Jospt 2014 5141
Jospt 2014 5141
OLIVERA DJORDJEVIC, MD1,2 • ALEKSANDAR DJORDJEVIC, MD1 • LJUBICA KONSTANTINOVIC, MD, DSc1,2
L
ow back pain (LBP) is one of the most prevalent medical and morphometry at rest and during con-
conditions, with a considerable socioeconomic impact.3-5,14,18 traction. Ultrasound imaging may be use-
ful for identifying patients with structural
Low back pain has been associated with an impaired function
Copyright © 2014 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
0.96). Intrarater ICC2,1 values across 3 consecutive surface of the transducer in B-mode are
intrarater reliability of ultrasound measure-
days were high for both raters across the 2 groups adequate for analyzing the structure and
ments of transverse abdominal (TrA) and lumbar
(LBP group, 0.95-1.00; healthy group, 0.93-1.00), diameter of the muscle and its surround-
multifidus (LM) muscle thickness, during rest and
albeit lower for the relative thickness change (0.79-
contraction, in subjects with low back pain (LBP) ings. The M-mode images are created
0.99). The 95% minimal detectable changes were
and healthy subjects over 3 consecutive days, from the data received from the mid-
less than 0.3 mm for the TrA and less than 2 mm
performed by an experienced and a novice rater.
for the LM (but, in most cases, less than 10% of point of the transducer. In the continu-
TTBACKGROUND: Previous reliability studies average thickness). ously moving image received in M-mode,
TTCONCLUSION: Both experienced and trained
of TrA or LM thickness did not simultaneously
hyperechoic fascial- or bone-to-muscle
account for muscle state, rater experience, and
novice raters provided reliable measurements boundaries become more visible, which
multiday assessment in large subject samples.
of TrA and LM thickness in participants with allows more precise determination of the
TTMETHODS: The 2 raters measured TrA and LBP and healthy participants, during rest and
LM thickness on 3 consecutive days in 42 healthy
distance between the muscle edges. Thus,
contraction. One-time measurements were similar
subjects and 56 subjects with LBP, during rest and to averaged measurements. Small absolute M-mode is more suitable for measuring
contraction, and calculated the percent thickness errors were observed. Public trial registry: muscle thickness.
change from rest to contraction. Intraclass correla- Australian New Zealand Clinical Trials Registry A comparison of USI against magnetic
tion coefficients (ICC2,k) and 95% minimal detect- ACTRN12613001077752. J Orthop Sports Phys resonance imaging,27 which is the gold
able change in thickness were derived for a single Ther 2014;44(12):979-988. Epub 3 November
standard for evaluating muscle struc-
measure (day 1) and an average measure (days 1-3). 2014. doi:10.2519/jospt.2014.5141
ture, showed that USI is a valid method
TTRESULTS: The interrater ICC2,1 values for single- TTKEY WORDS: assessment, rehabilitation,
for assessing the morphology of the LM
measure thickness (LBP group, 0.71-0.87; healthy reproducibility of measurements, trunk,
group, 0.94-1.00) were similar to those for average- ultrasonography muscle. In addition to validity, it is nec-
essary to establish the reliability of USI
Clinic for Rehabilitation Dr M. Zotovic, Belgrade, Serbia. 2Faculty of Medicine, University of Belgrade, Belgrade, Serbia. The research protocol was approved by the Medical
1
Research Ethic Committee of the Clinic for Rehabilitation Dr M. Zotovic. The authors certify that they have no affiliations with or financial involvement in any organization or entity
with a direct financial interest in the subject matter or materials discussed in the article. Address correspondence to Dr Olivera Djordjevic, Clinic for Rehabilitation Dr M. Zotovic,
Sokobanjska 13a, 11000 Belgrade, Serbia. E-mail: [email protected] t Copyright ©2014 Journal of Orthopaedic & Sports Physical Therapy®
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 | 979
tion, and the relative change from rest to FIGURE 1. The overview of subject flow, data collection, and data analysis. Abbreviations: ICC, intraclass
contraction; and the measurements are correlation coefficient; LBP, low back pain; USI, ultrasound imaging.
compared over time or between those
with and those without LBP. Thus, the ing multiple reliability estimates. and contraction, within a single day and
reliability study should account for all In terms of methodological limita- across 3 consecutive days, by an experi-
factors that may affect the equivalence tions, previous reliability studies have ex- enced and a novice rater, in a sample in-
and stability of measurements, which re- amined muscles in a relaxed state13,43,50,51 clusive of people with and without LBP.
quires larger study samples. and have not reported the reliability of We hypothesized that the intraclass cor-
The reliability of USI for assessing relative change in thickness from rest to relation coefficient (ICC) for the single-
Journal of Orthopaedic & Sports Physical Therapy®
thickness of deep stabilizing muscles has contraction. Occasionally, the selected measure (day 1) and average-measure
been investigated before as a primary outcomes were measures of association (days 1-3) thickness would be greater
objective8,12,26,32,35,38,48,51,53 or as part of a (coefficient of correlation) or dispersion than 0.75 across both muscles, during
larger study.13,31,39,43,46,49,50 In general, these (coefficient of variation)12,40 rather than rest and contraction, in both individu-
reports indicate good to excellent reliabil- measures of agreement. In addition, als with LBP and healthy individuals,
ity of USI measurements, but they ad- when thickness was measured more despite differences in rater experience,
dress in isolation the factors mentioned than once, the time interval was typically thus indicating excellent reliability.11,21
above, and the majority have targeted short,39,48,50,51 potentially leading to recall The corresponding estimates of absolute
either the TrA8,13,26,38,39,43,46,49 or LM.50,51 bias. agreement were also calculated (stan-
The comparative reliability of these 2 Ultrasound imaging requires prop- dard error of measurement [SEM], 95%
muscles is of interest, because they are er education and experience.54 To our minimal detectable change [MDC95]) to
typically both examined in clinical and knowledge, however, there are no specific indicate true change in muscle thickness
research settings due to the controversy recommendations for USI of the TrA and over time.
as to which muscle is more important in LM muscles. It is difficult to interpret
certain forms of LBP.22,24,36,55 reliability without knowing the skills of METHODS
Some studies have included only examiners, yet only a few groups have
asymptomatic individuals,8,26,39,51 yet ex- reported the duration of training and Participants
T
trapolation of reliability from healthy approximate number of USI studies per- he present study included 56
individuals to those with LBP is not formed by the examiners.26,32,35,48,51 subjects with LBP and 42 healthy
advised, because of pain or movement To overcome some of these limita- subjects. The subjects with LBP
limitations that may affect USI measure- tions, the objective of this study was to were recruited from patients referred for
ments. Also, previous samples typically determine interrater and intrarater reli- evaluation and treatment to the Clinic
ranged from 10 to 20 subjects,8,13,38,39,51 ability of USI measurements of thickness for Rehabilitation Dr M. Zotovic, and
which may introduce errors when deriv- in the TrA and LM muscles during rest the healthy subjects were recruited from
980 | december 2014 | volume 44 | number 12 | journal of orthopaedic & sports physical therapy
between the muscle and fascia (calibration, 1 cm between each dot). The values for the transverse abdominal are transversely just above the iliac crest in
4.7 mm (left) and 7.3 mm (right), and for the lumbar multifidus are 35.1 mm (left) and 41.9 mm (right).
the mid-axillary line.30,31 The participant
was instructed to breathe normally. The
the spouses, relatives, and acquaintances TrA and LM muscles on the same day. images were taken at rest and during
of the patients with LBP and the investi- The experienced rater was a medical muscle contraction. Because the TrA is
gators. None of the subjects declined to doctor board certified in physical medi- an auxiliary respiratory muscle actively
participate or dropped out of the study. cine and rehabilitation and in plastic involved in voluntary expiration,1,16 the
Low back pain was defined as pain and reconstructive surgery. The expe- relaxed-TrA image was acquired at the
between the 12th rib and buttocks, with rienced rater went through a 24-month end of quiet inspiration39 to prevent TrA
Journal of Orthopaedic & Sports Physical Therapy®
or without radiation into one or both training in abdominal USI, Doppler activation. For the acquisition of the ac-
legs, originating in the lumbar region.15 imaging, soft tissue and thyroid USI, tivated-TrA image, the participant was
The exclusion criteria for all participants breast USI with interventional pro- instructed to exhale and to slowly draw
were pregnancy, spinal fractures, surgery, cedures, and an additional 14-month the stomach in as much as possible, af-
infectious diseases, tumors, spina bifida, training in musculoskeletal USI. At the ter which the image was acquired. No
advanced forms of spinal deformity, hip time of the study, the experienced rater instructions, tactile cues, or verbal en-
diseases, and neuromuscular disorders. had performed USI for about 5.5 years couragement were given. The thickness
Additional exclusion criteria for partici- and averaged about 40 musculoskeletal of the TrA was measured as the distance
pants with LBP were self-reported pain USI examinations per week. The novice between the opposite hyperechoic fascial
levels of less than 3 on the numeric pain rater was a senior at medical school. layers, without taking into account fascia
rating scale (to account for patients com- He was given a basic training in USI thickness (FIGURE 2A).
monly encountered in clinical practice) techniques (5 hours) and performed 20 The assessment of LM thickness was
and inability to remain in prone and su- measurements of abdominal and back performed with the participant in prone,
pine hook-lying positions for 10 minutes muscle thickness under supervision of with a pillow under the participant’s
(to ensure compliance). The participants the experienced rater. The 2 raters were abdomen to reduce the lumbosacral
signed an informed-consent form, and the blind to group allocation and were not junction angle to less than 10°. The mea-
rights of human subjects were protected. involved in subject recruitment or clini- surement was taken by an electrogoni-
The study was approved by the Medical cal data collection. ometer connected to an ADU301 Angle
Research Ethics Committee of the Clinic Display Unit (Biometrics Ltd, Newport,
for Rehabilitation Dr M. Zotovic. Procedure UK). The transducer was positioned lon-
A study overview is presented in FIGURE gitudinally at the midline of the L4 spi-
Raters 1. Demographic data and past medical nal process, then moved slightly laterally
An experienced rater and a novice rater history were acquired by a nurse, who and angled medially to obtain the image
independently performed USI of the also administered the 11-point numeric of the L4-5 zygapophyseal joint.30,31 The
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 | 981
Body mass index, kg/m2 27.1 5.5 25.0 6.7 single measures (day 1) and ICC2,k was
ODI score 32.0 11.1 … used for interrater reliability of the aver-
NPRS last 24 h 4.1 1.8 … age measures across 3 consecutive days,
Duration of pain, d 81.4 7.1 … where k represented the number of thick-
Abbreviations: NPRS. numeric pain rating scale; ODI, Oswestry Disability Index. ness measurements averaged (ICC2,3).
*Values are mean SD unless otherwise indicated.
Intrarater reliability was assessed with
ICC2,1 based on 3 evaluations done by
each rater on 3 consecutive days. Based
Thickness of TrA and LM Muscles on ICC point estimates, reliability was
at Rest and During Contraction and considered poor (less than 0.40), mod-
TABLE 2 Relative Thickness Change From Rest erate (0.40-0.59), good (0.60-0.74), or
to Contraction by Rater in Subjects With excellent (greater than 0.75).11,21 The reli-
Copyright © 2014 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Low Back Pain and Healthy Subjects* ability was considered statistically differ-
ent if the ICC 95% confidence intervals
Low Back Pain Healthy (95% CIs) did not overlap.
Experienced Rater Novice Rater Experienced Rater Novice Rater The SEM was calculated as SD ≈
TrA √1 – ICC,41 and MDC95 as 1.96 × SEM.17,45,47
Rest, mm 4.3 1.3 4.4 1.0 4.1 1.0 4.3 0.9 Bland-Altman analysis6 was used to eval-
Contracted, mm 6.0 2.7 6.0 1.5 6.5 2.1 6.7 1.9 uate a systematic bias for each dependent
Change, % 39.5 6.8 36.9 8.8 59.0 8.5 55.6 10.6 variable by plotting the average measure-
LM ment for the 2 raters (x-axis) against the
Journal of Orthopaedic & Sports Physical Therapy®
Rest, mm 27.9 5.0 28.7 5.5 28.5 5.7 29.0 5.7 difference between the 2 raters (y-axis)
Contracted, mm 33.9 5.6 34.5 6.2 37.6 5.7 38.3 5.9 for each examined muscle, muscle state
Change, % 21 1.5 20 1.3 32 3 32 3.3 during the examination, and subject
Abbreviations: LM, lumbar multifidus; TrA, transverse abdominal. group.
*Values are mean SD averaged across 3 consecutive days for each subject.
RESULTS
distance between the dorsal edge of the Both raters performed USI of the TrA
joint and the thoracolumbar hyperechoic and LM once a day on 3 consecutive days. General Characteristics
A
fascia, which separates the muscle from Thus, each rater acquired and measured ge, body mass index, and distri-
the subcutaneous fat tissue, was con- thickness on 1176 images (98 partici- bution by sex in subjects with LBP
sidered the thickness of the LM muscle pants by 2 muscles by 2 muscle states by and healthy subjects are shown in
(fascia thickness not included) (FIGURE 3 days). Each rater was blinded to the TABLE 1. The subjects with LBP reported,
2B). The contraction image was acquired other’s results. on average, moderate levels of pain that
with the muscle activated, by lifting the lasted nearly 3 months prior to the en-
contralateral arm off the bed and holding Data Analysis rollment and caused moderate disability
it in 120° of shoulder abduction and 90° Statistical analyses were conducted in (TABLE 1). The average thicknesses of the
of elbow flexion.30,31 SPSS Version 21 (SPSS Inc, Chicago, IL). TrA and LM muscles measured by the ex-
For the subjects with LBP, the symp- The dependent variables were thickness perienced and novice raters during rest
tomatic side was included in analysis. If at rest, thickness during contraction, and and voluntary contraction are presented
the pain was evenly distributed between relative change in thickness from rest to in TABLE 2. On average, the thickness of
the left and right paravertebral regions, contraction (percent difference between the TrA increased from rest to contrac-
the side was chosen at random. The the contraction and resting thickness tion about 40% in subjects with LBP and
side for healthy controls was selected at divided by the resting thickness). For 60% in controls, whereas the thickness
random. descriptive purposes, each dependent of the LM muscle increased about 20%
982 | december 2014 | volume 44 | number 12 | journal of orthopaedic & sports physical therapy
with LBP and from 0.94 to 1.00 for Single Measure ICC2,1* SEM, mm MDC95, mm ICC2,1* SEM, mm MDC95, mm
healthy subjects (TABLE 3). The ICC values TrA
for relative thickness change were also Rest 0.71 (0.50, 0.83) 0.12 0.33 0.94 (0.89, 0.97) 0.11 0.30
lower in subjects with LBP (0.69-0.71) Contracted 0.87 (0.77, 0.92) 0.19 0.52 0.97 (0.95, 0.99) 0.21 0.59
than those in healthy subjects (0.93- Change, % 0.69 (0.48, 0.82) 0.02 0.07 0.93 (0.87, 0.96) 0.04 0.11
0.96). Overall, 9 of 12 single-measure LM
interrater ICC values were greater than Rest 0.85 (0.74, 0.91) 0.50 1.37 1.00 (0.99, 1.00) 0.62 1.71
0.75. The 2 ICC values less than 0.75 Contracted 0.78 (0.62, 0.87) 0.64 1.76 0.97 (0.95, 0.99) 0.62 1.72
were for the relative thickness change Change, % 0.71 (0.50, 0.83) 0.35 0.97 0.96 (0.93, 0.98) 0.02 0.04
in both the TrA and LM of subjects with Abbreviations: ICC, intraclass correlation coefficient; LM, lumbar multifidus; MDC, minimal
LBP. The 95% CIs for the TrA and LM detectable change; SEM, standard error of measurement; TrA, transverse abdominal.
muscles within each subject group and *Values in parentheses are 95% confidence interval.
Copyright © 2014 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
contraction. The interrater SEM for rela- Measure ICC2,3* SEM, mm MDC95, mm ICC2,3* SEM, mm MDC95, mm
tive thickness change was from 2% to 4% TrA
across both muscles and groups, except Rest 0.76 (0.59, 0.86) 0.13 0.36 0.93 (0.87, 0.96) 0.11 0.29
for the LM in subjects with LBP (35%). Contracted 0.73 (0.53, 0.84) 0.23 0.64 0.98 (0.97, 0.99) 0.21 0.60
Change, % 0.61 (0.33, 0.77) 0.03 0.09 0.94 (0.88, 0.97) 0.04 0.10
Average-Measure Interrater LM
Reliability (Days 1-3) Rest 0.84 (0.73, 0.91) 0.49 1.37 1.00 (0.99, 1.00) 0.62 1.72
The average-measure interrater ICC2,3 Contracted 0.78 (0.62, 0.87) 0.56 1.55 0.97 (0.95, 0.99) 0.63 1.73
ranged from 0.73 to 0.84 in subjects with Change, % 0.69 (0.47, 0.82) 0.01 0.04 0.96 (0.93, 0.98) 0.02 0.05
LBP and from 0.93 to 1.00 in healthy Abbreviations: ICC, intraclass correlation coefficient; LM, lumbar multifidus; MDC, minimal
subjects (TABLE 4). The ICC2,3 values for detectable change; SEM, standard error of measurement; TrA, transverse abdominal.
relative thickness change were lower in *Values in parentheses are 95% confidence interval.
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 | 983
SEM for relative thickness change was Single Measure ICC2,1* SEM, mm MDC95, mm ICC2,1* SEM, mm MDC95, mm
between 2% and 6%. TrA, experienced rater
Rest 0.97 (0.95, 0.98) 0.16 0.43 0.94 (0.91, 0.97) 0.16 0.44
Interrater and Intrarater MDC95 Contracted 0.98 (0.95, 0.99) 0.25 0.69 0.97 (0.96, 0.99) 0.33 0.92
The interrater MDC95 values were lower Change, % 0.88 (0.82, 0.93) 0.03 0.09 0.89 (0.83, 0.94) 0.06 0.16
for the TrA (less than 0.64 mm) than LM, experienced rater
those for the LM muscle (less than 1.76 Rest 0.95 (0.92, 0.97) 0.16 0.43 1.00 (0.99, 1.00) 0.88 2.44
mm), but similar between subjects with Contracted 0.97 (0.95, 0.98) 0.23 0.64 1.00 (0.99, 1.00) 0.86 2.38
LBP and healthy subjects (TABLES 3 and Change, % 0.98 (0.97, 0.99) 0.02 0.05 0.79 (0.67, 0.87) 0.03 0.07
4). Also, the single-measure interrater TrA, novice rater
MDC95 values were comparable to the Rest 0.99 (0.99, 1.00) 0.69 1.91 0.93 (0.89, 0.96) 0.14 0.39
average-measure MDC95 values. The in Contracted 1.00 (0.99, 1.00) 0.75 2.08 0.98 (0.97, 0.99) 0.30 0.82
Copyright © 2014 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
trarater MDC95 values for both TrA and Change, % 0.79 (0.69, 0.86) 0.04 0.10 0.92 (0.87, 0.95) 0.05 0.14
LM muscles were higher for the novice LM, novice rater
rater (less than 2.35 mm) than those for Rest 1.00 (0.99, 1.00) 0.73 2.03 1.00 (0.99, 1.00) 0.88 2.44
the experienced rater (less than 0.69 Contracted 1.00 (0.99, 1.00) 0.85 2.35 0.99 (0.99, 1.00) 0.91 2.52
mm) in subjects with LBP, but compa- Change, % 0.99 (0.98, 0.99) 0.02 0.05 0.82 (0.70, 0.90) 0.02 0.06
rable in healthy subjects (TABLE 5). Abbreviations: ICC, intraclass correlation coefficient; LM, lumbar multifidus; MDC, minimal
detectable change; SEM, standard error of measurement; TrA, transverse abdominal.
Bland-Altman Analysis *Values in parentheses are 95% confidence interval.
ness between the 2 raters (bias) was less reliable measurements to determine true ment and an average of 3 measurements).
than 1 mm, regardless of the muscle ex- change in thickness over time. However, no overlap in 95% CIs between
amined, muscle state during the exami- subjects with LBP and healthy subjects
nation, and subject group, except for the Interrater and Intrarater Reliability indicates significantly lower interrater
LM muscle at rest in healthy subjects The ICC2,k model applied here indicates reliability in subjects with LBP, particu-
(1.2 mm). The 95% limits of agreement relative agreement between measure- larly for the relative change in thickness
in all cases included zero. The inspection ments because it is influenced by the (TABLES 3 and 4). Consistently lower reli-
of plots indicated that data points were variability between the subjects and by ability for the relative thickness change
randomly distributed above and below the measurement error. In general, our is likely due to a compounding effect
zero along the y-axis, with no systematic study sample was representative of the of measurement errors at rest and dur-
bias along the x-axis (FIGURES 3 and 4). Ex- population that typically receives care ing contraction. More interaction may
cept in rare instances, all measurements for LBP at the Clinic for Rehabilitation be needed between the experienced
fell within the calculated 95% limits of Dr M. Zotovic. Also, characteristics of our and novice raters before their measure-
agreement. subjects with LBP were similar to those ments of relative thickness change can
included in previous reliability32,35 and be considered interchangeable. Because
DISCUSSION USI biofeedback studies,2,49 which sup- estimates of the interrater reliability
ports generalizability of our results. were similar between the single-measure
T
he overall results of this study The finding that most interrater ICC (TABLE 3) and average-measure (TABLE 4)
indicate good to excellent interrater values were above 0.75 indicates excellent data, we consider 1-time measurement
and intrarater reliability of USI mea- agreement between the experienced and of TrA and LM thickness sufficient for
sures of TrA and LM thickness during rest novice raters across the examined muscles longitudinal follow-up.
and contraction in LBP and healthy sub- (TrA and LM), muscle states during the Lower interrater reliability in subjects
jects for both an experienced and a novice examination (at rest and during contrac- with LBP may be due to disordered mo-
rater. Low measurement errors indicate tion), and data analyzed (for 1 measure- tor control of trunk muscles or incon-
984 | december 2014 | volume 44 | number 12 | journal of orthopaedic & sports physical therapy
1.5
ing, pressure applied, or tilt angle of the
1.0
transducer between the 2 raters,35 or lack
0.5
of a clear delineation of the double con-
0.0
tour of the zygapophyseal L4-5 joint in
Downloaded from www.jospt.org at Bird Lib/OUHSC Rm 490 on December 17, 2014. For personal use only. No other uses without permission.
–0.5
subjects with LBP, due to altered muscle
–1.0
echogenicity or degenerative changes in
–1.5
the lumbosacral spine.34
–2.0
1 2 3 4 5 6 7 Our results are in general agree-
Mean Thickness, mm ment with previous reliability stud-
B
4
ies,13,26,31,32,35,38,39,42,43,46,48-51 of which only
a few included both subjects with LBP
Difference (Rater 2 – Rater 1), mm
2
and healthy subjects.31,35,38 Although sub-
jects with LBP in the latter studies were
0
similar to ours, the results are not directly
comparable because those studies report-
–2
ed reliability for subjects with LBP and
Copyright © 2014 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
Absolute Agreement
–5 of Thickness Measures
The error of measurement (SEM) is of
–10 interest because it indicates precision,
15 25 35 45
that is, how close the results of repeated
Mean Thickness, mm
D measurements actually lie. We report
10 SEM of less than 1 mm (mostly less than
Difference (Rater 2 – Rater 1), mm
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 | 985
0.0
and the novice rater.
–0.5 Of primary clinical interest are the
MDC values (also known as smallest de-
–1.0
2 3 4 5 6 7 tectable change), because they reflect a
B
Mean Thickness, mm value beyond which a measured param-
2
eter may be considered truly changed (ie,
minimal change not due to error). Un-
Difference (Rater 2 – Rater 1), mm
986 | december 2014 | volume 44 | number 12 | journal of orthopaedic & sports physical therapy
T
1976). 1984;9:395-399.
he USI measurements of TrA
1. A be T, Kusuhara N, Yoshimura N, Tomita T, 15. Dawson AP, Schluter PJ, Hodges PW, Stewart S,
and LM thickness show good to ex- Easton PA. Differential respiratory activity of four Turner C. Fear of movement, passive coping,
cellent interrater reliability for both abdominal muscles in humans. J Appl Physiol manual handling, and severe or radiating pain
muscles during rest and contraction, and (1985). 1996;80:1379-1389. increase the likelihood of sick leave due to low
2. Anderson Worth SG, Henry SM, Bunn JY. Real- back pain. Pain. 2011;152:1517-1524. http://
regardless of the difference in rater expe-
time ultrasound feedback and abdominal hol- dx.doi.org/10.1016/j.pain.2011.02.041
rience, although the reliability was signif- lowing exercises for people with low back pain. 16. De Troyer A, Estenne M, Ninane V, Van Gansbeke
icantly lower in subjects with LBP than N Z J Physiother. 2007;35:4-11. D, Gorini M. Transversus abdominis muscle
in healthy subjects. Intrarater reliability 3. Andersson GB. Epidemiological fea- function in humans. J Appl Physiol (1985).
tures of chronic low-back pain. Lancet. 1990;68:1010-1016.
was nearly perfect for both the experi-
1999;354:581-585. http://dx.doi.org/10.1016/ 17. Eliasziw M, Young SL, Woodbury MG, Fryday-
Journal of Orthopaedic & Sports Physical Therapy®
enced rater and the novice rater across S0140-6736(99)01312-4 Field K. Statistical methodology for the concur-
the muscles, muscle states, and subject 4. Andersson GBJ. The epidemiology of spinal rent assessment of interrater and intrarater
groups. This applied to both the single- disorders. In: Frymoyer JW, Ducker TB, Hadler reliability: using goniometric measurements as
NM, Kostuik JP, Weinstein JN, Whitecloud TS, an example. Phys Ther. 1994;74:777-788.
day measures and the average measures
3rd, eds. The Adult Spine: Principles and Prac- 18. Engel CC, Von Korff M, Katon WJ. Back pain in
over 3 consecutive days. The provided tice. 2nd ed. Philadelphia, PA: Lippincott-Raven; primary care: predictors of high health-care
estimates of absolute measurement er- 1997:93-141. costs. Pain. 1996;65:197-204. http://dx.doi.
ror may serve as a guide for determining 5. Biering-Sørensen F. Low back trouble in a gen- org/10.1016/0304-3959(95)00164-6
eral population of 30-, 40-, 50-, and 60-year-old 19. Farrar JT, Berlin JA, Strom BL. Clinically impor-
true changes in TrA and LM thickness
over time. t
men and women. Study design, represen- tant changes in acute pain outcome measures:
tativeness and basic results. Dan Med Bull. a validation study. J Pain Symptom Manage.
1982;29:289-299. 2003;25:406-411. http://dx.doi.org/10.1016/
KEY POINTS 6. Bland JM, Altman DG. Measuring agreement S0885-3924(03)00162-3
in method comparison studies. Stat Meth- 20. Ferreira PH, Ferreira ML, Hodges PW. Changes
FINDINGS: Interrater and intrarater reli-
ods Med Res. 1999;8:135-160. http://dx.doi. in recruitment of the abdominal muscles in
ability of the ultrasound measurement org/10.1177/096228029900800204 people with low back pain: ultrasound mea-
of the studied trunk muscles was good 7. Brennan GP, Fritz JM, Hunter SJ, Thackeray A, surement of muscle activity. Spine (Phila Pa
to excellent at rest and during contrac- Delitto A, Erhard RE. Identifying subgroups of 1976). 2004;29:2560-2566. http://dx.doi.
patients with acute/subacute “nonspecific” low org/10.1097/01.brs.0000144410.89182.f9
tion for the experienced rater and the
back pain: results of a randomized clinical trial. 21. Fleiss JL. Statistical Methods for Rates and Pro-
novice rater, whether based on a single- Spine (Phila Pa 1976). 2006;31:623-631. http:// portions. 2nd ed. New York, NY: Wiley; 1981.
day or 3-day average measurement. The dx.doi.org/10.1097/01.brs.0000202807.72292.a8 22. Fritz JM, Delitto A, Erhard RE. Comparison
estimates of absolute errors were low. 8. Bunce SM, Moore AP, Hough AD. M-mode of classification-based physical therapy with
ultrasound: a reliable measure of transversus therapy based on clinical practice guidelines
IMPLICATIONS: Ultrasound is a reliable
abdominis thickness? Clin Biomech (Bristol, for patients with acute low back pain: a ran-
method for measuring thickness of the Avon). 2002;17:315-317. domized clinical trial. Spine (Phila Pa 1976).
TrA and LM muscles in subjects with 9. Campbell SE, Adler R, Sofka CM. Ultrasound 2003;28:1363-1371; discussion 1372. http://
and without LBP when performed by of muscle abnormalities. Ultrasound Q. dx.doi.org/10.1097/01.BRS.0000067115.61673.
2005;21:87-94; quiz 150, 153-154. FF
either an experienced rater or a prop-
journal of orthopaedic & sports physical therapy | volume 44 | number 12 | december 2014 | 987
Arch Phys Med Rehabil. 2005;86:1753-1762. 2013;5:678-687. http://dx.doi.org/10.1016/j. 47. Stratford PW, Goldsmith CH. Use of the standard
http://dx.doi.org/10.1016/j.apmr.2005.03.033 pmrj.2013.03.010 error as a reliability index of interest: an applied
25. Hides J, Wilson S, Stanton W, et al. An MRI 36. Long A, Donelson R, Fung T. Does it matter example using elbow flexor strength data. Phys
investigation into the function of the trans- which exercise? A randomized control trial
Ther. 1997;77:745-750.
versus abdominis muscle during “drawing- of exercise for low back pain. Spine (Phila
48. Teyhen DS, George SZ, Dugan JL, Williamson
in” of the abdominal wall. Spine (Phila Pa Pa 1976). 2004;29:2593-2602. http://dx.doi.
J, Neilson BD, Childs JD. Inter-rater reliability
1976). 2006;31:E175-E178. http://dx.doi. org/10.1097/01.brs.0000146464.23007.2a
of ultrasound imaging of the trunk muscula-
org/10.1097/01.brs.0000202740.86338.df 37. MacDonald DA, Dawson AP, Hodges PW. Behav-
ture among novice raters. J Ultrasound Med.
26. Hides JA, Miokovic T, Belavý DL, Stanton WR, ior of the lumbar multifidus during lower ex-
2011;30:347-356.
Richardson CA. Ultrasound imaging assessment tremity movements in people with recurrent low
of abdominal muscle function during drawing-in back pain during symptom remission. J Orthop 49. Teyhen DS, Miltenberger CE, Deiters HM,
of the abdominal wall: an intrarater reliability Sports Phys Ther. 2011;41:155-164. http://dx.doi. et al. The use of ultrasound imaging of the
study. J Orthop Sports Phys Ther. 2007;37:480- org/10.2519/jospt.2011.3410 abdominal drawing-in maneuver in subjects
486. http://dx.doi.org/10.2519/jospt.2007.2416 38. Mannion AF, Pulkovski N, Gubler D, et al. Muscle with low back pain. J Orthop Sports Phys Ther.
27. Hides JA, Richardson CA, Jull GA. Magnetic thickness changes during abdominal hollowing: 2005;35:346-355. http://dx.doi.org/10.2519/
jospt.2005.35.6.346
Copyright © 2014 Journal of Orthopaedic & Sports Physical Therapy®. All rights reserved.
@ MORE INFORMATION
611. http://dx.doi.org/10.2519/jospt.2009.3088 ment of severity of complaints, grip strength,
34. Langevin HM, Stevens-Tuttle D, Fox JR, et al. and pressure pain threshold in patients with
Ultrasound evidence of altered lumbar con- lateral epicondylitis. Arch Phys Med Rehabil. WWW.JOSPT.ORG
988 | december 2014 | volume 44 | number 12 | journal of orthopaedic & sports physical therapy