Professional Documents
Culture Documents
Van Stan 2019
Van Stan 2019
SPECIAL COMMUNICATION
Abstract
Despite significant advances in measuring the outcomes of rehabilitation interventions, little progress has been made in specifying the therapeutic
ingredients and processes that cause measured changes in patient functioning. The general approach to better clarifying the process of treatment
has been to develop reporting checklists and guidelines that increase the amount of detail reported. However, without a framework instructing
researchers in how to describe their treatment protocols in a manner useful to or even interpretable by others, requests for more detail will fail to
improve our understanding of the therapeutic process. In this article, we describe how the Rehabilitation Treatment Specification System (RTSS)
provides a theoretical framework that can improve research intervention reporting and enable testing and refinement of a protocol’s underlying
treatment theories. The RTSS framework provides guidance for researchers to explicitly state their hypothesized active ingredients and targets of
treatment as well as for how the individual ingredients in their doses directly affect the treatment targets. We explain how theory-based treatment
specification has advantages over checklist approaches for intervention design, reporting, replication, and synthesis of evidence in rehabilitation
research. A complex rehabilitation intervention is used as a concrete example of the differences between an RTSS-based specification and the
Template for Intervention Description and Replication checklist. The RTSS’s potential to advance the rehabilitation field can be empirically tested
through efforts to use the framework with existing and newly developed treatment protocols.
Archives of Physical Medicine and Rehabilitation 2019;100:146-55
ª 2018 by the American Congress of Rehabilitation Medicine
Over the past 50 years, the field of rehabilitation has seen as to which aspects of treatment contribute to patient outcomes
substantial advances in defining and measuring outcomes and and how they contribute.1-3 A significant barrier to identifying
quantifying patient characteristics associated with those outcomes. effective aspects of treatment is the lack of a comprehensive
However, treatment research seldom provides specific information system or framework for defining and describing the interventions
used in rehabilitation.4 Most often, treatments are defined by
either discipline (“X hours of occupational therapy”) or the
problem being treated (“gait training”), neither of which describes
Supported by the Patient-Centered Outcomes Research Institute (contract number ME-1403-
14083) awarded to John Whyte, MD, PhD, FACRM, at Moss Rehabilitation Research Institute and what the clinician actually does to affect functioning. Research
the National Institutes of Health (NIH) National Institute on Deafness and Other Communication reports that include detailed protocols often lack information
Disorders and Office of Behavioral and Social Sciences Research (grant numbers R21 DC016124
and OD016124).
about how a treatment was administered; for example, instead of
Disclosures: none. reporting what quantities of active treatment ingredients were
0003-9993/18/$36 - see front matter ª 2018 by the American Congress of Rehabilitation Medicine
https://doi.org/10.1016/j.apmr.2018.09.112
Improved rehab treatment research 147
provided, treatment dose descriptions simply state the duration or reported and are rarely reported to the level of detail needed for
number of sessions. Even published treatment manuals frequently replication.16
lack sufficient details to enable other researchers to replicate In summary, our ability to characterize rehabilitation treatments
findings or build on previous results or for clinicians to confidently has been challenging due to (1) a lack of clear guidance about
implement published treatments in everyday care. which details are directly related to changes in patient function,
The most comprehensive effort to date to characterize reha- making it difficult to determine what is important for research
bilitation treatments in multiple disciplines was a series of reporting; (2) a tendency to describe treatments by either the type
multicenter practice-based evidence studies that aimed to identify of therapist or the problem that was addressed instead of what was
effective rehabilitation methods for stroke,5,6 joint replacement,7 done in therapy; and (3) a lack of a uniform, standard, cross-
spinal cord injury,8 and traumatic brain injury.9 In these studies, discipline system for describing treatment. To address these chal-
front-line clinicians documented the contents of their treatments lenges, an interdisciplinary team of rehabilitation clinicians and
using point of care (POC) forms that included menus of activities researchers developed the Rehabilitation Treatment Specification
(eg, bed mobility, gait, community mobility) that clinicians could System (RTSS).* The purpose of this article is to describe how the
associate with intervention codes (eg, balance training, motor RTSS can advance the design, reporting, replication, and synthesis
learning, biofeedback).10 Entries listed on the POC forms, of evidence in rehabilitation research. In this issue, Hart et al17
however, suffered from the same limitations as noted for other provide a general introduction to the RTSS. Also, the Manual for
treatment studies: labeling interventions by the targeted impair- Rehabilitation Treatment Specification (which is available at http://
ments (eg, gait training), types of equipment (eg, parallel bars), or mrri/.org/innovations/manual-for-rehabilitation-treatment-specifica
modalities (eg, biofeedback), so the resulting data provide little tion/) describes specification* in detail. (Terms that are asterisked
information about specific actions clinicians performed to achieve when first used have RTSS-specific definitions that are provided in
the targets of treatment. Also, since the POC forms were the glossary included as supplemental appendix S1, available on-
developed by diagnosis- and discipline-specific work groups, line only at http://www.archives-pmr.org/.) We argue that the
cross-discipline and cross-diagnosis differences in labeling and RTSS, which uses a common language and systematic approach to
categorizing treatments obscured any common treatment themes. describing treatment, will offer solutions to the problems noted
To help improve the quality of intervention descriptions in above, encourage collaboration, permit aggregation of data across
clinical research, multiple individuals and committees have disciplines, and foster development of overarching treatment the-
developed reporting guidelines. Examples include the Consolidated ories that inform all of rehabilitation.
Standards of Reporting Trials extension for nonpharmacologic
treatment interventions,11 the Template for Intervention Descrip- The rehabilitation treatment specification
tion and Replication (TIDieR),12 the Guideline for Reporting
Evidence-Based Practice Education Interventions and Teaching,13
system
and the Consensus on Exercise Reporting Template.14 Guidelines Specification, as used here, refers to descriptions of the specific
typically list categories of information that should be described actions that clinicians take to achieve a particular change in
(eg, components of the intervention, procedures for tailoring the patient or client functioning. The RTSS endeavors to specify
intervention to individual patients), but do not explicitly require therapeutic interventions based on the smallest unit of treatment,
authors to identify the aspects of treatment that are thought to bring called a treatment component.* Most treatments are comprised of
about functional change. Reporting guideline authors presume that multiple treatment components, and each treatment component
clinicians and researchers can reliably identify which aspects of a has a tripartite structure*: (1) a singular treatment target,* the
treatment carry the intended effects and that research authors have precise proximal aspect of patient functioning that is to be
a standard method to articulate these aspects in a manner useful to changed by the ingredients provided; (2) one or more ingredients*
others. Because standard reporting methods are agnostic to thera- that the therapist selects or performs to achieve the target; and
pist actions, intervention describers could satisfy reporting guide- (3) a mechanism of action,* the causal chain through which the
line requirements by including anything that occurred during treatment is known or hypothesized to work (ie, how the
therapy regardless of its significance in achieving the desired ingredients affect the target).
change in patient functioning (eg, the color of the therapist’s scrubs The RTSS postulates that there are 3 broad groups of treatment
or the temperature of the room) while omitting the clinician’s components: Organ Functions,* Skills and Habits,* and Repre-
actions that are crucial (eg, the instructional methods used, how sentations.* Organ Functions treatment components are concerned
practice was structured, the type of feedback used). For example, with changes in the efficiency, functioning, or replacement of an
use of external memory aids is a recommended practice for patients organ or organ system (eg, exercise, habituation, prosthetics).
with traumatic brain injury because of strong evidence that using Skills and Habits components involve modifying mental or
these aids can improve relevant patient outcomes.15 Studies have behavioral skills through providing ingredients such as practice,
identified patient characteristics associated with successful device repetition, feedback, etc. Representations18,19 components are
use (eg, adequate dexterity and vision) and general principles of intended to change mental representations related to cognition,
intervention (eg, that it is individualized and includes practice), but affect, motivation, and volitional behavior. Table 1 outlines a few
methods for teaching patients to use these aids are inconsistently examples of each treatment component category from selected
common rehabilitation treatments.20-30
The RTSS also defines the concept of treatment aims,* which are
List of abbreviations:
distal (“downstream”) effects of treatment that may or may not
POC point of care result from achieving a single, or even multiple, targets. For
RTSS Rehabilitation Treatment Specification System
example, if an aim is to reduce frequency of falling, therapy might
TIDieR Template for Intervention Description and
include multiple targets that are thought to contribute to that aim
Replication
(eg, improved balance, increased use of fall prevention strategies,
www.archives-pmr.org
148
Table 1 Treatment group examples using the RTSS
Group Treatment Example Target for Treatment Ingredient(s) for Target
Organ Functions Prism adaptation training20 Improved line bisection or circle crossing 20-diopter, right-shifting by 11.4 degrees, goggle-mounted,
on the patient’s left side wedge prism lenses
Mallet finger treatment21 Increased passive range of motion of the Graded protective mobilization after 4 weeks of joint
distal interphalangeal joint immobilization via splinting
Circumlaryngeal manual therapy22 Decreased tension in suprahyoid muscles Apply light circular pressure via the thumb and middle finger
on the posterior horns of the hyoid bone, within the
thyrohyoid space, and posterior borders of the thyroid
cartilage.
Exercise training23 Increased strength in bilateral knee extensors 1-2 sets of 6-8 repetitions at 65% of the patient’s 1-repetition
maximum voluntary contraction.
Skills and Habits Constraint-induced movement therapy24 Improved accuracy in a functional reaching Perform a certain number of reaching repetitions with the more
task with the more impaired limb impaired limb while the less impaired limb is restrained
Cognitive Orientation to Occupation Formation of habit: use of the 4-step global Provide goal sheets for each goal trained in the session, which
Performance25 strategy “Goal, Plan, Do, Check” are to be completed in daily life to document successful or
nonsuccessful implementation of a plan.
Resonant Voice Therapy26 Formation of habit: use of forward focus resonance Perform a certain number of voicing trials, at various difficulty
levels (vowels, syllables, speech, etc.), with subsequent
clinician feedback on correctness.
Physical activity and fall prevention Formation of habit to increase physical activity Provide a practice schedule and provide feedback on daily
intervention27 activity with a Fitbit
Representations Frazier free water protocol28 Increased knowledge of oral hygiene protocol Oral and written information was provided to the patient
regarding how to properly perform oral hygiene.
Motivational interviewing29 Increased positive attitude toward behavior change Gather motivation-related information, cue patient to think
about meaning of change, and elicit change talk
Anger self-management training30 Modified beliefs regarding anger Verbally describe anger as a normal adaptive emotion and
discuss the specifics of the patient’s anger response in this
normalized context
Physical activity and fall prevention Patient to engage in recommended physical Tell the patient about specific individually tailored
intervention activity on a daily basis as directed opportunities to increase physical activity in the community
and provide website for searching for local active
www.archives-pmr.org
opportunities
increased leg strength). The distinction between targets and aims is action / target. Existing guidelines mainly encourage
critical because it explicitly differentiates changes in function that are researchers and research consumers to think critically about the
expected to result directly from an intervention from those that occur aspect of patient function they are trying to change with a
indirectly because of changes in one or more aspects of function. particular intervention but do not encourage hypothesis develop-
Another important emphasis in the RTSS is the concept of ment regarding how the ingredients directly or indirectly create
volition,* which can be roughly equated with effort expended by that change. For example, “improved gait” would need to be more
the treatment recipient.* It is important to consider volition in a specific when considered as a target in the RTSS. This is because
treatment specification system because the success of many one would likely use different ingredients to affect targets like
rehabilitation treatments depends on the patient’s voluntary “improved gait speed” versus “improved gait symmetry.”
actions as elicited, if necessary, by clinician actions31 (eg, in goal However, the TIDieR guidelines do not explicitly provide a means
setting,32 establishing rapport,33 using shared decision-making34). for asking key questions such as “What am I trying to change in
The RTSS posits that there are 2 sources of treatment success or the patient’s functioning with these specific ingredients?” or
failure for volitional treatments: (1) the degree to which “Which ingredients drive the effects demonstrated in the study?”
ingredients chosen affect the selected treatment target (eg, do By contrast, the RTSS-based box 1 specification allows the
tongue-hold swallow exercises decrease residual food in the answers to these questions to be expressed in an empirically
epiglottic/base-of-tongue valleculae after swallowing?); and testable manner. Making the treatment components and targets
(2) the degree to which ingredients result in the patient’s perfor- explicit allows researchers to generate informed hypotheses about
mance of the therapeutic activity as directed (eg, do instructional/ why their treatment worked or did not work. In reference to box 1,
motivational ingredients improve the probability that the patient if patients did not decrease their frequency of falls after the
will perform the tongue-hold exercise program the prescribed intervention (an aim of the treatment), was it because the practice
number of times per day, with correct execution?). That is, the schedule prescribed was insufficient to increase automaticity in
RTSS encourages clinicians and researchers to consider both the performing fall prevention strategies (target in row 4), because the
ingredients that are thought to change the patient’s functioning clinician’s explanation regarding the importance of using fall
and the ingredients that enhance the likelihood that the prescribed prevention strategies did not influence the patient’s volitional
therapeutic activities are done correctly. behavior enough to practice the strategies as directed (target in
row 5), or because the intervention failed to address other targets
Potential benefits of the RTSS for that are closely related to the aim of fall prevention (eg, prevention
of orthostatic hypotension)? Answers to these questions could
conducting research and disseminating direct the researcher to revise the approach to fall prevention
findings (eg, add more opportunities for practice) or add more treatment
ingredients to increase the likelihood the participants practice
Example of RTSS specification using the prescribed strategies (eg, add phone calls to query the
patient on their at-home practice). This closely relates to the issue
To illustrate the potential benefits of the RTSS for research, we of treatment appropriateness vs treatment adherence that is
chose an article by Tiedemann et al.27 We chose this article because frequently encountered in effectiveness studies.35-37 Treatment
it provided the most comprehensive description of a multicompo- appropriateness refers to whether the researcher-selected
nent rehabilitation intervention in our searches for publications that ingredients are likely to have a clinically meaningful effect on
used the TIDieR checklist. Box 1 shows the TIDieR elements the desired change in patient functioning, while treatment adher-
provided by Tiedemann27 for intervention as well as added treat- ence refers to whether the selected ingredients ever had a chance
ment information that was in the main text but missing from that to change patient behavior (ie, did the patient engage in the
article’s TIDieR table. Based on the RTSS guidelines, we identified therapeutic activities that would effect the desired change?).
treatment components in the protocol (table 2). As the TIDieR Use of the RTSS can improve the design of an intervention at
checklist does not ask for a tripartite structure, we used the RTSS the initial stages of protocol development. Before applying the
rules to guide our identification of treatment components and link- research treatment protocol to patients, protocol developers could
ing of ingredients with a target. In some cases, entire treatment use the RTSS framework to guide the development of their
components (numbers 4 and 5) were inferred because the TIDieR treatment methods, phrasing such questions as follows: should the
table listed fall prevention strategies that were not further explained protocol have additional ingredients for associated targets, do the
or described in the main text. We realize that the authors may have a ingredients match the target(s), should the protocol have addi-
more extensive protocol for use by their therapists, but we based our tional targets for associated aims, and how will the ingredient
analysis of the case example on what they published. Dosing dosages, targets, and aims be measured? Once the research is
parameters* for most treatment ingredients were missing from the completed, if treatment effects are weak, the stated relationships
TIDieR table, so we hypothesized parameters for individual treat- among ingredients, targets, and aims will be specific enough that
ment ingredients to complete the example. Note that these param- researchers can decide if (1) future studies require larger doses of
eters do not reflect the actual protocol implemented in the cited current ingredients or (2) future studies require different treatment
investigation but are included to provide examples of how relevant ingredients for specific targets or (3) additional targets (with
parameters would be articulated. All treatment components and associated ingredients) need to be added or (4) underlying theories
information we added are italicized in table 2. regarding the connections among ingredients, targets, and aims are
incorrect.38 Many rehabilitation treatments are considered
Design of experimental interventions complex interventions39 (ie, they contain multiple interacting
treatment components addressing different behaviors that can be
The most obvious benefit of the RTSS is explication of the 3 difficult to measure), so development of a set of theories
aspects of a treatment component: ingredient(s) / mechanism of connecting specific ingredients with their respective targets will
www.archives-pmr.org
150 J.H. Van Stan et al
Box 1 Intervention description using the TIDieR checklist provided by Tiedemann et al27
1. Brief name
Combined physical activity promotion and fall prevention intervention enhanced with health coaching and pedometers to increase
older adults’ physical activity levels and mobility-related goals.
2. Why
Physical inactivity and falls in older people are important public health problems. Health conditions that could be ameliorated with
physical activity are particularly common in older people. One in 3 people aged 65 years and over fall at least once annually, often
resulting in significant injuries and ongoing disability. These problems need to be urgently addressed, as the population proportion
of older people is rapidly rising.
3. What materials
Participants will receive:
The “Staying Active and On Your Feet” fall prevention booklet developed by the New South Wales Ministry of Health.
An assessment of their fall risk factors using the QuickScreen fall risk assessment.
A pedometer enhanced with a web interface (Fitbit) to give feedback on the amount of daily physical activity achieved. The FitbitsTM
will also be provided as a motivational tool to encourage ongoing physical activity participation.
4. What procedures
Telephone- or email-based health coaching will be used to identify barriers and facilitators to physical activity participation and to
provide education and support to assist participants in reducing their risk of falling and achieving their physical activity goals. The
health coach will . monitor and facilitate progress towards physical activity goals and assist participants to overcome any partici-
pation barriers that arise. Participants will be encouraged to wear the pedometer during waking hours on a daily basis for the whole
6 month intervention period to record their daily steps and provide feedback and motivation to increase their physical activity
participation. Participants will be encouraged to synchronise and download their data on a weekly basis or more often if desired. During
the home visit to implement the intervention, participants will be taught how to use the FitbitTM device and the associated internet
based feedback and monitoring technology. The research team will have access to all intervention participants’ FitbitTM data and will
monitor individual adherence with the intervention.
5. Who provided
Three health coaches with professional backgrounds in physiotherapy will deliver the intervention.
6. How
The fall risk assessment and tailored fall prevention and physical activity plan will be delivered during 1 face-to-face interview.
Health coaching will be delivered via telephone or email contact.
7. Where
The intervention will be delivered to community-dwelling people in Sydney, Australia, and the surrounding area.
8. When and how much
The face to face assessment and interview will occur at the beginning of the intervention period and will last for approximately
2 hours. The telephone-based health coaching will occur after the face-to-face assessment and interview once every 2 weeks for
approximately 20 minutes for a total duration of 6 months.
9. Tailoring
The fall prevention aspect of the intervention will be tailored to individual need with reference to the fall risk assessment results.
The physical activity plan will be tailored to participant goals, current physical ability, and preferences. Health coaches will also
enquire about the circumstances of any falls that participants may have experienced and they will discuss strategies for reducing the
risk of future falls. Intervention participants will also be assisted to find suitable local exercise opportunities (e.g. Tai Chi, balance and
strength training) that will be identified using the NSW Ministry of Health’s Active and Healthy online database (http://www.
activeandhealthy.nsw.gov.au/). If participants have not uploaded their FitbitTM data to their computer or internet-connected tablet
device in the past week, during the fortnightly contact their health coach will enquire about any problems encountered with the
pedometer and they will encourage participant compliance with the intervention protocol.
NOTE. Assessment has been identified with strikethrough font (because the RTSS does not address assessment), and added information
from the article’s narrative description has been identified with italic font. All italicized text is directly quoted from page 3, section
“Intervention Group,” of the Tiedemann article. TIDieR items 10 (modifications), 11 (how welleplanned), and 12 (how welleactual)
are not addressed by Tiedemann and are omitted here.
allow researchers to determine the sequence and combination of outcome of primary importance and additional secondary effects,
treatment components that optimize outcomes. the concept of treatment components directs him or her to identify
the outcomes that can be hypothetically achieved as a direct result
Selection of measures of the ingredients provided. The article by Tiedemann27 lists 3
primary outcomes of that intervention (physical activity, as
The RTSS concepts of targets and aims have advantages over the measured by ActiGraph over 7 days, and 2 individualized physical
current measurement focus on primary and secondary outcomes. activity goals based on goal attainment scaling) and 5 secondary
Specifically, instead of the researcher solely considering the outcomes (reduced falls and patient-reported measures of quality
www.archives-pmr.org
www.archives-pmr.org
151
152 J.H. Van Stan et al
of life, fear of falling, mood, and mobility). However, as is dosage with reference to theoretically important dimensions of the
common in studies of rehabilitation interventions, the outcome ingredient (eg, tension and duration for soft tissue stretch vs
measures do not match all of the listed targets: the 3 primary practice schedule for skill development). This is a significant
outcome measures would quantify only the row 1 target in box 1, conceptual and practical advance over the typical practice of
and all secondary outcomes appear to be aims that will require describing dose as total time or number of sessions, which con-
changes across multiple targets not directly addressed by this veys little meaningful information about the individual ingredients
treatment. This leaves all other targets (box 1 rows 2 through 5) provided.40-42 This practice is exemplified by TIDieR checklist
without an explicitly stated outcome measure. item 8: “Describe the number of times the intervention was
The RTSS can help researchers choose appropriate outcome delivered and over what period of time including the number of
measures. As the treatment target is an aspect of patient func- sessions, their schedule, and their duration, intensity, or
tioning to be directly changed by the ingredients, proof-of-concept dose.”12(p.6) The resulting TIDieR description does not tell readers
research should use a primary outcome measure aligned with which ingredients were provided during specific treatment ses-
change in that target. If the impact of treatment is weak on a sions, how much of each ingredient was provided per session
primary outcome measure aligned with a target, researchers (dose), and what the target was for which ingredients (eg, see our
should consider one or both of two possibilities: (1) the selected reworking of the Tiedemann27 protocol in box 1). The actual
outcome measure is a correct representation of the target, but the measurement of ingredient dosages is not necessarily a straight-
ingredients are not right, or are administered in insufficient dose; forward endeavor,43,44 especially ingredients for Skills and Habits
and (2) the selected outcome measure is either an aim (rather than targets and Representations targets that include notions such as the
a target) or something else weakly related to the target, and number of opportunities to practice, amount of feedback provided,
another primary outcome must be selected or developed. Later in type and amount of information conveyed in educational mate-
the progress of a translational research agenda, it may be impor- rials, and goal-setting parameters. Using the RTSS to specify
tant to address broader and more clinically meaningful aims of dosage, even if that specification is incomplete, will help unpack
treatment.38 Often, effective treatment of one target in isolation is the black box of rehabilitation treatments and ultimately improve
not sufficient to result in a meaningful functional impact on the replicability of research interventions.
broader treatment aims. If this is the case, or can be anticipated, The specification of ingredients and their doses is critically
the researcher should consider one or more of these possibilities: important for the concept of treatment progression, which is a key
(1) the ingredient dosage for one or more targets is not optimal feature of many treatments. The term progression refers to the
(eg, too much, too little); (2) the link between the treated targets clinician following a predetermined schedule of varying the quan-
and the clinical aim is too weak; (3) one or more targets that are tity of ingredients over time to change function by increasing
crucial to achievement of the aim have not been addressed physical or cognitive task difficulty. Like dosage in general, pro-
(ie, more causal links in the researcher’s theory of what targets gression is frequently underspecified. For example, in a recent
affect the aim are needed); and (4) a more narrow population of systematic review of intervention descriptions in exercise for breast
patients should be selected for this treatment, specifically those cancer survivors, only 29% of studies described progression of dose
patients where the treated target or targets are strongly linked to or intensity over the course of treatment.45 Furthermore, reporting of
the clinical aim. For example, strengthening the leg muscles may progression parameters did not improve over the time period be-
contribute to independent ambulation, but ambulation may also be tween the authors’ original systematic review in 2012 (Campbell
affected by factors such as balance deficits. Thus, a strengthening et al46) and their 2016 (Neil-Sztramko et al45) updated review.
treatment may be effective in terms of a change in the strength Lack of specificity regarding targets, ingredients, and doses is a
target but ineffective in achieving independent ambulation major barrier to the transition of knowledge into clinical practice.
(a possible aim of treatment). One could potentially achieve the Because ingredients are not explicitly identified or described in a
aim of ambulation by adding an effective treatment targeting standard fashion in published treatments (especially the ingredients
balance or by selecting only patients with weakness (but good that were vital to the reported improvements in patient functioning),
balance skills) for the strengthening treatment. Use of the RTSS front-line clinicians often struggle to pinpoint what he or she should
will help reveal problems like this by focusing the investigator on be doing to accurately implement these treatments. The RTSS
the specific functions that are targets of treatment and the clinician directly addresses this problem because research reporting would
actions that can directly affect the target. identify and describe ingredients in a standardized manner and tie
them directly to specific targets. Therefore, if a clinician has a pa-
Reporting, replication, and clinical translation tient who needs to be more physically active, box 1 row 1 can point
to specific treatment ingredients that may help achieve that target
Replication of treatment protocols for purposes of scientific (eg, the number and schedule of practice sessions needed to develop
validation requires that they are reported with sufficient specificity a habit, provision of a step-monitoring smartphone app). If the pa-
to allow those not involved in the research to implement them in tient is not pursuing opportunities to increase physical activity in the
their local setting. Reporting checklists are meant to help replication community, box 1 row 2 can point to specific treatment ingredients
by providing detailed descriptions of treatment protocols, but as that may help increase the probability that the patient will engage in
noted above, the type of details required by these checklists may not the recommended activity (eg, provide the patient with an individ-
improve either replicability or everyday clinical implementation of ually tailored list of activity opportunities in their community or
study protocols. As stated earlier, it is not simply more detail that is written materials on the dangers of inactivity).
needed but detail regarding administration of the active ingredients
and treatment dosage as well as clear identification of the function Evidence synthesis and meta-analyses
that these ingredients are hypothesized to change.
When researchers are identifying treatment ingredients for a Meta-analysis of clinical trials requires that all included studies are
given target, the RTSS requires them to also specify the ingredient the same or at least very similar with respect to population,
www.archives-pmr.org
Improved rehab treatment research 153
intervention, comparator, outcome, time after intervention, and specification system. Once the field can adequately describe its
setting of care.47 The terms intervention and comparator (the interventions, the development of treatment labels representing a
comparison intervention) would include ingredients as defined by host of ingredients and their associated targets becomes possible,
the RTSS, and outcome would correspond to measures of either which would be beneficial for meta-analysis.
targets or aims. Meta-analysis authors judge similarity of in-
gredients and outcome measures initially qualitatively, and any
presumed similarities are evaluated using a measure of heteroge- Next steps for implementing the
neity, such as I2. However, only infrequently does a high I2 lead to rehabilitation treatment specification
the decision not to perform a meta-analysis; typically the meta-
analysts judge that apples and oranges can be combined. The
system
RTSS emphasizes that substantial heterogeneity is hiding under the In this article, we described the major concepts of the RTSS and
labels we put on rehabilitation interventions (eg, memory therapy, the potential benefits of adopting this system for intervention
gait treatment), including in specific targets and in the nature and reporting, replication, knowledge translation, and evidence syn-
quantity of the ingredients used to achieve those targets. thesis. The RTSS provides a theory-based framework that is useful
The current approach to describing treatments is a significant across disciplines and diagnoses, provides a standard procedure
obstacle to evidence synthesis and meta-analysis because it groups for identifying treatment components, and links ingredients to
treatments that differ in its targets and active ingredients and fails to specific targets to facilitate investigation of the mechanisms by
group treatments that are very similar in targets and active in- which ingredients cause changes in a target (ie, mechanisms of
gredients that go by different names. Without a clear description of action). Categorization of treatments into 3 treatment groups
a treatment’s active ingredients, it is difficult to ensure that repli- (Organ Functions, Skills and Habits, Representations) helps
cation attempts actually deliver the same ingredients as the original emphasize commonalities across rehabilitation disciplines, and
study. As an example, grouping together executive function in- their use could have a significant positive effect on evidence
terventions for meta-analysis would be inappropriate, as recent synthesis. The recognition of the importance of volitional
clinical practice guidelines note at least 2 different subtypes, each behavior provides opportunities to assess the extent to which
with different targets: metacognitive strategy instruction targets the observed outcomes are driven by effects of the ingredients on the
consistent, accurate use of a strategy, and use of alerting or target vs the successful delivery of ingredients themselves
prompting aids targets the correct performance of a specific task.48 (ie, patient adherence). Finally, the distinction between targets and
Viewing these 2 types of executive function interventions through aims moves the field toward a closer match of outcome measures
the RTSS lens illustrates several benefits of using this system: to the targets of the therapy provided, which could help pose
(1) although these 2 treatments have different names, both meta- testable theory-based questions regarding whether and how
cognitive strategy instruction and use of alerting or prompting aids treatments exert their effects.
require ingredients related to practice or habit formation; and What would it take to successfully implement the RTSS in
(2) both treatments are likely to require ingredients to increase the rehabilitation research reporting? Practically, funding agencies
patient’s effort (the likelihood of using the strategy or aid as and journal editors would need to require that authors adopt the
directed). Thus, specification of treatments using the RTSS can shed specification system in their treatment grant proposals or manu-
light on the differences and commonalities amongst treatments to script submissions. However, we must first empirically demon-
guide appropriate grouping for analysis and potentially reduce the strate the potential benefits of the RTSS for research reporting. In
number of studies necessary to establish treatment benefits. collaboration with an advisory board consisting of rehabilitation
The grouping of targets into 3 categories (Organ Functions, stakeholders, we have initiated implementation projects that focus
Skills and Habits, Representations) is relevant to all rehabilitation on 2 main knowledge translation steps: (1) collaboration between
disciplines, and viewing targets from this perspective can enable RTSS specialists and developers of rehabilitation treatment
evidence synthesis across a wide variety of interventions and protocols to examine the impact and value of RTSS application
disciplines. These broad categories could facilitate novel questions and (2) development and implementation of training to help future
like “How are ingredient doses related to the achievement of users of the RTSS acquire skill in applying the framework.
habits across a range of different behaviors?” and “What The collaborative process to specify treatment protocols using
ingredients are associated with improved volitional engagement in the RTSS will entail a back-and-forth between the RTSS
a wide range of treatments?” For example, the skills of speaking specialists and treatment developers. An iterative approach is
with better voice quality, walking with a cane, and using adapted needed because identifying treatment components, discriminating
utensils all require the clinician to provide opportunities for the targets vs aims, and describing ingredients and targets require both
patient to practice (an ingredient from the Skills and Habits skill in using the RTSS rules and concepts (which the RTSS
group). In other words, use of the tripartite structure and target specialists have acquired) and knowledge of the hypothesized
groups of the RTSS might reveal general principles that govern relationship between ingredients and targets (which requires the
treatments across a wide variety of Skills and Habits targets and content expertise of the treatment developer). Creating and
Representations targets. Therefore, the effort to specify treatments refining the specification of research protocols will provide
using the RTSS can have major benefits in aggregating/integrating opportunities for qualitative assessments of the value of the RTSS.
information across studies, ultimately providing a stronger Protocol developers and clinicians who would use their protocols
evidence base. can be directly asked how the 2 treatment descriptions (original
Adoption of the RTSS does not guarantee that researchers will protocol vs its RTSS specification) differ in terms of replicability,
use the same treatments in their studies. However, the RTSS can opportunity for assessment of fidelity, and clarity of imple-
improve the field’s knowledge of what treatment type is being mentation instructions. RTSS specifications of research protocols
provided and how the treatment varies from study to study, which could be immediately impactful by allowing front-line clinicians
is currently impossible due to the lack of a standardized increased insight into how the protocols may be adapted for their
www.archives-pmr.org
154 J.H. Van Stan et al
www.archives-pmr.org
Improved rehab treatment research 155
educational interventions and teaching (GREET). BMC Med Educ results of a randomized controlled trial. J Head Trauma Rehabil 2017;
2016;16:237. 32:319-31.
14. Slade SC, Dionne CE, Underwood M, Buchbinder R. Consensus on 31. Whyte J, Dijkers MP, Hart T, et al. The importance of voluntary
Exercise Reporting Template (CERT): explanation and elaboration behavior in rehabilitation treatment and outcomes. Arch Phys Med
statement. Br J Sports Med 2016;50:1428-37. Rehabil 2019;100:156-63.
15. Sohlberg MM, Kennedy M, Avery J, et al. Evidence-based practice for 32. Wade DT. Goal setting in rehabilitation: an overview of what, why and
the use of external aids as a memory compensation technique. J Med how. Clin Rehabil 2009;23:291-5.
Speech Lang Pathol 2007;15:xv-li. 33. Sliwa JA, Makoul G, Betts H. Rehabilitation-specific communication
16. Powell LE, Glang A, Ettel D, Todis B, Sohlberg MM, Albin R. skills training: improving the physician-patient relationship. Am J
Systematic instruction for individuals with acquired brain injury: Phys Med Rehabil 2002;81:126-32.
results of a randomised controlled trial. Neuropsychol Rehabil 2012; 34. Joosten EA, DeFuentes-Merillas L, de Weert GH, Sensky T, van der
22:85-112. Staak CP, de Jong CA. Systematic review of the effects of shared
17. Hart T, Dijkers MP, Whyte J, et al. A theory-driven system for the decision-making on patient satisfaction, treatment adherence and
specification of rehabilitation treatments. Arch Phys Med Rehabil health status. Psychother Psychosom 2008;77:219-26.
2019;100:172-80. 35. Patti F. Optimizing the benefit of multiple sclerosis therapy: the
18. Johnson-Laird PN. Mental models in cognitive science. Cogn Sci importance of treatment adherence. Patient Prefer Adherence 2010;4:
1980;4:71-115. 1-9.
19. Kragel PA, Koban L, Barrett LF, Wager TD. Representation, pattern 36. Lester RT, Ritvo P, Mills EJ, et al. Effects of a mobile phone short
information, and brain signatures: from neurons to neuroimaging. message service on antiretroviral treatment adherence in Kenya
Neuron 2018;99:257-73. (WelTel Kenya1): a randomised trial. Lancet 2010;376:1838-45.
20. Goedert KM, Chen P, Foundas AL, Barrett AM. Frontal lesions predict 37. Jack K, McLean SM, Moffett JK, Gardiner E. Barriers to treatment
response to prism adaptation treatment in spatial neglect: a rando- adherence in physiotherapy outpatient clinics: a systematic review.
mised controlled study. Neuropsychol Rehabil 2018:1-22. Man Ther 2010;15:220-8.
21. Devan D. A novel way of treating mallet finger injuries. J Hand Ther 38. Whyte J. Contributions of treatment theory and enablement theory to
2014;27:325-8. rehabilitation research and practice. Arch Phys Med Rehabil 2014;
22. Roy N, Leeper HA. Effects of the manual laryngeal musculoskeletal 95(Suppl 1):S17-23.e2.
tension reduction technique as a treatment for functional voice 39. Hart T. Treatment definition in complex rehabilitation interventions.
disorders: perceptual and acoustic measures. J Voice 1993;7:242-9. Neuropsychol Rehabil 2009;19:824-40.
23. Binder EF, Brown M, Sinacore DR, Steger-May K, Yarasheski KE, 40. Waterschoot FP, Dijkstra PU, Hollak N, de Vries HJ, Geertzen JH,
Schechtman KB. Effects of extended outpatient rehabilitation after hip Reneman MF. Dose or content? Effectiveness of pain rehabilitation
fracture: a randomized controlled trial. JAMA 2004;292:837-46. programs for patients with chronic low back pain: a systematic review.
24. Winstein CJ, Miller JP, Blanton S, et al. Methods for a multisite Pain 2014;155:179-89.
randomized trial to investigate the effect of constraint-induced 41. Hayward KS, Brauer SG. Dose of arm activity training during acute
movement therapy in improving upper extremity function among and subacute rehabilitation post stroke: a systematic review of the
adults recovering from a cerebrovascular stroke. Neurorehabil Neural literature. Clin Rehabil 2015;29:1234-43.
Repair 2003;17:137-52. 42. Cifu DX, Stewart DG. Factors affecting functional outcome after
25. Dawson DR, Gaya A, Hunt A, Levine B, Lemsky C, Polatajko HJ. stroke: a critical review of rehabilitation interventions. Arch Phys Med
Using the cognitive orientation to occupational performance (CO-OP) Rehabil 1999;80(Suppl 5):S35-9.
with adults with executive dysfunction following traumatic brain 43. Zbogar D, Eng JJ, Miller WC, Krassioukov AV, Verrier MC. Move-
injury. Can J Occup Ther 2009;76:115-27. ment repetitions in physical and occupational therapy during spinal
26. Roy N, Weinrich B, Gray SD, Tanner K, Stemple JC, Sapienza CM. cord injury rehabilitation. Spinal Cord 2017;55:172-9.
Three treatments for teachers with voice disorders: a randomized 44. Roy N. Optimal dose-response relationships in voice therapy. Int J
clinical trial. J Speech Lang Hear Res 2003;46:670-88. Speech Lang Pathol 2012;14:419-23.
27. Tiedemann A, Paul S, Ramsay E, et al. What is the effect of a com- 45. Neil-Sztramko SE, Winters-Stone KM, Bland KA, Campbell KL.
bined physical activity and fall prevention intervention enhanced with Updated systematic review of exercise studies in breast cancer sur-
health coaching and pedometers on older adults’ physical activity vivors: attention to the principles of exercise training. Br J Sports Med
levels and mobility-related goals? Study protocol for a randomised 2017. https://doi.org/10.1136/bjsports-2017-098389.
controlled trial. BMC Public Health 2015;15:477. 46. Campbell KL, Neil SE, Winters-Stone KM. Review of exercise studies
28. Carlaw C, Finlayson H, Beggs K, et al. Outcomes of a pilot in breast cancer survivors: attention to principles of exercise training.
water protocol project in a rehabilitation setting. Dysphagia 2012;27: Br J Sports Med 2012;46:909-16.
297-306. 47. Chang SM, Matchar DB, Smetana GW, et al., editors. Methods Guide
29. Hsieh MY, Ponsford J, Wong D, Schönberger M, McKay A, Haines K. for Medical Test Reviews. Rockville (MD): Agency for Healthcare
Development of a motivational interviewing programme as a prelude Research and Quality (US); 2012.
to CBT for anxiety following traumatic brain injury. Neuropsychol 48. Kennedy MR, Coelho C, Turkstra L, et al. Intervention for executive
Rehabil 2012;22:563-84. functions after traumatic brain injury: a systematic review, meta-
30. Hart T, Brockway JA, Maiuro RD, et al. Anger self-management analysis and clinical recommendations. Neuropsychol Rehabil 2008;
training for chronic moderate to severe traumatic brain injury: 18:257-99.
www.archives-pmr.org