors: unclear methods of measurement given. Subjective methods of measurement. or methods of outcome measurement clear. Inadequate measurement of outcome that is relying on memory at follow-up alone methods of adjustment: Some adjustment and reporting. Low quality small sample < 200 low participation at baseline and follow-up < 70 per cent baseline measurement of risk factors: unclear methods of measurement given. Subjective methods of measurement. or methods of outcome measurement clear. Inadequate measurement of outcome that is relying on memory at follow-up alone methods of adjustment: adjusted variables not reported. Data abstraction Evidence table 1 (Appendix E, 2004) from the previous review formed the basis of data extraction, but further details of statistical methods were extracted from the original paper. Studies were quality assessed using the criteria above. For each risk factor, the following were extracted: Study reference, risk factor, summary statistic and 95 per cent confidence intervals, adjustment variables and method of multivariate analysis, quality of study.
Page 45 of 315
Results Twenty-four of the 31 risk factor studies had conducted multivariate analysis. The studies were characterised by heterogeneity, for example: different summary statistics were reported different methods of measurement of baseline characteristic were used different aspects of particular risk factors were measured. While this is useful to describe factors within domains, it was more difficult to combine for graphical representation falls outcome measurement included single fallers, two or more falls and recurrent fallers. Quality gradings of each study are shown in Evidence table 2 (Appendix E, 2004). Heterogeneity between studies prohibited aggregation of results and, where stated, crude estimate of the range of both RR and OR is provided. Evidence summary Evidence table 2 (Appendix E, 2004) describes the included prospective cohort studies in which multivariate analysis had been conducted. The results are reported for each risk factor and include both the statistically significant and non-significant summary statistics following multivariate analysis. Nonsignificant results were reported to avoid introducing reporting bias. Each factor is also reported by setting. The following (Table 4) summarises Evidence table 2 and provides a frequency count of significant and nonsignificant results, based on the multivariate.
Page 46 of 315
TABLE 4: FREQUENCY COUNT OF SIGNIFICANT AND NONSIGNIFICANT RESULTS FOR MULTIVARIATE RISK FACTOR STUDIES
Risk factor N = reporting statistical significance in multivariate analysis 10 2 3 4 3 3 3 2 2 N = reporting non statistically significant results in multivariate analysis 7 4 8 8 6 9 1 2 5
Falls history Mobility impairment Visual impairment Balance deficit Gait deficit Cognitive impairment Fear Environmental hazards Muscle weakness Incontinence
This further analysis indicated that the following factors were most predictive of falling and should be considered by clinicians responsible for assessing those at risk of falling: Community-dwelling older people Falls history Gait deficit Balance deficit Mobility impairment Fear Visual impairment
Cognitive impairment Urinary incontinence Home hazards. People cared for in extended care settings Falls history Gait deficit
Page 47 of 315
Page 48 of 315
3.3.3
3.3.3.1
Background
The purpose of assessment is to identify those at risk of falling in order to target effective intervention(s). There are many falls assessment instruments that have been developed for specific purposes and settings. Many have been developed for use by specific health care professionals for communitydwelling individuals and those receiving care in residential/extended care settings. Other assessment instruments, functional observations and clinical tests have been developed and tested with older people in different settings and vary in their detail and administration. Perell (2001) categorises such tools as follows: detailed medical examination and assessment of generic problems. nursing assessment by means of a scale with a scoring method. Low or high scores will trigger further investigation or planning of interventions. functional assessment or gait and balance limitation assessment to predict those likely to fall. The aim of the current review was to provide information on the most well developed and pragmatic tools available for use in community and extended care settings. Following methodological advice, key narrative reviews summarising assessment tools was used as a starting point for determining the scope of the review. These reviews suggested which tools were most advanced in their development and might be most useful for consideration in clinical practice. These tools were then profiled (see Evidence table 3, Appendix E, 2004), drawing on key primary studies with details provided of their development and properties. A systematic review was not undertaken because of the size of the literature associated with each tool. However, a range of key tools was identified, reviewed and presented. GDG input then assessed the value and utility of
Falls: NICE clinical guideline 161 (June 2013) Page 49 of 315
3.3.3.2
Objectives
The review sought to answer the following question: What assessment tool (or process) should be used to identify modifiable risk factors for falling and those at high risk of falling? 3.3.3.3 Selection criteria
Types of studies Narrative reviews were used as the principal source of evidence and further evidence was obtained from primary studies that described a particular tool. Narrative reviews were sought that provided information about currently available risk assessment instruments utilised in community dwelling and extended care settings. Primary studies describing the development of the most frequently cited risk assessment tools, the measurement properties and clinical utility of such tools were sought. Exclusion criteria Individual, newly developed and less pragmatic tools were excluded but referred to in the table of excluded studies (Appendix G 2004). Such tools include detailed analysis of gait requiring intensive training or specialist skills, and complex equipment for analysis. They are not useful as a generic tool for assessing and identifying risk. Inpatient assessment tools are excluded as this is beyond the scope of this section as it is covered in section 4. 3.3.3.4 Search strategy and sifting process
The search strategy, databases searched, dates and the sifting process are as for risk. See Sections 3.3.2.4 to 3.3.2.5.
Page 50 of 315
3.3.3.5
Data abstraction
Data were extracted by a single reviewer and evidence tables compiled. The following information was extracted: author, setting, population, objectives of tool, procedure, length of time to administer, training required, burden/acceptability to patients, measurement type, derivation of cut-off points for level of risk, further testing of the tool. 3.3.3.6 Appraisal of methodological quality
Narrative reviews and primary studies were included if they met the inclusion criteria. Where data were provided, this information was extracted. No clear quality criteria exist to appraise studies validating tools and tests for assessment. Whilst quality principles are defined for diagnostic studies (see Sackett 2000), these are not appropriate for assessing the quality of assessment tools or processes. 3.3.3.7 Data synthesis
No quantitative statistical analysis was conducted for this review. 3.3.3.8 Results of assessment evidence retrieval and appraisal
Table 5 details the sifting results and number of papers included. TABLE 5: SIFTING RESULTS
Initial search results N screened for relevance following sift N relevant N included 1396 223 46 17
Most of the evidence was extracted from identified narrative reviews (Evidence table 3, Appendix E, 2004). Supplementary evidence was obtained from included primary studies with large populations (greater than 50). Details are given of excluded studies (Appendix G, 2004). It was unrealistic to profile existing tools utilising all the original primary studies available on each tool.
Page 51 of 315
This was beyond the search scope and time limits of this review and there reached a point where no further studies could be included. Participants and settings Studies were conducted with older people in both community-dwelling settings and extended care. Assessment tools The categories of tools identified included: 1. Tests of balance and gait used in both community dwelling and extended care settings. 2. Multifactorial assessment instruments/processes administered by health care professionals for all settings, including: a) home hazard assessment instruments administered by health care professionals for community-dwelling people b) multifactorial falls risk assessment processes. 3. Minimum data set (MDS) for home care and residential settings for comprehensive assessment.
1. Tests of balance and gait used in both community-dwelling and extended care settings Table 6 illustrates the most frequently reported tools administered in community dwelling and extended care settings as identified by the review. For a full profile of each tool, readers should refer to the Evidence table 3, Appendix E (2004). TABLE 6: MOST FREQUENTLY USED TEST OF BALANCE AND GAIT
Timed up and go test Turn 180 Performance-oriented assessment of mobility problems (Tinetti scale) Functional reach Dynamic gait index Berg balance scale
Page 52 of 315
Methodological quality and type of studies Many studies reporting the development of new tools were identified, in addition to studies that tested existing tools tested on small populations. Other tests/tools exist but have limited information regarding further testing with large populations and are considered to be less useful in a clinical context. Such tools include detailed balance and gait analysis, examination of footwear and in-depth assessment of visual factors. These processes are more useful for diagnostic purposes, rather than identifying those at risk in community and extended care settings. The quality of reviews identified was variable and most were narrative with brief methods reported. Not all tests and instruments have undergone rigorous testing with large populations. Some studies use previous falls history as a reference frame and then examine whether the tool identifies the fallers from the non-fallers. Comments on the quality of information is given in the evidence table. However, it was not possible to quality assess individual references relating to each tool cited in the narrative reviews. Conclusion It is unclear which tool or assessment instrument is the most predictive and therefore useful. Many tools have undergone testing and exploration of measurement properties and predictive ability. The clinical utility, feasibility for clinicians and acceptability to patients often guides the choice of tools, but some appear more useful than others. For example, the timed up and go test (TUGT) as referred to in the AGS/BGS guidelines is both pragmatic and frequently cited, can be used in any setting, and its administration requires no special equipment. The turn 180 test is of similar value and can be administered in any setting. However, both these tests rely on clinical judgement and the value of timed cut-off values for the TUGT and number of steps for the turn 180 test need to be considered, if recommending their use. Other tests such as the Berg balance test, Tinetti scale, functional reach and dynamic gait test may offer more detailed assessment and be of diagnostic value, but take longer to administer and need both equipment and clinical
Page 53 of 315
expertise. These tests cannot be recommended for use in all settings and may be more useful during a comprehensive assessment by a multidisciplinary team. 2 & 3. Multifactorial instruments and minimum dataset instruments administered by health care professionals (all settings) There are many tools/instruments that can be administered by health care professionals. These can be categorised as follows: a) Home hazard assessment instruments, administered by health care professionals for community-dwelling population. b) Multifactorial falls risk assessment processes. c) Minimum data set (MDS) home care and residential assessment instrument for comprehensive assessment. a) Home hazard assessment instruments administered by health care professionals for community-dwelling population Home hazard assessment instruments have been developed for use by community nursing personnel, occupational therapists, and physiotherapists to identify hazards in the home that may contribute to or increase the risk of falling. The content validity of these tools has been established. Environmental hazards have been described as significant risk factors for selected individuals, but generalisability of the single most important risk factors for falling associated with home environment has not yet been established. The Perell (2001) review describes and details many nurse administered tools, but most are developed for use only in hospital settings. The benefit of home hazard assessment for community-dwelling people is difficult to extrapolate from available studies, as most include some kind of intervention such as either referral or home modification. It appears that benefit is only achieved if followed by such referral. The AGS/BGS (2001) guidelines recommended the following:
Page 54 of 315
When older people at increased risk of falling are discharged from hospital, a facilitated home hazard assessment should be considered (B). This is supported by level I evidence from a study by Cumming et al. (1999), which showed that a facilitated home/environmental hazard assessment and supervised modification programme after hospital discharge was effective in reducing falls: RR= 0.64(0.49-0.84). Sub-group analysis demonstrated a significant reduction in the number of participants falling in the group with a history of falling in the previous year: RR= 0.64(0.49-0.84), but not in those without a history of a previous fall RR=1.03(0.75-1.41). Five randomised controlled trials, reported in the AGS/BGS guidelines, demonstrated no benefit of home environment modification without other components of multifactorial interventions. Many off the shelf home hazard assessment tools are available and are being developed at local level. Those administering the instrument should decide the choice of tool (Evidence table 4, Appendix E, 2004 for further details). b) Multifactorial falls risk assessment processes Whilst the term multifactorial is frequently referred to in relation to falls assessment, there is disparity between studies of what factors are included within this process. The AGS/BGS (2001) guidelines describe different levels of assessment determined by an older persons falls risk status. Consequently, a brief assessment for those at low risk of falling is suggested, with a more comprehensive and detailed assessment for high-risk groups. Referral to a geriatrician may be needed for such comprehensive assessment. The Cochrane review (2001) on falls prevention reports that different details and levels of assessment are contained in the included studies. Components include: environmental, including home hazards medical
Page 55 of 315
functional psychosocial activities of daily living medication review. The review by Shekelle (2002) reports similar differences between studies. The most common domains included in relation to risk assessment were: medication review vision environmental hazards orthostatic BP. The results from Shekelle (2002) suggest that: Although not proven, it makes clinical sense that comprehensive post fall and falls risk assessment should be targeted to persons at high risk as they have most to gain. The benefit of multifactorial assessment for older people is difficult to extract from available sources, as it appears that benefit is only achieved if followed by referral and therefore specific intervention. The Shekelle review refers to randomised controlled trials in which multifactorial falls risk assessment and individually tailored follow-up and management programmes were most effective in preventing falls for community-dwelling older people. The pooled risk ratio of n=10 studies that included a multifactorial falls risk assessment and management programme was relative risk (RR) = 0.84 (0.73-0.97) for risk of falling and pooled incident ratio was 0.65 (0.49-0.85) for the number of falls (n=7 studies). The Cochrane review on falls prevention reported that multidisciplinary, multifactorial, health/environmental risk factor screening/intervention programmes were effective for both unselected community-dwelling people: three trials pooled RR= 0.73 (0.63-0.86) and those with a history of falling / or known risk factors two trials= RR 0.79 (0.67-0.94) (Gillespie et al. 2003).
Page 56 of 315
Nurse assessment, followed by physician referral for older people in extended care settings, was of no benefit in one study included in the Cochrane falls prevention review, RR= 0.97 (0.84-1.11) (Gillespie et al. 2003). c) The minimum data set home care and residential assessment instrument for comprehensive assessment
Glossary MDS: Minimum data set. HC: Home care (community dwelling). CAP: Client assessed protocol for home care. RAI: Residential assessment instrument (extended care). RAP: Residential assessed protocol for extended care.
While multifactorial assessment processes as described above are specific to falls, the implementation of the single assessment process (SAP) is driven by a holistic and individualistic approach to management and care of older people across a number of domains. MDS tools are referred to in the SAP and have been suggested as useful (DH 2001). Other tools are referred to in the DH single assessment process guidance (2002) and current existing tools are subject to accreditation. Details of such instruments are soon to be published on the SAP website (www.dh.gov.uk/scg/sap/). The MDS assessment instruments have undergone testing for reliability and validity in community-dwelling and extended care settings but details are not reported here. There are currently two principal instruments with others being developed. The first instrument MDS-RAI is aimed at older people in residential settings, while the second MDS-HC is for community-dwelling older people receiving home care. There is an assessment data collection form and software is available, which is used in conjunction with the appropriate MDS assessment manual. The RAI and HC both have a standardised form that provides an initial assessment of minimum data taken at various stages along the service users care pathway. The comprehensive design of the form will trigger 1-30 care protocols. These protocols provide a more focused assessment leading to suggested care plans. The RAI is associated with the RAP residential assessed protocol for extended care.
Page 57 of 315
The MDS-HC is associated with CAP a client-assessed protocol for home care. The MDS is a standardised multidisciplinary assessment system for assessing care needs for older people within residential care. This instrument was originally developed in the USA to enable an accurate assessment of the older people leading to planned quality care. However, it is now being used in many other countries such as the UK, China, Japan, Italy and Norway. The primary purpose of this tool is to provide a comprehensive assessment that is integrated with care planning. This includes identification and evaluation of potential problems; identification of requirements for rehabilitation; maintenance of client strengths and prevention of decline; and promotion of comprehensive well-being. It follows a pathway from identification and evaluation, to guidance on service provision and care planning. The instrument encompasses the following assessment domains: cognition, communication, activities of daily living, continence, social functioning, disease diagnosis, vision, physical functioning, health conditions and preventative health measures, informal supportive services, mood and behaviour, nutrition/hydration status, dental status, skin condition, environmental assessment, and service utilisation in the last seven days. The falls-related data are within different domains. Since 1997, it is compulsory for facilities in the US to complete this assessment instrument. This tool is suggested within Single assessment process: assessment tools and scales (DH 2002). Detailed examination of the MDS The content validity of the risk assessment of falls section of the MDS instrument was examined and information on the utility of the instrument in practice in relation to falls was also sought. This was done to see if the MDS HC and RAI instruments provide adequate information to identify those at risk of falling, and whether all the important risk factors for falls are included.
Page 58 of 315
Of particular interest was what factors within the associated protocols trigger either further assessment of falls or lead to targeted falls interventions. As indicated by the risk factor review prospective cohort studies, in which multivariate analysis with adjustment for confounding was undertaken, the risk factors below were shown to be most significant by setting. These were compared with those risk factors listed in the CAP and RAP protocols. Community-dwelling older people Falls history, gait deficit, balance deficit, mobility impairment, fear, visual impairment, cognitive impairment, urinary incontinence and home hazards. People cared for in extended care settings Falls history, gait deficit, balance deficit, visual impairment and cognitive impairment. The instruments (HC and RAI) contain falls-related data in various sections/domains and clear pathways exist for the trigger to the falls protocols. Triggers for falls CAP: home care instrument Within HC, the potential for repeated falls or risk of initial fall is suggested if one or more of the following factors below are present. This will lead to further detailed assessment and CAPs. Trigger factors for falls CAP Falls in the last 90 days Sudden change of mental functioning Being treated for dementia Being treated for Parkinsonism Has unsteady (abnormal) gait. Triggers for Falls RAP: Residential care instrument
Page 59 of 315
The potential for additional falls or risk of initial fall is suggested if one or more of the following factors outlined below are present. This will lead to further detailed assessment and the application of RAP (2000). Triggers for falls RAP Fall in the past month Fall in past one to six months Wandering Dizziness/vertigo Use of trunk restraint Anxiolytic drugs Antidepressants. These tools provide relevant information about potential intrinsic and extrinsic risk factors, for which there are beneficial interventions. Of particular interest is the information relating to the assessment of balance and gait, which provides detailed aspects of balance and gait abnormalities, with possible diagnoses and rehabilitative or environmental interventions. There are also suggested care pathways relating to home hazard assessment. However, although the instruments contain important risk factors for falling, no clear pathway exists to specifically identify patients at risk. In addition, the risk factors listed differ from those that emerged as significant in the risk factor evidence review. Each factor is within different domains and will lead to the falls care pathway. What is not clear is at what point an older person enters this process. Evaluation of performance of MDS instrument To see whether the MDS instrument improved the quality of care for older people at risk of falling, studies were sought evaluating its performance. Although as stated, this instrument is a comprehensive assessment tool that can provide information for the single assessment process, falls represents one protocol within this document with an associated range of items to act as a trigger for further assessment. For the purpose of this review and scope of the guideline, only studies focusing on falls-related information were reviewed.
Page 60 of 315
English language studies of the following designs: prospective cohort, quasi experimental/controlled before and after designs or pre and post were sought. In addition, these must have report fall-related information such as incidence rates, reduction in falls and the trigger of falls protocols. Appraisal of methodological quality The methodological quality of the studies was assessed using the following criteria: eligibility criteria stated appropriateness of design sampling method validation of measurements relevant to falls outcomes or the instruments ability to perform in relation to falls response rate statistical techniques used bias and confounding addressed. An overall subjective rating of quality was applied to each study as follows: High: all of above criteria met Medium: most of the criteria met Low: insufficient information given. Search strategy Eight electronic databases were searched between 1995 and April 2003 using a sensitive search strategy. The bibliographies of all retrieved and relevant publications were searched for further studies. The lower limit was selected because this instrument is relatively new. The major databases searched were MEDLINE, EMBASE, CINAHL, PSYCINFO, HMIC, AMED (Allied & Complementary Medicine Database), and BNI (British Nursing Index). The platform was Silver Platter Windows-based WINSPIRS. The Web of Science and Cochrane Library databases were also
Page 61 of 315
searched, using just the first part of the search strategy found in Appendix B (2004). Data abstraction The papers were screened for relevance and those papers that met the inclusion criteria were identified and quality appraised. Data were extracted by one reviewer and evidence tables compiled. The following information was extracted: Author and country of origin; aim and objective; population and setting; number of participants; study design and method; outcome measurements and summary statistics; and comments on methodological quality. Results An initial search strategy identifying UK only papers resulted in five papers, but they were not related to falls assessment or outcomes. The search was then broadened to include international papers. The following is the result of the search and sift for papers to meet the inclusion criteria. Table 7 provides information on the process of selecting papers for critical appraisal. TABLE 7: SIFTING RESULTS FOR STUDIES EVALUATING MD
Initial search results N screened for relevance N relevant N included 399 129 3 3
Methodological quality of studies Three studies met the inclusion criteria. Two studies were conducted in the US and the third was a multi-centre, cross-cultural study of five countries. The quality of the three included studies was medium. Two were prospective cohort and one a before/after study. Two were conducted with communitydwelling older people (HC) and one in extended care setting (RAI).
Page 62 of 315
Evidence summary The first study conducted by Fries et al. (1997) evaluated the effect of the implementation of the MDS:RAI system on selected conditions representing outcomes for nursing home residents. This was a simple before and after study design of medium quality. Measurements of the prevalence of falls 30 days prior to admission were taken at baseline and then at six months post intervention. The results were non-significant for prevalence of falls between pre and post administration of the RAI, although there was a slight increase in the percentage of residents who fell post-RAI (pre=10.5%, post=10.6%). The overall prevalence of falls was pre-RAI 6,597 and post-RAI 6,178. The second study included was conducted by Ritchie et al. (2002). The aim of this study was to evaluate the establishment of a co-ordinated care programme for community-dwelling older people to receive assessments that lead to effective treatments, referral or care-plans. The sample was 99.6 per cent male of which 83.65 per cent were married, mean age=78. A thorough screening process was undertaken to locate those elders deemed as at risk. Follow-up measurements were taken at first and subsequent assessments using the MDS-HC instrument. A total of 158 protocols were triggered out of a possible 226. There were four typical response activities to falls triggered protocols that patients received. 38.4 per cent received falls prevention education, 5 per cent received prosthetics, 3.8 per cent received rehabilitation referral and 1.3 per cent received adult protective services. It is unclear as to whether there was overlap between these services. The most fundamental problem with this study is that the sample was 99.6 per cent male. Finally, the third study (Morris et al. 1997) involved five volunteer countries: Australia, Canada, the Czech Republic, Japan and the US. The sample was randomly selected within facilities of community-dwelling people but did not represent a random sample within the population of the country. The study had two objectives, of which the one relevant to this review is reported. This examined the interaction between different client profiles measured by their cognitive performance measured on the Folstein mini-mental examination and the effect of these measurements on triggering the protocols. For a
Page 63 of 315
sample size of 780, the average number of protocols triggered was nearly 12 of which the falls protocol represents 79 per cent. Those mentally intact triggered 82.5 per cent of the falls protocols, whereas 65 per cent at the lower of the cognitive scale triggered falls protocols. Those with severe cognitive impairment more frequently triggered bowel management, incontinence, and pressure ulcer protocols. Further work needs to be done evaluating the impact of these instruments on patient care and outcomes. At this stage there is insufficient information to make recommendations regarding the use of these tools and protocols specifically for falls. This is a subject that should be reconsidered when the guidelines are updated.
3.3.4
Fear of falling as a risk factor and tools to measure fear of falling: methods and results
3.3.4.1
Background
Fear of falling is considered multifaceted in aetiology. While fear may result as a consequence of falling, anticipatory anxiety may also occur in those who have not fallen. Murphy et al. (1982) refers to the post fall syndrome that recognises fear as a consequence of falling. Ptophobia the phobic reaction to standing or walking is a term introduced by Bhala et al. (1982). Fear of falling has been further conceptualised as: encompassing activity limitation due to the residing fear fear resulting in loss of confidence in balance ability and low fall-related efficacy, which translates to low confidence at avoiding falls. Fear of falling is not necessarily limited to those with a history of falling nor is fear predictive of a future fall. Fear may also compromise quality of life by limiting mobility and social interaction. We conducted two evidence reviews on the area of fear of falling. Firstly, we reviewed the empirical evidence investigating associations of fear of falling
Page 64 of 315
with future falling. Secondly, we reviewed methods available to measure fear and their usefulness for patients and clinicians. 3.3.4.2 Aim of review
The aim of this review was to: 1) identify studies in which fear has been examined as a predictor of falling and/or a consequence of falling 2) ascertain whether fear of falling should be included in risk assessment 3) assess methods and tools available to measure fear of falling and to ascertain their clinical utility. 3.3.4.3 Selection criteria
Types of studies Prospective cohort studies, with fear and fall related data measured at baseline and follow-up, were preferred because we were interested in fear as a predictor of future or further falls. Systematic/narrative reviews describing methods for measuring fear of falling. Types of participants Older people aged 65 and above. Types of outcome Those studies which report falls as an outcome. Exclusion criteria Individual studies examining the psychometric properties of instruments used to measure fear of falling and related constructs this work was outside the resources available.
Page 65 of 315
3.3.4.4
Search strategy
The searches for both fear of falling as a risk factor and tools to measure fear of falling were combined, as this was the most efficient way of searching. Please refer to Appendix B (2004) for details of the search strategy and databases searched. Searches were confined to the period 1980 and December 2002/January 2003.The bibliographies of all retrieved and relevant publications were searched for further studies. The databases searched were MEDLINE, EMBASE, CINAHL, PSYCINFO, HMIC, AMED (Allied & Complementary Medicine Database), ZETOC and BNI using the Silver Platter Windows-based WINSPIRS platform. 3.3.4.5 Data abstraction
The following data were extracted and evidence tables compiled: Author, setting, number of participants at baseline and follow-up, methods and details of baseline and outcome measurement, results including summary statistics and 95 per cent confidence intervals, and comments on the quality of studies. Once individual papers were retrieved, the articles were checked for methodological rigour using quality checklists appropriate for each study design applicability to the UK and clinical significance. Assessment of study quality concentrated on dimensions of internal validity and external validity. Information from each study that met the quality criteria was summarised and entered into evidence tables. 3.3.4.6 Appraisal of methodological quality
The methodological quality of each trial was assessed by one reviewer, using the principles of quality referred to in the risk factor review (Section 3.3.2.7).
Page 66 of 315
3.3.4.7
Data synthesis
No quantitative analysis was carried out for this review. Summary statistics and reporting of statistical significance for each study are included in the evidence tables. 3.3.4.8 Details of studies included in the review
Sifting results The number of studies included is shown in Table 8. TABLE 8: SIFTING RESULTS ON DEAR OF FALLING
Initial search results N considered for inclusion N included 634 50 7 (inc. 2 reviews)
3.3.4.9
Generally, the quality of the prospective cohort studies on examining fear as a risk factor for falling and association of fear of falling with quality of life and health status was high. These studies were conducted on large samples of community-dwelling older people. No studies were identified that were specific to older people in extended care settings. Studies were excluded mainly because of small sample sizes. The studies identified within the reviews on measurement of fear of falling and related constructs were categorised as follows: examination of the psychometric properties of available instruments development of new tools for the measurement of fear modification and testing of internationally developed instruments for use in the UK for example, falls efficacy scale (FES). Generally, the two identified reviews (Nakamara 1998 and Legters 2002) were of limited value. Both were narrative with no details of methods used to identify and appraise studies.
Page 67 of 315
Characteristics of excluded studies are shown in Appendix G (2004). 3.3.4.10 Evidence summary
Fear of falling Three prospective cohort studies reported fear as a significant predictor of future falling (Arfken 1994; Cumming 2000; Friedman 2002). While it is clear that fear can be a predictor for falling and a consequence of falls, shared risk factors increase the likelihood of falling. Many studies examined specific factors that correlate with the fear of falling. Although such studies are not reviewed here, the literature refers to many correlates. For example: psychological indicators of balance confidence, (Powell 1995; Myers et al. 1996; Manning et al. 1997; and Parry 2000); lack of confidence leading to reduced activity and loss of independence, (Maki et al. 1991). Other correlates include chronic dizziness (Burker et al. 1995); fewer social contacts (Howland et al. 1998); lower quality of life (Lachman 1998), (see Evidence table 4, Appendix E, 2004 for further details of included studies). The findings from this review provided sufficient evidence that fear of falling is a significant predictor of future falling and should be considered in falls assessment of older people. Measurement of fear of falling Fear related to falling is an important consideration when assessing older people and planning interventions. How to elicit such information from older people has been the focus of much research. In the discussion paper by Legters (2002), details are given of existing methods of measuring fear.Early research focused on simple questions to establish if fear was present. Examples given were responses to questions of are you afraid of falling? in yes/no or fear/no fear format. Whilst this is a simple measure, it does not provide information of the degree of fear. Further development of such measures resulted in more sophisticated methods, such as verbal rating scales that provide ordinal levels of measurement of degrees
Page 68 of 315
of fear. Examples of verbal rating scales include responses such as: not afraid; slightly afraid; somewhat afraid; very afraid. Details of the study on the FES (Tinetti et al. 1990 USA), which appears to be the most widely used tool, are given in Evidence table 3, Appendix E (2004). This tool was designed for the purpose of measuring fear in a research context. The conceptual framework underpinning the development of this instrument is related to asking individuals about their feelings, within a variety of specific situations or activity. Perceptions of capability are referred to as self-efficacy. High efficacy relates to increased confidence. The FES measures the individuals degree of efficacy within a specific activity (Tinetti et al. 1991). Confidence in accomplishing each activity without falling is assessed on a 10-point scale, with a higher score equivalent to lower confidence or efficacy. The FES score is the sum of scores and possible scores range from 10-100. Other tools have been developed but none to the extent of FES. In terms of clinical utility, it is suggested that the FES could be an effective screening tool to determine if further evaluation is needed, particularly concerning balance (Legters 2002; Nakamura 1998). It is clear that fear of falling is related to future falling and this needs to be discussed with older people who are at risk of falling. However, whilst the FES does provide detailed information, this tool may, at this stage, only be useful for research purposes. What may be more important is that older people are asked if they are fearful of falling. If so, then the reason for this fear and the degree of fear should be assessed by an appropriate health care professional.
3.3.5
3.3.5.1
Background
Many preventive intervention programmes aimed at recognised risk factors have been established and evaluated. These have included exercise programmes designed to improve strength or balance, education
Page 69 of 315
programmes, medication optimisation, environmental modification in homes or institutions, and nutritional or hormonal supplementation. In some studies, interventions designed to reduce the impact of single risk factors have been evaluated. However, in the majority multiple interventions have been used. Interventions have been offered to older people at varying levels of fall risk, either as a standard package or individually tailored to target risk factors and impairments. Some are population-based approached programmes. The best evidence for the efficacy of interventions to prevent falling should emerge from large, well-conducted randomised controlled trials, or from metaanalysis of smaller trials. In July 2003, a Cochrane systematic review on Interventions for the prevention of falls in older people was updated (Gillespie et al. 2003). This was itself an update of a previous review (2001); has undergone peer review and is published in the Cochrane Library. This review has formed the basis for the evidence on effective interventions to prevent falls for this guideline. The review methods and results are summarised below from the updated systematic review (full details are available on www.cochrane.co.uk). 3.3.5.2 Objectives
The review sought to present the best evidence for effectiveness of programmes designed to reduce the incidence of falls in both communitydwelling older people and those in extended care settings among those at risk of falling and known fallers. This review has also provided evidence for rehabilitation interventions for the secondary prevention of falls (see Section 3.3.9). 3.3.5.3 Selection criteria
Types of studies RCTs, including those in which the method of allocation to treatment or control group was inadequately concealed for example, trials in which patients were allocated using an open random number list or coin toss.
Page 70 of 315
Subjects randomised to receive an intervention or group of interventions versus usual care to minimise the effect of, or exposure to, any risk factor for falling. Studies comparing two types of interventions were also included. Types of participants Older people of either sex, living in the community or extended care. Participant characteristics of interest included falling status at entry (for example, non-faller, single faller, multiple faller), residential status (for example, community, extended care), and where appropriate, associated comorbidity. While the review also included trials of interventions in hospital settings if the patients were elderly, those results are not reported here, as this is outside the scope of the guideline. Types of intervention Studies which evaluated the following interventions for falls prevention were included in the clinical effectiveness evidence review: 1 2 3 4 5 6 7 8 9 10 Exercise/physical therapy Home hazard modification Cognitive/behavioural interventions Medication withdrawal/adjustment Nutritional/vitamin supplementation Hormonal and other pharmacological therapies Referral for correction of visual deficiency Cardiac pacemaker insertion for syncope associated falls Exercise, visual correction and home safety Multidisciplinary, multifactorial health/environmental risk factor
Page 71 of 315
12
Types of outcome The main outcomes of interest were the number of fallers or falls, and severity of falls. Severity was assessed by the number of falls resulting in injury, medical attention, or fracture. Information was also sought on complications of the interventions employed, duration of effect of the interventions, and death during the study period. Trials that focused on intermediate outcomes, such as improved balance or strength, and did not report fall rates or number of fallers, were excluded. An improvement in a surrogate outcome does not provide direct evidence that an intervention can impact on the clinical outcome of interest (Gotzsche 1996) in this case, falls. Therefore only trials which reported falls or falling as an outcome were included. 3.3.5.4 Search strategy
The following databases were searched: MEDLINE (1966 to February 2003) EMBASE (1988 to 2003 Week 19) CINAHL (1982 to April 2003) The National Research Register, Issue 2, 2003 Current Controlled Trials (www.controlled-trials.com, accessed 11 July 2003) and reference lists of articles PsycLIT and Social Sciences Citation Index to May 1997 No language restrictions were applied and further trials were identified by contact with researchers in the field.
Page 72 of 315
The search strategies and the databases searched are presented in Appendix B (2004). All searches were comprehensive and included a large number of databases. A combination of subject heading and free text searches was used for all areas. Free text terms were checked on the major databases to ensure that they captured descriptor terms and their exploded terms. Further trials were identified by contact with researchers in the field. 3.3.5.5 Sifting process
From the title, abstract, or descriptors, two reviewers independently reviewed literature searches to identify potentially relevant trials for full review. Searches of bibliographies and texts were conducted to identify additional studies. From the full text, trials that met the selection criteria were quality assessed. Once articles were retrieved the following sifting process took place: First sift: for material that potentially meets eligibility criteria on basis of title/abstract by two reviewers. Second sift: full papers ordered that appear relevant and eligible and where relevance/eligibility not clear from the abstract by two reviewers. Third sift: full articles are appraised that met eligibility criteria by two reviewers. 3.3.5.6 Appraisal of methodological quality and data extraction
The methodological quality of each trial was assessed by two researchers independently. The following quality criteria were used (Appendix C, 2004): description of inclusion and exclusion criteria used to derive the sample from the target population description of a priori sample size calculation evidence of allocation concealment at randomisation description of baseline comparability of treatment groups outcome assessment stated to be blinded outcome measurement
Falls: NICE clinical guideline 161 (June 2013) Page 73 of 315
clear description of main interventions. The level of concealment of allocation at randomisation was assessed using the criteria in the Cochrane reviewers handbook (Clarke 2003b). Studies were graded A if it appeared that the assigned treatment was adequately concealed prior to allocation, B if there was inadequate information to judge concealment, and C if the assigned treatment was clearly not concealed prior to allocation (see Appendix C, 2004 for further details). Data were independently extracted by pairs of reviewers using a data extraction form, which had been designed and tested prior to use. Consensus or third party adjudication resolved disagreement. 3.3.5.7 Data synthesis
Statistical analysis of individually randomised studies was carried out using MetaView in Review Manager (RevMan 2003). Raw data from clusterrandomised studies were not entered, as the units of randomisation and analysis differed. For dichotomous data, the individual and pooled statistics were calculated, using the fixed effects model, and were reported as relative risk (RR) with 95 per cent confidence intervals (95% CI). For continuous data (reporting mean and standard deviation or standard error of the mean), pooled weighted mean differences (WMD) with 95 per cent confidence intervals were calculated. Heterogeneity between pooled trials was tested using a standard chi-squared test and was considered to be significant when P< 0.1. 3.3.5.8 Details of studies included in the review
Included in the updated review were 62 trials reporting a variety of settings,participants, and interventions. Four studies reported results of prevention interventions in hospital settings and are excluded from this report, as this is not within the scope of the guideline. Details are therefore given of the remaining 58 studies.
Page 74 of 315
Settings Of the 58 studies, 47 reported the effect of interventions in participants living in the community. Eight studies were set in long-term care facilities, including long-term care wards in hospital, or nursing homes. A further three studies included participants with specific conditions from a range of residential settings. Participants In 16 studies, eligibility for inclusion included a history of falling, or of a postulated risk factor other than general frailty, residence in long-term care, or age. General frailty, residence in long-term care, history of requiring admission to a rehabilitation facility for older people, use of home help services, or age at least 80 years defined eligibility in a further 14 studies. In the remaining 28 studies, participants were recruited from seniors centres, lists of older people, or through advertisement for volunteers. The mean age of participants at enrolment exceeded 80 years in 13 studies and was less than 70 years in four studies. In 10 studies, the participants were all women, and in one the participants were all men. The remaining studies recruited men and women in varying proportions. In most, the proportion of women was more than 70 per cent. Interventions Exercise/physical therapy interventions (22 studies) Fourteen studies compared a physical exercise or physical therapy intervention alone with a social meeting or visit, education only, or no intervention. In one study, self-paced brisk walking was compared with upper limb exercises. Another study compared an enhanced exercise programme
Page 75 of 315
that was offered to all other participants. The remaining six studies in this category examined complex interventions as follows: an exercise programme and a programme of medication withdrawal progressive resistance quadriceps exercises and the administration of oral vitamin D progressive strength training and conditioning with a Tai Chi programme, with a cognitive/behavioural component exercise programme and a cognitive intervention in a factorial design programme of exercise associated with management of urinary continence a cognitive/behavioural intervention either alone, or combined with: exercise, exercise and home safety screening, or exercise and home safety screening and medical assessment. Home hazard modification (nine studies) The following interventions were included in the studies: assessment of environmental hazards and supervision of home modifications by an experienced occupational therapist home safety assessment and facilitation of elimination of hazards comprehensive home visit that included assessment and modification of home hazards nurse-led home hazard assessment, free installation of safety devices, and an education programme exercise, correction of visual deficiency, and home hazard modification, each alone, and in combination. home hazard assessment as a component of two of four other intervention packages. Three other studies evaluated home hazard modification in combination with other interventions, using a cognitive/behaviour modification approach. Cognitive/behavioural interventions (seven studies) The following interventions were included within this category:
Page 76 of 315
comparison of two risk assessment interviews and a feedback/counselling interview, with a single baseline assessment interview only comparison of a one-hour fall prevention education programme, delivered to a group or individually, with a control group receiving only general health promotion information the remaining five studies in this category were complex interventions and were also included in the previous two categories. Medication withdrawal/adjustment (two studies) exercise programme and a placebo-controlled psychotropic medication withdrawal programme optimisation of medication along with home hazard modification medication withdrawal/adjustment was also included in the majority of the multifactorial intervention listed below. Nutritional/vitamin supplementation (six studies) Five studies were designed to evaluate the efficacy of vitamin D supplementation, either alone or with calcium co-supplementation, in fracture prevention. Each trial reported falls as a secondary outcome measure. One other studied the efficacy of a 12-week period of high-energy, nutrientdense dietary supplementation in older people with low body mass index, or recent weight loss. Hormonal and other pharmacological therapies (two studies) One reported incidence of falls as a secondary outcome after administration of hormone replacement therapy to calcium replete, post-menopausal women. Another studied the effect of administering a vaso-active medication (raubasine-dihydroergocristine) to older people presenting to their medical practitioner with a history of a recent fall. Referral for correction of visual deficiency (one study)
Page 77 of 315
This study compared a control group with groups receiving exercise, correction of visual deficiency, and home hazard modification, each alone, and in combination. Cardiac pacemaker insertion for syncope-associated falls (one study) One trial reported the effectiveness of cardiac pacing in fallers who were found to have cardioinhibitory carotid sinus hypersensitivity following a visit to a hospital emergency department. Exercise, visual correction and a home safety intervention (one study) This study reported the effects of exercise, vision improvement, home hazard modification or no intervention in a factorial design. Multidisciplinary, multifactorial, health / environmental risk factor screening and intervention (20 studies) These were complex interventions that differed in the details of the assessment, referral, and treatment protocols. In most studies, a health professional usually a nurse or other trained person made the initial assessment, assessing the participants, providing advice and arranged referrals. Multifactorial intervention in nursing home residents (one study) One cluster randomised trial assessed the effectiveness of staff and resident education, including advice on environmental adaptations. In addition, residents were offered progressive balance and resistance training and hip protectors, and could choose any combination for any length of time. 3.3.5.9 Methodological quality of studies
A summary of the methodological quality of each study of the trials is shown in Appendix F (2004). The quality of studies was variable. In 19 studies, it appeared that the assigned treatment was adequately concealed prior to allocation. In three the
Page 78 of 315
assigned treatment was not concealed prior to allocation. In the remaining 36, there was inadequate information to judge concealment. Losses from groups resulted from, for example, withdrawal from the study or death. In trials with community-dwelling subjects, the outcome of falling was selfreported and the subjects were often not blind to treatment assignment. Blinding was possible in four trials, by using placebos or identical tablets, when the intervention involved the administration of drugs. A number of studies did not define a fall, and a variety of definitions were used in those that did. A fall was most frequently defined as unintentionally coming to rest on the ground, floor or other lower level; excludes coming to rest against furniture, wall, or other structure. Active registration of falling outcomes, or use of a diary, was clearly indicated in 31 studies. In the remaining 27 studies ascertainment of falling episodes was by participant recall, at intervals during the study or at its conclusion, or was not described. TABLE 9: LENGTH OF FOLLOW-UP
Follow-up 3 months 4 months 5 months 6 months 8 months 44 weeks 49 weeks 2 years 3 years 4 years 10 years n = trials 5 3 1 6 1 1 1 4 2 1 1
Duration of follow-up varied both between and within studies. It was for a minimum of one year in 38 studies. Table 9 reports the length of follow-up for other trials.
Page 79 of 315
The period for which falls were recorded differed markedly between studies, and was not necessarily the same as the total period of follow-up described above. The characteristics of excluded studies table (Appendix G, 2004) lists 97 studies, which fall into two categories. Thirty-five non-randomised studies reporting falls or fall-related injuries as an outcome were excluded on the basis of non-randomisation. Sixty-two randomised trials originally identified by the search strategy either reported intermediate outcomes of preventive strategies for example, balance or muscle strength measures or did not describe an intervention designed to reduce the risk of falling. At the time of writing there were 14 trials waiting assessment and 29 ongoing trials identified. 3.3.5.10 Comparisons
Trials were included in which participants were randomised to receive an intervention or group of interventions, versus usual care to minimise the effect of, or exposure to, any risk factor for falling. Studies comparing two types of interventions were also included. 3.3.5.11 Summary of results
For full details of included studies see Evidence table 5, Appendix E (2004). The Cochrane review reports the following: Evidence for the effectiveness of home hazard management in people with a history of falling is somewhat strengthened by new data. Evidence for the effectiveness of exercise programmes and multifactorial assessment/ intervention programmes remains unchanged, despite the inclusion of a number of new trials. In a highly selected group of fallers with carotid sinus hypersensitivity, cardiac pacing is effective in reducing the frequency of syncope and falls. Interventions likely to be beneficial:
Page 80 of 315
A programme of muscle strengthening and balance retraining, individually prescribed at home by a trained health professional (three trials, 566 participants, pooled relative risk (RR) 0.80, 95 per cent confidence interval (95%CI) 0.66 to 0.98). A 15-week Tai Chi group exercise intervention (one trial, 200 participants, risk ratio 0.51, 95%CI 0.36 to 0.73). Home hazard assessment and modification that is professionally prescribed for older people with a history of falling (three trials, 374 participants, RR 0.66, 95% CI 0.54 to 0.81). Withdrawal of psychotropic medication (one trial, 93 participants, relative hazard 0.34, 95%CI 0.16 to 0.74). Cardiac pacing for fallers with cardioinhibitory carotid sinus hypersensitivity (one trial, 175 participants, WMD -5.20, 95%CI -9.40 to -1.00). Multidisciplinary, multifactorial,health/environmental risk factor screening/intervention programmes in the community, both for unselected population of older people (four trials, 1651 participants, pooled RR 0.73, 95%CI 0.63 to 0.85), and for older people with a history of falling, or selected because of known risk factors (five trials, 1176 participants, pooled RR 0.86, 95%CI 0.76 to 0.98). Multidisciplinary assessment and intervention programme in residential care facilities (one trial, 439 participants, cluster-adjusted incidence rate ratio 0.60, 95%CI 0.50 to 0.73). Interventions of unknown effectiveness: Group-delivered exercise interventions (nine trials, 1387 participants). Individual lower limb strength training (one trial, 222 participants). Nutritional supplementation (one trial, 46 participants). Vitamin D supplementation, with or without calcium (three trials, 461 participants). Home hazard modification in association with advice on optimising medication (one trial, 658 participants), or in association with an education package on exercise and reducing fall risk (one trial, 3182 participants).
Page 81 of 315
Pharmacological therapy (raubasinedihydroergocristine, one trial, 95 participants). Interventions using a cognitive/behavioural approach alone (two trials, 145 participants). Home hazard modification for older people without a history of falling (one trial, 530 participants). Hormone replacement therapy (one trial, 116 participants). Correction of visual deficiency (one trial, 276 participants). Interventions unlikely to be beneficial: Brisk walking in women with an upper limb fracture in the previous two years (one trial, 165 participants). The Cochrane review concluded the following: Prevention programmes that target an unselected group of older people with a health or environmental intervention on the basis of risk factors or age, are less likely to be effective than those that target known fallers. Even amongst known fallers, the risk reduction where significant is small, and the clinical significance remains less clear. Interventions that target multiple risk factors are marginally effective, as are targeted exercise interventions, home hazard modification and reducing psychotropic medications. Where important individual risk factors can be corrected, focused interventions may be more clearly effective. It appears that interventions with a focused intention may in fact be multifactorial. There is a lack of clarity about the optimum duration and intensity of interventions. Some interventions for example, brisk walking may increase the risk of falling. The outcome of interest falling was not always clearly defined in the studies and therefore the definition of falling used could alter the
Page 82 of 315
significance of the results. In addition, methods used for recording falls also varied widely between studies. The full summaries are included in Section 3.4. From these were derived evidence statements and recommendations.
3.3.6
3.3.6.1
Background
Ideally, all participants in a trial should complete the study and follow the protocol in order to provide data on every outcome of interest at all timepoints. However, in reality most trials have missing data. This may be because some of the participants drop out before the end of the trial; participants do not follow the protocol, either deliberately or accidentally; or some outcomes are not measured correctly, or cannot be measured at all, at one or more time-points. Regardless of the cause, inappropriate handling of the missing information can lead to bias. However, on occasions it is impossible to know the status of participants at the times when the missing information should have been collected. This could happen, for example, if participants move to different areas during the study or fail to contact the investigators for an unknown reason. Other reasons may include: inability to comply with the intervention, perhaps due to lack of motivation; the intervention being too difficult; or not acceptable to participants. Excluding these participants or specific outcome measurements from the final analysis can also lead to bias. The only strategy that can be confidently assumed to eliminate bias in these circumstances is called intention to treat analysis. This means that all the study participants are included in the analyses, as part of the groups to which they were randomised, regardless of whether they completed the study or not. This relies on the researcher having measurement of outcome, regardless of compliance to the intervention.
Page 83 of 315
The purpose of this analysis was to examine the drop out rates and/or losses to follow-up for each trial included in the Cochrane review, where reported. This was done to shed light on the acceptability and sustainability of clinically effective interventions and prevention programmes. 3.3.6.2 Aim
The aim was to assess patient compliance with clinically effective interventions, as measured by drop-out rates/losses to follow-up. 3.3.6.3 Methods
Losses to follow-up rates and drop-out rates were extracted from those RCTs that reported clinically effective interventions and were included in the updated Cochrane review Interventions for the prevention of falls in elderly people (Gillespie et al. 2003). Reasons for drop-out/loss to follow-up were recorded where reported. 3.3.6.4 Results
The total number of studies reporting drop out rates/losses to follow-up was 19 out of 58 studies. For each clinically effective intervention, where reported, details and reasons for drop out and losses to follow-up are presented in the table below. (Refer to Evidence table 6, for full details of the studies from which this information was extracted).
Page 84 of 315
TABLE 10: LOSSES TO FOLLOW-UP AND DROP-OUT RATES IN THOSE STUDIES REPORTING POSITIVE RESULTS
Muscle strengthening and balance training Study Drop-out rates/ losses to follow-up Campbell (1997, n = 622 invited to participate, n = 359 1999). chose not to participate, n = 30 not eligible. n = 233 at randomisation Communitydwelling women Intervention (I) = 116 aged 80 years and Control (C) = 117 older, individually At one year follow-up n = 213 (91%) tailored I = 103 (88%), C = 110 (94%) intervention. n = 153 (71%) agreed to continue for a further year: I = 71 C = 81 At two year follow-up n = 103 (67%): I = 41(57%), C = 62 (76%) Total losses/ drop-out rates at two years Intervention = 75 (64%) Control = 55 (47%) Robertson (2001). Communitydwelling aged 75 years and older, individually prescribed exercise programme. n = 590 invited to participate n = 284 chose not to participate n = 6 not eligible n = 240 at randomisation Intervention (I) = 121 Control (C) = 119 n = 13 (10%) withdrew from exercise intervention Withdrew from trial: n = 8 (1) n = 21 (c) At one year follow-up, falls monitored n = 211 (87%), I = 113 (93%), C = 98 (82%) For the intervention group, 43% (49 of 113) carried out their exercise programme three or more times per week, 72% (n = 81) carried it out at least twice a week, 71% (n = 80) walked at least twice a week during the years follow-up. Total losses/ drop-out rates: 10% Comments Falls were selfrecorded using a calendar, which was posted monthly to researcher, for both groups. The intervention group also recorded if they had completed the prescribed exercises. Intention to treat analysis.
Page 85 of 315
Tai Chi Study Wolf (1996). Community dwelling untargeted people, mean age 76 years.
Drop-out rates/ losses to follow-up Total losses/ drop-out rates: 40 of 200 (20%) 20 months
Page 86 of 315
Home hazard assessment and modification for those with a history of falling Study Drop-out rates/ losses to follow-up Comment Nikolaus (2003). n = 391 eligible n = 31 chose not participate Older people n = 360 at randomisation (mean age 81) Intervention (I) = 181 recruited from Control (C) = 179 geriatric hospital and assigned to At follow-up= comprehensive assessment, I = 140 (77%) followed by a C = 139 (77%) diagnostic home visit and home Total losses/ drop-out rates 23% intervention vs. recommendations and usual care. Compliance with intervention recommendations: Recommendation N (/%) Shower seat Emergency call Garb bars Night light (bed/ bathroom) Anti-slip mat bath Elevation of bed Rollator Elevation of toilet seat Removal of rugs Removal of obstructions in walkways Compliance rate 23 (82) 14 (78) 27 (77) 20 (70) 12 (66) 19 (63) 37 (56) 43 (54) 12 (41) 15 (33)
Day (2002). Untargeted communitydwelling 70 and over. Multi-faceted study including home hazard, exercise and vision referral interventions. Pardessus (2002).
Page 87 of 315
Home visit One year modification following hospital admission following a fall. Mean age 83 years. Withdrawal of psychotropic medications Study Drop-out rates/ losses to follow-up Campbell (1999). n = 547 invited to participate n = 400 chose not to Communityn = 54 not eligible dwelling people aged 65 years and n = 93 at randomisation over. Intervention (I) = 48 Gradual withdrawal Control (C) = 45 of psychotropic medications vs. continuing to take Falls monitored for 24 months medications. I = 33 (68%) C = 39 (86%) Total losses/ drop-out rates: I = 32% C = 14% Authors report that one month after completion of the study, 47% (8 of 17) of the participants from the medication withdrawal group who had taken capsules containing placebo only for the final 30 weeks had restarted taking psychotropic medication.
Comments This study also included a group receiving exercise. Data here is combined to illustrate compliance with the psychotropic programme.
Page 88 of 315
Cardiac pacing Study Kenny (2001). Older people presenting at A&E following a nonaccidental fall, mean age 73. Pacemaker vs. no pacemaker.
Drop-out rates/ losses to follow-up Total losses/ drop-out rates: n = 16 of 175 (9%)
On year
Comment 71,299 A&E attendees screened, n=1624 received carotid sinus massage, n=175 agreed to be randomised. Intention to treat analysis not possible. Intention to treat analysis not possible.
Jitapunkel (1998). Communitydwelling people, mean age 76. Newbury (2001). Communitydwelling, age range 75 91. Wagner (1994). Communitydwelling, mean age 72 years.
Total losses/ drop-out rates: 44 of 160 (28%) Three years Total losses/ drop-out rates: 11 of 100 (111%) 12 months Total losses/ drop-out rates: 89 of 1559 (6%) Two years
Not stated.
Page 89 of 315
Targeted multidisciplinary interventions Study Drop-out rates/ losses to follow-up Tinetti (1994). Total losses/ drop-out rates: 10 of 301 (3%) Communitydwelling, mean age One year 77 years with at least one risk factor present. Close (1999). Total losses/ drop-out rates: 93 of 397 (23%) Communitydwelling older people, mean age 78, presenting at A&E following a fall. Hogan (2001). Communitydwelling, aged 65 years and over, with a falls history in the previous 3 months. One year
n = 163 at randomisation Intervention (I) = 79 Control (C) = 84 Completed trial: I = 66 (83%) C = 73 (86%) Total losses/ drop-out rates: I = 17%, C = 14% One year Total losses/ drop-out rates: 17 of 109 (16%) 12 weeks
Kingston (2001). Communitydwelling, mean age 71 years, attending A&E following a fall. Lightbody (2002). Communitydwelling, median age 75, attending A&E following a fall. Van Hastregt (2000). Communitydwelling, mean age
n = 392 met inclusion criteria n = 316 at randomisation Intervention (1) = 159 Control (C) = 157
Page 90 of 315
N completed trial: I = 120 (75%) C = 115 (73%) Total losses/ drop-out rates: 81 of 316 (26%) 18 months
Multidisciplinary: extended care Study Drop-out rates/ losses to follow-up Jensen (2002). n = 439 residents (9 facilities) n = 402 assesses Extended care Intervention (I) = 194 (4 facilities) residents aged 65 Control (C) = 208 (5 facilities) and over. Follow-up and evaluation completed I = 157 (80%) C = 167 (80%) Total losses/ drop-out rates = 20% 8.5 months
3.3.6.5
Summary of results
Muscle strengthening and balance training: appears to be high participation with intervention at one-year follow-up. In one study, 57 per cent were carrying out the intervention at two years follow-up. Tai Chi: 20 per cent dropout at seven to 20 month follow-up Home hazard intervention: 2-28 per cent were not available at follow-up (one year-18 months). Psychotropic medication withdrawal: 68 per cent at follow-up (24 months). Cardiac pacing: 9 per cent were not available at follow-up (one year). Untargeted, multidisciplinary interventions: 6-28 per cent drop-out (one to three years).
Page 91 of 315
Targeted, multidisciplinary interventions: 3-26% drop-out (three to 18 months). Extended care, multidisciplinary intervention: 80 per cent participation at follow-up (34 weeks). Implications The intention of this analysis was to shed light on the factors affecting likely patient compliance and adherence to intervention packages and on sustainability. However, insufficient information on reasons for patient dropout was given in the studies. Drop-out/losses-to follow-up rates give a crude indication of possible participation rates. However, in everyday practice these could either be lower or higher. Factors influencing participation from the patients perspective is given in Section.3.3.8.
3.3.7
Interventions to reduce the psychosocial consequences of falling: review methods and results
3.3.7.1
Methods
Aim of the review To present findings on the effect of falls prevention interventions on psychosocial factors, such as confidence and fear of falling. No additional searching was conducted for this review as the source of results was extracted from those trials that reported effective falls prevention interventions and strategies in the Cochrane review (Gillespie et al. 2003). The review sought to answer the following question: Do effective falls prevention programmes also improve psychosocial factors related to fear of falling and the psychosocial consequences of falling?
Page 92 of 315
3.3.7.2
Selection criteria
Study designs RCTs from the Cochrane interventions for the prevention of falls systematic review that reported clinically effective interventions and that also investigated outcome in terms of psychosocial measures. Patients Older people, mainly more than 65 years of age but 60 acceptable. Settings All, including A&E; not relating to prevention of falls while a patient in hospital. Interventions Clinically effective prevention programmes/ interventions to reduce the incidence of falls that report psychosocial outcomes. Outcomes Number of falls. Measurement of fear, confidence, quality of life and other aspects of psychosocial consequences of falling. Mean change or summary statistics were extracted, with significance levels where reported. 3.3.7.3 Data synthesis
Synthesis of results was not appropriate. 3.3.7.4 Evidence tables and summary
The number of studies providing information on psychosocial outcomes was two out of the 19 studies reporting clinical effectiveness. The table below gives details of the psychosocial outcomes from the two studies.
Page 93 of 315
TABLE 11
Unselected, multidisciplinary interventions: please refer to evidence table 1 for details of interventions Drop-out rates/ losses Mean change or Study Comment to follow-up relative risk (RR) Baseline = 22 (50%) Newbury (2001). Communitydwelling, aged 75 and over. Self-rated health (n(%)) Very good or good for intervention group. Geriatric depression score (GDS). Follow-up = 30 (68%) p = 0.032 Mean change from baseline to follow-up = 0.5 (3.95 to 2.95) p = 0.05 Participants were less depressed as measured by the (GDS) following the intervention
Muscle strengthening and balance training Drop-out rates/ losses Mean change of Study to follow-up relative risk (RR) Mean (SD) falls Results reported for the Campbell efficacy score: (1997, 1999). intervention group, for those that continue with Continued = 89.4 the study at one year and (12.8) Communitythose who did not. Withdrew 83.3 (16.4) p dwelling Mean falls efficacy score. = 0.009 women aged 80 years and older, Results for those still individually exercising at two years Exercising = 93.3 (9.4) tailored and those not. Not exercising = 86.8 intervention. (15.1) p = 0.03
Comment
Participants reported increased confidence and reduced fear of falling in the intervention group.
As can be seen above, secondary outcomes relating to psychosocial variables are not routinely measured in all of the included trials. It is therefore difficult to extrapolate from the available limited published evidence. While it is important to determine if falls prevention programmes are effective in reducing the incidence of falls, other outcomes such as the reduction in fear of falling that are important for patients should also be measured. It is not clear from the available evidence which component of a prevention programme acts on reducing the incidence of falling and increasing confidence and other quality of life measures.
Page 94 of 315
These two trials, focused on those > 75 and 80 years of age, did show an improvement on psychosocial measures such as depression, confidence and fear of falling. However, in the absence of patient interviews it is difficult to know if it is the social benefits of participating in group programmes that exert a benefit, in addition to the benefit from reduction in falls.
3.3.8
3.3.8.1
Background
Information on patients views, compliance with and acceptability of falls prevention programmes is lacking within trials and systematic reviews. Accordingly, studies that investigated these factors were reviewed. Both this evidence and the evidence from the systematic review (Gillespie et al. 2003) are needed to enable the development of pragmatic recommendations on falls prevention. Frequently within trials the only indicator of compliance is the drop-out/losses to follow-up rate, which includes reasons relating to morbidity and mortality (see Section 3.3.6 above).While these are useful measures, it is likely that compliance rates in trials of falls interventions may be lower than in actual clinical practice. This is because of the advantage within trials of dedicated resources such as follow-up telephone calls etc to maximise participation. Information of patients views of the falls prevention trials in which they participated is similarly lacking. Therefore, it was thought useful to review and summarise evidence that captures the patient perspective on the likely barriers to and facilitators of participation in falls prevention programmes. This may indicate successful methods to promote compliance/adherence and participation in falls prevention programmes. Much of this evidence comes from studies conducted independently of trials of falls prevention, which therefore reflect a variety of designs, settings and participants. However, appraisal of this material enables a fuller consideration
Page 95 of 315
of some of the issues associated with falls prevention programmes from the perspective of the intended target group. All studies were quality assessed and the relevant data extracted and reported in evidence tables. The summarised results and conclusions were then condensed in a table on summary of barriers/facilitators relating to falls prevention programmes to give the GDG a breakdown of the key points arising from these studies. 3.3.8.2 Objective To review qualitative and quantitative studies published in the last 10 years, which examine older peoples views of falls prevention strategies. Inclusion criteria Study designs: All (systematic review qualitative). May include studies conducted concurrently with RCTs, Publication status: Not theses, letters, editorials Dates: 1990-May 2003 Language: English Patients: Older people (mainly more than 65 years of age) Settings: All, including A&E, except relating to preventing falls in hospital settings Outcomes: Measures and/or self-report/clinician report of: barriers to and benefits of participation in falls prevention programmes participant views and experiences of falls prevention strategies compliance/adherence with falls prevention strategies or components of falls prevention strategies, such as exercise. Methods
Page 96 of 315
Exclusion criteria Theses, letters, editorials, case studies. Studies with a focus on hospital-based falls prevention programmes for patients who have fallen whilst a hospital inpatient. 3.3.8.3 Search strategy
The search strategy was devised to be very broad in order to pick up qualitative studies for this review. The search strategies and the databases searched are presented in Appendix B (2004). All searches were comprehensive and included a large number of databases. All search strategies were adapted for smaller or simpler databases or for web-based sources, which did not allow complex strategies or multi-term searching. A combination of subject heading and free text searches was used for all areas. Free text terms were checked on the major databases to ensure that they captured descriptor terms and their exploded terms. 3.3.8.4 Data abstraction
Data from included trials were extracted by one reviewer into pre-prepared data extraction tables. The following data were extracted from each study: Qualitative Study, aim of study, methods, sample characteristics, setting, results, conclusions. Quantitative Study, objective, setting, population characteristics, methods, interventions, outcomes, results. All data were extracted into evidence tables (Evidence table 6, Appendix E, 2004). 3.3.8.5 Appraisal of methodological quality
All studies were quality assessed by one person, using study design specific quality assessment checklists developed by the Centre for Statistics in
Falls: NICE clinical guideline 161 (June 2013) Page 97 of 315
Medicine. The qualitative checklist was developed in-house, based on others, and then circulated to qualitative researchers for comment and refinement. See Appendix C (2004) for further details of quality for specific study designs. 3.3.8.6 Data synthesis
No quantitative analysis was undertaken. The data were presented in evidence tables and the main findings were qualitatively summarised. A table of barriers and facilitators was generated based on the findings of the studies. 3.3.8.7 Details of studies included in the review
Sifting results The numbers of studies obtained are detailed in Table 12. TABLE 12: RESULTS OF SEARCH/ SIFT FOR SUTDIES OF PATIENTS' VIEWS AND EXPERIENCES
Total number of hits Full articles ordered Final number of articles included 14576 31 24
Type of studies included Qualitative (two were unpublished) 10 Systematic review one Narrative review three Randomised controlled three Before/after three Cross-sectional four. Participants and settings Qualitative One study (Resnick 1999) investigated the views of nursing home residents. Three studies were conducted on hospital wards; one on people admitted to an orthopaedic trauma elderly care ward (Ballinger & Payne 2000); one on people admitted to an elder care ward after a fall sustained either in the community or hospital setting (Kong et al. 2002) and the other on patients
Page 98 of 315
admitted to an acute elderly care medical ward (reasons not given) (Simpson et al. 2003). The remaining studies were conducted on community-dwelling residents. Four studies examined the views of non-English speaking people (Aminzedah & Edwards 1998; Commonwealth of Australia 2000; Kong et al. 2002; Health Education Board 1999). Four studies were conducted in the UK. Quantitative All studies were based in the community, except Simpson (1995) who surveyed patients on a rehabilitation ward; and Wielandt (2002) and CulosReed (2000) who covered all settings. Most studies were conducted in the United Kingdom, USA or Australia. Outcomes Qualitative All studies examined peoples views or knowledge of falls prevention. Two examined perceptions, motivations and barriers to physical activity (Grossman et al. 2003; Stead et al. 1997). Outcomes were measured in various ways, including semi-structured interviews and focus groups. Commonly, the output from data collection was condensed into themes and categories. Quantitative These studies mainly measured or reviewed the following: predictors of increased exercise compliance, behaviour change, falls history, fear of falling, ability and confidence, self-efficacy, participation rates, or activity levels. Variables were categorical, ordinal or open-ended. 3.3.8.8 Methodological quality of studies
Page 99 of 315
Qualitative Ten qualitative studies of reasonable quality were found and reviewed. The results of the quality assessment are included in Appendix F (2004). Respondent validation where the analysis of the study is fed-back to the participants for validation was the one criterion for which studies exhibited the most variable quality. However, the studies mainly scored well on other criteria and all were considered worthy of data extraction. Qualitative methods used ranged from phenomenology a qualitative method used to gain information on patients experiences, in their own words to discourse analysis in which the output was subject to interpretation by the researcher. Many studies did not state a theoretical position. However, all papers appeared to be based on a similar framework, aiming to capture and analyse participant accounts and experiences of falls prevention or physical activity, using focus groups or unstructured/semi-structured interviews to collect data. Quantitative Overall the quality of the available studies was poor to fair. There were a limited number of review or summary papers and only one of these was done systematically (Hillsdon 1995). The conclusions authors drew and the recommendations they made very often did not flow from their own study results and/or from synthesising their results with previous work. None of the randomised trials had undertaken power calculations, so it is difficult to assess the reliability of these results. More details on the quality of each included study are included in the column comments/quality issues of the Evidence table 7a and 7b, Appendix E (2004). Characteristics of excluded studies are shown in Appendix G (2004). 3.3.8.9 Evidence summary
Studies focussed on patient views of either specific interventions, such as assistive/mobility aids (Aminzedah & Edwards 1998); or multiple separate interventions (Commonwealth of Australia 2000; Simpson et al. 2003) or a
single approach such as exercise (Health Education Board 1999; Stead et al. 1997; Grossman et al. 2003). There was no qualitative study that investigated older peoples views on multifactorial packages. A number of studies also focussed on the likelihood of adopting preventative practices and need for information on falls prevention (Ballinger & Payne 2000; Kong et al. 2002; Porter 1999; Resnick 1999). Table 13 summarises the facilitators to and barriers to falls prevention and physical activity from these studies. Most of the studies investigating potential participants views of falls prevention were conducted independently of trials of falls prevention. It is possible that if conducted concurrently as part of a trial the results may be different. Furthermore, it was not clear from many studies if any of the subjects had previously participated in falls prevention programmes. Nonetheless, important information is provided that requires consideration in addition to the clinical effectiveness evidence, when recommending which falls prevention programmes are suitable for whom and under what conditions.
Social stigma attached to programmes targeting older people. Low health expectation and low confidence in physical abilities. Differing agendas between older people and health professionals. Pain, effort and age (in relation to exercise programmes). Programmes with an emphasis on balance and strengthening.
Barriers Fear of falling; reluctance to walk; pain, effort and age (in relation to exercise programmes).
3.3.9
3.3.9.1
The focus of this review was on rehabilitation interventions following an injurious fall, which resulted in treatment within either primary or acute care. Rehabilitation involves a number of approaches from intensive training programmes from multifactorial interventions to single more targeted interventions that focus on balance or strength exercise training. These interventions can be given through specialist care from therapists or via a multidisciplinary team. Therefore the aim of this review was to determine the effectiveness of these programmes for rehabilitation, following a fall that resulted in hospitalisation. Definitions The following explains the differences between primary prevention, secondary prevention and rehabilitation for the purposes of this review. Also defined is injurious fall. Primary prevention interventions that are targeted at those at risk or high risk of a fall. Secondary intervention interventions that are targeted at those with a history of falls. Rehabilitation interventions that are targeted at those who have suffered an injurious fall. Injurious fall fall resulting in a fracture or soft tissue damage that required treatment. 3.3.9.2 Objectives
The review sought to answer the following questions: What are the most effective methods of rehabilitation/intervention/process of care, following an injurious fall?
3.3.9.3
Selection criteria
The Cochrane review on interventions for the prevention of falls was the principle source of evidence for this review, as this provided the most up-todate evidence of falls prevention programmes, including some specific to rehabilitation strategies. Data from the RCTs included in this review that met the selection criteria were extracted. A further search was conducted to ensure all relevant trials specific to rehabilitation had been identified. In addition, key relevant published documents relating to rehabilitation, such as guidelines and systematic reviews nominated by the GDG, were reviewed. Types of studies For individual studies we selected RCTs, controlled clinical trials, controlled before and after studies, and interrupted time series analyses. Included were studies in the Cochrane review, which had been conducted on participants who were selected on the basis of an injurious fall, and were given rehabilitation in residential settings or in the home. This included studies examining early discharge programmes. In addition, key documents such as clinical guidelines, health technology assessments, systematic reviews and other important policy documents relating to rehabilitation were sought. Participants Older people mainly more than 65 years of age but 60 acceptable who had sustained an injurious fall and received care/treatment from primary care, or acute care as an inpatient or outpatient. Settings Accident & Emergency, community-dwelling and extended care. Rehabilitation programmes implemented within inpatient discharge plans/programmes.
Interventions Any intervention that is implemented for the purposes of rehabilitation following an injurious fall. For example: exercise/strength training nurse/therapist interventions balance training home modification early discharge vs. hospital rehabilitation education assistive devices multidisciplinary and community support. Outcomes Reduction in number of falls/injurious falls. 3.3.9.4 Search strategy
A search was conducted to ensure all relevant papers were gathered for this review, in addition to those identified in the Cochrane review, and to identify key documents relating to rehabilitation. The first search was conducted in October 2002, and it was updated on all selected databases in July 2003. Seven electronic databases were searched between 1980 and October 2002, using a sensitive search strategy. The search strategies and the databases searched are presented in Appendix B (2004). All searches were comprehensive and included a large number of databases. All search strategies were adapted for smaller or simpler databases or for web-based sources, which did not allow complex strategies or multi-term searching. A combination of subject heading and free text searches was used for all areas. Free text terms were checked on the major databases to ensure that they captured descriptor terms and their exploded terms.
3.3.9.5
Data abstraction
No statistical analysis of inter-rater reliability of dual data extraction was performed. Differences were resolved by discussion. 3.3.9.6 Appraisal of methodological quality
Once individual papers were retrieved, the articles were checked for methodological rigour, using quality checklists appropriate for each study design (Appendix F, 2004), applicability to the UK and clinical significance. Assessment of study quality concentrated on dimensions of internal validity and external validity. Information from each study that met the quality criteria was summarised and entered into evidence tables. Quality appraisal for this review was based on the Cochrane review criteria of assessment of methodological quality (Appendix F, 2004).The two papers excluded from the Cochrane review relevant to rehabilitation (Tinetti 1999 and Crotty 2002) included here were quality appraised using the Cochrane quality criteria. 3.3.9.7 Data synthesis
Individual study results were reported in evidence tables. 3.3.9.8 Details of studies included
As detailed below, nine studies were relevant and included from the Cochrane review. There were nine trials from the Cochrane review relevant to rehabilitation (see table 14). Included studies were: Close et al. (1999); Crotty et al. (2002);
Ebrahim (1997); Kingston (2001); Lightbody (2001); Pardessus (2002); Rubenstein (1990); Shaw (2003); Tinetti (1999). TABLE 14: INCLUDED STUDIES FOR REHABILITATION REVIEW
Cochrane review: included studies Sifted relevant to this review 58 7 + 2 from the excluded 97 references related to review topic 9
Included
Two of these papers Tinetti (1999) and Crotty (2002) were excluded from the Cochrane review on the grounds that falls were only measured as adverse events, rather than as a primary outcome. However, they are relevant to this review as they evaluate rehabilitation programmes post-injurious fall. Data were extracted directly from the original paper and relative risks (RR) calculated. Results of the supplementary search for additional trials and key documents are shown in the table below. TABLE 15: RESULTS OF SUPPLEMENTARY RESEARCH
Total number of hits n screened n relevant Final number of articles included 1684 26 9 1 trial 7 relevant documents
The supplementary search conducted for this review elicited one further RCT for inclusion (Crotty 2002), which had not been included in the Cochrane review. Many studies were identified that had examined the effects of rehabilitation on intermediate outcomes for example: mobility, quality of life and psychosocial factors but these studies did not measure subsequent falls as an outcome. The key documents identified are listed below and summarised in Evidence table 9, Appendix E (2004).
Settings Five trials reported the effect of interventions in A&E settings. In two trials, the intervention was initiated within a hospital setting and continued in the community. Two further trials included participants from an extended care setting and community-dwelling. Participants Three trials recruited participants presenting to A&E, following a fall, who were discharged home following treatment. One trial set in an A&E setting, recruited cognitively impaired participants with a recent fall requiring treatment. Two trials recruited participants in a hospital setting following surgical treatment of a hip fracture, and one other recruited those who had been hospitalised following a fall. Two trials recruited participants with a history of falls in an extended care setting and community-dwelling. Interventions Close (1999) compared a multifactorial intervention with usual care in community-dwelling individuals presenting at A&E following a fall. The intervention involved medical and occupational therapy assessments and targeted interventions; medical assessment to identify primary cause of fall and other risk factors; with an intervention or referral as required, and home visit by occupational therapist. Participants were at least 65 years old with a
Falls: NICE clinical guideline 161 (June 2013) Page 108 of 315
history of falling, having presented at A&E with a subsequent fall. Falls data was obtained by a falls diary with four monthly follow-up for a period of one year. Crotty (2002) compared accelerated discharge and home-based rehabilitation, including home modifications with conventional treatment in those admitted for surgical treatment for a hip fracture. The intervention included a home visit by a physiotherapist, an occupational therapist, a speech pathologist, a social worker and a therapy aid, who negotiated short-term goals with a participant and their carer. The sample size was small (n=66) but no losses to follow-up were reported. Tinetti (1999) compared systematic multi-component rehabilitation with an aids to daily living strategy, with usual care with limited activities, in non demented older persons who underwent surgical repair of hip fracture and returned home within 100 days. This intervention included physical therapy involving assessment and exercise programmes individually tailored in strength, gait, balance, transfers and stair climbing. It also included functional therapy, based on principles of occupational therapy, to identify and improve performance of tasks of daily life. Both these programme elements involved tapered visits up to six months. Ebrahim (1997) compared general advice on health and diet, and encouraging brisk walking for 40 minutes, three times per week, with general advice on health and diet with upper limb exercises. Participants were post-menopausal women identified from A&E and orthopaedic fracture clinic records who had a fractured upper limb in the last two years. Kingston (2001) compared rapid health visitor intervention within five working days of index fall and multiple interventions, managed on an individual basis for 12 months, with usual post fall treatment in community-dwelling women attending A&E after a fall who were discharged directly home. The multiple interventions programme included pain control, getting up after a fall, education about risk factors, advice on diet and exercise to strengthen muscles and joints in an individualised programme.
Lightbody (2002) compared multifactorial assessment by a dedicated falls nurse, with usual care in consecutive patients attending A&E following a fall. The intervention included one home visit for assessment of medication, vision, hearing, balance mobility, feet and environmental assessment, with referral to a range of other services as required. Advice was also given and education about home safety. Pardessus (2002) compared a comprehensive two-hour home visit with specialist health care professionals of multifactorial interventions with usual care in those hospitalised with a recent fall, recruited in hospital. The intervention included assessment by specialist occupational therapist, rehabilitation doctor and physician prior to discharge. Environmental hazards were identified and modified and social support was given. Rubinstein (1990) compared nurse practitioner assessment within seven days of a fall with referral for intervention to physician for recommendations for action and referral for intervention with usual care in men and women in long-term residential care. Shaw (2003) compared multifactorial, multidisciplinary clinical assessment and intervention given for identified risk factors, with clinical assessment but no intervention in older people with cognitive impairment or dementia attending A&E after a fall. This intervention included medical, physiotherapy, occupational therapy and cardiovascular assessment with interventions for all identified risk factors. 3.3.9.9 Methodological quality of studies
A summary of the methodological quality of each study of the trials is shown in Appendix F (2004). In four studies, assignment of treatment was adequately concealed. (Crotty 2002; Ebrahim 1997; Rubinstein 1990; Shaw 2003). In the remaining five studies, information was inadequate to judge concealment. (Close 1999; Kingston 2001; Lightbody 2002; Pardessus 2002; Tinetti 1999). The overall quality scores were high-medium for two studies. (Crotty 2002; Shaw 2003).
They were medium in five studies. (Close 1999; Ebrahim 1997; Lightbody 2002; Pardessus 2002; Tinetti 1999). They were medium to low in one study (Rubinstein,1990) and low in one study (Kingston 2001). Losses to follow-up ranged from 0 (or not stated), (Crotty 2002; Rubinstein 1990) to 41 per cent (Ebrahim 1997) mostly the studies fell within the 20 per cent quality cut-off or just outside at 23 per cent (Close 1999). Five studies were based on intention to treat analysis (Crotty 2002; Pardessus 2002; Rubinstein 1990; Shaw 2003; Tinetti 1999) while for four studies, intention to treat analysis was not possible, as no outcome data were available (Close 1999; Ebrahim 1997; Kingston 2001; Lightbody 2002). Active registration of falling outcomes or use of a diary was clearly indicated in five studies (Close 1999; Crotty 2002; Lightbody 2002; Rubenstein 1990; Shaw 2003) or was by participant recall, at intervals during the study or at its conclusion (Ebrahim 1997), or was not described in three studies (Kingston 2001; Pardessus 2002; Tinetti 1999). Characteristics of excluded studies are shown in Appendix G (2004). 3.3.9.10 Comparisons
Trials were included in which participants were randomised to receive an intervention or group of interventions versus usual care to minimise the effect of, or exposure to, any risk factor for falling. Studies comparing two types of interventions were also included. 3.3.9.11 Evidence summary
The studies reporting significant results suggest that a multifactorial approach, including multidisciplinary assessment and targeted interventions, could have some impact on reducing the incidence of falling as part of a rehabilitation programme, following a fall resulting in medical attention. It is less clear from this evidence of the impact of these complex interventions on other factors such as confidence; quality of life and acceptability as limited data were available. There perhaps also needs to be consideration of the planned
withdrawal of such programmes and the ability of these individuals to sustain the improvement shown. It is less clear which specific mechanisms of this multifactorial approach to rehabilitation are effective, but the fundamental key to success may be through comprehensive discharge planning. This evidence is supported by key documents, in particular the expected standards of care outlined in the NSF for older people (standard six).
3.3.10
3.3.10.1
Background
Although hip protectors do not prevent falling, they do prevent one of the consequences of falling, that is hip fracture. Therefore, they can be considered as a secondary prevention/rehabilitation strategy in patients at risk of falling. The use of padding worn around the hip has been advocated as a measure of reducing the impact of the fall and thereby the chance of fracturing the hip. The fracture is usually the result of a fall. The fall usually occurs whilst standing or walking and the impact with the ground is usually on the side in the region of the hip (Hopkinson-W 1998). The rationale and development of such protectors has been summarised in Lauritzen (1977) and Lauritzen (1996).Various types of padded hip protectors have been developed. Most consist of plastic shields or foam pads, which are kept in place by pockets within specially designed underwear. A Cochrane review on the effectiveness of hip protectors has recently been updated (Parker et al. 2003). The methods and results of the review are summarised below and are taken from Parker et al. (2003). The full details are available at the Cochrane Library. 3.3.10.2 Objectives
Do hip pads or protectors worn about the hip reduce the risk of fracturing the hip?. 3.3.10.3 Selection criteria
Types of studies All randomised controlled trials comparing the incidence of hip fractures in those allocated to wearing hip protectors with the incidence in those not allocated to using protectors. Quasi-randomised trials were also considered for inclusion. Types of participants Older people of either gender living in the community or in institutional care. Types of intervention Allocation to wearing of hip protectors, or to not wearing hip protectors. Types of outcome Incidence of hip fractures over the study period Incidence of pubic rami and other pelvic fractures Incidence of other fractures Incidence of reported fall Mortality Compliance with protectors Reported complications of use of protectors, including skin damage/breakdown Cost effectiveness of the protectors. 3.3.10.4 Search strategy
The following sources were searched: Cochrane Musculoskeletal Injuries Groups specialised register (April 2003) Cochrane Central Register of Controlled Trials (The Cochrane Library issue 1, 2003) MEDLINE (1966 to April 2003)
EMBASE (1988 to 2003 Week 14) CINAHL (1982 to April 2003) reference lists of relevant articles trialists were contacted and ongoing trials identified in the National Research Register (http://www.updatesoftware.com/national/ accessed 20/01/03) and Current Controlled Trials (http://controlled-trials.com/ accessed 20/01/03). The search strategies and the databases searched are presented in Appendix B (2004). All searches were comprehensive and included a large number of databases. All search strategies were adapted for smaller or simpler databases or for web-based sources, which did not allow complex strategies or multi-term searching. 3.3.10.5 Sifting process
Once articles were retrieved the following sifting process took place: First sift: for material that potentially meets eligibility criteria on basis of title/abstract by two reviewers Second sift: full papers ordered that appear relevant and eligible and where relevance/eligibility not clear from the abstract by two reviewers Third sift: full articles are appraised that meet eligibility criteria by two reviewers. 3.3.10.6 Data abstraction
Data from included trials were extracted by two reviewers into pre-prepared data extraction tables. Discrepancies were discussed and resolved. The following data were extracted from each study: patient inclusion/exclusion criteria care setting key baseline variables by group description of the interventions and numbers of patients randomised to each intervention
description of any co-interventions/standard care duration and extent of follow-up outcomes acceptability and reliability if reported. If data were missing from reports then attempts were made to contact the authors to complete the information necessary for the critical appraisal. If studies were published more than once, the most detailed report was used as the basis of the data extraction. 3.3.10.7 Appraisal of methodological quality
The methodological quality of each trial was assessed by two researchers independently using a 10-item scale, with a total score for each trial. Full details of the principles of quality used in this review are reported in Appendix C (2004). The following quality criteria were used: description of inclusion and exclusion criteria used to derive the sample from the target population description of a priori sample size calculation evidence of allocation concealment at randomisation description of baseline comparability of treatment groups outcome assessment stated to be blinded clear description of main interventions intention to treat analysis timing of outcome measures reporting of loss to follow-up compliance of treatment. 3.3.10.8 Data synthesis
For each study, relative risk (RR)(fixed effect) and 95 per cent confidence limits (CI) were calculated for dichotomous outcomes. However, the authors of the review caution that the results must be considered as exploratory for the studies that used cluster randomisation. As cluster randomisation results in reduced effective sample size and statistical power, analysis using the
number of patients in each group gives inappropriately narrow confidence intervals (Parker et al. 2003). Results from individually randomised trials were pooled using the fixed effects model. Heterogeneity between comparable trials was tested using a standard chi-squared test. All statistical analysis was performed on Revman (v3.1.1) and conducted by the CWG. 3.3.10.9 Details of studies included in the review
Thirteen randomised controlled trials were included (seven in the previous review 1999). Settings The 13 included studies involved a total of 6,849 older people in residential settings or community dwelling. Within these, three studies were in a community-dwelling setting (one UK-based); the remaining 10 were conducted in a residential setting (one UK). Participants Mean age of participants in the individual studies, where reported, ranged from 80 to 86 years. Interventions Protective hip pads placed in the region of the greater trochanter were used in all trials. Ordinary underwear with no special fixation for the hip pad was used in Ekman (1997). The hip pads were fixed or sewn into special underwear in 12 studies (Birks 2003; Cameron 2001; Cameron 2003; Chan 2000; Harada 2001; Jantti 1996; Hubacher 2001; Meyer 2003; Kannus 2000; Lauritzen 1993; Van Schoor 2003; Villar 1998). All studies except two used an energy shunting design. In Jantti (1996) energy absorbing safety pants were used and for Chan (2000) the pads of local design for which it was not possible to say if they were energy absorbing or shunting. Outcomes: See Evidence table 11 (Appendix E, 2004) for details of other outcomes measured in the included trials.
3.3.10.10 Methodological quality of the studies A summary of the methodological quality of each study of the trials is shown in Appendix F (2004) and principles of quality assessment in Appendix C (2004). Eight studies were randomised by participant (Birks 2003; Cameron 2001; Cameron 2003; Chan 2000; Janitti 1996; Hubacher 2001; Van Schoor 2003; Villar 1998). In Birks (2003), randomisation was carried out by a remote randomisation service accessed by telephone. Cameron (2001), Cameron (2003) and Janitti (1996) randomised the patients individually by sealed envelopes.Van Schoor (2003) used computer generated random numbers. Chan (2000) stated that the method of randomisation was by taking draws literally. About half the participants in Hubacher (2001) were randomised by the head of the nursing home; the remainder were randomised by a computer. No details of the method of randomisation were provided by Villar (1998). The remaining five studies were cluster randomised. The unit of randomisation in Lauritzen (1993) was the nursing home ward occupied by the participants, selected by an independent physician drawing the number of the 28 nursing home wards. In Ekman (1997), residents of one of four nursing homes were offered the hip protectors with the other three homes acting as controls. Kannus (2000) used an independent physician drawing sealed envelopes to randomise treatment units within 22 community based health care centres. Losses within treatment units during the study were replaced from a waiting list. It is unclear how selection bias was avoided in this process. Harada (2001) used the even or odd digit of the patients room number to allocate participants. Each room had up to four patients. The unit of randomisation in Meyer (2003) was a nursing home or independently working wards in large nursing homes. Forty-nine clusters were randomised by phone from an external central location using computer-generated lists. Characteristics of excluded studies are shown in Appendix G (2004).
3.3.10.11 Comparisons The comparisons relevant to this guideline and able to be made on the basis of the included studies were: allocation to wearing of hip protectors, or to not wearing hip protectors. 3.3.10.12 Evidence summary Parker et al. (2003) report the following: Five studies involving 4,316 participants were cluster randomised by care unit, nursing home or nursing home ward rather than by the individual. Individually, each of these studies reported a reduced incidence of hip fractures within those units allocated to receive the protectors. Because of the use of cluster randomisation, pooling of results of these studies was not undertaken. Pooling of data from five individually randomised trials conducted in nursing/residential care settings (1,426 participants) showed no significant reduction in hip fracture incidence (hip protectors 37/822, controls 40/604, RR 0.83, 95% CI 0.54 to 1.29). Two individually randomised trials of 966 community-dwelling participants, reported no reduction in hip fracture incidence with the hip protectors (RR 1.11, 95% CI 0.65 to 1.90). No important adverse effects of the hip protectors were reported but compliance, particularly in the long-term, was poor. See Evidence table 10 and 11, Appendix E (2004) for further details of studies and outcomes. Implications for practice (Parker et al. 2003) Reported studies that have used individual patient randomisation, have provided insufficient evidence for the effectiveness of hip protectors when offered to older people living in residential care or in their own home. Data from cluster randomised studies provide some evidence of effectiveness of hip protectors in reducing the risk of hip fractures in those living in nursing homes and considered to be a high risk of hip fractures.
Reported adverse effects of hip protectors are skin irritation, abrasion and local discomfort. Compliance with wearing the protectors remains a problem. Full evidence reviews are included under the relevant recommendations in Section 3.4, along with evidence statements.
3.3.11
To fulfil the DH and Welsh Assembly Government remit, NICE requested that the cost effectiveness evidence of interventions for the assessment and prevention of falls in older people be assessed. In accordance with the objectives of the scope, cost effectiveness was addressed in the following way: a comparison of the cost and cost effectiveness of falls prevention interventions compared with usual care, other intentions or no intervention; and an investigation of which types of falls prevention programmes are the most cost effective. The aim of the review was twofold. Firstly, to identify economic evaluations that had been conducted alongside trials and secondly, to identify evidence that could be used in cost effectiveness modelling. Health economic evidence The searches for economic evidence were designed to identify information about the resources used in providing the existing service, and any additional resource use associated with increased interventions and the benefits that could be attributed. The searches were not limited to RCTs or formal economic evaluations. The search strategy is shown below and the number of papers, sorted by intervention. Identified titles and abstracts from the economics searches were reviewed by the health economist and full papers obtained as appropriate. The full papers
were critically appraised by the health economist. Consideration was given to each study design and the applicability of the results to the guideline context. Quality was assessed using the Drummond et al. (1999) economic evaluation checklist. An important issue in this respect is that much of the evidence on costs and benefits comes from health care systems outside a UK setting and are therefore of limited value to a UK guideline. Searching for health economics evidence The searching was carried out by an information scientist at the School of Health and Related Research (ScHARR), with guidance on the search terms from the health economist. Search strategy The search strategy used was as follows: Economic evaluations Fall or falls or falling or fallers Accidental-falls/all subheadings old or older or senior* or elder* or aged or geriatric* explode Aged/all subheadings Middle-Age/all subheadings 1 2 3 4 5 6 7 8 9 10 economics/ exp costs and cost analysis/ economic value of life/ exp economics, hospital/ exp economics, medical/ economics, nursing/ economics, pharmaceutical/ exp models, economic/ exp fees and charges/ exp budgets/
11 12 13 14 15 16 17 18 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.
ec.fs (cost or costs or costed or costly or costing$).tw (economic$ or pharmacoeconomic$ or price$ or pricing).tw or/1-13 exp quality of life/ quality of life.tw life quality.tw. hql.tw (sf 36 or sf36 or sf thirtysix or sf thirty six or short form 36 qol.tw. (euroquol or eq5d or eq 5d).tw. qaly$.tw quality adjusted life year$.tw hye$.tw health$ year$ equivalent$.tw. health utilitie$.tw. hui.tw. quality of well-being$.tw. quality of well being.tw. qwb.tw. (qald$ or qale$ or qtime$). Tw.
32. 32.
Searches were done from 1966 to the present (April 2003) and initially with no language restrictions. The following databases were searched Medline Embase NHS EED OHE HEED Databases were searched in April 2003 and from these searches there were 2,354 hits. In addition, reference lists from appraised papers were checked for further useful references. The systematic reviewer at the NCC also noted any potentially suitable references and passed them on to the health economist. Inclusion criteria The titles and, where available, the abstracts were screened to assess whether the study met the following inclusion criteria: Population: older people who had had a fall or were deemed at risk of a fall. Economic evidence: the study was an economic evaluation or included information on resources, costs or specific quality of life measures. Study design: no criteria for study design were imposed a priori. Exclusion criteria Papers were excluded if they did not contain cost effectiveness data, quality of life data or were simply a description of costs. An exception to this was made when examining papers that were of use in providing data on the costs of an intervention for any cost effectiveness modelling. Papers of this type needed to include a breakdown of resource use, unit costs, the source of the data, the year it was collected and the level of discounting applied.
Sifting was carried out by one assessor. Initially all papers that included the terms cost effectiveness,quality of life or costs were selected. The abstracts were checked where possible and those papers that were descriptive or commentary were excluded. Summary of results After reviewing titles, abstracts and CRD/OHE HEED commentaries (where available), 106 potentially useful papers were included. A small number of these papers included background information and more detailed input about the interventions and issues involved. Six papers were in languages other than English and were not obtained. Full papers were obtained and a significant number proved to be unhelpful. Papers had been ordered that contained at least one of the key words, costs, or quality of life and/or economics. On review, these papers were often found not to contain any data. This was particularly the case in papers that mentioned cost and quality of life in the title or abstract. This reduced the included papers to 14.Very few of these were good quality formal economic evaluations. Table 17 shows the areas directed by the GDG and the number of papers that were reviewed in each area. TABLE 17: COST EFFECTIVENESS PAPERS REVIEWED
Area Financial cost of falls to the NHS Pharmaceutical interventions Exercise programmes Tai Chi Home hazard assessment and modification Multifactorial interventions Hip protectors Numbers of papers reviewed 1 2 4 1 2 2 2
Table 18 below details the papers included, the methodology used in the studies and the cost effectiveness results
TABLE 18
Author, year & country Roberts on 2001a NZ Interventi on Client group age 80 Outco me measu re Fall reducti on Metho d e.g RCT Yes Costs included Cost per perso n NZ $432 Cost effectiven ess NZ $1,803 per fall prevented
Homebased exercise
Roberts on 2001b NZ
Homebased exercise
80
Fall reducti on
Yes
Centrebased exercise
68 - 65
Balanc e, gait, fall reducti on Overall health includin g falls Fall reducti on
Yes
All costs associate d with the interventi on. Treatmen t costs. All costs associate with the interventi on. Treatmen t costs. Not reported. Treatmen t costs. Not stated. Treatmen t costs. All costs associate d with interventi on. Treatmen t costs. All costs.
NZ$4 18
N/A
N/A
Yes
N/A
N/A
Yes
A$98
75
Fall reducti on
Multifacto rial
70
Fall reducti on
A$17 2
A$1721
$891
$2150
Refer to Section 3.4 for recommendations and cost effectiveness details for each intervention.
In addition to the evidence reported in Section 3.4, the report by Scuffham et al. (2003) was also considered as it contains information on the incidence and costs of falls in the UK. The authors accessed the dataset from the Department of Trade and Industry (DTI) to examine the data collected in the year 2000 from the participating A&E departments. They report the total cost of falls to the UK government as more than 1 billion. Just over half this cost was incurred by the NHS and rest by personal social services mainly in long-term care costs. They demonstrate the correlation between increasing age and less favourable outcomes after a fall. Excluded studies A study by Wilson and Datta (2001) Tai Chi for the prevention of fractures in a nursing home population: an economic analysis is a literature based costbenefit analysis. This study was reviewed and has been excluded on the grounds that the data used to populate this model was inappropriate. The data on the relative risk of falls is not compatible with the risk of hip fracture. In addition the GDG requested that we review the following studies. Two studies by Kenny were obtained and appraised. Neither of these studies met the inclusion criteria for economic evaluation. (Kenny 2001; Kenny 2002). A further abstract from the PROFET study (Close 1999) was appraised. This reports on an economic evaluation however, as an abstract, there is insufficient information to allow a full crucial assessment. Hip protectors Hip protectors are considered as a secondary prevention/rehabilitation strategy in patients at risk of falling. The use of padding worn around the hip has been advocated as a measure of reducing the impact of the fall and thereby the chance of fracturing the hip. The recent updated evidence on clinical effectiveness is inconclusive (Parker et al. 2003). For an intervention to be cost effective, it must first be clinically effective.
Summary of the health economics evidence Although clinical and cost effectiveness data exists for falls prevention, there are no UK studies. The quality of reporting in these studies is often patchy, as some costs and benefits are reported and not others. The above studies did not use the same costing methods or always report incremental costs or discounting. Those from countries other than the UK have limited applicability as the health care systems are often very different. Even in the small number of studies included, few comparisons can be made between studies due to the differences in methodology. Identifying those individuals who may benefit most from an intervention is not always reported.Who should be targeted for screening; when screening should take place; and at what intervals is an area of considerable uncertainty in terms of costs and benefits. There is a lack of cost effectiveness evidence in this area and therefore, we would recommend further research. The cost effectiveness of interventions to prevent falls in the elderly: modelling report Introduction Successive Government initiatives have identified falls in the elderly as a major cause of morbidity and mortality (DTI 2002). It has been estimated that between one-third and half of people above the age of 65 fall each year. Falls in the elderly result in extensive use of National Health Service resources. Scuffham and Chaplin report that in 2002, 400,000 A&E attendances per annum were attributable to accidents involving older people. There is also evidence of substantial mortality associated with such accidents. Interventions that reduce the likelihood of falling or injury in the event of a fall have the potential to save NHS resources and improve the health of the UKs increasingly elderly population.
A systematic review of the published literature up to August 2003 found no published cost effectiveness analyses of strategies for falls prevention in the elderly. In this chapter we report cost effectiveness analyses of two falls prevention strategies; exercise programmes for at risk individuals dwelling in the community and multifactorial interventions for at risk individuals dwelling in the community. For results and discussion please refer to Section 3.4. Methods A simple life table model was constructed for people aged 60 and over. The model starts with a cohort of 100 people aged 60 and runs on an annual cycle until all the members of the cohort are dead. In each year, each person faces a risk of death and a risk of experiencing a fall leading to a contact with local accident and emergency department. For each year of life there is a health related quality of life weight. This weight is on a scale between zero and one; where one is the value of full health and zero is the value given to a health state equivalent to being dead. The risk of a fall is taken from the report by Scuffham and Chaplin. The risk of mortality is taken from the all cause mortality statistics published in Office of National Statistics Population Trends. The quality of life weight is taken from the population norms for the EQ-5D published by Kind et al. (1998). Each fall incurs a cost of care and a reduction in quality of life. The cost of fall related injuries, except for hip fractures, is based upon the data reported by Scuffham and Chaplin. The cost reflects NHS and social service costs only that is the cost effectiveness analysis is from the NHS perspective, not that of society as a whole. Using the data from Scuffham and Chaplin, it is assumed that the severity of the injury determines the NHS services received. Thus all events lead to an attendance at A&E with an ambulance journey. It was necessary to make assumptions about the relationship between event and subsequent treatment, as no data was available. The assumptions were: Ordinary fractures are assumed to be treated at A&E with an outpatient follow-up.
Other fractures are assumed to be treated by hospital admission. Bruises, cuts, abrasions, and tenderness or swelling are assumed to be treated at A&E, with GP follow-up. Concussion and loss of consciousness are assumed to be treated by hospital admission. The follow-up from hospital admissions was modelled on the basis of the data reported in Table 3.8 of Scuffham and Chaplin. Currently, the expected cost of each injury varies by age group, but not by injury site, with the exception of hip fracture. The unit costs of these events were obtained from the unit costs of health and social care report (Netton et al. 2003). The direct cost of treating hip fractures is taken from a study by Parrot (2000), published by the UK Department of Trade and Industry. The utility reduction associated with injury was defined as a proportion of baseline utility, therefore it varied by age. The utility decrement for hip fracture was ranged from 0.166 (aged 60-69 years) to 0.146 (aged >=80 years) associated with hip. The utility decrement for all other fractures ranged from 0.074 (aged 60-69 years) to 0.065 (aged >=80 years). This model structure is used to estimate the total costs and QALYs accruing to treated and untreated cohorts for two interventions: 1. Exercise programme to prevent falls in at risk older people dwelling in the community; and 2. Multifactorial assessment and intervention programmes to prevent falls in at risk older people dwelling in the community. The relative risk of falling associated with each intervention is taken from the meta-analyses reported in Appendix H (2004) of the clinical practice guideline. Detailed descriptions of these interventions are included in Section 3.3 of this guideline. The cost of the exercise programme is taken from the work by Munro et al. (2002) and adjusted using the NHS Pay and Prices Indices to 2003 prices.
The cost of risk assessment for both interventions is taken from the work by Close et al. on the PROFET study. (Personal Communication J. Close). All costs and quality adjusted life years (QALYs) are discounted at 3.5 per cent per annum. This is based upon the recommendations in the National Institutes Guideline development methods technical manual. The incremental cost effectiveness of each of the interventions is calculated as the difference in the mean costs for the intervention and control cohorts, divided by the difference in the mean QALYs lived by each cohort. In line with current best practice, we undertook probabilistic sensitivity analysis of the incremental cost effectiveness ratio. For this purpose, we defined probability distributions for the effectiveness of each of the interventions, representing our uncertainty as to the actual effectiveness of these interventions in practice. In addition to modelling the uncertainty on the effectiveness of the interventions, we considered the uncertainty relating to the costs of the intervention, the costs of treating injuries and the probability that a fall will lead to a hip fracture rather than any other injury. Costs were assumed to have a log normal distribution that is there is a small chance that the actual cost is much higher than the reported mean cost. This characteristic of cost data has been routinely reported in economic evaluations in many different areas of health care. The effectiveness of each intervention is described using a beta distribution. The beta distributions are characterised to reflect the 95 per cent confidence intervals reported in the guideline meta-analysis (Appendix H, 2004). A Monte Carlo simulation, with 10,000 simulations, was then used to produce a probability distribution for the value of the mean costs and mean QALYs for an untreated cohort; a cohort receiving the exercise intervention; and a cohort receiving the multifactorial intervention. It is the mean value of the simulations that are used to estimate the incremental cost effectiveness ratios (ICERs). 95 per cent confidence intervals around the ICERs are then estimated using the bootstrap method. Table 20 gives the parameter values used for the cost effectiveness analysis.
Results The mean and incremental costs and QALYs and the ICERs for each of the interventions are given in Table 19. TABLE 19: ICERS FOR MULTIFACTORIAL INTERVENTION IN THE AT RISK POPULATION AND THE EXERCISE PROGRAMME IN THE COMMUNITY POPULATION
Increment al costs Increment al QALYs Increment al cost effectiven ess ratio, s per QALY - 980 9,559 Mean costs () Mean QALYs 8,766 8,915 8,893 - 146 1,214 0.149 0.127 Group Control Multifactorial intervention Exercise intervention
Note: negative ICERs must be interpreted with great caution as they can be produced by negative costs and positive QALYs, or positive costs and negative QALYs. These outcomes are clearly not equivalent. The ICERs are highly labile; that is the small changes in the mean QALY gain will have a large impact upon the ICER. Sensitivity analysis The bootstrapped 95 per cent confidence interval for the exercise is -184,828 to +187,149. Figure 1 is a scatter plot of the incremental costs and incremental QALYs for the exercise intervention. The bootstrapped 95 per cent confidence interval for the multifactorial intervention is -19,533 to +75,270. Figure 2 is a scatter plot of the incremental costs and incremental QALYs of the multifactorial intervention. Discussion The central estimates for the ICER for both the multifactorial and exercise intervention indicate that both interventions are cost effective, compared to doing nothing. However, these results most be interpreted with great caution. The bootstrapped confidence intervals around the ICERS are large, reflecting the great uncertainty surrounding the evidence for the effect, and indeed the
costs of providing the interventions and the costs of treating fall related injuries.
0.0324
0.0324
0.0324
0.0324
PSSRU
110
110
110
110
PSSRU
166
166
166
166
PSSRU
130
130
130
130
Parrot S Munro et al. PSSRU PROFET trail abstract NICE Guideline Development Methods Guidance
22,360 25,425
22,360 25,425
22,360 25,425
22,360 25,425
164
164
164
164
Discount rate
6% pa
6% pa
6% pa
6% pa
Time horizon
Lifetime
Lifetime
Lifetime
Lifetime
The ICERs are labile. The health gain from the interventions is small and small absolute variations in this gain lead to very large changes in the ICER. This is shown in Figures 1 and 2, which show that whilst there is no evidence that interventions will do any harm to the recipients (and are therefore better than many other health care interventions); the actual location of the intervention in the cost effectiveness plane is unclear. The intervention may save money and produce health or it may produce health at a substantial price. More evidence is needed about almost all the parameters considered in the model; inter alia: 1. 2. 3. the quality of life impact of the full range of fall-related injuries the cost of treating fall-related injuries and the cost of the interventions.
Section 3.4 contains the full results under recommendations for multifactorial interventions and strength and balance. Figure 1: Cost effectiveness of exercise intervention in falls prevention: plot on the cost effectiveness plane of 10,000 simulations
Figure 2: Cost effectiveness of multifactorial intervention in falls prevention: plot on the cost effectiveness plane of 10,000 simulations
3.3.12
In December 2002, stakeholders registered with NICE (Appendix D, 2004) were invited to submit a list of evidence for consideration to ensure that relevant material to inform the evidence base was not missed. The criteria for the evidence included: systematic reviews randomised controlled trials (RCTs) that examine clinical or cost effectiveness, and/or quality of life and economic analyses based on these findings representative epidemiological observational studies that have assessed the incidence of falls in the UK qualitative studies/surveys that examine patient/carer experiences of having fallen or fear of falling studies of any design which have attempted to formally assess the cost effectiveness of fall prevention programmes; assess the cost of falls or fall prevention programmes; assess quality of life in relation to falls. Information not considered as evidence included:
studies with weak designs when better studies are available commercial in confidence material unpublished secondary endpoint trial data,data -onfile and economic modelling promotional literature papers, commentaries or editorials that interpret the results of a published study representations or experiences of individuals not collected as part of properly designed research. Submissions were received from: Abbott laboratories Limited (BASF/Knoll) Alzheimerss Society Ambulance Service Association British Geriatric Society British Urological Institute BUPA Chartered Society of Physiotherapy College of Occupational Therapists Health Development Agency Help the Aged (Department of Trade and Industry) Limbless Association Medtronic Limited National Osteoporosis Society Novartis Pharmaceuticals UK Ltd Pfizer Limited Roche Products Limited Royal College of Physicians Shire Pharmaceuticals Limited Society of Chiropodists & Podiatrists Submitted material received included notification of published, unpublished and ongoing research related to falls prevention.All references were screened for relevance and design criteria and those considered eligible were checked
Falls: NICE clinical guideline 161 (June 2013) Page 135 of 315
with our databases to ensure our search had captured such studies. None of the submitted references provided relevant material additional to the studies we had already identified. A list of registered stakeholders is included in Appendix D (2004).
3.3.13
Evidence gradings were assigned to evidence statements that were derived from the evidence reviews. The evidence hierarchy used is shown below (Table 21) and was the hierarchy recommended at the time by NICE. (It should be noted that the hierarchy applies to questions of effectiveness, though it is used here to grade evidence other than clinical effectiveness).
Adapted from Eccles M, Mason J (2001) How to develop cost-conscious guidelines. Health Technology Assessment 5:16 The evidence tables and reviews were distributed to GDG members for comment and discussion.
3.3.14
In order for the GDG to formulate a clinically useful recommendation, it was agreed that the following factors be considered: the best evidence with preference given to empirical evidence over expert judgement where available, including: results of economic modelling
effectiveness data, taking into account the strength of evidence the level, quality, precision as well as the size of effect and relevance of the evidence where reported, data regarding additional outcomes such as adverse events, patient acceptability and patient views a comparison between the outcomes for alternative interventions where possible the feasibility of interventions including, where available, the cost of the intervention, acceptability to clinicians, patients and carers and appropriateness of intervention the balancing of benefits against risks including, where reported, all patient-relevant endpoints and the results of the economic modelling the applicability of the evidence to groups defined in the scope of the guideline, having considered the profile of patients recruited to the trials. This information was presented to the group in the form of evidence tables, accompanying evidence summaries and evidence statements, with associated level of evidence grading. Interpretations of the evidence were discussed at GDG meetings. Where the GDG identified issues that impacted on considerations of the evidence and the ability to formulate implementable and pragmatic guideline recommendations, these have been summarised in the GDG commentary sections under each recommendation, though not all recommendations required a GDG commentary section. Issues relating to interpretation of the evidence and the wording of recommendations were discussed by the GDG, until there was agreement on the wording and grading of recommendations. Where the GDG decided that hard evidence was essential before any recommendations could be considered, recommendations for future research were made using the NICE guidance on formulating recommendations. The grading of the recommendations was agreed at the GDG meeting prior to first stage consultation using the scheme below.
Adapted from Eccles M, Mason J (2001) How to develop cost-conscious guidelines. Health Technology Assessment 5:16. TABLE 22: RECOMMENDATION GRADING
A B C D Directly based on category I evidence Directly based on category II evidence or extrapolated recommendation from category I evidence Directly based on category III evidence or extrapolated recommendation from category I or II evidence Directly based on category IV evidence or extrapolated recommendation from category I, II or III evidence
The resulting recommendations with evidence statements, abbreviated evidence summaries and GDG commentary are presented in Section 3.4.
3.4
Below are the recommendations agreed by the GDG, with associated evidence statements, evidence summaries and, where relevant, GDG commentaries on the consideration and interpretation of the evidence.
1.2.1 Case/risk identification (please see Sections 3.3.1 and 3.3.2 for evidence review methods)
1.2.1.1 Recommendation Older people in contact with health care professionals should be asked routinely whether they have fallen in the past year and asked about the frequency, context and characteristics of the fall/s. [C]
Evidence statement Level III A previous fall is the most recently reported risk factor in prospective cohort studies, suggesting than an older person with a history of falling would be at high risk of a subsequent fall.
Evidence summary Falls history Falls history is a frequently reported significant risk factor and predictor of potential further falls. Ten studies reported falls history as statistical significant, among community-dwelling older people (Northridge 1996; Covinsky 2001;Tromp 2001; Friedman 2002; Stenbacka 2002; Wood 2002), and among residents of extended care facilities (Thapa 1996; Cavanillas 2000; Kallin 2002). For older people in community-dwelling settings, the range of summary statistics (OR/RR) reported was: 1.5-4.0. Three studies were of high quality; three of medium quality and one was low quality, with a reported OR of 4.0. Studies conducted in extended care settings reported significant results, of which one high quality study reported a incident density ratio of 2.23 (1.44.37). Two other studies, of low quality, reported an odds ratio range of 1.9-
4.65. Seven studies reported falls history as significant in bivariate analysis but not in multivariate. Heterogeneity between these studies hinders interpretation of the clinical relevance of this finding. GDG commentary There is good evidence from cohort studies that an older person who has had a previous fall would be at risk of a subsequent fall. The group was keen to recommend that an older person be asked about their falls history based on this evidence. The purpose of obtaining this history would be to establish where possible, the frequency of falling; context and circumstances of the fall; and severity or injuries sustained from the fall. There was debate within the group of the best approach to identifying older people at risk, based on their previous falls history. Some were in favour of an annual review based on screening. Others considered that a case finding approach was more appropriate, asking an older person if they had fallen in the last year when seen by a health care professional. The group was in support of this being done yearly but did not want to reflect this in the recommendation. 1.2.1.2 Recommendation Older people reporting a fall or considered at risk of falling should be observed for balance and gait deficits and considered for their ability to benefit from interventions to improve strength and balance. (Tests of balance and gait commonly used in the UK are detailed in Section 3.3.) [C]
Level of evidence Level III Evidence statements - Mobility impairment, gait disorders and balance deficits have frequently been reported as significant risk factors in prospective cohort studies. - Many tests for the assessment of balance and gait are available to support clinical skill and the choice of such a tool should be determined at local level. -Intervention trials focusing on gait and balance have shown a reduction in falls.
Level I
Evidence summary Mobility impairment, gait disorders and balance deficits have frequently been reported as significant predictors of future falling in prospective cohort studies
(Bueno-Cavanillas 2000; Cesari 2002; Covinsky 2001; Northridge 1996; OLoughlin 1993; Stalenhoef 2002). Tests are available for the assessment of an older persons balance and gait that can inform clinical judgement. A detailed list of such tests is provided in Appendix E (2004), Evidence table 3. These range from simple, pragmatic tests that require no special equipment, to those that require a trained health care professional with skill to administer. GDG commentary The group felt that assessment of older people who have fallen at least once should include observation for balance and gait deficits. This could be done on first contact by an appropriately trained health care professional in any setting. Clinical judgement should support the use of any test referred to in the clinical evidence and many other tests, developed by different disciplines, are likely to be available in trusts. However, a simple observation of a patients ability to stand, turn and sit is considered adequate as a first level assessment. Older people with observed gait or balance problems should be referred for targeted interventions. Identifying those most likely to benefit should also be considered. The group was unable to recommend specific tests for use in practice, as there was a lack of robust validation studies. A profile of tools and tests identified in the assessment review is provided in Appendix E (2004), Evidence table 3. The choice of tests should be determined at local level.
appropriate skills and experience, normally in the setting of a specialist falls service. This assessment should be part of an individualised, multifactorial intervention. [C] 1.2.2.2 Recommendation Multifactorial assessment may include the following: [C] identification of falls history assessment of gait, balance and mobility, and muscle weakness assessment of osteoporosis risk assessment of the older persons perceived functional ability and fear relating to falling assessment of visual impairment assessment of cognitive impairment and neurological examination assessment of urinary incontinence assessment of home hazards cardiovascular examination and medication review.
Evidence statements Many individual risk factors have been proven to be predictive of a subsequent fall; therefore presence of more than one of the factors listed below increases the risk of falling: - falls history - gait deficit - balance deficit - mobility impairment - fear of falling - visual impairment - cognitive impairment - urinary incontinence - home hazards - number of medications - psychotropic and cardiovascular medications - muscle weakness.
Three community-dwelling studies reported this risk factor as statistically significant with a range of OR: 1.96-2.2 (Koski 1998; Cesari 2002; Northridge 1996). Four studies in community-dwelling settings reported non-significance in multivariate analysis (Northridge 1996; Stalenhoef 2002; Wood 2002; Tinetti 1995). No studies in extended care settings reported gait deficit as significant (Cavanillas 2000; Kallin 2002) although one study carried out detailed gait analysis and found sitting down incorrectly as significant in multivariate analysis significant (Cavanillas 2000). In all of the above studies, the method of measuring gait and aspects of gait analysis differed between studies. Balance deficit Three studies conducted among community-dwelling participants, reported balance as statistically significant with a range of summary statistics of 1.833.9 (OLoughlin 1993; Stalenhoef 2002; Covinsky 2001). However, each study measured different aspects of balance including dizziness, unbalanced and postural sway. Eight studies did not find aspects of balance significant in multivariate analysis, two of which were conducted in extended care settings (BuenoCavanillas 2000; OLoughlin 1993; Tinetti 1995; Northridge 1996; Koski 1998; Wood 2001; Stalenhoef 2002; Kallin 2002). Again, different aspects of balance were analysed. Mobility impairment Two community-dwelling studies reported statistical significance: In study one: trouble walking 400m: IRR=1.6(1.2-2.4); trouble bending down: IRR=1.4(1.02.0) (OLoughlin 1993). Study two conducted statistical modelling adjusting for different variables and reported the range for both multivariate models: OR=2.64-3.06 for mobility impairment (Covinsky 2001).
Four studies reported non-significance but as discussed earlier, different methods and aspects of mobility were measured (Bueno-Cavanillas 2000; Kallin 2002; Cesari 2002; Stalenhoef 2002). Fear of falling Three community-dwelling studies reported statistical significance of this factor, with a range of summary statistics 1.5 3.2, although different methods of measuring fear were used (Arfken 1994; Cumming 2000; Friedman 2002). This included use of the falls efficacy scale (FES) to explore different cut-off values for determining risk and verbal rating scales to identify the degree of fear present. One study measured fear at baseline and reported nonsignificance in the results (Tromp 2001). Friedman et al. (2002) carried out a prospective cohort study to examine the temporal relationship between falls and the fear of falling with n=2212 community-dwelling participants aged between 65 and 84 years. Fear was measured at baseline and at one-year follow-up with a simple yes/no answers to whether they were worried or afraid of falling, with a further question relating to their activity limitation when afraid of falling. This study was of high quality with a large sample and detailed baseline measurement. Logistical regression with adjustment for other variables in the model was performed on the data and results as follows. Fear of falling at baseline was significantly predictive of falling at follow-up with OR=1.78 (1.41-2.24), as well as fear at baseline predictive of fear at follow-up OR=5.40 (4.23-6.91). In addition to this, a fall at baseline was predictive of fear at follow-up 1.58 (1.24-2.01). Shared predictors of both falls and fear at follow-up include female gender and history of stroke. Cumming et al. (2000) carried out a prospective study to assess the impact of fear of falling with n=418 community-dwelling aged 65 and over. This study was of medium quality with a smaller sample size than others. The FES was administered at baseline with a total score of 100 indicating high fall related self-efficacy and 0 low fall related self-efficacy. Cut-off points were tested for predictive ability of falling in the analysis. Adjusted hazard ratio for all study participants with a FES score of <75 =
2.09 (1.31-3.33). Tromp et al. (2001) conducted a prospective study to examine all predictors for falls with n=1285 community-dwelling participants aged 65 years and more. This was a high quality study, with detailed baseline measurement and fall events measured with falls calendars. Fear was determined using a modified FES where answers were rated on a scale 0 (no confidence) to 3 (completely confident). Odds ratio for 1 fall and recurrent falls were significant in bivariate analysis but non-significant in logistic regression analysis. Arfken et al. (1994) recruited patients from a prospective cohort study in which the purpose was to develop a screening tool for predicting falls in older people. The sample was 890 community-dwelling participants stratified in age groups ranging from 66 to 81+years. Baseline data were collected as part of the parent study and falls surveillance was conducted with participants reporting falls to a hotline plus monthly postcards reporting the incidence of falls.At one-year follow-up,the participants received a structured in-home assessment including demographics, health status, activity level, satisfaction with life, depressed mood and a brief physical assessment. Fear was determined with a three point verbal rating scale and dichotomised to summarise outcome as odds ratios: A= moderately fearful and not fearful, B= very fearful.Logistic regression models adjusted for gender and age. Results indicated that those who were moderately or not fearful predicted falling at least once: A= 1.52 (1.06-2.17) and very fearful participants: (B= 2.49 (1.48-4.20). Those experiencing frequent falls were more likely to be very fearful of falling: B=3.12 (1.61-6.06) than those moderately or not fearful A=1.71 (1.01-2.89). Visual impairment Two community-dwelling studies found that older people with a visual impairment were significantly at risk of falling, OR range=1.18-2.3 (Northridge 1996; Koski 1998). One extended care study of low quality reported OR =5.85 (Kallin 2002).
Eight studies reported non-significance in multivariate analysis, two of which were extended care setting studies (Tinetti 1995; Northridge 1996; Tromp 1998; Cesari 2002; Stalenhoef 2002; Wood 2002; Thapa 1996; BuenoCavanillas 2000). Different aspects of vision were measured in these studies and included: visual impairment, visual acuity, depth perception and others. Furthermore, there are a number of prospective cohort studies which we have been alerted to by stakeholders that demonstrate that visual impairment is an independent risk factor for falls and hip fractures (Felson et al. 1989;Cummings et al. 2003; Ivers et al. 2000 and 2004). Cognitive impairment Two studies in community-dwelling settings reported that older people with cognitive impairment were significantly at risk of falling OR=2.2-2.4 (Tinetti 1995; Van Schoor 2002). One low quality study in an extended care setting reported OR 6.2 (1.7-23.3) (Bueno-Cavanillas 2000). However, nine studies did not find older people with cognitive impairment significantly at risk of falling in both settings (Tinetti 1995; Northridge 1996; Tromp 1998; Cesari 2002; Stalenhoef 2002; Van Schoor 2002; Thapa 1996; Kallin 2002; Wood 2002). Urinary incontinence, including stress and urge incontinence Two studies reported that older people suffering from urinary incontinence were at risk of falling with OR range=1.26-1.8 (Tromp 1998, 2001; Brown 2000). Additional studies that support incontinence as a significant risk factor include Luukinen 1996 and Tinetti 1995. Five studies did not find incontinence a significant predictor of falling (Tinetti 1995;Koski 1998; Brown 2000; Cesari 2002; Thapa 1996). Home hazards Two studies reported that the presence of home hazards increased an older persons risk of falling, One study reported OR=1.51 (95% CI1.43-1.69)
(Cesari 2002). The other study (Gill 2000) carried out detailed analysis and reported that the following contributes to the risk of falls: Loose rugs and mats: hazard ratio=5.87(95% CI 1.4224.2) Carpet fold or tripping hazard: hazard ratio=3.45(95% CI 1.29-9.27). Multiple medications Seven studies were included in a systematic review and meta-analysis of cardiac and analgesic drugs (Leipzig et al. 1999a). All report that patients taking more than three to four medications were at risk of recurrent falls compared with patients taking fewer medications (range of results: OR 1.61 to 3.16). The studies included in this review were cohort, case control and cross sectional in design. Anti-arrhythmic medications In a meta-analysis of cohort, case control and cross-sectional studies (Leipzig 1999a), the following pooled results of 14 studies indicated that taking type 1A antiarrhythmic drugs increase the risk of falling (OR 1.22, 95% CI 1.05 to 1.42). Psychotropic medications In a systematic review and meta analysis of cohort, case control and crosssectional studies examining psychotropic drugs and falls (Leipzig 1999b) the pooled results for the association between taking any psychotropic drug and risk of falling was 1.73 (1.52 to 1.97). Muscle weakness Muscle weakness has been reported as a significant risk factor (Perell 2001). Our updated review did not identify any studies reporting statistical significance of this factor. One study conducted in extended care and one in community-dwelling setting reported non-significance in multivariate analysis (Bueno-Cavanillas 2000; Koski 1998). Discussion We have reported here risk factors that are associated with falling. These results were statistically significant in multivariable analyses. The evidence
suggests that although each factor can be a predictor of falls, in some population groups or settings some risk factors may be more important than others. This is illustrated by studies that have carried out multivariate analysis and reported non-significance for each factor. However, there was substantial heterogeneity between studies and within each risk factor. Many different methods of measurement of risk factors are reported and no one study replicates another. An important example of this is those studies examining gait, balance and mobility problems. There is substantial overlap between each studys definition of each domain and method of measurement. The possible synergism between different risk factors should also be considered. GDG commentary Assessment of older people with a history of falling and the presence of other risk factors should be undertaken. The identification of older people at risk will enable practitioners to refer older people for effective interventions targeted at specific factors. Multifactorial assessment is an important process but must be linked to interventions. The grading of this recommendation reflects both the evidence on risk factors and level I evidence of assessment linked to intervention(s). This multifactorial assessment should be done in the context of a comprehensive geriatric assessment where indicated.
1.2.3 Multifactorial interventions (please see Sections 3.3.5, 3.3.6, 3.3.9, 33.11 for evidence review methods)
1.2.3.1 Recommendation All older people with recurrent falls, or assessed as being at increased risk of falling, should be considered for an individualised multifactorial intervention. [A] In successful multifactorial intervention programmes the following specific components are common (against a background of the general diagnosis and management of causes and recognised risk factors): [A] strength and balance training
home hazard assessment and intervention vision assessment and referral medication review with modification/withdrawal. 1.2.3.2 Recommendation Following treatment for an injurious fall, older people should be offered a multidisciplinary assessment to identify and address future risk and individualised intervention, aimed at promoting independence and improving physical and psychological function. [A]
Level of evidence Level I Evidence statements Multidisciplinary, multifactorial, tailored interventions are effective in reducing falls in the following population groups and settings: - community-dwelling older people - older people in extended care settings - older people presenting at A & E following a fall Three trials suggest that multifactorial, multidisciplinary rehabilitation programmes are effective in reducing the incidence of further falling in older people who have suffered an injurious fall. This evidence is supported by key documents, in particular the expected standards of care outlined in the NSF for older people (standard six).
Level I
Evidence summary multifactorial interventions Community-dwelling, unselected (fallers and non-fallers in the population studied) Of the eight studies that evaluated a multifactorial screening and intervention programme in community-dwelling older people, who were recruited on the grounds of age and domestic circumstances, without a requirement for the presence of known risk factors, data were pooled from four (Fabacher 1994; Jitapunkul 1998; Newbury 2001; Wagner 1994) involving 1,651 participants. The pooled data are homogeneous and show that the interventions are effective in reducing the proportion of fallers in the intervention group (pooled RR 0.73, 95%CI 0.63 to 0.85). Data were not pooled from the other four studies in this category. In Carpenter (1990) (539 participants), which was cluster randomised by household, the
intervention involved an assessment by trained lay volunteers using a disability rating scale; an increase in disability score at a repeat visit was reported to the family medical practitioner. Only the total number of falls in each group in the month before the final interview was reported. The trialists reported significantly fewer falls in the experimental group during that period, but insufficient data were available to calculate an effect size. The fourth of the incremental interventions in Steinberg (2000) also cluster randomised, had a medical screen, home hazard assessment, and exercise. There was no significant difference in the incidence of falling between this group (59 participants) and the control group (63 participants) who received an information package alone. Van Rossum (580 participants) found no difference in the incidence of falls between the intervention and control groups,but no data were provided.Vetter (1992) (674 participants) was cluster randomised (by household). There were 95 of 350 fallers in the intervention group and 65 of 324 in the control group. Community-dwelling, targeted (population studied are known fallers or have identified risk factors prior to enrolment) Data from two studies in this category were not pooled as cluster randomisation was employed. Coleman (1999) (169 participants) reported that screening and intervention in a chronic care clinic provided no significant improvement in the incidence of falls at 12 or 24 months. Tinetti (1994) (301 participants) reported a significant reduction in the number of fallers in the intervention group, adjusting for age, sex, previous falls, and number of risk factors (adjusted incidence rate ratio 0.69, 95%CI 0.52 to 0.90). Data were pooled from the other five studies (Close 1999; Hogan 2001; Kingston 2001; Lightbody 2002; van Haastregt 2000). The pooled data show a significant reduction in the proportion of fallers in the intervention groups (pooled RR 0.86, 95%CI 0.76 to 0.98). Exercise, visual correction, and home safety intervention Day (2002), in a study of factorial design, examined the effect of exercise, visual correction and a home safety intervention. The impact of these three interventions combined was a significant reduction in the number of participants falling (RR 0.76, 95%CI 0.61 to 0.94). Further analysis was
carried out for the data for exercise plus vision correction (RR 0.76, 95%CI 0.62 to 0.95), and for exercise plus home hazard management (RR 0.84, 95%CI 0.69 to 1.03). These analyses are somewhat less favourable than the adjusted analyses presented by the authors in their original report. Extended care In Jensen (2002), a cluster randomised trial of an 11-week multidisciplinary programme, including general and resident-specific tailored strategies, reported a reduced incidence of falls in the intervention group (adjusted incidence rate ratio 0.60, 95%CI 0.50 to 0.73). McMurdo (2000) (133 participants), also a cluster randomised study in an institutional setting, reported no significant difference between intervention and control groups in the percentage of participants falling in the six-month period after completion of the intervention. Ray (1997) (482 participants) was also cluster randomised. Data were reported on recurrent falls and injurious falls. The reporting of the data provides insufficient detail to confirm whether the reduction in recurrent falls experienced in the intervention group was significant. Rubenstein (1990) (160 participants) found no benefit from nurse practitioner assessment and physician referral within seven days of a fall (RR 0.97, 95%CI 0.84 to 1.11). Vassallo (2001) evaluated a multidisciplinary fall assessment in a cluster randomised trial in a geriatric rehabilitation setting, and reported fewer fallers (39/275) in the intervention group, compared with 111/550 in the control group. Becker (2003), in a cluster randomised trial (N = 6) involving 981 long stay residents of community nursing homes, reported that the number of fallers was less in the intervention group (RR 0.75, 95%CI 0.57 to 0.98, trialists analysis). The incidence density rate of falls per 1,000 resident years was also reduced in the intervention group (RR 0.55, 95%CI 0.41 to 0.73, trialists analysis). Cognitively impaired (any residence) Shaw (2003), in a comparison of multifactorial assessment and intervention in 274 older people with cognitive impairment or dementia recruited from an A&E
department following a fall, could not confirm the effectiveness of this intervention (RR 0.92, 95%CI 0.81 to 1.05). There is a lack of evidence of effective interventions for this group of older people. Many trials specifically excluded older people with a cognitive impairment. Economic evidence Tinetti et al. (1994) and Rizzo et al. (1996) both report on the same study. Tinetti reported on the clinical effectiveness. Rizzo undertook the cost effectiveness analysis. This study reported that the intervention package was cost effective in the high risk individuals. The high cost of the intervention was offset against the treatment costs of the high risk individuals. However, in this study not many of the control group had costly hospital admissions and the data was skewed. They undertook sensitivity analysis. There still remained a number of individuals in the intervention group who required costly treatment. The overall effect of this was to reduce the expected benefit in the intervention group. The analysis presented in the cost effectiveness analyses chapter (see Section 3.3.11) assumes that the at risk population can be reliably identified. Clearly the specificity and the sensitivity of the assessment tools will impact upon the cost effectiveness of the interventions. The systematic review of assessment tools did not identify any information on the sensitivity and specificity of the existing assessment tools. In this context, there is even greater uncertainty about the true cost effectiveness of these interventions. The greater the ability of assessment tools to differentiate between those who are likely to fall without the intervention and the rest of the elderly population, the more cost effective the interventions will be. The figures presented in this chapter represent a best case, where the assessment is completely accurate. Nandy et al. (2004) report a high specificity (0.92) but a relatively low positive predictive value (0.57). Using this assessment tool, slightly more than 40 per cent of patients identified as being at high risk using this tool would not be expected to fall. This would have a significant upward impact upon the cost effectiveness results presented above. This evidence became available too late in the process for it to be incorporated directly in cost effectiveness modelling.
The existing evidence base for judging the cost effectiveness of these interventions is poor. If the at risk population can be identified, our analysis indicates that the multifactorial intervention is likely to be cost effective compared to conventional thresholds, although there is a large degree of uncertainty around the actual incremental cost effectiveness ratio. GDG commentary The evidence above suggests that multifactorial interventions targeted to risk factors are effective in reducing falls in older people. However, it is difficult to make a definite recommendation of the key effective components for specific settings and populations. It is sensible therefore to refer a patient for intervention(s) that target known risk factors. Illustrative examples of good practice were nominated from trials by Close (1999), Jensen (2002) and Tinetti (1994). Close (1999) identified older people living in the community who presented at A&E following a fall. The intervention included a detailed medical and occupational therapist assessment, with referral to relevant services for targeted interventions either by modification of risk factors where possible; referral to multidisciplinary team for further interventions; and drug medication review by the GP. Advice and education was given by the occupational therapist (OT) about safety in the home and modifications were made where appropriate. The OT supplied minor equipment or referral was made to social or hospital services as required. Jensen (2002) recruited older people from extended care settings who received assessment by a physician and physiotherapist. This assessment included a full clinical examination and medication review. Targeted interventions included staff education, environmental modifications, exercise, supply or repair of aids, medication review and hip protectors. Tinetti (1994) recruited older people living in the community with the presence of one of the following risk factors: postural hypotension; use of sedatives; use of at least four medications; impairment in arm or leg strength or range of motion, balance, and ability to move safely from chair to bed. Assessment
was conducted by a study nurse practitioner and physiotherapist. The intervention group was given either a combination of adjustment of their medications; or behavioural instructions and exercise programmes aimed at modifying their risk factors in the form of decision rules and intervention protocols for each risk factor. These trials provide an example of approaches to providing effective multifactorial interventions, but the fundamental element is to prescribe or refer for targeted interventions. Evidence summary: rehabilitation Two trials reported a significant reduction in the incidence of further falling in those who had received attention for a previous fall (Close 1999; Crotty 2002). The intervention in the trial by Tinetti (1999) did not show an effect on the risk of falling, but there was a significant reduction in the incidence of individuals hospitalised. The key components of these studies included medical, physiotherapy and occupational therapy assessments with follow-up interventions, medical assessment to identify primary cause of fall and other risk factors with intervention or referral as required. Interventions may involve individually tailored exercise programmes aimed at improving strength, gait, balance, transfers and stair climbing. Social care and support were also part of some programmes. The safety and efficacy of an exercise protocol designed to improve strength, mobility, and balance and to reduce subsequent falls in older patients with a history of injurious falls was examined in Hauer (2001). This RCT was a threemonth intervention trial, with an additional three-month follow-up in an outpatient geriatric rehabilitation unit. The participants included 57 female patients, above the age of 75 years, admitted to acute care or inpatient rehabilitation, with a history of recurrent or injurious falls, including patients with acute fall-related fracture. Fall incidence was reduced non-significantly by 25 per cent in the intervention group, compared with the control group (RR: 0.753 CI: 0.455-1.245).
The studies reporting significant results suggest that a multifactorial approach, including multidisciplinary assessment and targeted interventions, could have some impact on reducing the incidence of falling as part of a rehabilitation programme following a fall resulting in medical attention. It is less clear from this evidence of the impact of these complex interventions on other factors such as confidence;quality of life and acceptability as limited data were available. Perhaps there also needs to be consideration of the planned withdrawal of such programmes and the ability of these individuals to sustain the improvement shown. The evidence from geriatric hip fracture (GHFP) and early supported discharge (ESD) programmes suggest that they decrease the total length of hospital stay for older people who have suffered a hip fracture and inpatient treatment. In addition, these structured programmes of care achieve higher rates of return to previous residential status (Cameron et al. 2002). However, it is unclear what the effect these programmes have on reducing the incidence of a further fall. Furthermore, less is known about the impact on function, morbidity and quality of life for older people participating. Two trials suggest that a multidisciplinary, multifactorial approach to management of older people, who have suffered an injurious fall and who have received treatment in a primary care or acute care setting, is an effective intervention package. Important components include assessment and a targeted intervention(s), underpinned by detailed discharge planning. It is less clear which specific mechanisms of this multifactorial approach to rehabilitation are effective, but the fundamental key to success may be through comprehensive discharge planning. In addition, the overall aim of these programmes for older people should be to regain confidence and subsequently prevent further falling. However, practitioners need to assess the extent to which the older person is likely to cooperate with the intervention programme and the usefulness of the overall prevention strategies in the Cochrane review. GDG commentary
There was substantial overlap between secondary prevention interventions and rehabilitation strategies. It was not possible to review the evidence of the effect of these interventions on important rehabilitation outcomes such as improvement in function, mobility and psychosocial health as these outcomes were outside the scope of the guideline. In this guideline, rehabilitation is considered as part of the secondary prevention of falls, but users of the guideline need to be aware of the potential for improvement in outcomes other than falls prevention.
1.2.4 Strength and balance training (please see Sections 3.3.5, 3.3.6, 3.3.9, 3.3.11 for evidence review methods)
1.2.4.1 Recommendation Strength and balance training is recommended. Those most likely to benefit are older community-dwelling people with a history of recurrent falls and/or balance and gait deficit. A muscle strengthening and balance programme should be offered. This should be individually prescribed and monitored by an appropriately trained professional. [A]
Level of evidence Level I Evidence statement A programme of muscle strengthening and balance training, individually prescribed at home by a trained health care professional is effective in reducing falls (pooled results from three trials).
Evidence summary Exercise and/or physical therapy Community-dwelling: targeted interventions Pooled data from three studies from New Zealand, with a total of 566 participants (Campbell 1997; Campbell 1999; Robertson 2001a), using the same individually tailored programme of progressive muscle strengthening, balance retraining exercises and a walking plan, indicated that this intervention significantly reduced the number of individuals sustaining a fall over a one-year period (pooled RR 0.80, 95% CI 0.66 to 0.98). The number of people sustaining a fall resulting in injury was also significantly reduced
(pooled RR 0.67, 95%CI 0.51 to 0.89). Seventy-four per cent of participants in the control group and 69 per cent in the exercise group in Campbell (1997) continued for a second year.After two years,the rate of falls remained significantly lower in the exercise group (Campbell 1999a). The relative hazard for all falls in the exercise group was reported to be 0.69 (95%CI 0.47 to 0.97); the relative hazard for a fall resulting in a moderate or severe injury was 0.63 (95%CI 0.42 to 0.95). These three studies involved older participants, but the components of the successful intervention suggest that balance retraining may be an important component of successful exercise programmes. Economic evidence The two papers by Robertson et al. (2001a, 2001b) report on the trials of the same home-based exercise programme in different centres carried out by different health care professionals. The programme delivered by the practice nurse was less costly than that delivered by the physiotherapist. However, the study undertaken with the practice nurse did not reduce hospital costs overall between the control and intervention groups. The cost effectiveness analyses of exercise programmes for older people at risk of falling are reported. The exercise programme is likely to be cost effective but less cost effective than the multifactorial intervention. This said, exercise may produce other health benefits that have not been incorporated into the analysis presented in Section 3.3.11. However, in the absence of a sensitive and specific method for identifying those older people at high risk of falling, the cost effectiveness of exercise falls prevention strategies cannot be confirmed and any recommendation to implement such programmes should be treated with caution. GDG commentary The group agreed that strength and balance training should be administered by an appropriately trained professional. Although the evidence is relevant to community-dwelling older people with either a history of falls and/or a balance and gait deficit, this evidence could be generalised to other settings.At
present,individually prescribed exercise has been shown to be effective in falls prevention. Evidence of effectiveness of group exercise interventions is emerging and will be considered in the update of the guideline. In addition, the health benefits of exercise should be considered.
1.2.5 Exercise in extended care settings (please see Sections 3.3.5, 3.3.6, 3.3.11 for evidence review methods)
1.2.5.1 Recommendation Multifactorial interventions with an exercise component are recommended for older people in extended care settings who are at risk of falling. [A]
Level of evidence Level I Evidence statement The evidence suggests individually prescribed or group approached exercise interventions in extended care settings are not effective in reducing falls (pooled results from two trials and one single trial, non-significant). However, three trials report effectiveness of exercise as a component in multifactorial programmes.
Evidence summary Nowalk (2001) in a study in long-term care facilities - ranging from independent living to skilled nursing care - reported no significant difference in number of falls between a control group and two untargeted exercise groups (resistance endurance training or Tai Chi). Data were pooled from two studies. Donald (2000) studied the effect of a targeted physiotherapy programme in 54 patients in an elderly care rehabilitation ward. Mulrow (1994) studied elderly nursing home residents (194 participants), comparing a three times weekly exercise programme with a friendly visit of the same duration. The pooled data showed no evidence of effectiveness in this context (RR 1.02, 95%CI 0.74 to 1.41). Schnelle (2003) compared a low intensity functionally oriented exercise and incontinence care programme with usual care in 190 incontinent nursing home residents. There was a non-significant trend towards a reduction in the number of fallers in this study, which may have been underpowered (RR 0.62, 95%CI 0.37 to 1.06).
GDG commentary Whilst there is insufficient evidence to recommend exercise as a single intervention in extended care settings, multifactorial interventions in this setting with an exercise component have been shown to be effective. Please refer to page 66 for further details.
Level I
Evidence summary Evidence for the effectiveness of home hazard management in people with a history of falling is somewhat strengthened by new data from the updated Cochrane review. The association of domestic hazards with falls in the home has been controversial, despite its face validity (Clemson 1996; Gill 2000; McLean
1996; Northridge 1995; Parker 1996; Sattin 1998). However, six trials with a substantial home hazard modification component (Carter 1997; Cumming 1999; Day 2002; Hornbrook 1994; Nikolaus 2003; Pardessus 2002) have reported data that supports its effectiveness, particularly in those with a history of previous falls. Cumming (1999) cautioned that this effect is unlikely to be caused by home modifications alone since the reduction in falls was not confined to falls inside the home. This is true also of the reduction in the number of participants reporting two or more falls in Carter (1997), where falls in the yard/ garden associated with the dwelling were also eligible, and in the study reported by Stevens (2001). Hornbrook (1994) also used a complex intervention. While the evidence supports interventions designed to reduce home hazards, the exact mechanism of the effect remains unclear. Five studies evaluated home safety interventions alone (Cumming 1999; Day 2002; Nikolaus 2003; Pardessus 2002; Stevens 2001). Data for number of participants falling are available from four, (Cumming 1999; Day 2002; Nikolaus 2003; Pardessus 2002). Amongst those participants with a history of falling in the year prior to randomisation, there was a significant reduction in the number of participants sustaining two or more falls in the study period (RR 0.66,95%CI 0.54 to 0.81).An overall analysis including all participants, fallers and non-fallers prior to randomisation, showed a significant, but smaller, effect (RR 0.85, 95%CI 0.74 to 0.96). In those without a history of falls in the previous year (Cumming 1999) there was no evidence for the effectiveness of home hazard modification (RR 1.03, 95% CI 0.75 to 1.41). In Cumming (1999) the rate of falls away from home was reduced by a similar extent to the reduction in falls at home. Stevens (2001), in a population with mixed fall status, reported results of a cluster randomised study in which the individual household was the unit of randomisation. After one year there was no significant difference in the rate of falls (overall, and falls at home), the rate of fall injuries, or the proportion of fallers in the intervention group, compared with the control group. Economic evidence
In a well-conducted cost effectiveness analysis, Salkeld et al. (2000) recruited patients during hospital admission, a number of whom had a history of falls. The intervention was implemented by an experienced occupational therapist. There was little improvement in the falls in the intervention group as a whole, but there was a statistically significant reduction in the number of falls in those with a previous history. The cost effectiveness relates to the high risk groups of older people. Smith and Widiatmoko (1998) modelled the costs of fall with the costs of a home hazard intervention. Over the 10year period of the model, they demonstrated a cost saving of A$92 per person. However the various sources of the data used, and assumptions made, indicate that although useful, it is not necessarily a substitute for empirical evidence. GDG commentary It is clear from the evidence that providing a home hazard assessment with an intervention aimed at modification for older people with a history of falling is effective. It is not clear which component of this intervention has the most impact on preventing further falls. However, a combination of advice, education interventions aimed at increasing confidence, risk awareness and home modifications are effective. Cumming (1999) reported a significant reduction in two or more falls in older people with a history of falls. Assessment was carried out by an occupational therapist and recommendations for prevention supervised as necessary. This intervention not only reduced the incidence of falls within the home but also falls outside the home. There was debate about who should carry out home hazard assessments. The GDG acknowledged that in practice this may not always be carried out by a health care professional, but by a suitably trained member of the health care team. The personnel involved in assessment within the studies reviewed were trained health care professionals including a doctor, and occupational therapist (Pardessus 2002); nurses, physiotherapists, occupational therapists and social workers (Nikolaus 2003); occupational therapist (Cumming 1999); and a trained assessor (Day 2002).
1.2.7 Psychotropic medications (please see Sections 3.3.5, 3.3.6 for evidence review methods)
1.2.7.1 Recommendation Older people on psychotropic medications should have their medication reviewed, with specialist input if appropriate, and discontinued if possible to reduce their risk of falling. [B]
Level of evidence Level II Evidence statement One trial of older people above 65 years suggests that a psychotropic medication withdrawal programme, involving a gradual withdrawal of psychotropic medication over a 14week period, is effective in reducing the risk of falls.
Evidence summary Psychotropic drugs include neuroleptics, sedatives/hypnotics, antidepressants, and benzodiazepines. These can increase an older persons risk of falling, as can the use of multiple medications. Results of a systematic review and meta analysis to identify particular medications that may increase an older persons falls risk suggest that older people taking more than three to four medications were at risk of recurrent falls; and those taking psychotropic medications were also at risk of falling. Campbell (1999) reported the results of a study of factorial design, in which the interventions were an individually tailored exercise programme of progressive muscle strengthening and balance retraining; a walking plan (also used in Campbell 1997 and Robertson 2001a); and a placebo-controlled psychotropic medication withdrawal programme. This was gradual withdrawal of psychotropic medication over a 14-week period. Inclusion criteria included those above the age of 65 years who were currently taking benzodiazepine, any hypnotic, antidepressant or major tranquilliser. The analysis reported by the investigators, using a Cox proportional hazard regression model, showed that the overall risk of falls was lower for the medication withdrawal group (relative hazard 0.34, 95%CI 0.16 to 0.74). Economic evidence
One Australian and one US study (Andrews et al. 2001 and Coleman & Fox 2002) looking at the contribution of medication use were also assessed. These involved pharmacy reviews of medication, which may have resulted in falls. Neither of these studies provides strong economic evidence, but they highlight the importance of assessment following a fall. The costs detailed in the paper by Andrews (2001) show the relationship between medication and the outcomes for patients. The study by Coleman illustrates some potential cost savings in reviewing medications. GDG commentary In addition to the evidence for psychotropic medication review, polypharmacy was identified as a risk factor for falling and medication review should be part of a multifactorial assessment, as described in recommendation 1.1.2.1.
1.2.8 Cardiac pacing (please see Sections 3.3.5, 3.3.6 for evidence review methods)
1.2.8.1 Recommendation Cardiac pacing should be considered for older people with cardioinhibitory carotid sinus hypersensitivity who have experienced unexplained falls. [B]
Level of evidence Level II Evidence statement Cardiac pacing in fallers with cardioinhibitory carotid sinus hypersensitivity is effective in reducing falls and syncope (one trial).
Evidence summary Cardiac pacing in fallers with cardioinhibitory carotid sinus hypersensitivity (Kenny 2001) was associated with a statistically significant reduction in the number of participants who were not cognitively impaired, sustaining syncope (RR 0.48, 95%CI 0.32 to 0.73). In addition, the mean number of falls in 12 months in the intervention group was significantly reduced (WMD -5.2, 95%CI -1.0 to -9.4). GDG commentary This recommendation reflected the evidence for a standalone intervention for older people who have cardioinhibitory carotid sinus hypersensitivity. The
evidence is also reflected in recommendation 1.1.2.1, which indicates that the GDG considered it necessary that a cardiovascular assessment should be carried out as part of a multifactorial assessment, where appropriate.
1.2.9 Encouraging the participation of older people in falls prevention (please see Sections 3.3.7, 3.3.8 for evidence review methods)
1.2.9.1 Recommendation To promote the participation of older people in falls prevention programmes the following should be considered. [D] Health care professionals involved in the assessment and prevention of falls should discuss which changes a person is willing to make to prevent falls. Information should be relevant and available in languages other than English. Falls prevention programmes should also address potential barriers such as low self-efficacy and fear of falling, and encourage activity change, as negotiated with the participant. 1.2.9.2 Recommendation Practitioners who are involved in developing falls prevention programmes should ensure that such programmes are flexible enough to accommodate participants different needs and preferences, promoting the social value of such programmes. [D]
Evidence statement People may be reluctant to participate in falls prevention programmes if they have not previously exercised, do not perceive a risk of falling or perceived poor functional ability or have not been adequately consulted about what changes they are willing to make. Much of the current information provision on falls prevention programmes may alienate rather than encourage participation by sterotyping older people, not being available in languages other than English, not emphasizing that many falls may be preventable and not promoting the social value of falls prevention programmes.
Level IV
Evidence summary The review of the quantitative and qualitative evidence on older peoples views and experiences enabled the identification of factors that may promote the idea of falls prevention. Multiple barriers to participation in falls programmes were identified, the most significant of which are summarised in Table 16. Some studies indicate that much of the information on falls prevention alienates rather than encourages participation by stereotyping older people (Aminzedah & Edwards 1998; Ballinger & Payne 2000); and by not producing information in languages other than English (Aminzedah & Edwards 1998; Kong et al. 2002). Other information needs include giving special advice to older people about the benefits of physical activity and falls prevention and how to stay motivated in the face of multiple barriers (Commonwealth of Australia 2000; King 1995). Some studies also reported a mismatch between the strategies willingly accepted by older people for example, walking aids, home modification, low intensity exercise and those that are most effective (balance and strengthening training) (Commonwealth of Australia 2000; Health Education Board 1999; Stead et al. 1997; King 1998). Two studies pointed out that imposition of strategies thought most optimal by health professionals may alienate the target group (Simpson et al. 2003; Porter 1999) and that health professionals need to find out which characteristics people are willing to modify and what changes they are prepared to make (Porter 1999) before
suggesting strategies. This should be an ongoing process (Grossman et al. 2003). Some of the individual factors that were shown to increase participation in falls prevention programmes or specific components of these programmes were: high exercise self-efficacy, past exercise history and general good health and functional ability (Rejeski 1997; King 1995; Oman 1998; Resnick 2000).Aspects of the format of falls prevention programmes that appeared to improve participation and maintenance included: home-based, telephone supervised, peer role models, low intensity exercise for example, walking moderate frequency for example, two to three times per week and be perceived as relevant, beneficial and fun for the participants. The social aspects of falls prevention programmes are probably their strongest selling point (Health Education Board 1999; Kong et al. 2002), particularly to older people without a history of physical activity. Factors that appeared to be barriers to either initial participation or long-term maintenance of falls prevention programmes were mainly personal, rather than programme format issues. These included: low self-efficacy or lack of perceived ability to undertake components of the programme; fear of falling; fear of exertion; illness;denial or under-estimating personal risk of falling; embarrassment or increased inconvenience regarding use of assistive devices (Bruce 2003; King 1998; Yardley 2002). In addition, the economic systematic review identified two studies that used quality of life measures (SF36) to look at the impact of fear of falling. The paper by Cumming et al. (2000) showed a link between fear of falling, SF36 measures and the admission to a long-term care institution. The study by Suzuki et al. 2002 showed that those subjects who expressed a great deal of fear of falling had SF36 scores, reflecting their increased anxiety and depression. The most commonly occurring and consistent themes across all studies (observational and qualitative) were as follows: Preferred strategies
People may be reluctant to participate in falls prevention programmes that have an exercise-based component (including balance training), if they have not previously regularly exercised and in which the social value of participation is not promoted. This requires consideration in light of the Cochrane review findings that a) a programme of muscle strengthening and balance retraining, individually prescribed at home by a trained health professional and b) a 15week Tai Chi group exercise intervention are likely to be beneficial (Gillespie et al. 2003). Interventions not involving behaviour change, such as home modification and assistive aids, appear to be more readily accepted among potential participants. There was a fairly consistent finding across the reviewed studies that prevention programmes that were home-based, moderate or low intensity exercise with frequent professional contact were most acceptable and showed higher participation rates (Hillsdon 1995; King 1998; Oman 1998). Other single interventions reported as being beneficial in the Cochrane review (Gillespie et al. 2003) such as cardiac pacing and withdrawal of medicines similarly may be more acceptable to some people. Individual factors Although trials of multifactorial packages have reported beneficial results (Gillespie et al. 2003), in clinical practice there may need to be more emphasis on finding out what characteristics a person is willing to modify and what changes are they prepared to make at what stage in their lives. This somewhat concurs with the finding that individually tailored interventions delivered by a health professional are more effective than standard or group delivered programmes (Gillespie et al. 2003). There was also evidence that the following factors are associated with activity avoidance: increasing age, being female, increasing anticipation of loss of function (Yardley 2002), not facing up to the risk of falling, (Simpson 1995) lack of perceived ability (King 1998), fear of falling (Bruce 2003) and fear of exertion (Grossman et al. 2003). However, fall prevention programmes that address self-efficacy and encourage activity change may result in increased uptake of falls prevention programmes (Cheal 2001;
Resnick 2002). This suggests that consideration of these factors is important when devising falls prevention programmes to ensure practical and appealing interventions are developed. Barriers need to be addressed prior to participation in a falls prevention programme to ensure commitment to the strategies. Health promotion and information needs There is a need to inform and educate older people that many falls are preventable. Perceived relevance of falls prevention may be low until a fall has been experienced. The social value of falls prevention programmes, as well as the physical benefits, needs to be promoted to make them attractive to intended participants. Those from non-English speaking backgrounds may require targeted health promotion.
1.2.10 Education and information giving (please see Sections 3.3.8 for evidence review methods)
1.2.10.1 Recommendations All health care professionals dealing with patients known to be at risk of falling should develop and maintain basic professional competence in falls assessment and prevention. [D] 1.2.10.2 Recommendations Individuals at risk of falling, and their carers, should be offered information orally and in writing about: [D] what measures they can take to prevent further falls how to stay motivated if referred for falls prevention strategies that include exercise or strength and balancing components the preventable nature of some falls the physical and psychological benefits of modifying falls risk where they can seek further advice and assistance
how to cope if they have a fall, including how to summon help and how to avoid a long lie. Evidence summary See evidence summary above associated with encouraging participation in falls prevention programmes.
1.2.11 Interventions that cannot be recommended (please see Sections 3.3.5, 3.3.6 for evidence review methods)
1.2.11.1 Brisk walking
Level of evidence Level II Evidence statement There is no evidence that brisk walking reduces the risk of falling. One trial showed that an unsupervised brisk walking programme increased the risk of falling in post-menopausal women with an upper limb fracture in the previous year. However, there may be other health benefits of brisk walking by older people.
Evidence summary In one study (Ebrahim 1997), brisk walking in n=165 women with an upper limb fracture in the previous two years, reported RR 0.69,95% CI 0.12-4.03. This UK study included postmenopausal women identified from A&E and orthopaedic fracture clinic records, with a history of an upper limb fracture in the last two years. The intervention group received initial advice on general health/diet and then encouraged to build up to brisk walking 40 minutes, three times per week. The control group received initial advice on general health/diet and encouraged to perform upper limb exercises to improve postfracture function. Falls events were greater in the intervention group. GDG commentary The group had reservations about this trial. It was a small trial with a specific group of older women. Although there was a significant increase of falls (I=52/81 vs. C=50/84 and fractures (I=2/81 vs. C=3/84) in the intervention group, the GDG recognise the limitations of the generalisability of these findings. For some other groups of older people, walking may have health benefits and should not be discouraged.
1.2.12 Interventions that cannot be recommended because of insufficient evidence (please see Sections 3.3.5, 3.3.6 for evidence review methods)
We do not recommend implementation of the following interventions at present. This is not because there is strong evidence against them, but because there is insufficient or conflicting evidence supporting them. 1.2.12.1 Low intensity exercise combined with incontinence programmes
Level of evidence Level I Evidence statement There is no evidence that low intensity exercise interventions, combined with continence promotion programmes, reduces the incidence of falls in older people in extended care settings (one trial, non-significant).
Evidence summary Schnelle (2003) compared a low intensity functionally oriented exercise and incontinence care programme with usual care in 190 incontinent nursing home residents. There was a non-significant trend towards a reduction in the number of fallers in this study, which may have been underpowered (RR 0.62, 95%CI 0.37 to 1.06). Economic evidence The study by Schnelle et al. (2003) made a number of assumptions that were not all reported in the paper. They acknowledge that this was an expensive and labour intensive intervention. They do not detail the costs but refer to them in the discussion. This intervention resulted in no significant difference between the control and intervention groups in the costs of assessing and treating acute conditions. The only statistically significant result was the stable fall rate in the intervention group. However, the authors recommend caution when interpreting these results, as this was a post hoc decision to analyse the data in this way. 1.2.12.2 Group exercise (untargeted)
Evidence statement Exercise groups should not be discouraged as a means of health promotion, but there is little evidence that exercise interventions that were not individually prescribed for community-dwelling older people are effective in falls prevention.
Evidence summary Community-dwelling: untargeted interventions: Using the FICSIT definition of falling, participants (n=200) exposed to the 15week Tai Chi intervention had a lower rate of falling than controls in one trial (risk ratio 0.51, 95%CI 0.36 to 0.73) (Wolf 1996). Local advertisements and direct contact recruited the participants in this study. Inclusion criteria included ambulatory older people, above the age of 70 years, living in unsupervised environments. Eleven studies, involving a total of 1,480 participants, reported the results of exercise interventions offered to groups of older community-dwelling people, where exercise interventions were not individually prescribed. Pooled data from nine studies (Buchner 1997; Cerny 1998; Cornillon 2002; Day 2002; Ebrahim 1997; Lord 1995; McMurdo 1997; Pereira 1998; Rubenstein 2000) does not confirm the effectiveness of untargeted exercise interventions in community-dwelling older people based on number of fallers (pooled RR 0.89, 95%CI 0.78 to 1.01). Data from Wolf (1996) were reported as adjusted estimates from a Cox proportional hazards analysis, and raw data to allow pooling were unavailable. Carter (2002), in a comparison of a twice-weekly exercise class with no intervention, reported no difference between groups in the number of people falling. Means (1996) recruited 65 participants, with a history of falling, who all underwent a six-week supervised low to moderate intensity programme designed to improve balance and mobility. Thirty-one participants practised on an obstacle course, in addition to the exercise intervention, while 34 did not. No statistically significant difference in the mean number of falls was reported. There were three complex intervention studies that included exercise. In a factorial design, Day (2002) compared group-based exercise, home hazard
modification and management of reduced vision. Although group based exercise alone was the most potent single intervention in this study RR 0.82 (0.70-0.97), falls were also reduced when exercise was combined with home hazard management, or reduced vision management, or both. The remaining two trials were cluster randomised; their data could not be pooled. One (Reinsch 1992) evaluated the effectiveness of classes teaching exercise, relaxation and health and safety topics relating to fall prevention, and classes without the exercise component. Results did not demonstrate a statistically significant reduction in number of fallers for either intervention. The other (Steinberg 2000), using a cumulative intervention in which three out of four groups received a monthly one-hour exercise class and encouragement to exercise between classes, reported that the intervention strategies could achieve an 18 to 40 per cent reduction in the incidence of falling, but the hazard ratios were not significant. Conclusion The evidence for effectiveness of group exercise interventions remains limited, apart from the Tai Chi intervention of Wolf (1996) and Day (2002). However, the three trials from New Zealand (Campbell 1997; Campbell 1999; Robertson 2001a), which used an individually tailored exercise programme of progressive muscle strengthening, balance retraining and a walking plan, demonstrated effectiveness. These three studies involved older participants, but the components of the successful intervention suggest that balance retraining may be an important component of successful exercise programmes. However, there is no evidence of clinical effectiveness of other exercise interventions that was untargeted to specific older people at risk of falling. GDG commentary The GDG recognises the emerging positive evidence for group exercise with two studies published beyond the date of the literature review underpinning these guidelines (Lord et al. 2003; Barnett et al. 2003). This new evidence will need to be included in the guideline update. In addition the global health benefit of exercise needs to be emphasised.
Economic evidence The study by Buchner et al. (1997) reported a relative risk for falls in the control group of 0.61. This study also measured quality of life using the SF36. They note that the hospital use between the two groups was very similar and the length of stay for the control group was likely to be longer resulting in additional costs. 1.2.12.3 Cognitive/behavioural interventions
Level of evidence Level I Evidence statement There is no evidence of effect that cognitive/ behavioural interventions alone reduce the incidence of falls in community-dwelling older people of unknown risk status (two single trials, non-significant). Such interventions have included risk assessment with feedback and counselling and individual education discussions. There is no evidence that complex interventions - in which group activities included education, behaviour modification programme aimed at modifying risk, advice and exercise are effective in falls prevention with community-dwelling older people(four single trials, non-significant).
Evidence summary Cognitive/behavioural therapy alone In Gallagher (1996) (100 participants), comparison of the two risk assessment interviews and a feedback/counselling interview, with a single baseline assessment interview, showed that the intervention had no statistically significant impact on the main outcome measures. In Ryan (1996) (45 participants), analysis of the number of fallers at three months showed no evidence that individual education sessions provided by a trained nurse were more effective than the one-hour group discussion of intrinsic and environmental risk factors. Complex interventions including cognitive/behavioural intervention Carter (1997) (658 participants) and Hornbrook (1994) (3182 participants) used a behavioural approach after carrying out an environmental safety assessment. Data have not been pooled from these studies, as Hornbrook (1994) is cluster randomised (by household). Both had cointerventions. Hornbrook (1994) included group sessions designed to modify risk taking behaviour and an exercise component, and reported survival analyses for
Falls: NICE clinical guideline 161 (June 2013) Page 173 of 315
sustaining any fall, injury fall, medical care fall, fracture fall, and fall causing hospitalisation. Unadjusted rates for all falls were significantly lower among intervention participants; for other categories of fall (injury falls, medical care falls) there were no statistically significant differences between groups. In Carter (1997) advice on optimising medication was given to the two intervention groups; a low intensity intervention in which advice alone was given on home safety, and a high intensity intervention that included professional formulation of an action plan. There was no evidence of a difference in the number of individuals falling between the control group and either intervention group. However, both interventions were associated with a significant reduction in the number sustaining two or more falls (low intensity intervention RR 0.27, 95%CI 0.08 to 0.95; high intensity intervention RR 0.22, 95%CI 0.05 to 0.98). In a cluster randomised trial, Reinsch (1992) evaluated the effectiveness of classes teaching exercise, relaxation and health and safety topics relating to fall prevention, and classes without the exercise component. The trial did not identify a statistically significant reduction in number of fallers. In another cluster randomised trial (Steinberg 2000), a cumulative intervention in which three out of four groups received encouragement to exercise and a monthly one-hour exercise class, the intervention strategies achieved an 18 to 40 per cent reduction in the incidence of falling, but the hazard ratios were not significant in any group. 1.2.12.4 Referral for correction of visual impairment
Level of evidence Level I Evidence statement Exercise in groups should not be discouraged as a means of health promotion, but there is little evidence that exercise interventions that were not individually prescribed for community-dwelling older people are effective in falls prevention.
Evidence summary In Day (2002) there was no evidence that referral for correction of vision in community-dwelling older people was effective in reducing the number of people falling (RR 0.88, 95%CI 0.54 to 1.43). This study, using a factorial design, compared a control group with groups receiving exercise, correction of visual impairment, and home hazard modification, each alone, and in
combination. Results above reflect analysis for the visual correction alone group. GDG commentary Whilst there is insufficient evidence that single interventions targeting vision impairment are effective in reducing falls, referral for visual correction as part of a multifactorial intervention has a significant impact on falls reduction. Identifying older people with visual impairment and referral for intervention should be considered within a multifactorial intervention. 1.2.12.5 Vitamin D and oral supplementation
Level of evidence Level I Evidence statement There is evidence that vitamin D deficiency and insufficiency are common amongst older people and that when present they impair muscle strength and possibly also neuromuscular function via CNS-mediated pathways. In addition, the use of combined calcium and vitamin D3 supplementation has been found to reduce fracture rates in older people in residential/ nursing homes and sheltered accommodation. Although there is emerging evidence that correction of vitamin D deficiency or insufficiency may reduce the propensity for falling, there is uncertainty about the relative contribution to fracture reduction via this mechanism (as against bone mass) and on the dose and route of administration required. No firm recommendation therefore can currently be made on its use for this indication.5
Evidence summary There is no evidence from one small trial involving 50 participants (GrayDonald 1995), for the effectiveness of a programme of oral nutritional supplementation in this case, a high energy, nutrient-dense supplement in preventing falls in a group of frail elderly women RR 0.10 (0.01 to 1.69). Five studies (Bischoff 2003; Dawson-Hughes 1997; Latham 2003; Pfeifer 2000; Sato 1999) evaluated the effect of vitamin D on falling. Data were pooled from Bischoff (2003); Pfeifer (2000) and Latham (2003) (461
5
The following text has been deleted from the 2004 guideline: Guidance on the use of vitamin D for fracture prevention will be contained in the forthcoming NICE clinical practice guideline on osteoporosis, which is currently under development. As yet there is no NICE guidance on the use of vitamin D for fracture prevention.
participants). In these studies both intervention and control groups received calcium supplementation; the intervention group in each received oral vitamin D supplementation. Within this group of pooled studies, no evidence was produced of the effectiveness of vitamin D supplementation in reducing the number of people who fall amongst community-dwelling or hospitalised older people (RR 0.87, 95%CI 0.70 to 1.08). In Pfeifer (2000), the reduction in the number of falls resulting in fracture was not statistically significant (RR 0.48, 95%CI 0.02 to 11.84). In Sato (1999) (86 participants), the administration of 1alpha-hydroxyvitamin D alone to people with Parkinsons disease (Hoehn and Yahr Stage <5) significantly reduced the number of fracture falls (RR 0.12, 95%CI 0.02 to 0.98), but did not reduce the mean number of falls in the intervention group (WMD 0.10, 95%CI -0.71 to 0.91). In a placebo-controlled trial of administration of vitamin D and calcium supplementation to community-dwelling men and women over 65 years, Dawson-Hughes (1997) (445 participants) reported that the number of participants falling did not differ significantly between intervention and control groups. Data were not presented. Vellas (1991) (95 participants) reported that administration of the vaso-active medication raubasine-dihydroergocristine to older people presenting to their medical practitioner with a history of a recent fall, significantly reduced the numbers of the intervention group who reported falls in the six months of therapy (RR 0.48, 95%CI 0.29 to 0.78). A recent published meta-analysis of vitamin D supplementation suggests there is a reduction in falls (Bischoff-Ferrari, 2004). There results showed that vitamin D supplementation appears to reduce the risk of falls among ambulatory or institutionalised individuals with stable health by 20 per cent. However, although there is emerging evidence that correction of vitamin D deficiency or insufficiency may reduce the propensity for falling, there is uncertainty about the relative contribution to fracture reduction via this mechanism (as against bone mass) and on the dose and route of
administration required. No firm recommendation therefore can currently be made on its use for this indication. Guidance on the use of vitamin D for fracture prevention will be contained in the forthcoming NICE clinical practice guideline on osteoporosis that is currently under development. 1.2.12.6 Hip protectors (please see Sections 3.3.10 and 3.3.11 for evidence review methods)
Level of evidence Level I Evidence statement Reported trials that have used individual patient randomisation have provided no evidence for the effectiveness of hip protectors for the prevention of hip fractures when offered to older people living in extended care settings or in their own homes. Data from cluster randomised trials provides some evidence that hip protectors are effective in the prevention of hip fractures in older people living in extended care settings who are considered at high risk.
Evidence Summary Incidence of hip fractures Data from the five cluster randomised studies were not pooled with data from the individually randomised studies. Cluster randomisation methods were used in five studies (Ekman 1997; Harada 2001; Kannus 2000; Lauritzen 1993; Meyer 2003). However, an uncorrected exploratory analysis of the five cluster randomised studies was conducted by the trialists. In Kannus (2000), the exploratory analysis (RR 0.34, 95% CI 0.19 to 0.61) that uses the raw numbers of participants sustaining fracture in each group differs slightly from that in the primary report (relative hazard 0.4, 95% CI 0.2 to 0.8), which used Cox proportional hazards analysis adjusted for age, sex, body mass index, mental status, ability to walk, previous falls and previous fractures. The cluster randomised trial by Ekman (1997) reports RR 0.34 (0.12-1.01) for the incidence of hip fractures, randomised by unit or nursing home. Harada (2001) reported the number of hip and other fractures, number of falls and compliance with hip protectors. Results for the incidence of hip fractures was RR 0.11 (0.01-0.84) and the incidence of other fractures RR 4.33 (0.2188.74).
In the trial by Lauritzen (1993), the incidence of hip fractures, randomised by unit or nursing home, was RR 0.44 (0.20-0.93) and the incidence of pelvic fractures was RR 0.34(0.02-7.01). The incidence of other fractures was RR 1.02 (0.55-1.89) in this trial. Meyer (2003) reported the number of hip fractures, and other fractures; falls; mortality; compliance of wearing the hip protectors and the reasons for noncompliance. The incidence of hip fractures, randomised by unit or nursing home, was RR 0.53 (0.32-0.87) and the incidence of other fractures RR 1.14 (0.74-1.78). Pooling of data from the seven trials in which randomisation was by individual showed no significant reduction in the incidence of hip fracture in those allocation to wearing hip pads (64/1306 (4.9%) versus 64/1086 (5.9%), RR 0.94, 95% CI 0.67 to 1.31). Pooling of data from five individually randomised trials conducted in nursing/residential care settings (1,426 participants) (Cameron 2001; Chan 2000; Jantti 1996; Hubacher 2001; Van Schoor 2003) showed no statistically significant reduction in hip fracture incidence (hip protectors 37/822 (4.5%), controls 40/604 (6.6%), RR 0.83, 95% CI 0.54 to 1.24). The reviewers note that by the end of the one-year observation period, nearly half (16/36 versus 17/36) of the individuals in Jantti (1996) had been lost to follow-up through death or permanent hospitalisation. Two individually randomised studies recruited community-dwelling older people (Birks 2003; Cameron 2003). These studies did not achieve a statistically significant reduction in the incidence of hip fractures (27/484 (5.6%) versus 24/482 (5.0%), RR 1.11, 95% CI 0.65 to 1.90). Villar (1998) studied compliance with wearing hip pads in a study with a follow-up period of 12 weeks. As this study excluded mentally incapacitated patients, participants were at lower risk of hip fracture. No hip fractures occurred in either the 101 participants allocated to receive protectors or the 40
participants in the control group. Thus this study contributed no data to the meta-analysis. Incidence of pubic ramus and other pelvic fractures There is insufficient evidence to confirm whether the use of hip protectors significantly reduces the incidence of pelvic fractures. Data on the incidence of pubic ramus and other pelvic fracture were available in 10 studies. In the six studies that used individual randomisation there were 16/1266 (1.3%) in the protector group and 13/1055 (1.2%) in the control group (RR 1.15, 95% CI 0.58 to 2.31). Incidence of other fractures/injuries The use of hip protectors appears to have no effect on the incidence of other fall associated fractures. Data on the incidence of other fractures that occurred over the study periods were reported in 10 studies. Pooling of results from the individual randomised studies showed that 63/1266 (5.0%) occurred in the protector group and 56/1055 (5.3%) in the control group (RR 1.06, 95% CI 0.75-1.50). Compliance Amongst those who were assigned to their use, compliance with wearing of hip protectors was limited. It is not clear in some trials how compliance was measured, but for those that stated the method of measurement, the length of time wearing them was calculated. Chan (2000) reported a compliance of 50.3 per cent, with dementia given as a reason for non-compliance. Ekman (1997) reported an average compliance of 44 per cent, although it is not clear how this was calculated. Harada (2001) reported that 17/88 (19 per cent) of those allocated to the protectors refused to wear them. Complete compliance estimated by hours worn was 70 per cent and partial compliance 17 per cent. Jantti (1996) stated that, of the 19 participants available at one year, 13 (68 per cent) were still using hip protectors. Of the subgroup of 45 individuals allocated to hip pads monitored in Lauritzen (1993), only 11 (24 per cent) wore the protectors regularly. In Kannus (2000), 31 per cent of those eligible declined to participate in the
study, while a further 71 of 446 patients discontinued use during the study. Compliance in those who agreed to participate in the study assessed as the number of days the protector was worn as a percentage of all available followup days was 48 per cent (29%, range <1 to 100%).Van Schoor (2003) used random visits to assess compliance and found that, at one month, 39 per cent were not compliant with wearing the protectors. This figure had risen to 55 per cent at six months and 63 per cent at one year. Hubacher (2001) reported that for 384 allocated to the protector group, 138 were regular wearers, 124 discontinued wearing them and 122 refused to wear them. Even the 138 regular wearers only wore the pads 49.1 per cent of the time. Birks (2003) gave an overall compliance figure of 34 per cent. Cameron (2001) stated total compliance was 57 per cent. At the end of the study only 37 per cent were still regular wearers of the protectors. Meyer (2003) reported that the hip protectors were worn by 34 per cent of the intervention group participants. Cameron (2003) approached 1,807 potential subjects living in their own homes and 34 per cent of these agreed to participate. By two years, the end of this study, only 33-38 per cent of participants were wearing the protectors all the time. In Villar (1998), of the 288 individuals approached only 141 consented to participate. Of the 101 who received the protectors only 27 (27 per cent) wore them throughout the 12week study period. In a breakdown of the reasons for non-compliance presented by Villar (1998), discomfort and poor fit were the most common reasons for discontinued use. Other evidence reporting compliance problems is also worth summarising, as these sources of evidence also confirm many of the Cochrane review findings reported above. A systematic review of the literature reported that the acceptance of, and compliance with, of hip protectors (Van Schoor 2002) ranged from 37 per cent to 72 per cent (median 68 per cent) for acceptance and 20 per cent and 92 per cent (median 56 per cent) for compliance. No details were given of specific settings or populations.
In a randomised controlled trial (Cameron 2000), the effect of hip protectors on fear of falling was examined in 131 women aged 75 and above who had two or more falls in the previous year. The results of this study report that hip protector users had greater improvement in falls self-efficacy at follow-up. In a prevalence study (Villar 1998), which aimed to assess compliance with the use of hip protectors in a residential setting, only 27 per cent wore the hip protectors for the full 12-week study period and half of the women wore them for less than one week. The reasons for non-compliance were poor fit or discomfort. Pakkari (1998) conducted a before and after study assessing the acceptability and compliance with hip protectors in 19 ambulatory residents in a nursing home. The small sample size for this study prevents generalisability, but results indicated that the tight fit of the hip protectors reduced the ability for independent toileting. Complications (including skin damage/breakdown) Ekman (1997) mentioned that the occurrence of skin irritation was used as a reason for non-compliance.Villar (1998) reported three individuals who were unable to tolerate the special undergarments during a heatwave and also mentioned discomfort as the prime reason for non-compliance. Kannus (2000) reported skin irritation or abrasion in 15 cases. In addition, one person reported the protector caused swelling of the legs and another that it caused bowel irritation. Hubacher (2001) reported that aches and pains and an uncomfortable feeling with wearing the protectors was given as a reason for noncompliance. Minor skin irritation was reported in Cameron (2001), and Cameron (2003) reported minor skin irritation or infection caused by hip protectors in 16 users (5 per cent). Meyer (2003) reported five cases of skin irritation. In addition some of the care homes reported increased dependency of some of the residents at toileting, more difficulty in dressing and discomfort from wearing the protectors. For the results of other outcomes measured in this review, see Evidence table 11 (Appendix E, 2004).
Summary The cluster randomised studies, which formed the bulk of the evidence from the previous review (2001), supported a significant beneficial effect of hip protectors in reducing the incidence of hip fracture (Parker et al. 2003). However, this significant protective effect was not confirmed by pooling of data from studies using individual randomisation in the updated version (Parker et al. 2003). For those living in their own homes, the review authors suggest there is insufficient evidence from randomised trials to support any benefit of hip protectors. The authors note that in a number of the cluster randomised studies, although allocation was by institution, analysis was by individual, without allowing for the effect of clustering. This leads to an estimation of the treatment effect in which the confidence intervals are inappropriately narrow. Thus there is a risk that a statistically significant effect appears to exist, when in fact it may not. This may have encouraged inappropriate interpretation of the strength of the evidence. The authors of the Cochrane review also noted other shortcomings such as evidence of heterogeneity amongst the populations studied in respect of baseline risk of fracture; that most of the individually randomised studies were underpowered; that the use of protectors appears to have varied between trials and within trials; and that initial acceptance of, and later compliance with, wearing the hip protectors were reported as problems in all of the studies. The reader is referred to the Cochrane report for full details. Finally, the studies included in the Cochrane review (Parker et al. 2003) and additional studies on hip protectors involve the use of a number of different designs of hip protector. It is not possible to be sure that the different types of hip protector used had equal effectiveness. A variety of different types of hip protectors have now been produced and clinical studies will be required to see if these new designs of protector are equally effective in reducing the risk of hip fracture. In addition, the compliance may vary for the different types of hip protector. Economic evidence
Two studies were identified as being relevant to the use of hip protectors. The first paper by Kumar and Parker (2000) looked at the cost effectiveness of hip protectors using the audit data from an English hospital and the Cochrane review of musculoskeletal injuries (Parker et al. 2001). The intervention was the wearing of hip protectors and the control was no intervention. The outcome measure was the number of hip fractures prevented.As the cost and benefit period was calculated over one year, discounting was not necessary. Direct costs only were used in the analysis and the number of protectors needed per person was obtained from previous studies and communication with the authors. The cost per item was obtained from the manufacturer. The authors use a previously published paper to estimate the average cost of a hip fracture the data updated to their cost year (1998). The cost results showed that the three hip protectors required for each person cost 113 per year. The average cost of treating a hip fracture was 7,200. The results were presented by age group. The cost of fracture prevented in the 50 to 59 age group was 508,500. The cost per fracture prevented in the above 85 age group was 2,485. The authors conclude that the use of hip protectors in the above 85 age group appears to be cost effective. However, there are a number of assumptions made in this study that may influence the results shown. The costs were calculated for those people who complied. They did not cost the supply of protectors to people who did not comply. They report a compliance rate of 36 per cent, which suggests that there is a problem. In addition, no sensitivity analysis was carried out on the price of the protectors. No indirect costs for hip fracture were included. The results of this study should be treated with caution. The second study by Segui-Gomez, Keuffel and Frick (2002) was a state transition model. This models the movement of patients through the probability of sustaining a fall resulting in a hip fracture, not falling or dying from any cause. That is to say the patient is in one of three states: well, hip fracture or dead. Data for models are obtained from published literature, epidemiological data, quality of life data or utility data. The data driving the model was obtained from published literature of trials. The authors state that
they made some assumptions concerning the effectiveness of the protectors, which is normal when modelling. However, these assumptions need to be explicit in order to give validity to the model. This model was populated by two hypothetical groups of 500,000 65-year-old men and women in the USA. The model was run for 35 years. As with other studies it is difficult to generalise between health care systems. However, they did include a cost utility analysis. They obtained QALY data from expert opinion (a sample of gerontologists) and a sample of older people using a VAS scale. The authors do not give information about the sample other than it being one of convenience. There have been recent concerns about the use of VAS scales in deriving QALY data and this does raise some questions about their results (Brazier et al. 2003). The authors showed that hip protectors are cost effective in the above 85 age group. The QALY data they collected showed that women gained QALYs overall, but with men there was a decrement. This is attributed to the inconvenience for men of wearing the protector. There is considerable uncertainty about some of the sensitivity analysis. Compliance is an issue, as the authors state that hip protectors only result in cost savings when compliance is 70 per cent. The literature illustrates that there are problems with compliance and achieving 70 per cent would be difficult. There are methodological questions with this model that make it difficult to use the results to inform practice. Both of these studies have no intervention that is doing nothing is the comparator. It is likely that this may not be the case in some areas where prescribed vitamin D and calcium or bisphosphonates may occur as part of a fracture prevention programme. In view of recent effectiveness data, which show fewer benefits than previously anticipated, these two flawed cost effectiveness studies demand that their results be treated with caution.
GDG commentary The GDG acknowledged that the evidence is less convincing of the effectiveness of hip protectors in the prevention of falls, following the update of the Cochrane systematic review on hip protectors. There was discussion about the benefit of hip protectors for high risk groups of older people. Older people at high risk might include those with the presence of multiple risk factors. However, the GDG felt that it was not possible, on the basis of the current clinical effectiveness evidence, to make a potentially expensive recommendation about their use until there are trials evaluating the newer types of hip protectors and national standards for their manufacture and safety are made.
3.5
The following research gaps were identified by the GDG. Following NICE requirements, the first five are those prioritised by the GDG. Further analysis of existing trial data to identify which components of multifactorial interventions are important in different settings and amongst different patient groups. Future trials designed and analysed with the intention of identifying cost effective components of multifactorial programmes for particular groups of older people in different settings. Evaluation of multi-agency falls prevention programmes to measure the impact of these programmes on reducing falls, injurious falls and fractures in older people. Falls prevention trials with a focus on injury reduction, such as fracture outcomes and fall related outcomes. Research on the optimal methods of risk assessment for falls in older people and evaluation of whether fall-prone individuals can be risk stratified, in terms of whom will most benefit from assessment and intervention. Trials investigating the most effective strategy for preventing falls in older people with cognitive impairment and dementia. UK-based cost effectiveness studies of falls prevention interventions. Trials to investigate the effectiveness of hip protectors compared with other fracture prevention interventions in older people at high risk of falling.
3.6
Validation
Section 3 of the guideline was validated through two stakeholder consultation processes. The first and second drafts were submitted to NICE in January and April 2004. They obtained and collated stakeholders comments, which were considered by the GDG.
4.1
Background
A scheduled review of NICE clinical guideline 21 identified that older people who are admitted to hospital should be included in the scope of that guideline to make the guideline more comprehensive, because effective interventions for falls prevention cannot simply be transferred to inpatient settings. Thus the Internal Clinical Guidelines Programme at NICE was commissioned to develop new recommendations on the assessment and prevention of falls in older people during a hospital stay (inpatients). Clinical need Falls in hospitals are the most common patient safety incidents reported in hospital trusts in England, and treating inpatient falls alone costs the NHS more than 15 million per year (National Patient Safety Agency 2007). Patients in hospital have a greater risk of falling than people who are not in hospital. This is partly because newly acquired risk factors (such as acute illness, delirium, cardiovascular disease, impaired mobility, medication and syncope) and unfamiliar surroundings can increase the risk of falling.
4.2
See appendix J (2013) for details of how this part of the guideline was developed.
4.3
4.3.1
Review question
What risk prediction tool(s) or process(es) should be used to identify modifiable and non-modifiable risk factors for falling for patients in hospital? Does this method vary by inpatient setting?
4.3.2
Evidence review
This question focused on establishing which, if any, risk prediction tools should be used to predict which patients are at risk of falling in hospital and whether the risk prediction tool should vary depending on the inpatient setting. The Guideline Development Group (GDG) wanted to make it clear that they were specifically referring to risk prediction tools rather than assessment tools. This is because they felt that the term assessment is often used interchangeably to refer to predicting or evaluating risk and to refer to a multifactorial risk assessment. In the context of falls, the GDG defined risk prediction as a process applied to the entire inpatient population to predict which patients are at risk of falling during their hospital stay. The GDG defined multifactorial assessment as a more in-depth and possibly on-going process of identifying falls risk factors that can be treated, managed or improved during the individual patients hospital stay, with the aim of reducing the patients risk of falling in hospital. The initial search retrieved 1552 studies, and another 20 studies were identified through other sources. Studies were included in the review if they related to predicting which patients are at risk of experiencing a fall during their hospital stay. Only studies using a cross-sectional, cohort, casecontrol or randomised controlled trial design, and that were available in full text, were eligible for inclusion. Studies that did not relate to inpatient risk prediction tools, or that related to predicting fracture risk, were excluded. In addition, the GDG felt that further exclusion criteria should be applied. Published literature has suggested that fall risk prediction tools can be considered to be of high predictive value when they demonstrate sensitivities and specificities above
70% (Oliver et al. 2004; Perell et al. 2001; Scott et al. 2007). The GDG agreed with this, and only risk prediction tools or processes that demonstrated sensitivity and specificity values above 70% in at least one study were included. If a risk prediction tool or process met the threshold in one study, all studies regarding this tool/process were included in the analysis, even if they didnt achieve the threshold. This was done to ensure completeness in the evidence review. For a full list of excluded papers for this review question, see appendix G (2013). Thirteen full text articles met the eligibility criteria and were included in the final review. The quality of the included studies was assessed using appropriate NICE methodology checklists, and the GDG categorised each study according to the setting in which it was conducted (acute; non-acute; mixed or unclear). GRADE (Grading of Recommendations Assessment, Development and Evaluation; see appendix J, 2013, for details of the methods used) was applied to the studies using an approach developed by (Schunemann et al. 2008). In this approach prospective studies are regarded as high quality, and studies based on retrospective data as moderate quality. Any other design is considered low quality. Downgrading then takes place as appropriate with reference to study limitations, inconsistency, imprecision and indirectness as used in the standard GRADE approach, but with different criteria. Meta-analysis was not conducted for this question, as the studies related to different risk prediction tools and it was not appropriate to combine the results. The summary table and the GRADE tables below present information for risk prediction tools or processes which demonstrated sensitivity values greater than 70% together with specificity values greater than 70% in at least one study. Appendix E (2013) contains the full summary and GRADE tables for the included studies.
Clinical Judgement
Clinical Judgement
NA
54
29
Clinical assessment (using Timed up and go, St Thomas's risk assessment tool in falling elderly inpatients, Barthel index, and sensory data) Clinical judgement
Unclear
n=226 Aged care and rehabilitation wards. n=136 Mean age=80 years. Geriatricpsychiatric inpatients.
NA
Fall Risk Assessment in GeriatricPsychiatric Inpatients to Lower Events (FRAGILE) Falls Assessment Questionnaire plus behavioural impulsivity measure. Clinical Judgement
0.05
n=32 Participants were all Males (mean age=62 years, range=47-74) who were non-ambulatory and had sustained Right hemisphere stroke n=200 Mean age=81 years Rehabilitation hospital
0.55
Population n=130 Mean age=75 years (range 29-97). Acute medical and surgical wards.
Threshold 13
See Appendix E (2013) for summary tables for all included studies. Table 2 GRADE table summary for inpatient risk prediction 2a) Acute setting
Studies N Index TP FP FN TN Sens (95% CI) Spec (95% CI) Quality
Hendrich Fall Risk Model Threshold: 3 1 Hendrich et al. (1995) Score 10 1 Walsh et al. (2010) Score 13 1 Walsh et al. (2010) TP= True positive FP= False positive FN= False negative TN= True negative Sens= Sensitivity Spec= Specificity V LOW= Very low 130 WHeFRA score 13 6 10 1 113 86 (42100) 92 (86-96) V LOW 130 WHeFRA score 10 6 28 1 95 86 (42100) 77 (69-84) V LOW 338 HFRM 3 79 67 23 169 77 (68-85) 72 (65-77) V LOW
Berg Balance Scale Score 29 1 Maeda et al. (2009) 72 Berg Balance Scale score 29 22 10 5 35 82 (65-98) 78 (65-91) LOW
Falls Assessment Questionnaire Score >0.49 1 Rapport et al. (1993) Score >0.55 1 Rapport et al. (1993) FAQ plus behavioural impulsivity measure. Risk >0.55 82 (57-96) 32 Risk >0.49 15 7 0 10 100 (78100) 59 (33-82) LOW
32
12
14
80 (52-92)
LOW
Clinical Observation/Assessment Observation of wandering behaviour 1 Vassallo et al. (2008) 1 Eagle et al. (1999) 200 Observation of wandering behaviours Clinical judgement 22 14 29 135 43 (29-58) 76 (56-90) 91 (85-95) 49 (37-62) LOW
98
22
35
34
V LOW
Peter James Centre Falls Risk Assessment Tool (PJC-FRAT) Recommendation of an Alert Card 1 Haines et al. (2006) 122 316 Alert card Alert card 52 41 61 83 19 30 184 162 73 (61-83) 58 (45-68) 75 (69-80) 66 (60-71)
LOW
TP= True positive FP= False positive FN= False negative TN= True negative Sens= Sensitivity Spec= Specificity V LOW= Very low
Falls Risk Assessment in Geriatric-Psychiatric Inpatients to Lower Events (FRAGILE) Score0.05 1 Nanda et al. (2011) 225 FRAGILE 0.05 125 15 11 74 92 (86-96) 83 (74-90) LOW
Clinical Assessment/observation Clinical Judgement 1 Myers and Nikoletti (2003) 1 Haines et al. (2009) Clinical judgement 88 (73-97) 61 (54-67) 26 (20-33) 82 (80-85)
226
30
142
50
LOW
1123
Clinical judgement
125
161
81
756
LOW
Clinical assessment using TUG, STRATIFY and Barthel index 1 Marsholleck et al. (2009) Clinical assessment using TUG, STRATIFY and Barthel index 38 (20-59) 97 (92100)
110
10
16
82
LOW
Clinical assessment and sensory measurement data 1 Marscholleck et al. (2009) Clinical assessment and sensory measurement data 58 (37-77) 100 (96100)
110
15
11
84
LOW
Clinical risk factors Chu et al. (1999) 102 Clinical risk factors 25 5 26 46 49 (35-63) 90 (79-97) V LOW
Clinical risk factors and functional performance Chu et al. (1999) 102 Clinical and functional performance 43 12 8 39 84 (71-93) 76 (65-88) V LOW
Hendrich Falls Risk Model Score 3 Heinze et al.(2008) Score >11 Heinze et al. (2008) 560 HFRM 11 47 263 16 234 75 (62-85) 47 (43-52) LOW 560 HFRM 3 61 449 2 48 97 (89100) 10 (7-13) LOW
TP= True positive FP= False positive FN= False negative TN= True negative Sens= Sensitivity Spec= Specificity V LOW= Very low
4.3.3
Evidence statements
For details of how the evidence is graded, see The guidelines manual (2009). 4.3.3.1 Although the falls risk prediction tools and processes examined in this question demonstrated sensitivities and specificities above the desired threshold (>70%), none of the tools or processes were adequately replicated in the relevant setting, meaning their validity and reliability are uncertain. The quality of this evidence was low or very low.
4.3.4
No existing literature was found to support this question (see section 4.4.4). In the economic plan, it was proposed that this question would be addressed using the same health economic model as for the question relating to inpatient fall prevention interventions (see section 4.4.4). However, based on the clinical effectiveness evidence, the GDG made a recommendation that fall risk prediction tools or processes should not be used, so no health economic modelling was undertaken for this question.
4.3.5
Evidence to recommendations
The sensitivity and specificity of falls risk prediction tools and processes were the only outcome(s) considered in this review, because the GDG agreed that sensitivity and specificity are the most appropriate measures for prediction. Ideally, a perfect risk prediction tool would demonstrate 100% sensitivity (that is, all people who are at risk will test positive) and 100% specificity (that is, all people who are not at risk will test negative), but in reality it is not possible to predict falls with such a high degree of accuracy. The GDG discussed the minimum sensitivity and specificity values that it would tolerate. The GDG agreed that sensitivity should be as high as possible to ensure that as many people who are at risk of falling are identified as such. For specificity, lower values could be tolerated, since this would mean that patients would get additional falls prevention resources directed to them, which may seem advantageous. However, the GDG felt that directing falls prevention resources at patients who do not need them would come at a cost from both the patients perspective and a resource use perspective. For example, a patient who has been asked to call for assistance when going to the toilet to mitigate their false positive falls risk may feel that their independence is being impacted on, and the staff member accompanying the patient is not able to provide care to another patient with a true positive falls risk. Thus, the GDG felt that the sensitivity and specificity values are equally as important. The GDG took guidance from published literature and agreed that both values should be 70% or higher (Oliver et al. 2004; Perell et al. 2001; Scott et al. 2007). The evidence review presented fall risk prediction tools and processes in terms of their sensitivity and specificity. The majority of risk prediction tools and processes that the GDG reviewed did not achieve the agreed 70% threshold. Those that did were from single studies that had not been replicated, so their validity and reliability are uncertain. Thus the GDG acknowledged that there is currently no way of accurately predicting the risk of falling for an older inpatient. Furthermore, risk prediction tools that provide a falls risk score or rating were felt to simplify the complex issues surrounding falls and were thought to potentially lead healthcare professionals to intervene with patients in a non-targeted and ineffective way. The GDG agreed that a 'do not do' recommendation was needed to prevent healthcare professionals from using and relying on the results of fall risk prediction tools. Instead, the GDG felt that all inpatients aged 65 and older, and all inpatients aged 50 to 64 years who are judged by a clinician to be at a higher risk of falling, should have their care managed as if they are at risk of falling. This is because inpatients will often have newly acquired risk factors (such as acute illness, delirium, cardiovascular disease, impaired mobility, medication or syncope syndrome) and are exposed to unfamiliar surroundings, which puts them at increased risk of falling during their inpatient stay. The GDG felt that a good routine clinical assessment would highlight a patients individual risk factors for falling
Other considerations
that could be treated, managed or improved, and this assessment should be linked to related interventions to address the identified risk factors. Discussions around this were continued in a discussion of the evidence relating to inpatient fall prevention interventions (see section 4.4), where the GDG considered further evidence to enable it to elaborate on what this process should involve. Health economics were not considered as a priority for this review question. No economic studied were found. Numerous fall risk prediction tools are available and used by healthcare professionals. However, most risk prediction tools were not validated for different data sets or study populations. From the systematic searches, 13 studies relating to 12 different risk prediction tools and processes were identified that met the inclusion criteria. These 13 studies were of low or very low quality and therefore needed cautious interpretation. The GDG wanted to emphasise that all older people who are admitted to hospital should be considered for an individualised multifactorial risk assessment for their inpatient falls risk, and people living in the community who have risk factors for falls (see recommendations 1.1.1.1 and 1.1.1.2) should be considered for an individualised multifactorial falls risk assessment for their community-based risk. Patients who have had a multifactorial risk assessment for their inpatient risk should still be considered for multifactorial falls risk assessment for their community-based risk, if appropriate (see recommendation 1.1.1.2). The GDG also wanted to emphasise that all people aged 65 or older who attend hospital for any reason should be considered for a multifactorial falls risk assessment for their community-based risk, in accordance with recommendation 1.1.1.1, regardless of whether they are admitted or not.
4.3.6
Recommendations Recommendation 1.2.1.1 Do not use fall risk prediction tools to predict inpatients risk of falling in hospital. Recommendation 1.2.1.2 Regard the following groups of inpatients as being at risk of falling in hospital and manage their care according to recommendations 1.2.2.1 to 1.2.3.2: all patients aged 65 years or older patients aged 50 to 64 years who are judged by a clinician to be at higher risk of falling because of an underlying condition.
4.4
4.4.1
Review question
What interventions reduce older patients' risk of falling and/or the severity of a fall in hospital, compared with usual care? Which interventions are the most effective? Does the intervention vary by inpatient setting?
4.4.2
Evidence review
This question focused on identifying the best interventions or strategies for reducing the risk and/or severity of a fall in hospital. The search strategy for this question faced difficulties in using the term falls as it is such a broad term. Using the term falls on its own would mean the
search would be too inclusive, as it would retrieve literature using the term fall in contexts that were not relevant to the research question, such as falls in glomerular filtration rates or falls in cancer rates. It would also have been difficult to comprehensively list all the possible terms for falls prevention interventions and capture them in the search strategy. Therefore after discussions with the GDG members it was decided to take the approach of combining the search terms for the falls concept with terms such as prevention, reduction, intervention and management for the intervention concept, instead of trying to list every intervention term that could be used to reduce inpatient falls. The performance of this strategy was also checked against the known relevant references to double-check the validity of this approach and it was found that most of the relevant references were being identified. It was hoped that searching in this way would capture the most relevant references needed to answer the review question. Using this approach 2880 studies were identified, and another 2 studies were identified through other sources. All studies were subject to screening, which led to the retrieval of 109 full texts which were considered in the review. Studies were included in the review if they related to interventions which aimed to reduce the risk of a patient falling during their inpatient stay. Only comparative studies such as randomised controlled trials, cohort, case control and before/after studies which were available in full text were eligible for inclusion. Studies that did not deliver the intervention in the inpatient setting, and studies that were non-comparative, were excluded. For a full list of excluded papers for this review question, see appendix G (2013). Thirty-seven full text articles met the eligibility criteria and were included in the final review. The quality of the 37 included studies was assessed using appropriate NICE methodology checklists, and the GDG categorised each study according to the setting in which it was conducted (acute; non-acute; mixed or unclear). There was a lack of clarity in the terminology used to describe the type of intervention by the included studies, and so the ProFaNE (Prevention of falls network Europe) classification of falls prevention interventions (Lamb et al.
2005) was adopted to assist with synthesising the evidence. The system categorises interventions as follows: Single. Only one major subdomain of the intervention is provided to the participants. Multiple. Two or more subdomains of the intervention are given to every participant of the falls prevention programme Multifactorial. Two or more subdomains of the intervention can be given to participants, but the interventions are linked to each individuals risk profile and, unlike multiple interventions, not all participants receive the same combination of subdomains. Using this system, some studies were placed in a different category to that described by the author. For example, the intervention in Haines et al. (2004) was described by the authors as a multi-intervention, but was classified as a multifactorial intervention for the purposes of this review, as participants received different combinations of two or more subdomains of the intervention. Incidence rate ratios (IRRs) were calculated whenever it was possible to identify the number of falls from the number of occupied bed days. Relative risks (RRs) were calculated when it was only possible to identify the number of patients who fell from the total number of patients included. In studies in which data were provided for intervention and control groups for preimplementation and post-implementation time points, a test of interaction (Altman and Bland 2003) was performed to compare the effect size for the pre-implementation period to the post-implementation period, giving a ratio of incidence rate ratio (RIRR) or ratio of relative risk (RRR). It was considered appropriate to combine IRR with RIRR, and RR with RRR, in subsequent meta-analysis. Meta-analysis was conducted for multifactorial interventions only. The GDG had made an a priori decision to split the evidence according to the setting in which it was conducted, and so separate meta-analyses were performed for each setting (acute; non-acute; mixed or unclear). Several limitations were
identified which could have an impact on the results of the meta-analyses. Firstly, multifactorial interventions are individually tailored to each patients falls risk assessment and it was clear that there would be some clinical heterogeneity in the evidence, as different patients in a particular study could receive different components of the multifactorial intervention. Similarly, different studies used different types of multifactorial intervention depending on the resources available and the falls risk assessment that was used. Thus, the true treatment effect may be different for different studies. Secondly, methodological heterogeneity was also present in the pool of evidence as different study designs were utilised. Different study designs suffer from different types and degrees of bias, which can lead to some studies over or under estimating intervention effects. Meta-analyses performed when there is substantial clinical and methodological heterogeneity run the risk of obscuring important differences and providing misleadingly precise results. Thus the limitations associated with preforming a meta-analysis on heterogeneous evidence were discussed by the GDG and the technical team and it was debated whether it was appropriate to perform meta-analysis at all. It was expected that clinical heterogeneity could not be significantly reduced by exploring the causes of it as there was limited available data on subgroups to perform the necessary analyses. The GDG considered methodological heterogeneity and whether data from non-randomised trials should be included in the meta-analysis or not, given that non-randomised trials are likely to over-estimate the intervention effect. The GDG agreed that only evidence from randomised studies should be used to drive recommendations, but it would be useful to examine the effects of non-randomised studies too. Thus, all study designs were included in the analysis, and the effects of randomised studies were examined separately from non-randomised studies by stratifying the analysis according to study design. This was done because exploring both randomised and non-randomised studies would provide the GDG with a comprehensive view of this complex and heterogeneous evidence base. The GDG acknowledged that progressing with a meta-analysis in this way would limit the results and may not be reflective of the actual effects that might be
achieved in future inpatient multifactorial falls prevention interventions. After considering these limitations it was finally agreed that for this review question a meta-analysis with significant heterogeneity and limitations would be better than no meta-analysis at all. Rather than performing a narrative review or using the meta-analysis to identify the true effect amongst the pool of studies, a random-effects model was used to estimate an average effect amongst this heterogeneous body of evidence. GRADE was used to appraise the quality of the evidence, and the overall quality of the evidence was low. The tables below are the summary table and the summary GRADE tables. Appendix E (2013) contains the full summary and full GRADE tables for the included studies.
Unclear
Multifactorial
1,200 mg calcium carbonate once daily Pre intervention audit 1,200 mg calcium carbonate once daily
3.7 days
43 days
(Capan and Lynch 2007) USA (Cumming et al. 2008) Australia (Donald et al. 2000) England
Unclear
7 days
Multifactorial
(Donoghue et al. 2005) Australia (Dykes et al. 2010) USA (Fonda et al. 2006) Australia (Giles et al. 2006)
Unclear
Pre intervention audit Wards receiving usual care Pre intervention audit Pre intervention
3 days
21 and 18 days
Multifactorial
Unclear
Companion observers
Age=unclear Geriatric wards n=626 Mean age=80 Rehabilitation and care of the elderly wards n=226 Mean age=82 Rehabilitation and care of the elderly wards (subgroup of Haines 2004) n=173 Mean age=81 Rehabilitation and care of the elderly wards (subgroup of Haines 2004) 18 hospitals Mean age=unclear 18 hospitals who had not had access to low-low beds n=1,206 Mean age=74 Acute and geriatric rehab/assessment units All admissions Mean age=81 Care of the elderly wards All admission Age=unclear All admitted patients to a medical centre n=unclear Age=unclear Acute care telemetry unit n=52 Mean Age=84 Long term care facility n=unclear Age=unclear Patients on unit with higher than the hospital average fall rate who had previously fallen in the hospital n=1122 Age=unclear Medical, surgical 29 and 30 days Multifactorial
Unclear
Patient education
Usual care
Unclear
Exercise
Usual care
Unclear
Low-Low beds
Usual care
Usual care
(Healey et al. 2004) England (Huda and Wise 1998) USA (Jeske et al. 2006) USA (Kato et al. 2008) Japan (Kilpack et al. 1991) USA
18 and 21 days
Multifactorial
Usual care
Unclear
Multifactorial
Unclear
Educational Poster
1 year
Multifactorial
Unclear
Multifactorial
>365 days
Multifactorial
and geriatric patients from two acute care hospitals (Krauss et al. 2008) USA n=unclear Mean age=65 General medicine floors of a tertiary care hospital n=292 Age range=21-99 Patients from medicalsurgical/critical care at a large community hospital n=2106 Age=unclear Geriatric inpatients n=71 Age >60 Medical inpatients with confusion and behavioural disturbances n=134 Mean age=70 Specialist physical rehabilitation hospital n=58 Age range=38-92 Medical ward All admissions Age=unclear All inpatients admitted to a short term care facility on a unit with the highest rate of falls n=34,972 Mean Age=67 All inpatients n=409 Mean age=71 Department of Internal Medicine n=199 Age >70 Patients with fractured neck of femur n= 10 hospital wards Age18 All patients admitted to internal medicine Unclear Multifactorial
Unclear
Multifactorial
Unclear
Multifactorial
Unclear
Unclear
Unclear
Multifactorial
Unclear
Multifactorial
(Schwendimann et al. 2006a) Switzerland (Schwendimann et al. 2006b) Switzerland (Stenvall et al. 2007) Sweden
12 days
Multifactorial
11 and 12 days
Multifactorial
28 and 38 days
>5 days
Usual care
(van Gaal 2011a) Netherlands (van Gaal 2011b) Netherlands (Vassallo et al. 2004) England
n=825 Mean age=86 Elderly care wards in a community rehabilitation hospital n=4272 Mean age=80 Geriatric clinic n=217 Mean age=80 Medical inpatients n=1357 Median age=79 Tertiary teaching hospital
21 and 27 days
Usual care
(von RentelnKruse and Krause 2007) Germany (Wald et al. 2011) USA (Williams et al. 2007) Australia
21 and 19 days
Multifactorial
3 days
7 days
Table 4 GRADE table summary for inpatient interventions 4a) Acute setting
No of studies Design Intervention Comparison Effect (95% CI) Quality Geriatric Consultation team compared with routine care Implementation of recommendations by staff (Mean difference) 1 Allen et al. (1986) Randomised trials 313/446 (70.4%) 102/377 (27.1%) MD=2.59 (2.17 to 3.19) MOD
Hospital Acute Care of the Elderly Service compared with Usual Care Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Wald et al. (2011) Randomised trials (4.8) (6.4) V LOW
Safe or Sorry patient safety programme Falls (Ratio of Incidence Rate Ratio) 1 van Gaal Randomised trials IRR=0.98 (0.32 to 2.96) LOW
Non-pharmacological behaviour management strategies compared with usual care Falls (Relative Risk- Number of inpatients who fell as a proportion of number of inpatients) 1 Mador (2004) Randomised trials 10/36 4/35 RR=2.43 (0.84 to 7.03) LOW
Companion observers in the rooms of high risk patients compared with no observers on the ward Falls in the intervention rooms and no intervention wards (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 2 Donoghue et al. (2005); Giles et al. (2006) Nonrandomised trials
111/8755 (12.68)
135/8770 (15.39)
V LOW
Falls in the intervention rooms only (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Donoghue et al. (2005) Nonrandomised trials 2/3455 (0.57) 10/3972 (2.52) IRR=0.22 (0.06 to 0.93) V LOW
Educational Poster for patients/relatives, compared with no educational poster Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Jeske et al. (2006) Nonrandomised trials (4.7) (4.4) V LOW
Multifactorial interventions, compared with no multifactorial interventions Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 NonHuda and Wise randomised trials (1998) 2 Cumming et al. Randomised (2008), Dykes trials et al. (2010), 1 Controlled (3.7) (5.4) V LOW
MOD
V LOW
Intervention Comparison
Quality
7 Brandis et al. (1999), Krauss et al. (2008), Lieu et al. (1997), Mitchell Nonand Jones randomised (1996), Rainville et al. trials (1984), Schwendimann et al. (2006b); Von RentelnKruse et al. (2007) Falls rate 1 Capan and Lynch (2007) 1 Kilpack et al. (1991) Nonrandomised trials Nonrandomised trials
V LOW
3.20
4.50
V LOW
4.4
4.7
V LOW
8.9
9.1
V LOW
Any Injury (Incidence Rate Ratio- Number of falls resulting in any injury as a proportion of occupied bed days) 1 Koh et al. (2009) Controlled pre/post RIRR= 0.64 (0.33 to1.27) V LOW
3 Brandis et al. (1999), NonSchwendimann randomised et al. (2006b), trials Von RentelnKruse et al. (2007)
V LOW
Any Injury (Relative Risk- Number of inpatients who fell and sustained any injury as a proportion of number of inpatients) 1 Dykes et al. (2010), Randomised trials 7/2755 9/2509 RR=0.71 (0.26 to 1.90) MOD
548/805
495/763
V LOW
Severe Injury (Incidence Rate Ratio- Number of falls resulting in severe injury as a proportion of occupied bed days) 3 Brandis et al. Non(1999), Randomised Schwendimann trials et al. (2006b), Von Renteln-
V LOW
Intervention Comparison
Quality
Severe Injury (Relative Risk- Number of inpatients who fell and sustained severe injury as a proportion of number of inpatients) 1 NonSchwendimann randomised et al. (2006a) trials 31/805 19/763 RR=1.55 (0.88 to 2.71) V LOW
Staff knowledge (Mean difference- Post intervention compared with pre intervention) 1 Krauss et al. (2008) Nonrandomised trials 90.7 71.3 MD=19 (16.70 to 21.73) V LOW
RR= Relative Risk IRR= Incidence rate ratio MD= Mean difference V LOW= Very low
Flooring- Carpet flooring compared with Vinyl flooring Falls (Relative Risk- Number of inpatients who fell as a proportion of number of inpatients) 1 Donald et al. (2000) Randomised trials 7/28 (25.0%) 1/26 (3.8%) RR=6.50 (0.86 to 49.30) V LOW
Physiotherapy- Enhanced (2x daily standard physiotherapy plus specific strengthening exercises) compared with Standard physiotherapy alone Falls (Relative Risk- Number of inpatients who fell as a proportion of number of inpatients) 1 Donald et al. (2000) Randomised trials 2/30 (6.7%) 6/24 (25.0%) RR=0.27 (0.06 to 1.20) V LOW
Education for patients (including 1:1 sessions) delivered in combination with another intervention, compared with no education Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 26/3190 (8.2) 48/3007 (16.0) IRR=0.51 (0.32 to 0.82) 4/1026 (3.9) Randomised trials 11/1964 (5.6) 15/1219 (12.3) 9/652 (13.8) 24/2201 (10.9) 24/805 (8.9) IRR=0.28 (0.09 to 0.86) IRR=0.51 (0.26 to 1.03) IRR=0.41 (0.22 to 0.78)
10
11
12
HIGH
13
Exercise (45 min 3x per week) compared with no exercise Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Haines et al. (2007) Randomised trials 26/2596 (10.0) 47/2215 (21.2) IRR=0.47 (0.29 to 0.76) HIGH
Bracelets worn by high risk patients, compared with no bracelet Falls (Relative Risk- Number of inpatients who fell as a proportion of number of inpatients) 1 Mayo et al. (1994) Randomised trials 27/65 (41.5%) 21/69 (30.4%) RR=1.36 (0.86 to 2.16) V LOW
Proactive MDT approach (Weekly assessment by all MDT members) compared with standard MDT approach Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 NonVassallo et al. randomised trials (2004) 72/5855 (12.3) 170/14791 (11.5) IRR=1.07 (0.81 to 1.41) V LOW
Multifactorial interventions compared with no multifactorial intervention Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 2 Randomised Cumming et al. trials (2008), Haines IRR=0.78 (0.60 to 1.01) MOD
Intervention Comparison
Quality
V LOW
Falls (Relative Risk- Number of inpatients who fell as a proportion of number of inpatients) 1 Barry et al. (2001) Nonrandomised trials 26/149 39/156 RR=0.70 (0.45 to 1.09) LOW
Any Injury (Incidence Rate Ratio- Number of falls resulting in any injury as a proportion of occupied bed days) 1 Haines et al. (2004) 1 Kato et al. (2008) Randomised trials Controlled pre/post IRR=0.71 (0.42 to 1.20) MOD
V LOW
Any Injury (Relative Risk- Number of inpatients who fell and sustained any injury as a proportion of number of inpatients) 1 Barry et al. (2001) Nonrandomised trials 4/149 27/156 RR=0.16 (0.05 to 0.43) LOW
Severe Injury (Incidence Rate Ratio- Number of falls resulting in severe injury as a proportion of occupied bed days) 1 Haines et al. (2004) Randomised trials 2/9356 2/9239 IRR=0.99 (0.14 to 7.01) LOW
Severe Injury (Relative Risk- Number of inpatients who fell and sustained severe injury as a proportion of number of inpatients) 1 Barry et al. (2001) Nonrandomised trials 0/149 8/156 RR=0.06 (0.01 to 1.06) LOW
RR=Relative Risk IRR=Incidence Rate Ratio RRR=Ratio of Relative Risk MOD= Moderate V LOW= Very low 10=Any participant recommended Education 11=Participants only recommended Education 12=Any participant recommended education with Mini Mental State Exam >23 13=Any participant recommended education with MMSE <23 (cognitively impaired)
7 8
LOW
Any injury (Incidence Rate Ratio- number of falls with any injury as a proportion of occupied bed days) 1 Haines et al. (2011) 32/9174 Randomised trials 22/2941 10/6234 25/8737 10/3465 15/5275 IRR=1.22 (0.72 to 2.06) IRR=2.59 (1.28 to 5.47) IRR=0.56 (0.25 to 1.26)
6 7 8
LOW
Severe injury (Incidence Rate Ratio- number of falls with severe injury as a proportion of occupied bed days) 1/9174 1 Haines et al. (2011) Randomised trials 1/2941 0/6234 2/8737 0/3465 2/5275 IRR=0.48 (0.04 to 5.25)
6 7
LOW
Education for patients (Model 2- written materials only) compared with no education Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Haines et al. (2011) 96/11149 Randomised trials 35/3695 61/7457 81/8737 35/3465 46/5275 IRR=0.92 (0.69 to 1.25) IRR=0.94 (0.59 to 1.50) IRR=0.94 (0.64 to 1.38)
6 7 8
LOW
Any injury (Incidence Rate Ratio- number of falls with any injury as a proportion of occupied bed days) 1 Haines et al. (2011) 40/11149 Randomised trials 15/3695 25/7457 25/8737 10/3465 16/5275 IRR=0.39 (0.27 to 0.57) IRR=1.04 (0.63 to 3.13) IRR=1.17 (0.62 to 2.24)
6 7 8
LOW
Severe injury (Incidence Rate Ratio- Number of falls with severe injury as a proportion of occupied bed days) 1 Haines et al. (2011) 2/11149 Randomised trials 1/3695 1/7457 2/8737 0/3465 2/5275 IRR=0.78 (0.11, 5.56)
6 7 8
LOW
Education (Model 1- including 1:1 session) compared with education (Model 2- written materials only) Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Haines et al. (2011) 70/9174 (7.36) 45/2941 25/6234 96/11149 (8.6) 35/3465 61/7457 IRR=0.89 (0.65 to 1.12) IRR=1.62 (1.04 to 2.51) IRR=0.49 (0.30 to 0.78)
6
Randomised trials
7 8
LOW
Any injury (Incidence Rate Ratio- Number of falls with any injury as a proportion of occupied bed days) 1 Haines et al. (2011) 32/9174 Randomised trials 22/2941 10/6234 40/11149 15/3695 25/7457 IRR=0.97 (0.61 to 1.55) IRR=1.84 (0.96 to 3.55) IRR=0.48 (0.23 to 1.00)
6 7 8
LOW
Severe injury (Incidence Rate Ratio- Number of falls with severe injury as a proportion of occupied bed days) 1 Randomised 1/9174 2/11149 IRR=0.60 (0.06 to 6.70)
6
LOW
Quality
Low-Low beds (1 for every 12 standard beds) compared with usual care Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Haines et al. (2010) Randomised trials 186/35441 (5.25) 114/30228 (3.77) RIRR=1.01 (0.74 to 1.37) LOW
Any injury (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Haines et al. (2010) Randomised trials 85/35411 (2.4) 51/30228 (1.69) RIRR=1.29 (0.80 to 2.07) V LOW
Severe injury (Incidence Rate Ratio- Number of falls with severe injury as a proportion of occupied bed days) 1 Haines et al. (2010) Randomised trials 3/35441 (0.08) 7/30228 (0.23) RIRR=2.06 (0.36 to 11.70) V LOW
Vitamin D plus calcium compared with calcium alone (Number of inpatients who fell as a proportion of number of inpatients) Falls (Relative Risk- Number of inpatients who fell as a proportion of number of inpatients) 1 Burleigh et al. Randomised (2007) trials
36/100
45/103
MOD
Severe injury (Relative Risk- Number of inpatients who fell and sustained severe injury as a proportion of number of inpatients) 1 Burleigh et al. Randomised (2007) trials 1/100 (1.0%) 3/103 (2.91%)
MOD
Adherence amongst all participants to drugs (Mean Difference) 1 Burleigh et al. Randomised (2007) trials Multifactorial Interventions Falls (Incidence Rate Ratio- Number of falls as a proportion of occupied bed days) 1 Williams et al (2007) Nonrandomised trials (8.0) (9.5) V LOW 89/100 (89%) 87/103 (87%)
MOD
2 Randomised Healey et al. (2004) Stenvall trials et al. (2007) 2 NonFonda et al. randomised (2006), Lane et trials al. (1997),
MOD
V LOW
Any Injury (Incidence Rate Ratio- Number of falls resulting in any injury as a proportion of occupied bed days)
Quality MOD
Severe Injury (Incidence Rate Ratio- Number of falls resulting in severe injury as a proportion of occupied bed days) 1 Fonda et al. (2006) NonRandomised trials 7/41013 (0.17) 27/37133 (0.73) IRR=0.23 (0.10 to 0.53) V LOW
RR=Relative Risk IRR=Incidence Rate Ratio RIRR=Ratio of Incidence Rate Ratio MD=Mean Difference MOD= Moderate V LOW= Very low 6=Analysis of all participants 7=Analysis of participants only with cognitive impairment 8 Analysis of participants who are cognitively intact
4.4.3
Evidence statements
For details of how the evidence is graded, see The guidelines manual (2009). Acute setting 4.4.3.1 For most of the single interventions included in the acute setting the evidence was low or very low in quality and there was uncertainty about whether the interventions actually reduced or increased fall rates. Moderate and low quality evidence from two single interventions (geriatric consultation team; companion observers) showed that these single interventions significantly reduced falls rates, but there was no further supporting evidence about these interventions. 4.4.3.2 For multifactorial interventions, high quality evidence from a subgroup of randomised trials indicated there was uncertainty about whether the multifactorial interventions reduced or increased falls. Additional very low quality evidence from non-randomised studies suggested that multifactorial interventions significantly
Falls: NICE clinical guideline 161 (June 2013) Page 214 of 315
reduced falls rates. The results from non-randomised studies need interpreting with caution, since these results may be spurious. Non-acute setting 4.4.3.3 Again, most of the single interventions included in the non-acute setting had uncertainty about whether the interventions actually reduced or increased fall rates. High quality evidence on two single interventions (education in combination with another intervention; exercise) showed that the interventions significantly reduced falls rates, but there was no further supporting evidence from different settings about these interventions. 4.4.3.4 For multifactorial interventions, the meta-analysis results provided moderate quality evidence showing that, for the subgroup of randomised trials, there was uncertainty about whether the multifactorial interventions reduced or increased falls. Additional very low quality evidence from non-randomised studies also showed that there was uncertainty about whether the multifactorial interventions reduced or increased falls. Mixed or unclear setting 4.4.3.5 Again, for single interventions included in the mixed or unclear setting there was uncertainty about whether the interventions actually reduced or increased fall rates. The only exception to this was low quality evidence from subgroup analysis of one study which showed that fall rates were lower for people who were cognitively intact and were offered written and video based educational materials with 1:1 follow up in comparison with those receiving materials only and the no intervention control. There was no further supporting evidence about these interventions. 4.4.3.6 For multifactorial interventions, the meta-analysis results provided high quality evidence showing that, for the subgroup of randomised trials, multifactorial interventions significantly reduced the incidence of falls. Additional very low quality evidence from non-randomised
studies showed that there was uncertainty about whether the multifactorial interventions reduced or increased falls.
4.4.4
This is a summary of the modelling carried out for this review question. See appendix K (2013) for full details of the modelling carried out for the guideline. A search of published health economic analyses addressing the question of interest yielded a total of 1432 unique citations. Only one study (Haines et al., 2009) analysed both the costs and outcomes of measures to prevent inpatient falls but this study did not report outcomes in terms of QALYs and was therefore excluded. In the absence of relevant published literature, an original health economic model was constructed. Decision problem The health economic analysis addressed one question from the guideline scope, based on GDG prioritisation: What is the cost effectiveness of multifactorial interventions to reduce the risk and/or severity of inpatient falls for patients aged 65+, compared with usual care (assumed to be no actions to prevent falls)? The GDG recommended that single interventions should not be used to reduce the risk and/or severity of inpatient falls. Therefore, only multifactorial interventions were considered for health economic modelling. De novo model: model structure and methods The model was constructed in Microsoft Visual Basic for Applications, using Microsoft Excel 2010 as a front-end in which parameters can be specified and results collected and analysed. Costs and benefits were discounted at 3.5% per annum each and all costs are based on 201011 financial year. Model structure The model used a discrete event (or individual patient) structure, capturing the costs and benefits associated with a series of events and discrete health states. A discrete event model was used to accurately model each patients
falls history, variable lifespan and differing time in each state (for more details, see appendix K [2013]). Figure 1 illustrates the model structure. A patient could be in one of five states in the model and a fall could occur in any state (apart from the dead state, the only absorbing state in the model). The model did not have a time cycle, but measured time continuously (in days).
Change in state?
Entry to Non Acute Acute Entry to Care Hospital Discharge home Fall Death
full-time care
Acute Hospitalisation Fall Death
dead
+ home help
The GDG requested that the hospital state was split into acute and non-acute settings because the clinical evidence was split in this manner, because they felt that costs and resource use may differ between settings and because doing so allowed modelling of a number of scenarios or settings. The settings modelled were: Acute: All patients start in the acute hospital setting but can be transferred to or admitted to acute or non-acute settings later in the model. Intervention is only ever applied in acute settings
Non-Acute:
All patients start in the non-acute hospital setting but can be transferred to or admitted to acute or non-acute settings later in the model. Intervention is only ever applied in non-acute settings
The GDG agreed that the care state could be a generic state that represented any residential care facility (nursing or otherwise). The GDG also agreed to the simplifying assumptions that patients could not return home from the care state and patients cannot be directly admitted to the non-acute hospital setting from care (see appendix K [2013] for more details). The GDG discussed the complex array of social care arrangements available but no evidence was found to support any further division of the care process, apart from the addition of home help to those who fell in the home state or who fell in hospital and then returned to the home state. Model transitions All patients began the model in the hospital state (acute or non-acute), with their age, underlying residence (home or care), sex (all based on hospital episode statistics [HES] data, NHS Information Centre for Health and Social Care, 2012), falls history (falls within last 12 months, Vass et al., 2013) and risk of death generated probabilistically. Times for each next possible event were generated based on these factors and the event with the soonest occurrence simulated, and the model updated. Patients were then modelled for the remainder of their lives. Some patients may only have their initial hospitalisation and no falls, whilst others may have repeated falls in different settings and many hospitalisations. Whenever patients were simulated to experience a fall, their chances of injury were calculated and sampled probabilistically. The probability of hospital admission from home or care was taken from the Health Survey for England 2000 (Department of Health, 2002) and was split by age, sex and underlying residence. The type of hospitalisation (acute or non-acute) was determined once a hospitalisation occurred, using an approximation calculated using NHS reference costs (Department of Health,
2011) and HES-online data for the rehabilitation specialty (specialty code 314, NHS Information Centre for Health and Social Care, 2012). Hospital discharge time was determined by the length of stay, which was calculated on admission. A patient whose underlying residence was home could be discharged to care. In the acute setting, the probability of discharge to care reflected the age and gender of the simulated patient (NHS Information Centre for Health and Social Care, 2011); in addition, the likelihood of discharge to care was increased by the incidence of inpatient falls (Vass et al., 2013). In the non-acute setting, this transition was dependent on the same age and gender rates, but a different odds ratio was applied to account for inpatient falls (Aditya et al., 2003). The probability of transfer from acute to non-acute hospital settings was modelled separately for patients whose original admission was from home and care states. Different underlying rates were taken from HES data (NHS Information Centre for Health and Social Care, 2011). Rates for admissions from both settings were adjusted to take account of the increased risk associated with inpatient falls (Vass et al., 2013). The only time when patients could transfer from non-acute to acute settings in the model was as a result of a serious injury fall in the non-acute setting. The likelihood of entering care from home was calculated from a variety of sources (NHS Information Centre for Health and Social Care, 2011, Darton et al., 2006, Wang et al., 2001). As risk is linked to age, the risk of entering care is recalculated on every fifth birthday. Patient mortality was calculated as predicted time of death, based on standard Office for National Statistics life tables for 2008-2010 (Office for National Statistics, 2011). Hazard ratios were applied for the increased risk of death when entering the care state (McCann et al., 2009) and for experiencing a serious fall (in any setting, see appendix K (2013) for further discussion, Goldacre et al. 2002).
Inpatient fall consequences: the causal multiplier A major challenge in modelling falls prevention interventions is accounting for the relationship between a fall and subsequent events. The GDG noted it would be misleading to assume that this relationship is directly causative; it will be confounded by a wide range of known and unknown patient characteristics. Consequently it is fallacious to assume that preventing an individual from falling will make the patient entirely immune from all consequences that are known to be associated with falls. On the other hand, preventing falls can be expected to result in some benefits in these areas. The unknown factor is the extent of these benefits. To account for this uncertainty in the model, potential falls were simulated at the same rate in both the control and intervention arms but, whereas all potential falls were assumed to occur in the control arm, some potential falls were randomly selected to be averted in the intervention arms (with the proportion of averted falls derived from the incidence rate ratio defining the effectiveness of the intervention). The averted falls were then associated with some of the consequences of falls; the proportion of those consequences assumed to occur became known as the causal multiplier. A causal multiplier of 0 is equivalent to assuming that none of the consequences of falling occur when a fall is averted (that is, all post-fall negative events are directly ascribable to the fall). Setting the causal multiplier to 1 is equivalent to assuming that all of the consequences of falling occur, even when a fall is averted (that is, averting falls has no benefit in attenuating subsequent disadvantages). In the base case, this multiplier was assumed to be 0.5, and the impact of varying this assumption was tested in sensitivity analysis. The causal multiplier is discussed in more detail in appendix K (2013) and was applied to the following consequences of having an inpatient fall: Increase in length of hospital stay observed in fallers Utility decrement seen in fallers Increase in mortality seen in those experiencing a hip fracture Days of home help received on discharge following a fall
Increase in probability of discharge from acute to non-acute hospital Increase in probability of discharge from hospital to full-time care Fall rates and severity Hospital fall rates The underlying rate of falls in hospital settings is taken from Healey et al. (2008), a comprehensive analysis of 12 months data from the National Patient Safety Agency (NPSA). A rate of 6.7 falls per 1000 bed days was calculated for patients aged 65 and over. On GDG advice, this underlying rate was varied by age (Healey et al., 2008), sex (Healey et al., 2008) and falls history (falls within last 12 months, Vass et al., 2013). The same Healey et al. (2008) paper also reported on the severity of hospital falls, categorising them into five categories (see table 5). These probabilities were applied to all simulated inpatient falls in both acute and non-acute hospital settings. Table 5: Inpatient fall severities (Healey et al., 2008) Fall severity Percent of Inpatient falls
No harm Low Moderate Severe Fatal 65% 31% 4% 1% 0.01%
The effectiveness of the modelled intervention was based on the metaanalysed RCT evidence of multifactorial falls interventions discussed above (see table 4a and forest plots in Appendix E [2013]). In the acute setting, an incidence rate ratio of 0.75 was applied, implying multifactorial interventions have the capacity to reduce fall rates by 25%. A similar rate (0.77) was used for the non-acute setting. These rates were the pooled means from the metaanalyses of RCTs only (2 studies in the acute setting, 2 studies in the nonacute setting); rates from meta-analyses with alternative study groupings were covered in the sensitivity analyses. Whilst these rates were very slightly different to the rates estimated when pooling all study designs, sensitivity
analysis demonstrated that model results were trivially different if those rates were used instead (see below). Home state fall rates The fall rate for people living at home was taken from the Health Survey for England 2005 (NHS Information Centre for Health and Social Care, 2007). This gives a fall rate of 0.46 per person per year (split by age and sex). An incidence rate ratio of 2.6 for people with a history of falling (within last 12 months) was derived from OLoughlin et al. (1993) and applied to the underlying rate. Fall severities by age, sex and underlying residence were taken from Watson et al. (2009). Whilst based in Australia and on a slightly younger population, this study is one of the few to report non-injurious falls as well as injurious falls. In Watson et al., underlying falls incidence across all older people of both sexes is estimated at 0.27 per person per year, of which 28% result in injury. In order to treat mortality from a fall as a separate category, the probability of mortality from a fall leading to an A&E attendance or hospitalisation from Scuffham et al. (2003) was applied to moderate and serious injury rates from Watson et al. (2009). Care state fall rates The rate and severity of falls in care homes were based on pooled data from 3 UK-based studies (OHalloran et al., 2004, Dyer et al., 2004, and McMurdo et al., 2000). These resulted in an overall fall rate of 3.9 falls per resident per year. An odds ratio was calculated from Delbaere et al. (2008) to reflect the increased risk of falling in those care home residents who had fallen within the past year. Delbaeres regression model found that age and sex were not significant predictors of falling, so care home fall rates were not adjusted for these factors. Fall severity probabilities were derived from the same sources as used for falls at home (see above).
Resource use and unit costs State resource use and costs The unit cost of a day in an acute hospital setting was based on the weighted average of all NHS reference cost elective and non-elective activity, including excess bed days (524.01). The unit cost of a day in a non-acute hospital setting was based on all inpatient rehabilitation categories of NHS reference costs (588.01, Department of Health, 2011). Length of acute hospital stay was based on data from Vass et al. (2013), because it was the only UK study available to us that provided details of length of stay relative to incidence and severity of falls. In this study, patients who did not experience a fall (n=1695) had an average length of stay of 11.6 days (95%CI: 11.36, 11.84); for people who had a non-injurious fall (n=93), the same figure was 21.0 days (95%CI: 19.70, 22.30) and, for people who fell and were injured (n=34), the figure was 24.4 (95%CI: 21.88, 26.92). Length of stay in non-acute hospital settings was based on HES data for those transferred to other NHS settings (NHS Information Centre for Health and Social Care, 2012). Length of stay in nonacute hospital settings was split by age only as an inspection of the data showed length of stay did not vary by gender. The length of each hospital stay was probabilistically sampled. When a fall occurred during the simulated patients admission, the length of stay was adjusted by sampling an additional period reflecting the difference between a stay without falls and one in which a fall (non-injurious or injurious, as appropriate) occurred. This difference, calculated from Vass et al.s data, was assumed to be the same for both acute and non-acute settings, in the absence of any source of data specific to the latter. Extended length of stay was calculated in exactly the same way for all averted falls, but was subject to the causal multiplier (see above). The change in length of stay was only applied to the first fall experienced in any given stay, as Vass et al. did not differentiate between single and repeat fallers. The cost of a day in care was calculated from standard sources (PSSRU, 2011). Following the approach taken in previous NICE guidelines on delirium (CG103, National Clinical Guidelines Centre, 2010) and hip fracture (CG124,
National Clinical Guidelines Centre, 2011) unit costs for different care home settings were weighted according to Netten et al. (1998). These combine to give a daily cost of being in care of 103.78 (see appendix K [2013] for details). Resource use for patients receiving home help, stratified by fall severity, was agreed by the GDG. A fall leading to minor injury received 7 days help, moderate injury 21 days help and severe injury 42 days help. The cost per day of receiving home help following a fall was estimated at 38.97 (PSSRU, 2011). Not all residential care and home help is NHS/PSS funded. As in previous delirium (CG103) and hip fracture (CG124) guidelines and the Department of Health Fracture Plan (Department of Health, 2009) an assumption that 60% of such care is NHS/PSS funded was used. Hospital fall resource use and costs Inpatient fall resource use and costs are based on costs previously calculated by the NPSA (NPSA, 2007), uplifted from 200506 prices to 201011 prices (PSSRU, 2011). No evidence was found to assign different resource or costs by hospital setting, so the GDG agreed to assume that inpatient falls incurred the same cost, regardless of hospital setting (see table 6). The GDG agreed to assume that fatal inpatient falls incurred no cost. Whilst this may not be true, the number of fatal inpatient falls that occurred in the model was so few that the costs will be negligible. Table 6: Inpatient fall treatment costs by fall severity (based on NPSA, 2007) Inpatient Fall Severity Cost (201011)
No harm Low Moderate Severe 47.41 76.63 371.21 2,290.94
Home and care fall resource use and costs Watson et al. (2009) assumed levels of health service use following each fall severity, these were combined with GDG advice and assumptions about further likely follow up and transport resource use (see table 7). The unit cost of home and care falls were based on standard sources (PSSRU, 2011) for each activity. As with fatal inpatient falls, due to the small numbers of fatal falls at home or in care the GDG agreed to assume that these incurred no cost. Table 7: Home and Care Fall Treatment Costs (PSSRU, 2011) Fall Resource use (unit cost) Total cost per fall severity Home Care Home Care
No harm Low Moderate No NHS/PSS resource use GP clinic attendance (36) A&E attendance (106) GP clinic follow up (36) A&E attendance(147) via ambulance (253), hospitalisation (simulated separately), outpatient follow up (T&O, 93) No NHS/PSS resource use GP home visit (121) A&E attendance (106) via ambulance (253) GP home visit follow up (121) A&E attendance(147) via ambulance (253), hospitalisation (simulated separately), outpatient follow up (T&O, 93) 0 36 142 0 121 480
Severe
492.89
492.89
Intervention resource use and costs As no economic evaluations of inpatient falls prevention programmes currently exist (Davis, 2011), there were no examples of costed fall prevention interventions. One GDG member agreed to provide unpublished resource use estimates from a published trial that was included in the guideline metaanalysis and therefore the intervention costs are based on Healey et al. (2004). All the percentages and staff time requirements are assumptions and were varied in the sensitivity analysis.
The intervention first conducts a multifactorial assessment of a patients risk factors. The assessment is assumed to be undertaken by a nurse and require 20 minutes staff time. It is assumed the proportion of patients receiving the assessment will vary by setting (30% in acute; 80% in non-acute). Depending on the outcome of the assessment, patients then receive the necessary components of the multifactorial intervention. Assumed proportions of patients receiving each intervention component and associated resource use (staff time and consumables) are shown in table 8.
Nurse
Eyesight ophthalmology referral Medication - extra BP checks if CVD drugs changed Bed height alteration and bedrail removal Blood pressure check - referral to medical staff if high/low
35
More complex referral, hence staff time needed Extra daily checks for 1 week
Assess and remove if necessary Referrals sticky label based, so no burden. No extra staff resource, just allowed prioritisation Referrals sticky label based, so no burden Also laboratory costs (estimated 1 per test, 2010-11 costs) Phone call to relatives. Also 100 pairs slippers purchased (4 each at 2001-02 costs, 5.35 at 2010-11 costs) 2x HCA 5min each
Unknown
Healthcare Assistant
Mobility physiotherapy referral Urine test - send sample for analysis Footwear check and advise relatives on replacements
80%
25%
10
10%
Patient position in ward - move close to nurses Call bell and hazard education (assumed grade) Doctor Optician Medication review Optician referral for glasses
10%
10
100%
60% Unknown
7 Unknown
No additional cost, ought to happen as part of routine practice Review 2min, explanation 5min Referral sticky label based, so no burden
The intervention cost is calculated as the cost per admitted patient. Given the different proportion of patients assessed, the acute cost per admitted patient is 7.83 and the non-acute cost per admitted patient is 21.81. Utilities State Utilities Utility values for people in the home state were taken from the standard source for UK population norms (Kind et al., 1999). Polynomial (quadratic) regression was used to allow more accurate estimates of utility values in older age groups (75+). All subsequent changes to utility were applied as decrements to this baseline. No published literature was found that detailed the utility decrement experienced as a result of being in hospital. Therefore, unpublished data from a falls prevention trial were used (Vass et al., 2013) to estimate utility decrements of 0.720 for men (that is, the quality of life of men in hospital is, on average, 72% of that experienced by men of the same age at home) and 0.714 for women. The GDG agreed that utility decrements were unlikely to differ between acute and non-acute hospital settings. No source of evidence was found that detailed the utility decrement experienced as a result of being in the care state. The GDG agreed to use an assumption of 0.8, which was a decrement compared with being at home, but less than that for being in hospital. Fall utilities Falls in any setting are assumed to have a detrimental impact on quality of life. A search for published studies containing utility values related to inpatient falls yielded a total of 3460 unique citations. 91 papers were retrieved at title and abstract search. However, none were found to meet the NICE reference case. Therefore, relative utility decrements for no, minor and moderate injury inpatient falls were taken from unpublished data (Vass et al., 2013). This study reports EQ-5D-based utility estimates of 0.52 (95%CI: 0.507, 0.533) for people who did not fall during their hospitalisation (n=1567), 0.49.
Falls: NICE clinical guideline 161 (June 2013) Page 228 of 315
(95%CI: 0.449, 0.531) for people who experienced a non-injurious fall during their hospitalisation (n=84) and 0.38 (95%CI:0.261, 0.499) for people who fell and were injured during their hospitalisation (n=26). As Vass et al.s data did not allow differentiation between different severities of injurious falls, the relative decrements for home/care injurious falls (see below) were applied to these data. Due to the small number of severe injuries sustained in Vass et al., utility decrements for inpatient falls resulting in severe injuries were assumed to be the same as those associated with a hip fracture (in any setting, see below). The inpatient utility decrement was assumed to last until discharge, at which point the patient reverted to the relevant injury severity decrement for the home/care states for the remainder of the specified decrement duration. Vass et al. did not differentiate between single and repeat fallers, so fall utility decrements could not be applied repeatedly. If more than one fall occurred whilst in hospital, the decrement associated with the most serious fall was applied. Once a patient was discharged, they reverted to the utility decrement associated with the same severity of fall in the home or care state. Falls resulting in serious injury (in any setting) were assumed to be similar in utility loss and duration to hip fractures; falls resulting in moderate injury were assumed to be similar in utility loss and duration to wrist fractures (Peasgood et al., 2009). In line with Iglesias et al. (2009), falls resulting in minor injury were assumed to have half the impact on utility of moderate falls. Home and care falls resulting in no injury were assumed to have no impact on utility (see table 9). All fall utility decrements were assumed to last 1 year, within the exception of falls resulting in severe injury which were spread over 5 years, in accordance with Peasgood et al. (2009). Fall utility decrements from multiple falls are applied multiplicatively in the home and care states. Finally, the causal multiplier was applied to utility decrements associated with averted falls in the inpatient setting.
Model Assumptions The health economic model of interventions to prevent inpatient falls relies on a number of assumptions. These assumptions tend to arise for two reasons either to reduce the model complexity or because no data point could be found in the evidence base. All the assumptions were discussed with and agreed by the GDG and are listed in appendix K (2013) the most important are considered in the discussion section. Where possible, a range of values for assumed inputs was tested in the sensitivity analyses. Sensitivity Analysis Sensitivity analysis was conducted to explore the various areas of uncertainty and their impact on the model, including one-way, two-way and threshold analyses (using point estimates of parameters only). Note that the use of a discrete event simulation model accounts for first order (patient level) uncertainty (OHagan et al., 2007), Given the large number of inputs, it would be of benefit to perform probabilistic sensitivity analysis (PSA) to explore the effects of parameter uncertainty of model outputs. However, at the time of reporting, this has not been feasible because of the additional computational burden introduced by the discrete event simulation approach. Model outputs results In order to verify the face validity of the health economic model of multifactorial interventions to prevent inpatient falls, various model outputs were checked and presented to the GDG. All results are taken from a model run of 200,000 patients through an acute setting.
Falls: NICE clinical guideline 161 (June 2013) Page 230 of 315
Patients were slightly more likely to be women (51.1%) than men, which was in line with the source data (HES, 51.5% women). The mean starting age of patients in the model was 77.2 years (range 65.0 to 109.3 years). Patients spent an average of 10.1 years in the model (range 0 to 43.5 years) and the average age at death was 87.3 years (range 65.2 to 111.0 years). 0.3% of patients started the model in full-time care (source data HES 0.2%) and another 23.1% of patients entered care at some point. Patients who started in or entered full-time care spent an average of 2.6 years in care. Throughout their modelled lifetimes, patients had an average of 3.9 hospitalisations (including the first, which everyone incurs). 20.1% had no more hospitalisations whilst 10.5% had 8 or more hospitalisations. The initial acute hospitalisation had a mean length of stay of 12.3 days (range 0.0 to 157.0 days). Subsequent non-acute hospitalisations had a mean length of stay of 20.4 days (range 0.0 to 244.7 days). Length of stay was higher for fallers and recurrent fallers than non-fallers.
The model produced an overall average inpatient fall rate of 7.2 falls per 1000 bed days. This is higher than the input rate (6.6 falls per 1000 bed days,
Healey et al. 2008), but this is to be expected, as the model is a cohort model in which simulated patients age and become subject to greater risks, whereas the source study is a cross-sectional snapshot of the same population. As evidence of this, it can be seen that the fall rate varied by age in a similar manner to the source data (Healey et al., 2008; see figure 2). The majority (93.1%) of patients had no falls and very few (1.0% of patients) fell more than once during their hospital stay. Six patients fell six times and two patients fell seven times during a single hospital episode. The severity of inpatient falls compared very closely with the source data (Healey et al. 2008), as this input was not related to age. Patients had a mean of 1.0 falls per year at home and 4.8 falls per year in care. Injuries were more likely to occur in care, as in the source data. Costutility results base-case analysis The health economic model to assess the cost effectiveness of a multifactorial inpatient fall prevention intervention (compared with no action) was run with 500,000 patients per arm. Results are tabulated in table 10. Table 10: Base-case cost and QALY results (all costs and QALYs discounted at 3.5% per annum, 500,000 patients per arm)
Outcome Lifetime costs Arm Control Intervention Difference Lifetime QALYs Control Intervention Difference Cost per QALY (ICER) Incremental net monetary benefit (20k threshold) Acute 32,440 32,202 238 5.446 5.448 0.002 Dominant 268 Non-acute 36,853 36,725 128 5.419 5.422 0.003 Dominant 189
Acute setting In the acute setting, the multifactorial inpatient falls prevention intervention reduced costs and increased QALYs so is said to be dominant over the control arm.
The QALY difference (0.002 extra QALYs) generated was small equivalent to less than 1 extra quality-adjusted day over the average 10-year lifetime of a patient in the model. The QALY gain arose from a small increase in time spent at home and a decrease in time spent in care. In the acute setting, the cost difference was less than 1% of lifetime costs. The difference in costs was largely generated by a saving in the hospital state (on average, 167 per patient over a lifetime). This comprises a slight reduction in length of stay (average 0.06 days shorter following the intervention) and reduced costs of treating falls. These savings are more than enough to offset the cost of implementing the intervention. There were also savings from a reduction in time spent in care (2.3 days per person, equating to an average lifetime saving of 136). Non-acute setting In the non-acute setting, the multifactorial inpatient falls prevention intervention also reduced costs and increased QALYs so is said to be dominant over the control arm. The QALY difference (0.003 extra QALYs) was similar to the acute setting and was again driven by a slight increase in time spent at home. The cost difference was less than 0.5% of lifetime (discounted) costs and was due to a decrease in the average length of hospital stay.
Change in Cost ()
400
-200
-400
Sensitivity analysis Given the large number of inputs to this model, it is important to assess whether any inputs have a large influence on the outcomes generated. All deterministic sensitivity analyses were run in the acute setting with 500,000 patients using a fixed seed. Regression lines were fitted to the threshold analyses to minimise remaining sampling variation. One-way sensitivity analysis A number of input parameters had some impact on the cost effectiveness of the intervention (see appendix K (2013) for a tornado plot showing all input parameters varied). The only parameter that impacted the cost effectiveness of the intervention to such an extent as to make the intervention not cost effective was the intervention effect (IRR for falls with intervention compared with control). The cost of the intervention per patient was explored between 0 and 100 (base case 7.83 per patient) and the intervention remained cost effective; however it is inevitable that an even higher intervention cost would make the intervention not cost effective.
Varying individual parameters reflecting the costs and utilities associated with falls within plausible ranges did not affect the apparent cost effectiveness of the intervention. Threshold Analysis As it was the only parameter that appeared capable of influencing model results independently, the IRR associated with multifactorial inpatient falls prevention intervention was examined further in threshold analysis (see figure 4). This showed that, as long as some falls are prevented (IRR<1), the intervention is likely to be cost effective.
1,000
500
0
-500
-1,000
-1,500 0.5 1.0 Intervention Effect (IRR) 2.0
Figure 4: Threshold analysis of the incidence rate ratio for the multifactorial falls preventions intervention
Two-way sensitivity analysis The GDG indicated it would be useful to consider the relationship between intervention effectiveness and cost. This two-way sensitivity analysis could either consider the cost worth paying for a known effectiveness, or the effectiveness required for a known cost. The two-way sensitivity analysis suggests that, if the intervention effectiveness IRR is less than 0.85 (that is, if it reduces the incidence of inpatient falls by 15% or more), the intervention remains cost effective (at the
20,000 per QALY threshold) even when the intervention cost is 100 per admitted patient. The meta-analysis gave an IRR of 0.76, for which the twoway sensitivity analysis implies the intervention is cost effective even when the intervention costs 100 per admitted patient. Similarly, if the intervention is known to cost 7.83, then it remains cost effective as long as the IRR is less than 1. Discussion Principal findings It would appear that, if inpatient falls can be prevented and/or their severity reduced, this is very likely to be a cost-effective course of action. However, in both settings, the difference in QALYs is small, as can be seen in figure 3. The changes represent less than 0.06% of lifetime QALYs in the model. The cost differences are slightly bigger, but still less than 0.7% of an average patients total lifetime costs. A difference of 0.003 QALYs or less (1 quality-adjusted day or less) may be viewed as clinically not relevant over an average lifetime of ten years. However, the average experience of the simulated cohort contains a great heterogeneity of experience, and some simulated patients will have derived very appreciable benefit from having falls averted during their hospitalisation(s). Certainly, the 25% reduction in fall rates (taken from the meta-analysis) was thought to be clinically significant by the GDG. Cost differences were small but generally indicated savings in both settings. Therefore, the intervention can probably be said to be cost effective in the acute and non-acute settings, compared with no action to reduce the rate and/or severity of inpatient falls. Strengths of the analysis This is one of the first health economic analyses attempting to capture both the costs and benefits of interventions to prevent inpatient falls and is the first to consider outcomes in terms of QALYs (Davis et al., 2011). It was developed
with a high degree of expert input from the GDG members that resulted in significant model changes. The model structure reflects many of the complexities of the patient pathway, and is able to capture all important potential benefits of the intervention over a patients lifetime. The discrete event approach enables a realistic simulation of a heterogeneous population, and allows the detailed exploration of the history of simulated patients and their likelihood of experiencing subsequent events. Although the model contains a great many parameters, each of which is subject to a degree of uncertainty, conclusions are relatively robust when these values are varied within plausible ranges. Limitations of the analysis The model relies heavily on estimates of the cost of the intervention, the cost of treating inpatient falls and the ability of the multifactorial interventions to reduce fall rates the true values of each remain unknown. The multifactorial intervention is deliberately generic but is costed based on informed estimates from one RCT and the estimated costs of treating inpatient falls are based on a number of assumptions about treatment received by each patient. Despite being based on a meta-analysis of reasonably powered RCTs, the intervention effect has relatively wide confidence intervals and, as discussed in section 4.4.2, using different groupings of evidence in meta-analyses produced different intervention effect sizes. However, the analysis suggests that the intervention as modelled is likely to be cost effective as long as (a) some reduction in falls is achieved and (b) the costs of the intervention do not outweigh this benefit. The model relies on a linear combination of data on inpatient fall risk factors and is not able to account for the interactions that are likely to exist between these risk factors. The splitting of hospital states into acute and non-acute is unlikely to reflect the complex array of arrangements that exist in the NHS. It was based on a simplifying decision by the GDG and is likely to be a source of structural uncertainty in the model. In some instances, the non-acute setting inputs were
Falls: NICE clinical guideline 161 (June 2013) Page 237 of 315
based on approximations to the true rates and, whilst they were shown not to impact the results, the parameters were not ideal. It is debatable whether splitting the hospital state into acute and non-acute settings increased the value of the health economic model or just increased the uncertainty within the model. The causal multiplier was a necessary but limiting assumption. No evidence exists to quantify the causal relationship between an inpatient fall and subsequent events. However, sensitivity analysis shows that the model is relatively robust to this uncertainty: as long as it can be assumed that some of the negative experiences associated with falls are avoided by averting the falls themselves, some value can be anticipated from the intervention. Patients in the meta-analysis are unlikely to have received the inpatient falls intervention more than once. However, the model generates around four hospitalisations in each patients lifetime and applies the intervention to each hospitalisation the intervention may be more or less effective on subsequent applications and, accordingly, the model may under- or overestimate the value for money it provides. The modelling of the social care process is a gross simplification of reality. The focus of the model was preventing falls during a patients stay in hospital and, for this reason, the structure of the model was focused primarily on the hospital episode. However, the care state has a nontrivial influence on cost savings and QALY gains and therefore perhaps more modelling time should have been given to refining this state. This is set against the low quality of the evidence base and parameters on which to base the existing parameters any further refinement of the care state would have probably introduced more uncertainty into the model. The costs and benefits of interventions to prevent inpatient falls are likely to be borne by different parts of the NHS/PSS system. It is likely co-ordinated commissioning arrangements will be needed to implement them. The intervention modelled only applies to the hospital setting. No costs or utilities associated with community falls prevention (CG21) are included within
this model. The crossover between the previous falls prevention in the community guideline (CG21) and this guideline may be an area for potential future health economic research. There remains a lack of direct evidence on the utility experienced by patients in hospital and the utility decrement and duration experienced following an inpatient fall. Similarly, the utility of people in care homes and the impact of falls therein were based on assumed values. It is a significant weakness of this analysis that it has not proved computationally feasible to undertake full probabilistic sensitivity analysis, to explore the implications of parameter uncertainty for decision-making. A wide range of one-way sensitivity analyses was undertaken; this enables a fair degree of inference on the impact of such second-order uncertainty and, in the light of these analyses, it is possible to state with some confidence that the intervention would be associated with a greater than 50% probability of cost effectiveness in a fully probabilistic analysis. However, it is not possible to quantify this probability accurately, nor to explore the potential value of further research, in the absence of such an analysis. Finally, it is acknowledged that the model is extremely complex. A bigger model carries more structural uncertainty and requires more computational time. More sampling error than would be desired remains in the model, not least because inpatient falls are rare events, and inpatient falls with substantial consequences are rarer still. Therefore, it is necessary to simulate a very large number of patients to provide stable model results. Evidence Statement An innovative discrete event health economic model has been built that showed that, if it is possible to reduce inpatient fall rates, then this appears to be a cost effective course of action in the acute and non-acute hospital settings. However, the gains in QALYs and savings in costs were, for an average patient, very small.
4.4.5
Evidence to recommendations
Studies were grouped according to the setting in which they were conducted, because the GDG felt that there could be important differences between settings and that some interventions may be effective in some settings but not all. Thus the GDG categorised each study using their judgement based on the description of the hospital, type of patients and mean length of stay stated in each study. If this information was not available or was unclear, the GDG classified the study in the mixed or unclear setting. Interventions were grouped using the fall prevention interventions classification system of single, multiple or multifactorial, developed by the Prevention of Falls Network Europe (ProFaNE). The GDG prioritised number or rate of falls as the outcome of interest, and this is what its deliberations were based on. In addition to number or rate of falls, the GDG also considered injuries to be important, because ideally a successful intervention should result in statistically significant reductions in both falls and injuries. However, the GDG acknowledged that injurious falls are relatively rare and most studies are underpowered to detect meaningful changes in injury rates. Thus falls were prioritised over injuries. An original costutility model was built to assess the cost effectiveness of multifactorial interventions to prevent or reduce the number and/or severity of inpatient falls, compared with no action. The model results suggested that, if falls can be prevented, then this is likely to be a cost-effective course of action. Although the changes in costs and quality-adjusted life years (QALYs) were small, this conclusion was robust under sensitivity analyses. For interventions for which evidence was available, it was generally provided by only one study, and so meta-analysis was not conducted. In general the evidence was of low quality, and reported no statistically significant effects, meaning that there was uncertainty around whether the intervention actually worked or not. Some evidence did report statistically significant effects, but these results came from single studies, so the GDG was not confident that the results could be replicated. Since there was no meta-analysis for the GDG to base its deliberations on, and the GDG was not confident in the results of single studies, the GDG members used their experience and expertise to develop recommendations. The GDG felt that most older inpatients will have multiple risk factors for falling, and so single interventions are unlikely to work in isolation. The GDG felt that healthcare professionals should not offer a blanket intervention to whole groups of patients in the hope that it will reduce the risk of falls, as this may increase the likelihood that this type of intervention will be the only one used. Because of this, the GDG felt that single interventions that are not linked to an individual patients multifactorial falls risk assessment should not be used, and so a do not do recommendation was developed to reflect this. No evidence relating to multiple interventions was identified in this review. For multifactorial interventions, there were numerous studies, some of which provided high-quality evidence. Meta-analysis was conducted
Economic considerations
Quality of evidence
separately for each setting and the analyses were stratified by study design. Some significant heterogeneity remained in the analysis and the GDG debated whether it was appropriate to continue with the meta-analysis. It was agreed that in this case a meta-analysis with heterogeneity was better than no meta-analysis at all, as the GDG had no reason to identify a study or group of studies that were better or more believable than the others. So, rather than using the metaanalysis to identify the true effect among the pool of studies, a random-effects model was used to estimate an average effect among this heterogeneous body of evidence. For acute and non-acute settings, the meta-analyses of randomised studies were non-significant, indicating that there is uncertainty about whether the intervention actually reduced or increased fall rates in these settings. For the mixed or unclear setting, the meta-analysis of randomised studies was statistically significant and showed that multifactorial interventions reduced the rate of falls in this setting. The GDG discussed these results at length. The GDG felt that the a priori decision to separate studies depending on the setting in which it was conducted may not have been useful, as dividing the evidence up in this way may have contributed to a lack of statistical power in the meta-analyses. On reflection the GDG felt that had all randomised studies been included in a single meta-analysis, the overall results were likely to have been statistically significant in support of multifactorial fall prevention interventions. The GDG felt that in reality most hospital wards in the UK will accommodate a mixture of acute and non-acute patients, and patients will have a mixture of acute and non-acute needs as their condition changes. The GDG felt that although the body of evidence is heterogeneous and lacking in statistical power, the meta-analysis for the mixed and unclear category that most closely represents UK inpatient settings is supportive of multifactorial interventions in general, and evidence from non-randomised studies also supports this finding. Thus the GDG agreed that multifactorial fall prevention interventions could be recommended for all older inpatients. The GDG felt that there were two elements to most multifactorial interventions that were reviewed: general improvements to the inpatient environment (such as adequate lighting, flooring, furniture etc.), and targeted multifactorial interventions that link to each patients own multifactorial assessment. Thus the discussions linked back to those in section 4.3, where the GDG felt that a good multifactorial clinical assessment would highlight the patients individual risk factors for falling that could be treated, managed or improved during their stay, and this assessment should be linked to a related multifactorial intervention to address the identified factors. The group felt that it was not necessary to have a single assessment document for this; rather, the documentation could come from one or more existing assessments that are currently used. The GDG used their expertise and consensus to identify core elements that should be covered in the assessment as a minimum. They felt that the assessment process should focus on factors that could be treated, managed or improved during the inpatient stay, and listed nine example elements that a multifactorial assessment could cover. The
Other considerations
nine examples were not felt to be exhaustive, but the GDG agreed that these were the most common. The GDG agreed that the aim of inpatient falls prevention interventions is to reduce the risk of the patient falling during their hospital stay, not to reduce their risk of falling once they leave hospital. This is because the risk factors in hospital are different from those in the community and require a different identification and assessment process. Recommendations for falls prevention when a patient returns to their normal place of residence already exist (see section 3 of this guideline). The GDG wanted to stress that staff should take into account the length of a patients expected hospital stay when considering which interventions to offer, as short-stay patients may only benefit from short-term interventions, whereas longer-stay patients could benefit from more enduring interventions. The GDG felt that healthcare professionals are usually able to predict whether a patient will have a short or long inpatient stay, but acknowledged that this is not always possible. The GDG also felt that some interventions, such as the provision of appropriate walking aids, may only be available during Monday to Friday in some hospitals, with no provision over the weekend. The GDG felt that this was unacceptable; as potentially some patients admitted on Friday and discharged the following Monday may not have access to appropriate interventions. Thus the GDG wanted to stress that healthcare professionals should promptly address the patients individual risk factors for falling that are identified. Although the search strategy included all subgroups of the older inpatient population, evidence relating to specific subgroups (such as amputees or wheelchair users) that met the inclusion criteria was not identified. This lack of evidence meant that the GDG did not make specific recommendations relating to the specific subgroups, but the GDG wanted to emphasise that these groups were considered in their deliberations and are implicitly included in the guideline recommendations. For example, the multifactorial falls risk assessment that is recommended for all older inpatients may cover different areas of assessment for a person with an amputation than that for an older inpatient with dementia. Similarly, the GDG was unable to recommend any single interventions (such as hip protectors, telecare or cushioned flooring) because of a lack of evidence that met the inclusion criteria or poor-quality evidence. The GDG also felt that single interventions are unlikely to work in isolation. However, any intervention that aims to reduce or prevent falls can be used as part of a multifactorial intervention to address specific risk factors highlighted during each patients multifactorial assessment. For example, the linked multifactorial intervention for a person with a high risk of hip fracture may include hip protectors as one of the components, and the linked multifactorial intervention for a person with an amputation may include prosthetic adjustment as one of the components, even though no specific evidence relating to these components was identified in the evidence review.
4.4.6
Recommendations Recommendation 1.2.2.1 Ensure that aspects of the inpatient environment (including flooring, lighting, furniture and fittings such as hand holds) that could affect patients risk of falling are systematically identified and addressed. Recommendation 1.2.2.2 For patients at risk of falling in hospital (see recommendation 1.2.1.2), consider a multifactorial assessment and a multifactorial intervention. Recommendation 1.2.2.3 Ensure that any multifactorial assessment identifies a patients individual risk factors for falling in hospital that can be treated, improved or managed during their expected stay. These may include: cognitive impairment continence problems falls history, including causes and consequences (such as injury and fear of falling) footwear that is unsuitable or missing health problems that may increase their risk of falling medication postural instability, mobility problems and/or balance problems syncope syndrome visual impairment. Recommendation 1.2.2.4 Ensure that any multifactorial intervention: promptly addresses the patients identified individual risk factors for falling in hospital and
takes into account whether the risk factors can be treated, improved or managed during the patients expected stay. Recommendation 1.2.2.5 Do not offer falls prevention interventions that are not tailored to address the patients individual risk factors for falling.
4.5
4.5.1
Review question
What are the education and information needs of patients and their family members and carers after a hospital-based falls risk assessment, or a fall in hospital?
4.5.2
Evidence review
This question focused on identifying what information patients and their family members and carers should receive after a falls risk assessment or a fall in hospital. This review specifically did not address education and information that can be provided to patients with the aim of reducing their risk of falling, as this was considered to be a type of educational intervention for inclusion in section 4.4. Only information that aimed to enhance the patient experience was included in this question. The initial search identified 2441 studies and all were subject to screening. This led to the retrieval of 13 full texts which were considered in the review. Studies were included in the review if they related to patient or carer experience of the management of inpatient falls risk, or if they identified patient or carer needs for information during the management of inpatient falls risk. No restrictions on study design were imposed. Studies that sought patient experience of a fall that did not occur in the inpatient setting were excluded, as were studies that were not focused on patient or carer experience or needs. For a full list of excluded papers for this review question, see appendix G (2013). Three full text articles met the eligibility criteria and were included in the final review. The quality of the 3 included studies was assessed using appropriate NICE methodology checklists.
Quality assessment The studies all had limitations, mainly because of inadequate reporting of the researchers role and the impact this could have on participant responses, and a lack of reliability concerning the study findings. Indirect evidence from one quantitative study (Haines and McPhail 2011) provided some qualitative data for this review question (see table 11 and evidence tables in appendix E [2013] for further details). Since GRADE is not currently developed for use with qualitative evidence, a key themes matrix was used to present the key themes from the included studies.
3 months
Unclear
Cross sectional survey of patients willingness to pay for falls prevention interventions
Indirect evidence
Patients often forgot that they were at risk of falling or to call for help, because of medication that clouded their memory or because they urgently needed to move.
Carers accepted the use of side rails, but were aware they may also cause injury/harm
Participants who were provided with visual cues about the intervention were less likely to misunderstand the intervention
Participants valued targeted multifactorial interventions the most, followed by falls consultation, exercise, and face to face education. Participants valued hip protectors, and booklet and video education the least.
4.5.3
Evidence statements
For details of how the evidence is graded, see The guidelines manual (2009). Awareness 4.5.3.1 Evidence from 3 studies with some methodological limitations conducted in the USA, Ireland or Australia showed that: There is a lack of awareness about falls risk and why interventions such as bed rails are being used. Patients receive inconsistent messages about their falls risk from staff.
Visual cues about falls prevention interventions can reduce patients misunderstanding of the intervention. Patients may forget they are at risk of falling, or may forget to ask for help when mobilising. Patients do not want to be a burden to staff by asking for help when mobilising. Some interventions such as bed rails can cause harm, even though they are meant to reduce harm. Multifactorial fall prevention interventions are valued the most by patients, and booklet/leaflet education is valued the least.
4.5.4
Health economic modelling was not considered a priority for this review question. An economic literature search specific to this question was not undertaken.
4.5.5
Evidence to recommendations
The GDG felt that one study (Simpson, 1995) should be removed from the analysis because it was out of scope. This is because the intervention examined whether patients at risk of falling wanted to, and could, be taught how to get up from the floor if they fell when they were at home. Thus the effects of this intervention would not be seen during the inpatient stay. The GDG discussed the need to provide information to patients (and to their family members and carers if the patient agrees) that is relevant and useful. However, it recognised that the ability of some patients (such as those with memory problems or cognitive impairment) to understand and retain information may be compromised. Qualitative evidence identified that patients and their families and carers are often unaware of the patients fall risk, and that some patients who are aware of their increased falls risk feel they are burdening staff if they ask for help. The GDG felt that this was an accurate reflection of the inpatient experience and wanted to emphasise the need for healthcare professionals to provide consistent explanations about the patients individual risk factors for falling and encourage them to ask for help when moving around the hospital. To do this, healthcare professionals also need to show the patient how to use the nurse call system. This also led to a discussion about bed rails, as some GDG members felt that family members and carers often do not know how to use them correctly. Since incorrect use can lead to patient injuries, the GDG wanted to include a bullet point about bed rails even though no evidence was identified in relation to them. Health economics were not considered as a priority for this review question. The evidence reviewed originated from two qualitative studies, both of which had limitations, and one quantitative study from which qualitative data were extracted. Thus the data can be considered to be of low quality. The GDG agreed that the themes identified were an accurate representation of the literature, and of the experiences of some patients and carers. NICE clinical guideline 138 Patient experience in adult NHS services makes recommendations about continuity of care across NHS services. The GDG felt that it was important to refer to this guideline since older people who have fallen or who have been identified as being at higher risk of falling may have contact with multiple healthcare professionals. The GDG felt it was necessary to emphasise that relevant information should be shared across services to facilitate good, consistent care. Recommendation 1.2.3.2 was developed to do this.
Other considerations
4.5.6
Recommendations Recommendation 1.2.3.1 Provide relevant oral and written information and support for patients, and their family members and carers if the patient agrees. Take into account the patients ability to understand and retain information. Information should include: explaining about the patient's individual risk factors for falling in hospital showing the patient how to use the nurse call system and encouraging them to use it when they need help informing family members and carers about when and how to raise and lower bed rails. providing consistent messages about when a patient should ask for help before getting up or moving about helping the patient to engage in any multifactorial intervention aimed at addressing their individual risk factors. Recommendation 1.2.3.2 Ensure that relevant information is shared across services. Apply the principles in Patient experience in adult NHS services (NICE clinical guideline 138) in relation to continuity of care.
4.6
The Guideline Development Group has made the following recommendations for research, based on its review of evidence, to improve NICE guidance and patient care in the future.
population underpins the development of more effective and better targeted multifactorial assessments and interventions.
5
5.1
5.1.1
Further information
Guideline development group information
2013 Guideline Development Group membership and acknowledgements
The Guideline Development Group Damien Longson (Chair) Consultant Liaison Psychiatrist, Manchester Mental Health & Social Care Trust Harry Allen Consultant Psychiatrist for the Elderly, Manchester Mental Health and Social Care Trust Senel Arkut Strategic & Operational Social Care Manager, London Borough of Brent Elizabeth Caroline Brown Principle Physiotherapist in Medicine, University Hospital of North Staffordshire Harm Gordijn Falls Prevention Co-ordinator, South Warwickshire NHS Foundation Trust Frances Healey Head of Clinical Review and Response, National Patient Safety Agency Ray Jankowski Deputy Director of Public Health, NHS Hertfordshire PCT Rosemary Leaf Patient and carer member JoAnne Panitzke-Jones Senior Commissioning Manager, Torbay Care Trust
Opinder Sahota Professor in Orthogeriatric Medicine & Consultant Physician Lindsay Smith GP, East Somerset Research Consortium Cameron Swift Emeritus Professor of Health Care of the Elderly & Honorary Emeritus Consultant Physician in General Internal Medicine John Taylor Patient and carer member Julie Windsor Clinical Nurse Specialist, Portsmouth Hospitals NHS Trust Internal Clinical Guidelines Technical Team An Internal Clinical Guidelines Technical team was responsible for this guideline throughout its development. It prepared information for the Guideline Development Group, drafted the guideline and responded to consultation comments. Nicole Elliott Associate Director Michael Heath Programme Manager Dylan Jones Technical Adviser Jenny Kendrick Information Specialist Stephanie Mills Project Manager Gabriel Rogers Technical Adviser (Health Economics)
Falls: NICE clinical guideline 161 (June 2013) Page 255 of 315
Toni Tan Technical Adviser Steven Ward Technical Analyst (Health Economics) Sheryl Warttig Technical Analyst NICE Centre for Clinical Practice Phil Alderson Associate Director Ben Doak Guideline Commissioning Manager (from January 2013) Elaine Clydesdale Guideline Coordinator (from February 2013) Laura Donegani Guideline Coordinator (until February 2013) Jasdeep Hayre Technical Lead (Health Economics) Lyn Knott Editor Rachel Ryle Guideline Commissioning Manager (until January 2013) Nichole Taske Technical Lead Erin Whittingham Project Manager, Public Involvement Programme Guideline development group declarations of interest can be found in Appendix A (2013).
5.1.2
Professor Gene Feder (group leader) St Bartholomews and the London Queen Marys School of Medicine and Dentistry Miss Margaret Clark Alzheimers Society Dr Jacqueline Close Royal College of Physicians Dr Colin Cryer Centre for Health Services Studies, University of Kent at Canterbury Ms Carolyn Czoski-Murray School of Health and Related Research, University of Sheffield Mr David Green Royal Pharmaceutical Society of Great Britain Dr Steve Illiffe Royal College of General Practitioners Professor Rose Anne Kenny Institute for Health of the Elderly, University of Newcastle upon Tyne Dr Chris McCabe School of Health and Related Research, University of Sheffield Mrs Eileen Mitchell Clinical Effectiveness Forum for Allied Health Professionals Dr Sarah Mitchell Clinical Effectiveness Forum for Allied Health Professionals Dr Peter Overstall
Falls: NICE clinical guideline 161 (June 2013) Page 257 of 315
British Geriatrics Society Mrs Mary Preddy National Osteoporosis Society Professor Cameron Swift Kings College Hospital - also link GDG member for the osteoporosis guideline. Dr Deirdre Wild Royal College of Nursing National Collaborating Centre for Nursing and Supportive Care Staff at the National Collaborating Centre for Nursing and Supportive Care who contributed to this guideline were: Ms Jacqueline Chandler-Oatts Research associate Ms Elizabeth Gibbons R&D fellow Dr Gill Harvey Director Ms Jo Hunter Information specialist Ms Elizabeth McInnes Senior R&D fellow Mr Robin Snowball Information specialist (seconded from Cairns Library, John Radcliffe Hospital,Oxford) Mr Edward Weir Centre Manager
Additional assistance Dr Phil Alderson Cochrane Centre, UK Dr Lesley Gillespie Cochrane, UK Ms Colette Marshall National Institute for Clinical Excellence Dr Martyn Parker Cochrane UK Dr Lesley Smith Centre for Statistics in Medicine
5.2
NICE guidelines are developed in accordance with a scope that defines what the guideline will and will not cover. The scope for section 3 of this guideline (all older people) is given in section 3.2. The scope of section 4 of this guideline (inpatients) is given in appendix I (2013).
5.3
Implementation
5.4
5.4.1
The NICE guideline contains all the recommendations, without the information on methods and evidence.
5.4.2
NICE pathway
The recommendations from this guideline have been incorporated into a NICE pathway.
5.4.3
A summary of the recommendations is available for the public (Information for the public). We encourage NHS and voluntary sector organisations to use this text in their own information about falls.
5.5
Published Osteoporosis. NICE clinical guideline 146 (2012) Patient experience in adult NHS services. NICE clinical guidance 138 (2012) Service user experience in adult mental health. NICE clinical guidance 136 Hip fracture. NICE clinical guideline 124 (2011) Delirium. NICE clinical guideline 103 (2010) Medicines adherence. NICE clinical guideline 76 (2009) Mental wellbeing and older people. NICE public health guidance 16 (2008) Stroke. NICE clinical guideline 68 (2008) Head injury. NICE clinical guideline 56 (2007) Dementia. NICE clinical guideline 42 (2006) Parkinson's disease. NICE clinical guideline 35 (2006) Under development NICE is developing the following guidance (details available from the NICE website): Head injury (update). NICE clinical guideline. Publication expected January 2014. Older people with multiple morbidities. NICE public health guidance. Publication expected September 2014.
5.6
References
Abrahamsen B, van Staa T, Ariely R, Olson M, Cooper C (2009). Excess mortality following hip fracture: a systematic epidemiological review. Osteoporos International. Oct; 20(10):1633-50 Abreu N, Hutchins J, Matson J, Polizzi N, Seymour CJ (1998) Effect of group versus home visit safety education and prevention strategies for falling in community-dwelling elderly persons, Home Health Care Management & Practice, 10(4):57-65. Ades PA, Ballor DL, Ashikaga T, Utton JL, Nair KS (1996) Weight training improves walking endurance in healthy elderly persons, Annals of Internal Medicine, 124:568-72. Aditya BS, Sharma JC, Allen SC, Vassallo M (2003). Predictors of a nursing home placement from a non acute geriatric hospital. Clinical Rehabilitation. 17: 108-113 Allen CM, Becker PM, McVey L et al. (1986) A randomized, controlled clinical trial of a geriatric consultation team. Journal of the American Medical Association 255: 2617-21. Allen CM, Becker PM, McVey LJ, Saltz C, Feussner JR, Cohen HJ (1986) A randomized, controlled clinical trial of a geriatric consultation team. Compliance with recommendations, Journal of the American Medical Association, 255:2617-21. Alpini D, Kohen RR, Braun R, Burstin A, Tesio L, Pugnetti L, Mendozzi L, Sambataro G, Cesarani A, Giuliano DA (2001) Falls in the elderly: The development of a risk questionnaire and posturographic findings, The International Tinnitus Journal, 7:2-108. Altman DG, Bland JM (2003) Interaction revisited: the difference between two estimates. BMJ 326:219
American Geriatric Society, British Geriatrics Society and American Academy of Orthopaedic Surgeons Panel on Falls Prevention (2001) Guideline for the prevention of falls in older persons, Journal of the American Geriatric Society, 49:664-672. American Psychiatric Association (1996) Diagnostic and statistical manual of mental disorders, DSM-IV, Washington DC. Aminzadeh F, Edwards N (1998) Exploring seniors views on the use of assistive devices in fall prevention, Public Health Nursing, 15:297-304. Andresen EM, Gravitt GW, Aydelotte ME, Podgorski CA (1999). Limitations of the SF-36 in a sample of nursing home residents. Age Ageing. Oct; 28(6):562-6 Andrews HP, Gilbar PJ, Wiedmann RJ, Groves CF (2001) Fall related hospital admission in elderly patients: contribution of medication use, The Australian Journal of Hospital Pharmacy, 31:3: 183-187. Anemaet WK, Motta-Trotter E (1997) The user friendly home care handbook, Learn Publications. Arfken CL, Lach HW, Birge SJ, Miller JP (1994) The prevalence and correlates of fear of falling in elderly persons living in the community, American Journal of Public Health, 84(4):565-570. Armstrong AL, Oborne J, Coupland CAC, Macpherson MB, Bassey EJ, Wallace WA (1996) Effects of hormone replacement therapy on muscle performance and balance in postmenopausal women, Clinical Science, 91:685-90. Ballinger C, Payne S (2000) Falling from grace or into expert hands? Alternative accounts about falling in older people, British Journal of Occupational Therapy, 63:573-9.
Barnett A, Smith B, Lord SR, Williams M, Baumond A (2003) Community based group exercise improves balance and reduces falls in at-risk people: A randomised, controlled trial, Age and Ageing, 32(4):407-14. Barry E, Laffoy M, Matthews E et al. (2001) Preventing accidental falls among older people in long stay units. Irish Medical Journal 94: 172-6. Bates DW, Pruess K, Souney P, Platt R (1995). Serious falls in hospitalized patients: correlates and resource utilization. American Journal Medicine. Aug;99(2):137-43 Bean J, Herman S, Kiely DK, Callahan D, Mizer K, Frontera WR, et al. (2002) Weighted stair climbing in mobility-limited older people: a pilot study, Journal of the American Geriatric Society, 50(4):663-70. Becker C, Kron M, Lindemann U, Sturm E, Eichner B, Walter-Jung B, et al. (2003) Effectiveness of a multifaceted intervention on falls in nursing home residents, Journal of the American Geriatric Society, 51(3):306-13. Behrman AL, Light KE, Flynn SM, Thigpen MT (2002) Is the functional reach test useful for identifying falls risk among individuals with Parkinsons disease? Archives of Physical Medicine and Rehabilitation, 83(4):538-542. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B (1992) Measuring balance in the elderly: validation of an instrument, Canadian Journal of Public Health, 83 (Suppl) 2:S7-11. Bergland A, Jarnlo GB, Laake K (2002) Validity of an index of self-reported walking for balance and falls in elderly women, Advances in Physiotherapy, 4(2):65-73. Bhala RP, ODonnell J, Thoppil E (1982) Ptophobia. Phobic fear of falling and its clinical management, Physical Therapy, 62(2):187-190. Biderman A, Cwikel J, Fried AV, Galinsky D (2002) Depression and falls among community-dwelling elderly people: A search for common risk factors, Journal of Epidemiology and Community Health, 56:631-636.
Binder EF, Brown M, Craft S, Schechtman KB, Birge SJ (1994) Effects of a group exercise program on risk factors for falls in frail older adults, Journal of Ageing and Physical Activity, 2:25-37. Birks Y, Hildreth R, Campbell P, Sharpe C, Torgerson D, Watt I (2003) A randomised controlled trial of hip protectors for the prevention of second hip fractures, Age and Ageing, 32(4):442-4. Bischoff HA, Stahelin HB, Dick W, Akos R, Knecht M, Salis C, et al. (2003) Effects of vitamin D and calcium supplementation on falls: A randomized controlled trial, Journal of Bone & Mineral Research, 18(2):343-51. Bischoff-Ferrari HA, Dawson-Hughes B, Willett WC, Staehelin HB, Bazemore MG, Zee RY, Wong JB (2004) Effect of Vitamin D on falls: a meta-analysis, Journal of American Medical Association, 291(16): 1999-2006. Blake AJ, Morgan K, Bendall MJ, Dallosso H, Ebrahim SB, Arie TH, Fentem PH, Bassey EJ (1988). Falls by elderly people at home: prevalence and associated factors. Age Ageing. Nov; 17(6):365-72 Bliuc D, Nguyen ND, Milch VE, Nguyen TV, Eisman JA, Center JR (2009). Mortality risk associated with low-trauma osteoporotic fracture and subsequent fracture in men and women. JAMA. Feb 4;301(5):513-21 Bloem BR,Grimbergen YA,Cramer M,Valkenburg VV (2000) Stops walking when talking does not predict falls in Parkinsons disease, Annals of Neurology, 48(2)268. Bowling A, Formby J (1992) Accidents in elderly care: a randomised controlled trial (Part 1), Nursing Standard, 6(29):28-30. Bowling A, Formby J, Grant K (1992) Accidents in elderly care: a randomised controlled trial (Part 2), Nursing Standard, 6:28-31. Bowling A, Formby J, Grant K (1992) Accidents in elderly care: a randomised controlled trial (Part 3), Nursing Standard, 6:25-7.
Brand CA, Sundararajan V (2010). A 10-year cohort study of the burden and risk of in-hospital falls and fractures using routinely collected hospital data. Quality and Safety in Health Care; Dec;19(6): e51 Brandis S (1999) A collaborative occupational therapy and nursing approach to falls prevention in hospital inpatients. Journal of Quality in Clinical Practice 19: 215-20. BrazierJ, Green C, McCabe C, Stevens K (2003) Use of visual analogue scales in economic evaluation. Expert Review, Pharmacoeconomics Outcomes Research, 3(3), 293-302. Brown JS,Vittinghoff E,Wyman JF,Stone KL,Nevitt MC,Ensrud KE, Grady D (2000) Urinary incontinence: does it increase risk for falls and fractures? Study of Osteoporotic Fractures Research Group, Journal of the American Geriatric Society, 48(7)721-725. Bruce DG, Devine A, Prince RL (2002) Recreational physical activity levels in healthy older women: the importance of fear of falling, Journal of the American Geriatric Society, 50:84-9. Buchner DM, Cress ME, de Lateur BJ, Esselman PC, Margherita AJ,Price R, et al. (1997) The effect of strength and endurance training on gait, balance,fall risk,and health services use in community-living older adults, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 52(4):M218-24. Bueno-Cavanillas A, Padilla-Ruiz F, Jimenez-Moleon JJ, Peinado-Alonso CA, Galvez-Vargas R (2000) Risk factors in falls among the elderly according to extrinsic and intrinsic precipitating causes, European Journal of Epidemiology, 16(9)849-859. Burleigh E, McColl J, Potter J (2007) Does vitamin D stop inpatients falling? A randomised controlled trial. Age & Ageing 36: 507-13. Cameron ID, Crotty M, CurrieC, Finnegan T, Gillespie W, Gillespie L, Handoll H, Kurrle S, Madhol R, Murray G, Quinn K, Torgerson D (2000) Geriatric
rehabilitation following fractures in older people: a systematic review, Health Technology Assessment, 4(2). Cameron ID, Cumming RG, Kurrle SE, Quine S, Lockwood K, Salkeld G, et al. (2003) A randomised trial of hip protector use by frail elderly older women living in their own homes, Injury Prevention, 9(2):138-41. Cameron ID, Handol HHG, Finnegan TP, Madhok R, Langthorne P (2002) Co-ordinated multi-disciplinary approaches for inpatient rehabilitation of older patients with proximal femoral fractures, (Cochrane review) in The Cochrane Library,issue 3, Oxford. Cameron ID,Venman J,Kurrle SE,Lockwood K,Bicks C, Cumming RG, et al. (2001) Hip protectors in aged-care facilities: a randomized trial of use by individual higher-risk residents, Age and Ageing, 30(6):477-81. Campbell AJ, Reinken J, Allan BC, Martinez GS (1981). Falls in old age: a study of frequency and related clinical factors. Age Ageing. Nov;10(4):264-70 Campbell AJ, Robertson MC, Gardner MM, Norton RN, Tilyard MW, Buchner DM (1997) Randomised controlled trial of a general practice programme of home based exercise to prevent falls in elderly women, British Medical Journal, 315:1065-9. Campbell AJ, Robertson MC, Gardner MM, Norton RN, Tilyard MW, Buchner DM (1999) Psychotrophic medication withdrawal and a home-based exercise program to prevent falls: a randomised, controlled trial, Journal of the American Geriatric Society, 47(7):850-3. Capan K, Lynch B (2007) A hospital fall assessment and intervention project. Journal of Clinical Outcomes Management 14: 155-60. Caplan GA, Ward JA, Brennan NJ, Coconis J, Board N, Brown A (1999) Hospital in the home: a randomised controlled trial, Medical Journal of Australia, 170(4):156-60.
Carpenter GI, Demopoulos GR (1990) Screening the elderly in the community: controlled trial of dependency surveillance using a questionnaire administered by volunteers, British Medical Journal, 300:1253-6. Carroll DL, Dykes PC, Hurley AC (2010) Patients' perspectives of falling while in an acute care hospital and suggestions for prevention. Applied Nursing Research 23: 238-41. Carter ND, Khan KM, McKay HA, Petit MA, Waterman C, Heinonen A, et al. (2002) Community-based exercise program reduces risk factors for falls in 65to 75-year-old women with osteoporosis. Randomized controlled trial, Canadian Medical Association Journal, 167(9):997-1004. Carter S, Campbell E, Sanson-Fisher R, Tiller K, Gillespie WJ (1997) A randomised controlled trial of two strategies aimed at reducing falls and other unintentional events through home modification and medication review, (unpublished work). Cerny K, Blanks R, Mohamed O, Schwab D, Robinson B, Russo A, Zizz C (1998) The effect of a multidimensional exercise program on strength, range of motion, balance and gait in the well elderly (abstract), Gait and Posture, 7(2): 18. Cesari M, Landi F, Torre S, Onder G, Lattanzio F, Bernabei R (2002) Prevalence and risk factors for falls in an older community-dwelling population, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 57(11):M722-M726. Chan DK, Hillier G, Coore M, Cooke R, Monk R, Mills J, Hung W (2000) Effectiveness and acceptability of a newly designed hip protector: a pilot study, Archives of Gerontology and Geriatrics, 30(1):25-34. Chandler JM, Prescott B, Duncan PW (2001) Special feature: The home assessment profile- a reliable and valid assessment tool, Top Geriatric Rehabilitation, 16(3) 77-88.
Charette SL, McEvoy L, Pyka G, Snow Harter C, Guido D, Wiswell RA,et al. (1991) Muscle hypertrophy response to resistance training in older women, Journal of Applied Physiology, 70(5):1912-6. Chartered Physiotherapists Working with Older people, Association of Chartered Physiotherapists in the Community, Occupational Therapy for Elderly People (2000) Guideline for the collaborative, rehabilitative management of elderly people who have fallen. Chartered Society of Physiotherapy. Cheal B,.Clemson L (2001) Older people enhancing self-efficacy in fall-risk situations, Australian Occupational Therapy Journal, 48:-80. Cheng P-T, Wu S-H, Liaw M-Y, Wong AM, Tang F-T (2001) Symmetrical body-weight distribution training in stroke patients and its effect on fall prevention, Archives of Physical Medicine and Rehabilitation, 82:1650-4. Chin A, Paw MJ, de Jong N, Schouten EG, Hiddink GJ, Kok FJ (2001) Physical exercise and/or enriched foods for functional improvement in frail, independently living elderly; a randomized controlled trial, Archives of Physical Medicine and Rehabilitation, 82:811-7. Cho CY, Kamen G (1998) Detecting balance deficits in frequent fallers using clinical and quantitative evaluation tools, Journal of the American Geriatric Society, 46(4):426-430. Chu LW, Pei CK, Chiu A et al. (1999) Risk factors for falls in hospitalized older medical patients. Journals of Gerontology Series A-Biological Sciences & Medical Sciences 54: M38-M43. Clark BA,Wade MG,Massey BH,Van Dyke R (1975) Response of institutionalized geriatric mental patients to a twelve-week program of regular physical activity, Journals of Gerontology, 30:565-73. Clarke M,Oxman AD,editors.Optimal search strategy for RCTs, Cochrane Reviewers Handbook 4.1.6 [updated January 2003]; Appendix 5c. in: The
Cochrane Library [database on disk and CDROM]. The Cochrane Collaboration. Oxford: Update Software; 2003, issue 2. Clemson L (1997) Home fall hazards and the Westmead home safety assessment, West Brunswick: Coordinates Publications. Clemson L, Cumming RG, Roland M (1996) Case-control study of hazards in the home and risk of falls and hip fractures, Age and Ageing, 25:97-101. Close J, Ellis M, Hooper R, Glucksman E, Jackson S, Swift C (1999) Prevention of falls in the elderly trial (PROFET): a randomised controlled trial, Lancet, 353(9147):93-7. Close JCT, Patel A, Hooper R, Glucksman E, Jackson SHD, Swift CG (1999) PROFET improved clinical outcomes at no additional cost (abstract). Coleman E, Fox PD (2002) Translating evidence-based geriatric care into practice: lessons from managed care organisations. Part ll: falls and medication-related complications, Annals of Long Term Care, 10: 10: 42-48 Coleman EA, Grothaus LC, Sandhu N, Wagner EH (1999) Chronic care clinics: a randomized controlled trial of a new model of primary care for frail older adults, Journal of the American Geriatric Society, 47(7):775-83. Colon-Emeric CS, Laing. (2002) Falls in older adults: assessment and intervention in primary care, Hospital Physician, 38:55-66. Commonwealth of Australia (2000) Step out with confidence national falls program, Canberra. Conley D, Schultz AA, Selvin R (1999) The challenge of predicting patients at risk for falling: development of the Conley scale, Medsurg Nursing, 8(6)348354. Cooper C, Atkinson EJ, Jacobsen SJ, et al. (1993) Population-based study of survival after osteoporotic fractures, American Journal of Epidemiology, 137:1001-1005. Cornillon E, Blanchon MA, Ramboatsisetraina P, Braize C, Beauchet O, Dubost V et al. (2002) Effectiveness of falls prevention strategies for elderly
subjects who live in the community with performance assessment of physical activities (before-after), Annales de Readaptation et de Medecine Physique, 45(9):493-504. Covinsky KE, Kahana E, Kahana B, Kercher K, Schumacher JG, Justice AC (2001) History and mobility exam index to identify community-dwelling elderly persons at risk of falling, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 56(4)M253-M259 Crilly RG, Willems DA, Trenholm KJ, Hayes KC, Delaquerriere Richardson LF (1989) Effect of exercise on postural sway in the elderly, Gerontology, 35:13743. Crotty M, Whitehead CH, Gray S, Finucane PM (2002) Early discharge and home rehabilitation after hip fracture achieves functional improvements: a randomised controlled trial, Clinical Rehabilitation, 16(4):406-13. Cryer C (2001) What works to prevent accidental injury amongst older people, London: Health Development Agency Culos-Reed S, Rejeski J, McAuley E, Ockene J, Roter D (2000) Predictors of adherence to behaviour change interventions in the elderly, Controlled Clinical Trials, 21:200S-5S. Cumming RG, Salkeld G, Thomas M, Szonyi G (2000) Prospective study of the impact of fear of falling on activities of daily living, SF-36 scores, and nursing home admission, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 55(5):M299-M305. Cumming RG, Sherrington C, Lord SR et al. (2008) Cluster randomised trial of a targeted multifactorial intervention to prevent falls among older people in hospital. BMJ 336: 758-60. Cumming RG, Thomas M, Szonyi G, Salkfield G, ONeill E, Westbury C, Frampton G (1999) Home visits by an occupational therapist for assessment and modification of environmental hazards: A randomised trial of falls prevention, Journal of American Geriatric Society, 47(12):1471-2.
Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, Cauley J, Black D Vogt TM Oestoporotic fractures research group, Journal of the American Geriatric Society, 51(3):356-63. Cwikel J, Fried AV, Biderman A, Galinsky D (1998) Validation of a fall risk screening test, the elderly fall screening test (EFST) for community-dwelling elderly, Disability Rehabilitation, 0(5):161-7. Cwikel J, Fried AV, Galinsky D, Ring H (1995) Gait and activity in the elderly: implications for community falls-prevention and treatment programmes, Disability Rehabilitation, 17(6): 277-280. Darton R, Forder J, Bebbington A, Netten A, Towers A, Williams J (2006). Analysis to Support the Development of the Relative Needs Formula for Older People: Final Report. PSSRU Discussion Paper 2265/3 Davis J.C., Roberston M.C., Comans T., Scuffham P.A (2011). Guidelines for conducting and reporting economic evaluation of fall prevention strategies. Osetoporosis International; 22:2449-2459 Dawson-Hughes B, Harris SS, Krall EA, Dallal GE (1997) Effect of calcium and vitamin D supplementation on bone density in men and women 65 years of age or older, New England Journal of Medicine, 337(10):670-6. Day L, Fildes B, Gordon I, Fitzharris M, Flamer H, Lord S (2002) Randomised factorial trial of falls prevention among older people living in their own homes British Medical Journal, 325:128-31. Deery HA, Day LM, Fildes BN (2000) An impact evaluation of a falls prevention program among older people, Accident Analysis and Prevention, 32:427-33. Delbaere K, Close JC, Menz HB, Cumming RG, Cameron ID, Sambrook PN (2008). Development and validation of fall risk screening tools for use in residential aged care facilities. Med J Aust 2008;189(4):193-6
Department of Health (2001) National service frameworks, older people, London: Stationery Office. Department of Health (2002). Health Survey for England 2000 (online). Available at http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsSt atistics/DH_4008793 (accessed 24 August 2012) Department Of Health (2009). Fracture Prevention Services: An Economic Evaluation of Fracture Plan. London. Department of Health (2011) NHS Costing and Cost Collection (online). Available at http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPol icyAndGuidance/DH_131140 (accessed 24 August 2012) Di Fabio RP, Seay R (1997) Use of the fast evaluation of mobility, balance, and fear in elderly community dwellers: validity and reliability, Physical Therapy, 77(9):904-917. Donald IP, Pitt K, Armstrong E, Shuttleworth H (2000) Preventing falls on an elderly care rehabilitation ward, Clinical Rehabilitation, 14(2):178-85. Donoghue J, Graham J, Mitten-Lewis S et al. (2005) A volunteer companionobserver intervention reduces falls on an acute aged care ward. International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 18: 24-31. Downton JH, Andrews K (1991). Prevalence, characteristics and factors associated with falls among the elderly living at home. Ageing 3(3) 219-228 Drummond M, Jefferson T (1996) Guidelines for authors and peer reviewers of economic submissions to the BMJ, British Medical Journal, 313:275-283. Duncan PW, Weiner DK, Chandler J, Studenski S (1990) Functional reach: a new clinical measure of balance, Journal of Gerontology 45(6):M192-7.
Dyer CA, Taylor GJ, Reed M, Dyer CA, Robertson DR, Harrington R (2004). Falls prevention in residential care homes: a randomised controlled trial. Age Ageing Nov;33(6):596-602 Dykes PC, Carroll DL, Hurley A et al. (2010) Fall prevention in acute care hospitals: a randomized trial. JAMA 304: 1912-8. Eagle DJ, Salama S, Whitman D et al. (1999) Comparison of three instruments in predicting accidental falls in selected inpatients in a general teaching hospital. Journal of Gerontological Nursing 25: 40-5. Ebrahim S, Thompson PW, Baskaran V, Evans K (1997) Randomized placebo-controlled trial of brisk walking in the prevention of postmenopausal osteoporosis, Age and Ageing, 26(4):253-60. Eccles M and Mason J (2001) How to develop cost conscious guidelines, Health Technology Assessment, 2001; 5 (16). Egger M, Davey Smith G, Altman DG (editors) (2001) Systematic Reviews in Health Care-Meta-analysis in context, London: British Medical Journal Books. Ekman A, Mallmin H, Michaelsson N, Ljunghall S (1997) External hip protectors to prevent osteoporotic hip fractures, Lancet, 350:563-4. Ensrud KE, Blackwell TL, Mangiere C, Bowman PJ, Whooley MA et al. (2002) Central nervous system-Active medications and risk for falls in older women, Journal of the American Geriatric Society; 50:1629-37. Fabacher D, Josephson K, Pietruszka F, Linderborn K, Morley JE, Rubenstein LZ (1994) An in-home preventive assessment program for independent older adults: a randomized controlled trial, Journal of the American Geriatric Society, 42(6):630-8. Felson DT, Anderson JJ, Hannan MT, Milton RC. Wilson PW, Kiel DP Impaired vision and hip fracture. The Framingham study, Journal of the American Geriatric Society, 37(6):495-500.
Fiatarone MA, ONeill EF, Doyle RN, Clements K (1997) Efficacy of home based resistance training in frail elders in Andrews GR (editor) Abstracts of the 16th Congress of the International Association of Gerontology, Bedford park, South Australia: World Congress of Gerontology Inc, 323 Abstract 985. Fiatarone MA, ONeill EF, Ryan ND, Clements KM, Solares GR, Nelson ME, Roberts SB, Kehayias JJ, Lipsitz LA, Evans WJ (1994) Exercise training and nutritional supplementation for physical frailty in very elderly people, New England Journal of Medicine, 330(25):1769-75. Finlay O, van dM, Beringer TRO (1999) Use of gait analysis to demonstrate benefits of footwear assessment in elderly people, Physiotherapy; 85:451456. Folstein MF, Folstein SE, McHugh PR (1975) Mini-mental state. A practical method for grading the cognitive state of patients for the clinician, Journal of Psychiatric Research, 12(3):189-98. Fonda D, Cook J, Sandler V et al. (2006) Sustained reduction in serious fallrelated injuries in older people in hospital. Medical Journal of Australia 184: 379-82. Forder J, Fernandez J-L (2011). Length of stay in care homes, Report commissioned by Bupa Care Services, PSSRU Discussion Paper 2769. Canterbury: PSSRU Friedman SM, Munoz B, West SK, Rubin GS, Fried LP (2002) Falls and fear of falling: which comes first? A longitudinal prediction model suggests strategies for primary and secondary prevention, Journal of the American Geriatric Society, 50(8):1329-1335. Fries BE, Hawes C, Morris JN, Philips CD, Mor V, Park PS Effect of the national resident assessment instrument on selected health conditions and problems, Journal of American Geriatric Society, 45(8):994-1001. Galindo-Ciocon DJ, Ciocon JO, Galindo DJ (1995) Gait training and falls in the elderly, Journal of Gerontological Nursing, 21(6):10-7,54-5.
Gallagher EM, Brunt H (1996) Head over heels: impact of a health promotion program to reduce falls in the elderly, Canadian Journal on Ageing,15(1):8496. Gallinagh R, Nevin R, Campbell L et al. (2001) Relatives' perceptions of side rail use on the older person in hospital. British Journal of Nursing 10: 391-2. Giles LC, Bolch D, Rouvray R et al. (2006) Can volunteer companions prevent falls among inpatients? A feasibility study using a pre-post comparative design. BMC Geriatrics 6: 11. Gill TM, Williams CS, Tinetti ME (2000) Environmental hazards and the risk of non-syncopal falls in the homes of community living older persons, Medical Care, 38:1174-1183. Gillepse LD, Gillepse WJ, Robertson MC, Lamb SE, Cumming RG, Rowe BH (2003) Interventions for preventing falls in the elderly, (Cochrane review),in The Cochrane Library,issue 2. Goldacre MJ, Roberts SE, Yeates D (2002). Mortality after admission to hospital with fractured neck of femur: database study. BMJ; 325:8689 Goodgold S, Kiami S, Ule D, Schoenberg A, Forman, G (2001) Applicability of the functional reach and timed up and go tests for elderly individuals with Alzheimers Disease: Pilot investigation, Physical and Occupational Therapy in Geriatrics, 19:1-36. Gotzsche PC, Liberati A, Torri V, Rossetti L. (1996) Beware of surrogate outcome measures, International Journal of Technology Assessment in Health Care, 12(2):238-46. Graafmans WE, Ooms ME, Hofstee HMA, Bezemer PD, Bouter LM, Lips P (1996) Falls in the elderly: a prospective study of risk factors and risk profiles, American Journal of Epidemiology, 143(11):1129-36.
Gray-Donald K, Payette H, Boutier V (1995) Randomized clinical trial of nutritional supplementation shows little effect on functional status among freeliving frail elderly, Journal of Nutrition, 125(12):2965-71. Green J,Forster A,Bogle S,Young J (2002) Physiotherapy for patients with mobility problems more than 1 year after stroke: a randomised controlled trial, Lancet, 359(9302):199-203. Greendale GA,Salem GJ,Young JT,Damesyn M,Marion M,Wang MY, et al. (2000) A randomized trial of weighted vest use in ambulatory older adults:strength,performance,and quality of life outcomes, Journal of the American Geriatric Society, 48(3):305-11. Grossman, MD, Stewart AL (2003) You arent going to get better by just sitting around:Physical activity perceptions, motivations and barriers in adults 75 years of age or older, American Journal of Geriatric Cardiology, 12(1):33-37. Gunter KB, White KN, Hayes WC, Snow CM (2000) Functional mobility discriminates non-fallers from one-time and frequent fallers, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 55:M672M676. Hagberg JM, Graves JE, Limacher M, Woods DR, Leggett SH, Cononie C, et al. (1989) Cardiovascular responses of 70 to 79 yr old men and women to exercise training, Journal of Applied Physiology, 66(6):2589-94. Haines T., Kuys,S.S., Morrison,G., Clarke,J. Bew,P (2009). Cost-effectiveness analysis of screening for risk of in- hospital falls using physiotherapist clinical judgement. Medical Care; 47(4):448-56 Haines TP, Bell RA, Varghese PN (2010) Pragmatic, cluster randomized trial of a policy to introduce low-low beds to hospital wards for the prevention of falls and fall injuries. Journal of the American Geriatrics Society 58: 435-41. Haines TP, Bennell KL, Osborne RH et al. (2004) Effectiveness of targeted falls prevention programme in subacute hospital setting: randomised controlled trial. BMJ 328: 676.
Haines TP, Bennell KL, Osborne RH et al. (2006a) A new instrument for targeting falls prevention interventions was accurate and clinically applicable in a hospital setting. Journal of Clinical Epidemiology 59: 168-75. Haines TP, Hill AM, Hill KD et al. (2011) Patient education to prevent falls among older hospital inpatients: a randomized controlled trial. Archives of Internal Medicine 171: 516-24. Haines TP, Hill KD, Bennell KL et al. (2006b) Patient education to prevent falls in subacute care. Clinical Rehabilitation 20: 970-9. Haines TP, Hill KD, Bennell KL et al. (2007) Additional exercise for older subacute hospital inpatients to prevent falls: benefits and barriers to implementation and evaluation. Clinical Rehabilitation 21: 742-53. Haines TP, McPhail S (2011) Patient preference for falls prevention in hospitals revealed through willingness-to-pay, contingent valuation survey. Journal of Evaluation in Clinical Practice 17: 304-10. Hale WA, Delaney MJ, McGaghie WC (1992) Characteristics and predictors of falls in elderly patients, Journal of Family Practice, 34:5-581. Halfon P, Eggli Y, Van Melle G, Vagnair A (2001). Risk of falls for hospitalized patients: a predictive model based on routinely available data. Journal of Clinical Epidemiology. Dec; 54(12):1258-66 Hall N, De Beck P, Johnson D, Mackinnon K, Gutman G, Glick N (1992) Randomized trial of a health promotion program for frail elders, Canadian Journal of Ageing, 11:72-91. Hansen FR, Spedtsberg K, Schroll M (1992) Geriatric follow-up by home visits after discharge from hospital. A randomized controlled trial, Age and Ageing, 21:445-50. Harada A, Mizuno M, Takemura M, Tokuda H, Okuizumi H, Niino N (2001) Hip fracture prevention trial using hip protectors in Japanese nursing homes, Osteoporosis International, 12(3):215-21.
Harada N, Chiu V, Damron RJ, Fowler E, Siu A, & Reuben DB (1995) Screening for balance and mobility impairment in elderly individuals living in residential care facilities, Physical Therapy, 75:6-469. Hauer K, Rost B, Rutschle K, Opitz H, Specht N, Bartsch P, et al. (2001) Exercise training for rehabilitation and secondary prevention of falls in geriatric patients with a history of injurious falls, Journal of the American Geriatric Society, 49(1):10-20. Healey F, Monro A, Cockram A, Adams V, Heseltine D (2004). Using targeted risk factor reduction to prevent falls in older in-patients: a randomised controlled trial. Age Ageing. Jul;33(4):390-5. Healey F, Scobie S, Oliver D, Pryce A, Thomson R, Glampson B (2008). Falls in English and Welsh hospitals: a national observational study based on retrospective analysis of 12 months of patient safety incident reports. Quality and Safety in Health Care. Dec; 17(6):424-30 Health Education Board for Scotland (2001) The construction of the risks of falling in older people: lay and professional perspectives. Scotland: Health Education Board for Scotland. Hbert R,Robichaud L,Roy PM,Bravo G,Voyer L (2001) Efficacy of a nurseled multidimensional preventive programme for older people at risk of functional decline. A randomized controlled trial. Age and Ageing, 30(2):14753. Heikinheimo RJ, Jantti PL, Aho HJ, Maki-Jokela PL (1996) To fall but not to break safety pants. 3rd International Conference on Injury Prevention and Control; Feb 18-22; Melbourne. 576-8. Heinze C, Halfens R, Dassen T (2008) The reliability of the Hendrich Fall Risk Model in a geriatric hospital. International Journal of Older People Nursing 3: 252-7. Hendrich A (1988) An effective unit-based fall prevention plan, Journal of Nursing Quality Assurance, 3(1):28-36.
Hendrich A, Nyhuis A, Kippenbrock T et al. (1995) Hospital falls: development of a predictive model for clinical practice. Applied nursing research : ANR 8: 129-39. Hendriksen C, Lund E, Stromgard E (1984) Consequences of assessment and intervention among elderly people: a three-year randomised controlled trial, British Medical Journal, 89:1522-4. Hendriksen C, Lund E, Stromgard E (1989) Hospitalization of elderly people. A 3-year controlled trial, Journal of the American Geriatric Society, 37:117-22. Hill KD, Vu M, Walsh W (2007). Falls in the acute hospital setting--impact on resource utilisation. Australian Health Review. Aug;31(3):471-7 Hillsdon M, Thorogood M, Anstiss T, Morris J (1995) Randomized controlled trials of physical activity promotion in free living populations: a review, Journal of Epidemiology and Community Health, 49:448-453. Hofmeyer MR, Alexander NB, Nyquist LV, Medell JL, Koreishi A (2002) Floorrise strategy training in older adults, Journal of the American Geriatric Society, 50(10):1702-6. Hogan DB, MacDonald FA, Betts J, Bricker S, Ebly EM, Delarue B, et al. (2001) A randomized controlled trial of a community-based consultation service to prevent falls, Canadian Medical Association Journal, 165(5):537-43. Holmqvist LW, von Koch L, Kostulas V, Holm M, Widsell G, et al. (1998) A randomized controlled trial of rehabilitation at home after stroke in southwest Stockholm, Stroke, 29:591-7. Hopman-Rock M, Staats PGM, Tak ECPM, Droes RM (1999) The effects of psychomotor activation programme for use in groups of cognitively impaired people in homes for the elderly, International Journal of Geriatric Psychiatry, 14(8)633-42.
Hornbrook MC, Stevens VJ, Wingfield DJ, Hollis JF, Greenlick MR, Ory MG (1994) Preventing falls among community-dwelling older persons: results from a randomized trial, Gerontologist, 34(1):16-23. Howland J, Lachman ME, Peterson EW, Cote J, Kasten L, Jette A (1998) Covariates of fear of falling and associated activity curtailment, Gerontologist, 38(5): 549-555. Hu MH, Woollacott MH (1994) Multisensory training of standing balance in older adults: I. Postural stability and one-leg stance balance, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 49:M52-61. Hu MH, Woollacott MH (1994) Multisensory training of standing balance in older adults: II. Kinematic and electromyographic postural responses, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 49:M62-71. Hubacher M, Wettstein A (2001) Acceptance of hip protectors for hip fracture prevention in nursing homes, Osteoporosis International, 12:794-9 Huda A, Wise LC (1998) Evolution of compliance within a fall prevention program. Journal of Nursing Care Quality 12: 55-63. Iglesias CP, Manca A, Torgerson DJ (2009). The health-related quality of life and cost implications of falls in elderly women. Osteoporos Int. Jun;20(6):869-78 Iver RQ, Cumming RG, Mitchell P, Simpson JM, Peduto AJ (2000) Visual risk factors for hip fracture in older people, American Journal of Epidemiology, 152(7):633-9. Jannink-Nijlant JM, Diederiks JP, Brouwers MA, Metsemakers JF (1999) Screening for mobility disorders by the Mobility Control subscale of the short version of the Sickness Impact Profile, Clinical Rehabilitation, 13(6):492-497.
Jantti P, Aho H, Maki-Jokela L (1996) Protector trousers in prevention of hip fractures, Turvahousut lonkanseudun murtumien ehkisyss Suomen Lkrilehti, 51(32):3387-9 Jensen J, Lundin-Olsson L, Nyberg L, Gustafson Y (2002) Fall and injury prevention in older people living in residential care facilities: A cluster randomized trial, Annals of Internal Medicine, 136(10):733-41. Jeske L, Kolmer V, Muth M et al. (2006) Partnering with patients and families in designing visual cues to prevent falls in hospitalized elders. Journal of Nursing Care Quality 21: 236-41. Jitapunkul S (1998) A randomised controlled trial of regular surveillance in Thai elderly using a simple questionnaire administered by non-professional personnel, Journal of the Medical Association of Thailand, 81(5):352-6. Joo JH, Lenze EJ, Mulsant BH, Begley AE, Weber EM, Stack JA, Mazumdar S, Reynolds III CF, Pollock BG (2002) Risk factor for falls during treatment of late-life depression, Journal of Clinical Psychiatry, 63(10):936-941. Judge JO, Underwood M, Gennosa T (1993) Exercise to improve gait velocity in older persons, Archives of Physical Medicine and Rehabilitation, 74(4):4006. Kallin K, Lundin-Olsson L, Jensen J, Nyberg L, Gustafson Y (2002) Predisposing and precipitating factors for falls among older people in residential care, Public Health, 116(5):263-271. Kanis JA, Oden A, Johnell O et al (2003) The components of excess mortality after hip fracture. Bone 32:468473 Kannus P, Parkkari J, Niemi S, Pasanen M, Palvanen M, Jarvinen M,Vuori I (2000) Prevention of hip fracture in elderly people with use of a hip protector, New England Journal of Medicine, 343(21):1506-13.
Karnon J, Stahl JE, Brennan A, Mar J, Moller J (2012). Modelling using discrete event simulation: A report of the ISPOR-SMDM modelling good research practices task force-4. Medical Decision Making 32: 701 Kato M, Izumi K, Shirai S et al. (2008) Development of a fall prevention program for elderly Japanese people. Nursing & Health Sciences 10: 281-90. Kemoun G, Thoumie P, Boisson D, Guieu JD (2002) Ankle dorsiflexion delay can predict falls in the elderly, Journal of Rehabilitation Medicine, 34(6):278283. Kempton A, van Beurden E, Sladden T, Garner E, Beard J (2000) Older people can stay on their feet: Final results of a community-based falls prevention programme, Health Promotion International, 15(1):27-33. Kenny RA, OShea D, Walker HF (2002) Impact of a dedicated syncope and falls facility for older adults on emergency beds, Age and Ageing, 31:272-275. Kenny RA, Richardson DA, Steen N, Bexton RS, Shaw FE, Bond J (2001) Carotid sinus syndrome: a modifiable risk factor for non-accidental falls in older adults (SAFE PACE), Journal of the American College of Cardiology, 38(5):1491-6. Kerschan-Schindler K, Uher E, Kainberger F, Kaider A, Ghanem AH, Preisinger E (2000) Long-term home exercise program: Effect in women at high risk of fracture, Archives of Physical Medicine and Rehabilitation, 81(3):319-23. Kilpack V, Boehm J, Smith N, Mudge B (1991) Using research-based interventions to decrease patient falls, Applied Nursing Research, 4:50-6. Kind P, Hardman J, Macran S (1999). UK Population Norms for EQ-5D. Centre for Health Economics Discussion Paper 172. University of York King AC, Rejeski J, Buchner D (1998) Physical activity interventions targeting older adults, American Journal of Preventative Medicine, 15(4):316-333.
King,AC.,Haskell,W.,Young,D.,Oka,R.,Stefanick,M (1995) Long-term effects of varying intensities and formats of physical activity on participation rates, fitness and lipoproteins in men and women aged 50-60 years, Circulation; 91(10):154-162. Kingston P, Jones M, Lally F, Crome P (2001) Older people and falls: A randomized controlled trial of a health visitor (HV) intervention, Reviews in Clinical Gerontology, 11(3):209-14. Koh SL, Hafizah N, Lee JY et al. (2009) Impact of a fall prevention programme in acute hospital settings in Singapore. Singapore Medical Journal 50: 425-32. Kong KS, Lee FK, Mackenzie AE, Lee DT (2002) Psychosocial consequences of falling: the perspective of older Hong Kong Chinese who had experienced recent falls, Journal of Advance Nursing, 37:234-42. Koski K, Luukinen, H, Laippala P, Kivela SL. (1998) Risk factors for major injurious falls among the home-dwelling elderly by functional abilities. A prospective population-based study, Gerontology, 44(4):232-238. Krauss MJ, Tutlam N, Costantinou E et al. (2008) Intervention to prevent falls on the medical service in a teaching hospital. Infection Control & Hospital Epidemiology 29: 539-45. Krishna KM,Vancleave RJ (1983) Decrease in the incidence of patient falls in a geriatric hospital after educational programmes [Letter],Journal of the American Geriatric Society, 31(3):187. Krishnan LV, OKane KS, Gill-Body, KM (2002) Reliability of a modified version of the dynamic gait index-a pilot study, Neurol Rep, 26(1):8-14. Kuipers HMM,Hoefnagels WHL,Van Lier HJJ (1993) Reduction of falling of hospitalised patients by using a risk index and taking preventive measures. Nederlands, Tijdschrift voor Geneeskunde, 137(40):2043-8. Kumar BA, Parker MJ (2000) Are hip protectors cost effective? Injury,31: 693695.
Kustaborder MJ, Rigney M (1983) Interventions for safety, Journal of Gerontological Nursing, 9(3):158-62,173,182. Lachman ME, Howland J, Tennstedt S, Jette A, Assmann S, Peterson EW (1998) Fear of falling and activity restriction: the Survey of Activities and Fear of Falling in the Elderly (SAFE), Journals of Gerontology. Series B Psychological Sciences and Social Sciences, 53B(1):43.50 Laferriere RH (1998) Rural research: piloting a tool to identify home care clients risk of falling, Home Care Provider, 3 (3):162169 Laird RD, Studenski S, Perera S, Wallace, D (2001) Fall history is an independent predictor of adverse health outcomes and utilization in the elderly, American Journal of Management Care, 7(12):1133-1138. Lajoie Y, Girard A, Guay M (2002) Comparison of the reaction time, the Berg scale and the ABC in non-fallers and fallers, Archives of Gerontological Geriatrics, 35(3):215-225. Lamb SE, Jrstad-Stein EC, Hauer K et al. (2005) Development of a common outcome data set for fall injury prevention trials: the Prevention of Falls Network Europe consensus. Journal of the American Geriatrics Society 53: 1618-22. Lambert C, Sterbenz KA, Womack DE, Zarrinkhameh LT, Newton RA (2001) Adherence to a fall prevention program among community-dwelling older adults, Physical Occupational Therapy of Geriatrics, 18:18-43. Lamoureux EL, Murphy A, Sparrow A, Newton RU (2003) The effects of progressive resistance training on obstructed-gait tasks in community-living older adults, Journal of Ageing & Physical Activity, 11(1):98-110. Lane AJ (1999) Evaluation of the fall prevention program in an acute care setting. Orthopaedic Nursing 18: 37-43. Latham NK, Anderson CS, Lee A, Bennett DA, Moseley A, Cameron ID (2003) A randomized, controlled trial of quadriceps resistance exercise and
vitamin D in frail older people: The frailty interventions trial in elderly subjects (FITNESS), Journal of the American Geriatric Society, 51:291-9. Latham NK, Stretton C, Ronald M (2001) Progressive resistance strength training in hospitalised older people: a preliminary investigation, New Zealand Journal of Physiotherapy, 29(2):41-8. Lauritzen JB (1996a) Hip fractures: incidence, risk factors, energy absorption, and prevention, Bone, 18:Suppl 1:65S-75S. Lauritzen JB, Hindso K, Petersen MM, Lund B (1996b) Effect of hip protectors in falls on the hip and compliance, Osteoporosis, 331-336. Lauritzen JB, Petersen MM, Lund B (1993a) Effect of external hip protectors on hip fractures, Lancet, 341:11-3. Lauritzen JB, Petersen MM, Lund B (1993b) Hip fractures prevented by external hip protectors. A randomized study, Proceedings of the 4th International Symposium on Osteoporosis, 109-11. Lawrence JI, Maher PL (1992) An interdisciplinary falls consult team: a collaborative approach to patient falls, Journal of Nursing Care Quality, 6:219. Lee V, Ross B, Tracy B (2001) Functional assessment of older adults in an emergency department, Canadian Journal of Occupational Therapy, 68(2):121-131. Legters K (2002) Fear of falling, Physical Therapy, 82(3):264-272. Leipzig RM, Cumming RG, Tinetti ME (1999a) Drugs and falls in older people: A systematic review and meta-analysis: I Psychotrophic drugs, Journal of the American Geriatric Society, 47(1):30-39. Leipzig RM, Cumming RG, Tinetti ME (1999b) Drugs and falls in older people: A systematic review and meta-analysis: II Cardiac and analgesic drugs, Journal of the American Geriatric Society, 47(1):40-50.
Leveille SG, Bean J, Bandean-Roche K, Jones R, Hochberg M, Guranlnik JM (2002) Muscoskeletal pain and risk for falls in older disabled women living in the community, Journal of the American Geriatric Society, 50:671-78. Lichtenstein MJ, Shields SL, Shiavi RG, Burger MC (1989) Exercise and balance in aged women: a pilot controlled clinical trial, Archives of Physical Medicine and Rehabilitation, 70:138-43. Lieu PK, Ismail NH, Choo PW et al. (1997) Prevention of falls in a geriatric ward. Annals of the Academy of Medicine, Singapore 26: 266-70. Lightbody E, Watkins C, Leathley M, Sharma A, Lye M (2002) Evaluation of a nurse-led falls prevention programme versus usual care: a randomized controlled trial, Age & Ageing, 31(3):203-10. Lipsitz LA, Nakajima I, Gagnon M, Hirayama T, Connelly CM, Izumo H, Hirayama T (1994) Muscle strength and fall rates among residents of Japanese and American nursing homes: An international cross-cultural study, Journal of the American Geriatric Society, 42:9-959. Lord SR, Castell S, Corcoran J, Dayhew J, Shan A, Williams P (2003) The effect of group exercise on physical functioning and falls in frail older people living in retirement villages: A randomised,controlled trial, Journal of the American Geriatric Society, 51:1685-92. Lord SR, Dayhew J (2001) Visual risk factors for falls in older people, Journal of the American Geriatric Society, 49(5):508-515. Lord SR, Lloyd DL, Nirui M, Raymond J, Williams P, Stewart RA (1996a) The effect of exercise on gait patterns in older women: a randomized controlled trial, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 51A(2):M64-70. Lord SR, Rogers MW, Howland A, Fitzpatrick, R (2000) Lateral stability, sensori-motor function and falls in older people, Journal of the American Geriatric Society, 47(9):1077-1081.
Lord SR, Ward JA, Williams P (1996b) Exercise effect on dynamic stability in older women: a randomized controlled trial, Archives of Physical Medicine and Rehabilitation, 77:232-6. Lord SR, Ward JA, Williams P, Anstey KJ (1994) Physiological factors associated with falls in older community-dwelling women., Journal of the American Geriatric Society, 42:10-1117. Lord SR, Ward JA, Williams P, Strudwick M (1995) The effect of a 12-month exercise trial on balance, strength, and falls in older women: a randomized controlled trial, Journal of the American Geriatric Society, 43:1198-206. Lowery K, Buri H, Ballard C (2000) What is the prevalence of environmental hazards in the homes of dementia sufferers and are they associated with falls? International Journal of Geriatric Psychiatry, 15(10):883-6. Lundin-Olsson L, Nyberg L, Gustafson Y (2000) The mobility interaction fall chart, Physiother Res Int, 5(3):190-201. Mackenzie L, Byles J, Higginbotham N (2000) Designing the home falls and accidents screening tool (HOME FAST): selecting the items, British Journal of Occupational Therapy, 63(6):260-269 MacRae PG, Asplund LA, Schnelle JF, Ouslander JG, Abrahamse A, Morris C (1996) A walking program for nursing home residents: effects on walk endurance, physical activity, mobility, and quality of life, Journal of the American Geriatric Society, 44(2):175-80. Maeda N, Kato J, Shimada T (2009) Predicting the probability for fall incidence in stroke patients using the Berg Balance Scale. J Int Med Res 37: 697-704. Maki BE (1997) Gait changes in older adults: Predictors of falls or indicators of fear? Journal of the American Geriatric Society, 45:3-320.
Malmivaara A, Heliovaara M, Knekt P, Reunaren A, Aromaa A (1993) Risk factors for injurious falls leading to hospitalisation or death in a cohort of 19,500 adults, American Journal of Epidemiology, 138:384-94. Manning J, Neistadt ME, Parker S (1997) The relationship between fear of falling and balance and gait abilities in elderly adults in a subacute rehabilitation facility, Physical and Occupational Therapy in Geriatrics, 15(2):33-47. Marschollek M, Nemitz G, Gietzelt M et al. (2009) Predicting in-patient falls in a geriatric clinic: a clinical study combining assessment data and simple sensory gait measurements. Zeitschrift fur Gerontologie und Geriatrie 42: 31721. Mayo NE, Gloutney L, Levy AR (1994) A randomized trial of identification bracelets to prevent falls among patients in a rehabilitation hospital. Archives of Physical Medicine & Rehabilitation 75: 1302-8. McCabe F (1985) Mind you dont fall, Nursing Mirror, 160(26):S2-6. McCann M, O'Reilly D, Cardwell C (2009). A Census-based longitudinal study of variations in survival amongst residents of nursing and residential homes in Northern Ireland. Age Ageing. Nov;38(6):711-7 McCarty CA, Fu CL, Taylor HR (2002) Predictors of falls in the Melbourne visual impairment project, Australia and New Zealand Journal of Public Health, 26 (2):116-119. McEwan RT, Davison N, Forster DP, Pearson P, Stirling E (1990) Screening elderly people in primary care: a randomized controlled trial, British Journal of General Practice, 40:94-7. McIntosh A, Hutchinson A, Home PD, Brown F, Bruce A et al. (2002b) Clinical guidelines and evidence review for type 2 diabetes: management of blood glucose, Sheffield: ScHARR, University of Sheffield.
McLean D, Lord S (1996) Falling in older people at home: transfer limitations and environmental risk factors, Australian Occupational Therapy Journal, 43(1):13-8. McMurdo ME, Millar AM, Daly F (2000) A randomized controlled trial of fall prevention strategies in old peoples homes, Gerontology, 46(2):83-7. Mcmurdo ME, Mole PA, Paterson CR (1997) Controlled trial of weight bearing exercise in older women in relation to bone density and falls, British Medical Journal, 314:596. Mcmurdo ME, Rennie L (1993) A controlled trial of exercise by residents of old peoples homes, Age and Ageing, 22(1):11-5. Means KM, Rodell DE, OSullivan PS, Cranford LA (1996) Rehabilitation of elderly fallers: pilot study of a low to moderate intensity exercise program, Archives of Physical Medicine and Rehabilitiation, 77:10306. Menz HB, Sherrington C (2000) The footwear assessment form: a reliable clinical tool to assess footwear characteristics of relevance to postural stability in older adults, Clinical Rehabilitation, 14(6):657-664. Meyer G, Warnke A, Bender R, Muhlhauser I (2003) Effect of on fractures of increased use of hip protectors in nursing homes: cluster randomised controlled trial, British Medical Journal, 326:76-8. Mills EM (1994) The effect of low intensity aerobic exercise on muscle strength, flexibility, and balance among sedentary elderly persons, Nursing Research, 43(4):207-11. Mitchell A, Jones N (1996) Striving to prevent falls in an acute care setting-action to enhance quality. J Clin Nurs 5: 213-20. Mohide EA, Tugwell PX, Caulfield PA, Chambers LW, Dunnett CW, Baptiste S, et al. (1988) A randomized trial of quality assurance in nursing homes, Medical Care, 26:554-65.
Morganti CM, Nelson ME, Fiatarone MA, Dallal GE, Economos CD, Crawford BM, et al. (1995) Strength improvements with one year of progressive resistance training in older women, Medicine and Science in Sports and Exercise, 27(6):906-12. Morris JN, Fries BE, Steel K, Ikegami N, Carpenter R, Bernabel GI, Gilgen R, Hirdes JP, Topinkova E (1997) Comprehensive clinical assessment in community setting: Applicability of the MDS-HC, Journal of the American Geriatric Society, 45(8):10171024. Morse JM, Tylko SJ, Dixon HA (1987). Characteristics of the fall-prone patient. Gerontologist. Aug;27(4):516-22 Morton D (1989) Five years of fewer falls, American Journal of Nursing, 89:204-5. Mulrow CD, Gerety MB, Kanten D, Cornell JE, DeNino LA, Chiodo L, et al. (1994) A randomized trial of physical rehabilitation for very frail nursing home residents, Journal of the American Medical Association, 271(7):519-24. Murphy J, Isaacs B (1982) The post-fall syndrome. A study of 36 elderly patients, Gerontology, 28(4);265-270. Myers H, Nikoletti S (2003) Fall risk assessment: a prospective investigation of nurses' clinical judgement and risk assessment tools in predicting patient falls. International Journal of Nursing Practice 9: 158-65. Najafi B, Aminian K, Loew F, Blanc Y, Robert PA (2002) Measurement of stand-sit and sit-stand transitions using a miniature gyroscope and its application in fall risk evaluation in the elderly. IEEE transactions on biomedical engineering. Journal of the American Geriatric Society, 49(8):843851. Nanda S, Dey T, Gulstrand RE, Jr. et al. (2011) Fall Risk Assessment in Geriatric-Psychiatric Inpatients to Lower Events (FRAGILE). Journal of Gerontological Nursing 37: 22-30.
Nandy S, Parsons S, Cryer C, Underwood M, Rashbrook E, Carter Y, Eldridge S, Close J, Skelton D , Taylor S, Feder G, on behalf of the falls prevention pilot steering group (2004) Development and preliminary examination of the predictive validity of the falls risk assessment tool (FRAT) for use in primary care, Journal of Public Health, 26 (2); 138-143. Naso F, Carner E, Blankfort Doyle W, Coughey K (1990) Endurance training in the elderly nursing home patient, Archives of Physical Medicine and Rehabilitation, 71(3):241-3. National Clinical Guideline Centre (2010). Delirium: Diagnosis, Prevention and Management. London: National Clinical Guideline Centre. Available at www.nice.org.uk/cg103 (accessed 24 August 2012) National Clinical Guideline Centre (2011). The Management of Hip Fracture in Adults. London: National Clinical Guideline Centre. Available at www.nice.org.uk/cg124 (accessed 24 August 2012) National Institute for Clinical Excellence (2001) Information for national collaborating centres and guideline development groups, London: NICE. National Institute for Clinical Excellence (2003) Draft guidance to manufacturers and stakeholders on the methods for NICE appraisal submissions, London: NICE. National Patient Safety Agency (2007). Slips, Trips and Falls in Hospital (online). Available at http://www.nrls.npsa.nhs.uk/resources/?EntryId45=59821 (accessed 24 August 2012) National Patient Safety Agency (2010). Slips, Trips and Falls data update June 2010 (online). Available at http://www.nrls.npsa.nhs.uk/resources/?EntryId45=74567 (accessed 24 August 2012)
Netten A, Bebbington A, Darton R, Forder J, Miles K. (1998) Survey of Care Homes for Elderly People. Final Report. Canterbury, Kent: Personal Social Services Research Unit Netten, A., and Curtis, L (2003) Unit costs of health and social care, Kent: University of Kent Personal Social Services Research Unit. Newbury JW, Marley JE, Beilby J (2001) A randomised controlled trial of the outcome of health assessment of people aged 75 years and over, Medical Journal of Australia, 175:104-7. NHS Information Centre for Health and Social Care (2007). Health Survey for England 2005: Health of Older People (online). Available at: http://www.ic.nhs.uk/statistics-and-data-collections/health-and-lifestylesrelated-surveys/health-survey-for-england/health-survey-for-england-2005:health-of-older-people-%5bns%5d (accessed 24 August 2012) NHS Information Centre for Health and Social Care (2011). National Adult Social Care Intelligence Service (online). Available at https://nascis.ic.nhs.uk/ (accessed 24 August 2012) NHS Information Centre for Health and Social Care (2012). Hospital Episode Statistics (online). Available at http://www.hesonline.nhs.uk/ (accessed 24 August 2012) Nichols JF, Omizo DK, Peterson KK, Nelson KP (1993) Efficacy of heavy resistance training for active women over 60 -muscular strength, body composition,and program adherence, Journal of the American Geriatric Society,41(3):205-10. Nikolaus T, Bach M (2003) Preventing falls in community-dwelling frail older people using a home intervention team (HIT): Results from the randomized falls-HIT trial, Journal of the American Geriatric Society, 51:300-5. Northridge ME, Nevitt MC, Kelsey JL (1996) Non-syncopal falls in the elderly in relation to home environments, Osteoporosis International, 6:249-55.
Northridge ME, Nevitt MC, Kelsey JL, Link B (1995) Home hazards and falls in the elderly: the role of health and functional status, American Journal of Public Health, 85(4):509-15. Nowalk MP, Prendergast JM, Bayles CM, DAmico FJ, Colvin GC (2001) A randomized trial of exercise programmes among older individuals living in two long-term care facilities: the falls FREE program, Journal of the American Geriatric Society, 49(7):859-65. Nyberg L, Gustafson Y (1997) Using the Downton index to predict those prone to falls in stroke rehabilitation, Stroke, 27(10):1821-1824. OBrien K, Pickles B, Culham E (1998) Clinical measures of balance in community-dwelling elderly female fallers and non-fallers, Physiotherapy Canada, 50:212-7,221. OHagan A, Stevenson M, Madan J (2007) Monte carlo probabilistic sensitivity analysis for Patient level simulation models: efficient Estimation of mean and variance using anova. Health Econ. 16: 10091023 (2007) OLoughlin JL, Robitaille Y, Boivin JF, Suissa S (1993) Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly, American Journal of Epidemiology, 137:3-354. Obonyo T, Drummond M, Isaacs B (1983) Domiciliary physiotherapy for old people who have fallen, International Rehabilitation Medicine, 5:157-60. Office for National Statistics (2010). Population Estimates for UK, England and Wales, Scotland and Northern Ireland, mid 2009 (online). Available at http://www.ons.gov.uk/ons/rel/pop-estimate/population-estimates-for-uk-england-and-wales--scotland-and-northern-ireland/2009/index.html (accessed 24 August 2012) Office for National Statistics (2011). Interim life tables 2008-10 (online). Available at http://www.ons.gov.uk/ons/rel/lifetables/interim-life-tables/20082010/index.html (accessed 24 August 2012)
O'Halloran PD, Cran GW, Beringer TR, Kernohan G, O'Neill C, Orr J, Dunlop L, Murray LJ (2004). A cluster randomised controlled trial to evaluate a policy of making hip protectors available to residents of nursing homes. Age Ageing. Nov;33(6):582-8 Oliver D, Daly F, Martin FC et al. (2004) Risk factors and risk assessment tools for falls in hospital in-patients: A systematic review. Age and Ageing 33: 122-30. O'Loughlin JL, Robitaille Y, Boivin JF, Suissa S (1993). Incidence of and risk factors for falls and injurious falls among the community-dwelling elderly. Am J Epidemiol. Feb 1;137(3):342-54 Oman RF, King AC (1998) Predicting the adoption and maintenance of exercise participation using self-efficacy and previous exercise participation rates, American Journal of Health Promotion, 12:154-162. Pakkari J, Heikkila J, Kannus P (1998) Acceptability and compliance with wearing energy-shunting hip protectors: A six-month prospective follow up in a Finnish nursing home, Age and Ageing, 27(2):225-230. Pardessus V, Puisieux F, Di P, Gaudefroy C, Thevenon A, Dewailly P (2002) Benefits of home visits for falls and autonomy in the elderly: A randomized trial study, American Journal of Physical Medicine & Rehabilitation, 81(4):24752. Parker MJ, Anand JK (1991). What is the true mortality of hip fractures? Public Health 105:443446 Parker MJ, Gillespie LD, Gillespie WJ (2001) Hip protectors for preventing hip fractures in the elderly, Nursing Times, 4;97(26):41. Parker MJ, Gillespie LD, Gillespie WJ (2003) Hip protectors for preventing hip fractures in the elderly (Cochrane review), in The Cochrane Library,issue 4, Oxford.
Parker MJ, Handoll HHG, Dynan Y (2002) Mobilisation strategies after hip fracture surgery in adults (Cochrane review), in The Cochrane Library,issue 4, Oxford. Parker MJ, Twemlow TR, Pryor GA (1996) Environmental hazards and hip fractures, Age and Ageing, 25(4):322-5. Parry SW, Steen N, Galloway SR, Kenny RA, Bond J (2001) Falls and confidence related quality of life outcome measures in an older British cohort, Postgraduate Medical Journal, 77(904):103-108. Pathy MSJ, Bayer A, Harding K, Dibble A (1992) Randomised trial of case finding and surveillance of elderly people at home, Lancet, 340:890-3. Peasgood T, Herrmann K, Kanis JA, Brazier JE (2009). An updated systematic review of Health State Utility Values for osteoporosis related conditions. Osteoporos Int. Jun;20(6):853-68 Pereira MA, Kriska AM, Day RD, Cauley JA, LaPorte RE, Kuller LH (1998) A randomized walking trial in postmenopausal women: effects on physical activity and health 10 years later, Archives of Internal Medicine, 158(15):1695701. Perell KL, Nelson A, Goldman RL, Luther SL, Prieto-Lewis N, Rubenstein LZ (2001) Fall risk assessment measures: an analytic review, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 56(12):M761-M766. Pfeifer M, Begerow B, Minne HW, Abrams C, Nachtigall D, Hansen C (2000) Effects of a short-term vitamin D and calcium supplementation on body sway and secondary hyperparathyroidism in elderly women, Journal of Bone and Mineral Research, 15(6):1113-8. Plautz B, Selmar C, Beck DE, Radetsky M (1996) Modifying the environment: a community-based injury-reduction program for elderly residents, American Journal of Preventive Medicine, 12(Suppl 4):33-8.
Ploeg J, Black ME, Hutchison BG, Walter SD, Scott EAF, Chambers LW (1994) Personal, home and community safety promotion with communitydwelling elderly persons: response to a public health nurse intervention, Canadian Journal of Public Health, 85(3):188-91. Podsiadlo D, Richardson S (1991) The timed up and go: A test of basic functional mobility for frail elderly patients, Journal of the American Geriatric Society, 39:142-148. Pomeroy VM, Warren CM, Honeycombe C, Briggs RS, Wilkinson DG,Pickering RM et al. (1999) Mobility and dementia: is physiotherapy treatment during respite care effective? International Journal of Geriatric Psychiatry, 14(5):389-97. Porter EJ (1999) Getting up from here: frail older womens experiences after falling, Rehabilitation Nursing, 24:201-6,211. Posner JD, Gorman KM, Gitlin LN, Sands LP, Kleban M, Windsor L, et al. (1990) Effects of exercise training in the elderly on the occurrence and time to onset of cardiovascular disease, Journal of the American Geriatric Society, 38(3):205-10. Poulstrup A, Jeune B (2000) Prevention of fall injuries requiring hospital treatment among community-dwelling elderly, European Journal of Public Health, 10(1):45-50. Prudham D, Evans JG (1981). Factors associated with falls in the elderly: a community study. Age Ageing. Aug;10(3):141-6 PSSRU (2011) Unit Costs of Health and Social Care 2011. University of Kent. Rainville NG (1984) Effect of an implemented fall prevention program on the frequency of falls, Quality Review Bulletin, 10:287-91. Rantz MJ, Popejoy L, Petroski GF, Madsen RW, Mehr DR, ZwygartStauffacher M, et al. (2001) Randomized clinical trial of a quality improvement intervention in nursing homes, Gerontologist, 41(4):525-38.
Rapport LJ, Webster J, Flemming KL et al. (1993) Predictors of falls among right-hemisphere stroke patients in the rehabilitation setting. Archives of Physical Medicine & Rehabilitation 74: 621-6. Ray WA, Taylor JA, Meador KG, Thapa PB, Brown AK, Kajihara HK, et al. (1997) A randomized trial of a consultation service to reduce falls in nursing homes, Journal of the American Medical Association, 278(7):557-62. Ray WA, Thapa, PB, Gideon P (2000) Benzodiazepines and the risk of falls in nursing residents home, Journal of the American Geriatric Society, 48:6-685. Reinsch S, MacRae P, Lachenbruch PA, Tobis JS (1992) Attempts to prevent falls and injury: a prospective community study, Gerontologist, 32:450-6. Rejeski WJ, Brawley LR, Ettinger W, Morgan T, Thompson C (1997) Compliance to exercise therapy in older participants with knee osteoarthritis: implications for treating disability, Medicine and Science in Sports and Exercise, 29:977-85. Resnick B (1999) Motivation to perform activities of daily living in the institutionalized older adult: can a leopard change its spots? Journal of Advance Nursing, 29:792-9. Resnick B (2001) A prediction model of aerobic exercise in older adults living in a continuing-care retirement community, Journal of Ageing Health, 13:287310. Resnick B (2002) Testing the effect of the WALC intervention on exercise adherence in older adults, Journal of Gerontological Nursing, 28:40-9. Resnick B, Spellbring AM (2000) Understanding what motivates older adults to exercise, Journal of Gerontological Nursing, 26:34-42. Reuben DB, Borok GM, Woldetsadik G, Ershoff DH, Fishman LK, Ambrosini VL, et al.. (1995) A randomized trial of comprehensive geriatric assessment in the care of hospitalized patients. New England Journal of Medicine; 332:134550.
Ritchie C, Wieland D, Tully C, Rowe J, Sims R, Bodner E (2002) Coordination and advocacy for rural elders (CARE): A model of rural case management with veterans, Gerontologist, 42(3):399-405. Rizzo JA, Baker D, McAvay G, Tinetti M (1996) The cost-effectiveness of a multifactorial targeted prevention program for falls among community elderly persons, Medical Care, 34; 9: 954-969 Robbins S, Gouw GJ, McClaran J (1992) Shoe sole thickness and hardness influence balance in older men, Journal of the American Geriatric Society, 40:1089-94. Robertson MC (2001c) Development of a falls prevention programme for elderly people: evaluation of efficacy, effectiveness, and efficiency [PhD thesis], Dunedin, New Zealand: University of Otago. Robertson MC, Devlin N, Gardner MM, Campbell AJ (2001a) Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 1: Randomised controlled trial, British Medical Journal, 322(7288):697-701. Robertson MC, Gardner MM, Devlin N, McGee R, Campbell AJ (2001b) Effectiveness and economic evaluation of a nurse delivered home exercise programme to prevent falls. 2: Controlled trial in multiple centres, British Medical Journal, 322(7288):701-4. Robinson B, Gordon J, Wallentine S (2002) Effectiveness of physical therapy intervention in decreasing the risk for falls in a community-dwelling ageing population, Orthopaedic Nursing, 21(1):55-69. Ross JER, Maas ML, Huston JC, Kundel CJ, Woodward PJ, Gyldenward TA, et al. (1992) Evaluation of two interventions to reduce falls and fall injuries: the challenge of hip pads and individualized elimination rounds, in Funk SG, Tornquist EM, Champagne MT, Weise RA, (editors) Key aspects of elder care: managing falls, incontinence, and cognitive impairment.New York: Springer Publishing Company, 97-103.
Royal College of Nursing (2001) Clinical practice guidelines: pressure ulcer risk assessment and prevention. Recommendations, London: RCN. Rubenstein LZ, Josephson KR, Robbins AS (1994). Falls in the nursing home. Annals of Internal Medicine; 121:44251 Rubenstein LZ, Josephson KR, Trueblood PR, Loy S, Harker JO, Pietruszka FM, et al. (2000) Effects of a group exercise program on strength, mobility, and falls among fall-prone elderly men, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 55(6):M317-21. Rubenstein LZ, Robbins AS, Josephson KR, Schulman BL, Osterweil D (1990) The value of assessing falls in an elderly population. A randomized clinical trial, Annals of Internal Medicine, 113:308-16. Ryan JW, Spellbring AM (1996) Implementing strategies to decrease risk of falls in older women, Journal of Gerontological Nursing, 22(12):25-31. Sackett D, Straus S, Scott Richardson W, Rosenberg W, Haynes B (2000) Evidence-based medicine: how to practice and teach EBM (2nd edition), Edinburgh: Churchill Livingston. Salkeld G, Cameron ID, Cumming RG, et al. (2000). Quality of life related to fear of falling and hip fracture in older women: a time trade off study, British Medical Journal, 320:341-345. Salkeld G, Cumming RG, ONeill E, Thomas M, Szonyi G, Westbury C (2000)The cost effectiveness of a home hazard reduction programme to reduce falls among older persons, Australia and New Zealand Journal Public Health, 24(3):265-271. Sato Y, Manabe S, Kuno H, Oizumi K (1999) Amelioration of osteopenia and hypovitaminosis D by 1alpha-hydroxyvitamin D3 in elderly patients with Parkinsons disease, Journal of Neurology, Neurosurgery and Psychiatry, 66(1):64-8.
Sattin RW, Rodriguez JG, DeVito CA, Wingo PA (1998) Home environmental hazards and the risk of fall injury events among community-dwelling older persons. Study to assess falls among the elderly (SAFE) group, Journal of the American Geriatric Society, 46(6):669-76. Sauvage LR, Myklebust BM, Crowpan J, Novak S, Millington P, Hoffman MD, et al. (1992) A clinical trial of strengthening and aerobic exercise to improve gait and balance in elderly male nursing home residents, American Journal of Physical Medicine and Rehabilitation, 71(6):333-42. Schlicht J, Camaione DN, Owen SV (2001) Effect of intense strength training on standing balance, walking speed, and sit-tostand performance in older adults, Journals of Gerontology. Series A-Biological Sciences & Medical Sciences, 56(5):M281-6. Schmid NA (1990) Reducing patient falls: a research-based comprehensive fall prevention program, Military Medicine, 155:202-7. Schnelle JF, Kapur K, Alessi C, Osterweil D, Beck JG, Al Samarrai NR, Ouslander JG (2003) Does an exercise and incontinence intervention save healthcare costs in a nursing home population? Journal of the American Geriatric Society, 51(2):161-8. Schnelle JF, MacRae PG, Giacobassi K, MacRae HSH, Simmons SF,Ouslander JG (1996) Exercise with physically restrained nursing home residents: maximising benefits of restraint reduction, Journal of the American Geriatric Society, 44:507-12. Schunemann A. HJ, Oxman A. D, Brozek J. et al. (2008) GRADE: grading quality of evidence and strength of recommendations for diagnostic tests and strategies. BMJ 336: 1106-10. Schwendimann R, Buhler H, De GS et al. (2006a) Falls and consequent injuries in hospitalized patients: effects of an interdisciplinary falls prevention program. BMC Health Services Research 6: 69.
Schwendimann R, Milisen K, Buhler H et al. (2006b) Fall prevention in a Swiss acute care hospital setting Reducing multiple falls.[Erratum appears in J Gerontol Nurs. 2006 Jun;32(6):56]. Journal of Gerontological Nursing 32: 1322. Scott V, Votova K, Scanlan A et al. (2007) Multifactorial and functional mobility assessment tools for fall risk among older adults in community, home-support, long-term and acute care settings. [Review] [50 refs]. Age & Ageing 36: 130-9. Scottish Intercollegiate Guidelines Network (2001) A guideline developers handbook, Scotland: Scottish Intercollegiate Guidelines Network Scottish Intercollegiate Guidelines Network. (2002) Prevention and management of hip fracture in older people, Scotland: Scottish Intercollegiate Guidelines Network. www.sign.ac.uk. Scuffham P, Chaplin S(2002) The incidence and costs of accidental falls in the UK. Final Report, York: York Health Economic Form Consortium. The University of York. Scuffham P, Chaplin S, Legood R (2003) Incidence and costs of unintentional falls in older people in the United Kingdom, Journal of Epidemiology and Community Health, 57; 9 740-744. Segui-Gomez M, Keuffel E, Frick KD (2002) Cost and effectiveness of hip protectors among the elderly, International Journal of Technology Assessment in Health Care, 18:1, 55-66. Shaw FE, Bond J, Richardson DA, Dawson P, Steen IN, McKeith IG,et al. (2003) Multifactorial intervention after a fall in older people with cognitive impairment and dementia presenting to the accident and emergency department: randomised controlled trial, British Medical Journal, 326(7380):73-5. Shekelle P, Woolf S, Eccles M, Grimshaw J (1999) Developing guidelines, British Medical Journal, 318:593-596.
Shekelle PG, Maglione M, Chang J, Mojica W, Morton SC, Suttrop MJ, Roth EA, Rhodes S, Wu S-Y, Newberry S (2002) Falls prevention interventions in the medicare population -evidence report and evidenced based recommendations, USA: RAND. Sherrington C, Lord S (1997) Home exercise to improve strength and walking velocity after hip fracture: a randomized controlled trial, Archives of Physical Medicine and Rehabilitation, 78:208-12. Shiffman R (1997) Representation of clinical practice guidelines in conventional and augmented decision tables, Journal Medical Informatics Association, 4(5):382-393. Shumway-Cook A, Gruber W, Baldwin M, Liao S (1997) The effect of multidimensional exercises on balance, mobility, and fall risk in communitydwelling older people, Physical Therapy, 77(1):46-57. Simmons D, Hansen PD (1996) Effectiveness of water exercise on postural mobility in the well elderly: an experimental study on balance enhancement, Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 51A:M233-8. Simpson JM, Darwin C, Marsh N (2003) What are older people prepared to do to avoid falling? A qualitative study in London, British Journal of Community Nursing, 8:152-9. Simpson JM, Mandelstam H (1995) Elderly people at risk of falling: do they want to be taught how to get up again? Clinical Rehabilitation, 9:65-9. Simpson JM, Worsfold C, Reilly E, Nye N (2002) A standard procedure for using TURN180: testing dynamic postural stability among elderly people, Physiotherapy, 88(6):342-353. Sinaki M, Lynn SG (2002) Reducing the risk of falls through proprioceptive dynamic posture training in osteoporotic women with kyphotic posturing: a randomized pilot study, American Journal of Physical Medicine & Rehabilitation, 81(4):241-6.
Skelton DA, Dinan SM (1999) Exercise for falls management: Rationale for an exercise programme aimed at reducing postural instability, Physiotherapy Theory and Practice, 15:10520. Speltz DE, McCausland M, Neumann C, McDonald PE (1987) A risk/falls program, Health Matrix, 5:34-9. Stalenhoef PA, Diedericks JPM, Knottnerus JA, Kester ADM, Crebolder HFJM (2002) A risk model for the prediction of recurrent falls in community-dwelling elderly: A prospective cohort study, Journal of Clinical Epidemiology, 55:10881094. Stead M, Wimbush E, Eadie D, Teer P (1997) A qualitative study of older peoples perceptions of ageing and exercise: the implications for health promotion, Health Education Journal, 56:3-16. Steinberg M, Cartwright C, Peel N, Williams G (2000) A sustainable programme to prevent falls and near falls in community-dwelling older people: results of a randomised trial, Journal of Epidemiology & Community Health, 54(3):227. Stenbacka M, Jansson B, Leifman A, Romelsjo A (2002) Association between use of sedatives or hypnotics, alcohol consumption, or other risk factors and a single injurious fall or multiple injurious falls: a longitudinal general population study, Alcohol, 28(1):9-16. Stenvall M, Olofsson B, Lundstrom M et al. (2007) A multidisciplinary, multifactorial intervention program reduces postoperative falls and injuries after femoral neck fracture. Osteoporosis International 18: 167-75. Stevens M, Holman CD, Bennett N, De Klerk N (2001) Preventing falls in older people: Outcome evaluation of a randomized controlled trial, Journal of the American Geriatric Society, 49(11):1448-55. Stevenson M, Lloyd Jones M, De Nigris E, Brewer N, Davis S, Oakley J (2005). A systematic review and economic evaluation of alendronate,
etidronate, risedronate, raloxifene and teriparatide for the prevention and treatment of postmenopausal osteoporosis. Health Technol Assess; 9(22) Svanstrom L, Ader M, Schelp L, Lindstrom A (1996) Preventing femoral fractures among elderly: the community safety approach, Safety Science, 21(3):239-46. Tennstedt S, Howland J, Lachman M, Peterson E, Kasten L, Jette A (1998) A randomized, controlled trial of a group intervention to reduce fear of falling and associated activity restriction in older adults, Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 53(6):P384-92. Thapa PB, Brockman KG, Fought RL, Ray WA (1996) Injurious falls in nonambulatory nursing home residents: A comparative study of circumstances, incidence, and risk factors, Journal of American Geriatric Society, 44:273-78 The AGREE Collaboration (2001) Appraisal of guidelines for research and evaluation (AGREE) instrument, www.agreecollaboration.org Thompson PG (1996) Preventing falls in the elderly at home: a communitybased program, Medical Journal of Australia, 164:530-2. Thompson RF, Crist DM, Marsh M, Rosenthal M (1988) Effects of physical exercise for elderly patients with physical impairments, Journal of the American Geriatric Society, 36(2):130-5. Thorbahn LD (1998) Value and limitations of the Berg balance test to predict risk of falls in nursing home residents, Ann Long Term Care, 6:49-53. Thorbahn LDB, Newton RA, Chandler J (1996) Use of the Berg balance test to predict falls in elderly persons, Physical Therapy, 76:6-585. Tideiksaar R (1990) Slips, stumbles, and falls: pedestrian footwear and surfaces in: Everett Gray B (editor) The biomedical and environmental characteristics of slips, stumbles, and falls in the elderly, Philadelphia: American Society for Testing Materials, pp.17-27.
Tideiksaar R (1992) Falls among the elderly: a community prevention program, American Journal of Public Health, 82:892-3. Tinetti ME, Baker DI, Gottschalk M, Williams CS, Pollack D, Garrett P, et al. (1999) Home-based multicomponent rehabilitation program for older persons after hip fracture: a randomized trial, Archives of Physical Medicine and Rehabilitation, 80:916-22. Tinetti ME, Baker DI, McAvay G, Claus EB, Garrett P, Gottschalk M, et al. (1994) A multifactorial intervention to reduce the risk of falling among elderly people living in the community, New England Journal of Medicine, 331:821-7. Tinetti ME, Doucette J, Claus E, Marottoli R ( 1995) Risk factors for serious injury during falls by older persons in the community, Journal of the American Geriatric Society, 43:1274-21. Tinetti ME, Inouye SK, Gill TM, Doucette JT (1995) Shared risk factors for falls, incontinence, and functional dependence, Journal of the American Medical Association, 273(17)1348-1353. Topp R, Mikesky A, Dayhoff NE, Holt W (1996) Effect of resistance training on strength, postural control, and gait velocity among older adults, Clinical Nursing Research,5(4):407-27. Topp R, Mikesky A, Wigglesworth J, Holt W, Edwards JE (1993) The effect of a 12-week dynamic resistance strength training program on gait velocity and balance in older adults, Gerontologist, 33(4):501-6. Torgerson DJ, Iglesias CP, Reid DM (2001) The economics of fracture prevention, in The effective management of osteoporosis, London: Aesculapius Medical Press, pp. 111-121. Tromp AM, Pluijm SM, Smit JH, Deeg DJ, Bouter LM, Lips P (2001) Fall-risk screening test: a prospective study on predictors for falls in communitydwelling elderly, Journal of Clinical Epidemiology, 54(8):837-844.
Tromp AM, Smit JH, Deeg DJH, Bouter LM, Lips P (1998) Predictors for falls and fractures in the longitudinal ageing study Amsterdam, Journal of Bone Mineral Research, 13:12-1939. Tynan C, Cardea JM (1987) Home health hazard assessment, Journal of Gerontological Nursing, 13:25-8. Urton MM (1991) A community home inspection approach to preventing falls among the elderly, Public Health Reports, 106(2):192-5. Van Haastregt JC, Diederiks JP, van Rossum E, de Witte LP, Voorhoeve PM, Crebolder HF (2000) Effects of a programme of multifactorial home visits on falls and mobility impairments in elderly people at risk: randomised controlled trial, British Medical Journal, 321:994-8. Van Rossum E, Frederiks CM, Philipsen H, Portengen K, Wiskerke J, Knipschild P (1993) Effects of preventive home visits to elderly people, British Medical Journal, 307:27-32. Van Schoor NM, Smit JH, Pluijm SMF, Jonker C, Lips P (2002) Different cognitive functions in relation to falls among older persons: Immediate memory as an independent risk factor for falls, Journal Clinical Epidemiology, 55(9):855-862. Van Schoor NM, Smit JH, Twisk JWR, Bouter LM, Lips P (2003) Prevention of hip fractures by external hip protectors: a randomized controlled trial, Journal of the American Medical Association, 289(15):1957-62. Van Swearingen JM,Paschal KA,Bonino P,Yang JF (1996) The modified gait abnormality rating scale for recognizing the risk of recurrent falls in community-dwelling elderly adults, Physical Therapy, 76:9-1002. Vass C, Sahota O, Drummond A, Kendrick D, Grainge M, Sach T, Gladman J, Avis M (2013). REFINE-REducing Falls in In-patieNt Elderly using Bed and Chair Pressure Sensors in Acute Hospital Care: A Randomised Controlled Trial. Age and Ageing 2013; 42: ii20ii22
Vassallo M, Poynter L, Sharma JC et al. (2008) Fall risk-assessment tools compared with clinical judgment: an evaluation in a rehabilitation ward. Age & Ageing 37: 277-81. Vassallo M, Vignaraja R, Sharma JC et al. (2004) The effect of changing practice on fall prevention in a rehabilitative hospital: the Hospital Injury Prevention Study. Journal of the American Geriatrics Society 52: 335-9. Vassallo M, Vignaraja R, Sharma JC, Briggs R, Allen S (2005). The relationship of falls to injury among hospital in-patients. International Journal of Clinical Practice. Jan;59(1):17-20 Vassallo M,Sharma JC, Briggs RSJ, Allen S (2003). Characteristics of early fallers on elderly patient rehabilitation wards. Age Ageing; 32(3): 338-342 Vassallo M,Vignaraja R,Sharma JC,Briggs RS,Allen SC (2001) Can intervention prevent falls and injury in geriatric wards? Hospital injury prevention (HIP) study [abstract], Age and Ageing,30(suppl 2):15 Vassallo, M (2000) Fall risk assessment, Cme Bulletin of Geriatric Medicine, 2:3-105. Vellas B, Albarede JL A randomised clinical trial on the value of raubasinedihdroergocristine (Iskedyl) in the prevention of post fall syndrome [Effect de lassociation raubasinedihdroergocristine (Iskedyl) sur le syndrome post-chute et sur la prevention de la chute chez le sujet age], Psychologie Medicale, 23(7):831-9. Verghese J,Buschke H.Viola L,Katz M,Hall C,Kuslansky G, Lipton R (2002) Validity of divided attention tasks in predicting falls in older individuals: A preliminary study, Journal of the American Geriatric Society, 50(9):1572-1576. Vetter NJ, Lewis PA, Ford D (1992) Can health visitors prevent fractures in elderly people? British Medical Journal, 304:888-90. Villar MTA, Hill P, Inskip H, Thompson P, Cooper C (1998) Will elderly rest home residents wear hip protectors? Age and Ageing, 27:195-8.
Von Koch L, Pedro-Cuesta J, Kostulas V, Almazan J, Widen HL (2001) Randomized controlled trial of rehabilitation at home after stroke: one-year follow-up of patient outcome, resource use and cost, Cerebrovascular Diseases, 12(2):131-8. von Renteln-Kruse W, Krause T (2007) Incidence of in-hospital falls in geriatric patients before and after the introduction of an interdisciplinary teambased fall-prevention intervention. Journal of the American Geriatrics Society 55: 2068-74. Wagner EH, LaCroix AZ, Grothaus L, Leveille SG, Hecht JA, Artz K, et al. (1994) Preventing disability and falls in older adults: a population-based randomized trial, American Journal of Public Health,84:1800-6. Wald HL, Glasheen JJ, Guerrasio J et al. (2011) Evaluation of a hospitalistrun acute care for the elderly service. Journal of Hospital Medicine 6: 313-21. Walsh W, Hill KD, Bennell K et al. (2011) Local adaptation and evaluation of a falls risk prevention approach in acute hospitals. International Journal for Quality in Health Care 23: 134-41. Wang JJ, Mitchell P, Smith W, Cumming RG, Leeder SR (2001). Incidence of nursing home placement in a defined community. Med J Aust. Mar 19;174(6):271-5 Ward D, Severs M, Dean T, Brooks N (2003) Care home versus hospital and own home environments for rehabilitation of older people (Cochrane review), in The Cochrane Library, issue 3. Oxford. Watson W, Clapperton A, Mitchell R (2009). The cost of fall injuries among older people in NSW, 200607. NSW Injury Risk Management Research Centre, University of New South Wales White D (1991) Old age is not a reason to fall, Nursing Standard, 6:20-1. Wielandt T, Strong J (2000) Compliance with prescribed adaptive equipment: a literature review, British Journal of Occupational Therapy, 63:65-75.
Williams TA, King G, Hill AM et al. (2007) Evaluation of a falls prevention programme in an acute tertiary care hospital. Journal of Clinical Nursing 16: 316-24. Wolf SL, Barnhart HX, Kutner NG, McNeely E, Coogler C, Xu T (1996) Reducing frailty and falls in older persons: an investigation of Tai Chi and computerized balance training, Journal of the American Geriatric Society, 44:489-97. Wolf-Klein GP, Silverstone FA, Basavaraju N, Foley CJ, Pascaru A, Ma PH (1988) Prevention of falls in the elderly population, Archives of Physical Medicine and Rehabilitation, 69:689-91. Wolfson L, Whipple R, Derby C, Judge J, King M, Amerman P, et al. (1996) Balance and strength training in older adults: intervention gains and Tai Chi maintenance, Journal of the American Geriatric Society, 44:498-506. Woo J,Sum C,Yiu HH,Ip K,Chung L,Lo L (2003) Efficacy of a specially designed hip protector for hip fracture prevention and compliance with use in elderly Hong Kong Chinese, Clinical Rehabilitation, 17:2:203-5. Wood BH, Bilclough JA, Bowron A, Walker RW (2002) Incidence and prediction of falls in Parkinsons disease: a prospective multidisciplinary study, Journal of Neurology Neurosurgery, and Psychiatry, 72(6):721-5. Wortberg W (1998) In-vivo studies with geriatric safety pants for preventing hip fractures in the elderly. (In-vivo-untersuchungen mit einer seniorensicherheitshose zur verhinderung von oberschenkelhalsbruchen bei alteren menschen), Geriatrie Forschung, 8(1):42-6. Yardley L, Smith H (2002) A prospective study of the relationship between feared consequences of falling and avoidance of activity in community-living older people, Gerontologist, 42:17-23. Yates SM, Dunnagan TA (2001) Evaluating the effectiveness of a homebased fall risk reduction program for rural community-dwelling older adults,
Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 56(4):M226-30. Ytterstad B (1996)The Harstad injury prevention study: community based prevention of fall-fractures in the elderly evaluated by means of a hospital based injury recording system in Norway, Journal of Epidemiology and Community Health, 50(5):551-8.
5.7
5.7.1
Acute setting A setting with onsite availability of the full range of diagnostic and therapeutic capabilities needed to diagnose and treat acute physical illnesses. Assessment An in-depth, and possibly ongoing, process of identifying risk factors. Carer Where the term carer is used, this refers to unpaid carers, not paid carers such as care workers. Cognitive impairment A person is defined as having cognitive impairment if they have a mini-mental state examination (MMSE) score of lower than 24 (Folstein et al. 1975). Dementia The diagnostic and statistical manual of mental disorders, fourth edition (DSM-IV 1994), expresses the internationally prevailing view of the concept of dementia being a form of memory disturbance, with at least one of the following disturbances of aphasia, apraxia, agnosia and disturbance in executive functioning. Extended care A care setting such as a nursing home or supported accommodation. Fall An event whereby an individual comes to rest on the ground or another lower level with or without loss of consciousness (AGS/BGS 2001). Home hazard assessment The assessment of an older persons home environment and the identification of any hazards that may contribute to that person being at risk of falling.
Injurious fall A fall resulting in a fracture or soft tissue damage that needs treatment. Multidisciplinary More than one healthcare professional from different disciplines. Multifactorial assessment or multifactorial falls risk assessment An assessment with multiple components that aims to identify a persons risk factors for falling. Multifactorial intervention An intervention with multiple components that aims to address the risk factors for falling that are identified in a persons multifactorial assessment. Non-acute setting A setting focused on recovery and rehabilitation, symptom control or palliative care. Older people In sections 1.1 and 3, older people are people aged 65 years and older. In sections 1.2 and 4, older people are people aged 50 years and older. Older people living in the community Older people living in their own homes or in extended care. Primary prevention Interventions that aim to prevent the first fall in a person who is vulnerable to falling because of, for example, unsteady gait, but who has not yet fallen. Rehabilitation Interventions that are targeted at people who have suffered an injurious fall. Risk prediction tool A tool that aims to calculate a persons risk of falling, either in terms of at risk/not at risk, or in terms of low/medium/high risk, etc.
Secondary intervention Interventions that are targeted at a person who has a history of falls. Self-efficacy A persons perception of their capability. High self-efficacy relates to increased confidence. This term is referred to in relation to reducing an older persons fear of falling. Tailored Intervention packages or programmes that are planned to meet the needs of the particular person. Targeted Interventions that are aimed at modifying a particular risk factor or factors. Please see the NICE glossary for an explanation of terms not described above.
5.7.2
Abbreviations
NCC-NSC NICE
activities of daily living absolute relative risk client assessed protocol confidence intervals Department of Health falls efficacy scale Guideline Development Group home care health technology assessment Incidence rate ratio Medicines and Healthcare Products Regulatory Agency (formerly Medical Devices Agency) National Collaborating Centre for Nursing and Supportive Care National Institute for Clinical Excellence (until March 2013); National Institute for Care Excellence (from April 2013) number needed to treat residential assessment instrument resident assessed protocol Royal College of Nursing randomised controlled trial Ratio of incidence rate ratio Relative risk Ratio of relative risk School of Health and Related Research