RESEARCH

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
Efficacy of standardised manual therapy and home exercise
programme for chronic rotator cuff disease: randomised
placebo controlled trial
Kim Bennell, professor,1 Elin Wee, physiotherapist and research assistant,1 Sally Coburn, physiotherapist,1,2
Sally Green, professor,3 Anthony Harris, professor,4 Margaret Staples, biostatistician,5,6 Andrew Forbes,
professor,6 Rachelle Buchbinder, professor5,6

1
Centre for Health, Exercise and ABSTRACT 1.39, 0.94 to 2.03). Several secondary outcomes
Sports Medicine, Department of Objective To investigate the efficacy of a programme of favoured the active group, including shoulder pain and
Physiotherapy, School of Health
Sciences, University of Melbourne, manual therapy and exercise treatment compared with disability index function score, muscle strength,
Parkville 3010, Victoria, Australia placebo treatment delivered by physiotherapists for interference with activity, and quality of life.
2
Department of Physiotherapy, people with chronic rotator cuff disease. Conclusion A standardised programme of manual therapy
Royal Melbourne Hospital,
Melbourne
Design Randomised, participant and single assessor and home exercise did not confer additional immediate
3
Monash Institute of Health
blinded, placebo controlled trial. benefits for pain and function compared with a realistic
Services Research, Monash Setting Metropolitan region of Melbourne, Victoria, placebo treatment that controlled for therapists’ contact
University, Melbourne Australia. in middle aged to older adults with chronic rotator cuff
4
Centre for Health Economics, Participants 120 participants with chronic (>3 months) disease. However, greater improvements were apparent
Monash University
5 rotator cuff disease recruited through medical at follow-up, particularly in shoulder function and
Monash Department of Clinical
Epidemiology, Cabrini Hospital, practitioners and from the community. strength, suggesting that benefits with active treatment
Melbourne Interventions The active treatment comprised a manual take longer to manifest.
6
Department of Epidemiology and therapy and home exercise programme; the placebo Trial registration Clinical trials NCT00415441.
Preventive Medicine, School of
Public Health and Preventive treatment comprised inactive ultrasound therapy and
Medicine, Monash University application of an inert gel. Participants in both groups INTRODUCTION
Correspondence to: K Bennell received 10 sessions of individual standardised Shoulder disorders are a common cause of persistent
k.bennell@unimelb.edu.au treatment over 10 weeks. For the following 12 weeks, the musculoskeletal morbidity,1 2 particularly in the mid-
Cite this as: BMJ 2010;340:c2756 active group continued the home exercise programme dle to older age groups.3 Pain and compromised
doi:10.1136/bmj.c2756 and the placebo group received no treatment. shoulder function have a substantial impact on tasks
Main outcome measures The primary outcomes were pain essential to daily living, as well as on sleep.4 Shoulder
and function measured by the shoulder pain and disorders are a common reason for seeking medical
disability index, average pain on movement measured on care and may require surgical intervention in up to
an 11 point numerical rating scale, and participants’ 28% of cases.5-7 Shoulder disorders can thus lead to
perceived global rating of overall change. considerable disability, reduced health related quality
Results 112 (93%) participants completed the 22 week of life, absenteeism from work, and use of healthcare
trial. At 11 weeks no difference was found between resources.4 8 9
groups for change in shoulder pain and disability index Although definitions of different diagnostic cate-
(3.6, 95% confidence interval −2.1 to 9.4) or change in gories of shoulder pain are controversial, a large pro-
pain (0.7, −0.1 to 1.5); both groups showed significant portion of shoulder problems can be classified as
improvements. More participants in the active group “rotator cuff disease,” the most common cause of
reported a successful outcome (defined as “much shoulder pain in primary care.10 The term, or its var-
better”), although the difference was not statistically iants such as impingement syndrome, may include a
significant: 42% (24/57) of active participants and 30% spectrum of pathologies of rotator cuff disease (such
(18/61) of placebo participants (relative risk 1.43, 0.87 to as subacromial bursitis, partial rotator cuff tears, and
2.34). The active group showed a significantly greater bicipital tendinosis), but they are characterised clini-
improvement in shoulder pain and disability index than cally by pain with abduction (painful arc) and signs of
did the placebo group at 22 weeks (between group impingement.11 Although standard criteria have not
difference 7.1, 0.3 to 13.9), although no significant been established for use in clinical trials, most trials
difference existed between groups for change in pain that have assessed interventions for rotator cuff disease
(0.9, −0.03 to 1.7) or for the percentage of participants have used variations of these features to select their
reporting a successful treatment outcome (relative risk study populations.12-14 Rotator cuff disease differs
BMJ | ONLINE FIRST | bmj.com page 1 of 10
 RESEARCH

from other major diagnostic categories of shoulder including corticosteroid injection, arthrographic dis-
pain such as adhesive capsulitis, osteoarthritis, and cal- tension of the glenohumeral joint with corticosteroid
cific tendinitis, which are known to have different pre- and saline (hydrodilatation), or physiotherapy; anti-

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
sentations, underlying causes, prognoses, and inflammatory drugs in the previous two weeks; and
responses to treatment. inability to understand written and spoken English.
A combination of modalities of physiotherapy, such
as manual therapy and exercise, is often used in the Procedures
management of rotator cuff disease.15 These aim to cor- We did a randomised, participant and assessor
rect modifiable physical impairments thought to con- blinded, controlled trial. Participants had a baseline
tribute to pain and dysfunction rather than to treat the assessment and were randomised in permuted blocks
specific pathology. These impairments include rotator of six and eight, stratified by treating physiotherapist,
cuff and scapular muscle weakness and dysfunction, to receive either active manual therapy and home exer-
tightness of the posterior capsule and other soft tissues, cise treatment or placebo treatment according to a
and postural abnormalities.16 Little conclusive evi- computer generated table of random numbers created
dence supports or refutes the efficacy of different by the study biostatistician (AF). Allocations were
physiotherapy programmes given the variable metho- sealed in opaque and consecutively numbered envel-
dological quality of the trials, including a lack of pla- opes kept in a central locked location. An independent
cebo control and the fact that many tested a single administrator opened the envelopes in sequence and
modality despite multimodality treatment being the then revealed the group allocation to the relevant
most common way in which physiotherapists treat physiotherapist by facsimile just before the participant
shoulder disorders.13 17-20 The conclusions and recom- presented for treatment.
mendations of recent systematic reviews support the
need for further clinical trials.21-23 The primary aim of Interventions
this trial was, therefore, to determine whether a Details about the interventions have been published
10 week programme of standardised manual therapy previously.26 Fourteen musculoskeletal physio-
and home exercise delivered by a physiotherapist therapists (all with more than four years of relevant
improves shoulder pain and function more than pla- clinical experience) from 12 centres (two public hospi-
cebo treatment does in people with chronic rotator tal physiotherapy departments and 10 private physio-
cuff disease. therapy clinics) were trained to provide both
interventions. Therapists attended initial training ses-
METHODS sions and were given a detailed treatment manual. We
Participants could not blind the therapists to treatment group. Both
Between March 2004 and November 2007, we interventions were standardised and comprised indivi-
recruited people with chronic rotator cuff disease dual sessions twice weekly for the first fortnight, once a
through medical practitioners and from the commu- week for the next four weeks, then once a fortnight in
nity through print and radio media. We required all the last four weeks (10 visits, 30-45 minutes each). To
participants to have a plain radiograph of the shoulder minimise the risk of participants meeting, appoint-
to check for exclusions (see below), and we required ments were scheduled at different times. We assessed
potential participants recruited directly from the com- therapists’ adherence to the protocol by completion of
munity to have the diagnosis of rotator cuff disease a treatment log. Simple analgesia was permitted, but
confirmed by a medical practitioner. After an initial participants were asked to refrain from seeking other
screen by telephone, an experienced physiotherapist forms of treatment during the trial. Treatment in both
(EW or SC) did a physical examination. Inclusion cri- groups was provided at no cost to the participant.
teria were age over 18 years, shoulder pain for more We based the active intervention on the literature
than three months, severity of pain on movement rated and on the results of a formal written survey of 16 Aus-
greater than 3/10 on an 0-10 numerical rating scale, tralian musculoskeletal physiotherapists with expertise
pain on active abduction or external rotation, and a in treating shoulder conditions.27 The intervention was
positive quick test for shoulder impingement.24 Exclu- directed at improving dynamic scapular control,
sion criteria were resting severity of shoulder pain strengthening scapular stabiliser and rotator cuff mus-
greater than 7/10; reason to suspect a complete rotator cles, improving shoulder and thoracic posture, and
cuff tear (for example, substantial shoulder weakness, a increasing range of motion of thoracic extension. The
positive drop-arm sign, or a high riding humerus on intervention had five components comprising soft tis-
plain radiograph); previous shoulder surgery; radio- sue massage, passive mobilisation of the glenohumeral
logical evidence of shoulder osteoarthritis, calcifica- joint, scapular retraining and postural taping, spinal
tion, or previous fracture; systemic pathology mobilisation (to assist in improving shoulder girdle
including inflammatory joint disease or neoplastic dis- posture and spinal range of motion), and home exer-
orders; more than 50% restriction of passive range of cises (table 1). 26 We incorporated behavioural strate-
motion in two or more planes; shoulder pain referred gies, including education, goal setting, motivation, and
from vertebral structures diagnosed by spinal clearing positive reinforcement. Home exercises were done
tests25; symptoms of complex regional pain syndrome; daily, except during the first week of treatment when
active intervention in the previous three months, exercises were completed twice daily (web appendix).
page 2 of 10 BMJ | ONLINE FIRST | bmj.com
 RESEARCH

Table 1 | Components of active physiotherapy intervention

Treatment component Description Dosage

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
Soft tissue massage Anterior and posterior shoulder tissues, in supine and side-lying positions respectively 6 minutes each position
Glenohumeral joint mobilisation Anteroposterior and inferior joint glides in supine position with shoulder at 45° and 90° 4×30 seconds each position
abduction respectively
Thoracic spine mobilisation (T1-8) In prone position, using central posteroanterior technique Grade IV on each level: 4 minutes in total
Cervical spine mobilisation (C5-7) In prone position using unilateral posteroanterior technique on both sides Grade IV on each level: 4 minutes in total
Scapular retraining In side-lying position, therapist passively moves shoulder through range from elevation/ Weeks 1 and 2 only; 15 repetitions × 5 repetitions with
protraction to retraction/depression, then assisted by participant, then independently by 10 second holds
participant; isometric holds in retraction/depression
Postural taping Taping of shoulders and scapula to encourage scapular retraction and depression and Continuous (day and night) for two weeks; re-applied after
thoracic extension one week by therapist
Home exercises Supervised and done as home programme Home programme: twice daily in first two weeks; once a day
thereafter

After the 10 week programme, participants in the Participants’ perceived global rating of change over-
active group were instructed to maintain their daily all and in pain, strength, and stiffness (from baseline)
home exercise programme for 12 weeks. were recorded on separate five point Likert-type scales
Participants in the placebo group attended the same (1=much worse, 2=slightly worse, 3=no change,
number of treatments as did those in the active treat- 4=slightly better, 5=much better).35 We defined a suc-
ment group but received sham ultrasound therapy and cessful outcome for each a priori as “much better” on
light application of a non-therapeutic gel to the the rating scale.
shoulder region for 10 minutes each. They received We measured generic health related quality of life
no instruction in exercise techniques and no manual with the Medical Outcomes Study 36-item short form
therapy. We have successfully used this same placebo (SF-36) (eight subscales scaled from 0-100, in which a
protocol in previous studies.28-30 During the 12 week higher score represents better health, summarised into
follow-up period, placebo participants did not receive physical function and mental health scales) and the
any intervention and were not instructed to do any assessment of quality of life (AQoL) instrument.36 37
home exercises. The latter instrument comprises 15 items covering
five dimensions (illness, independent living, social
Outcome measures relationships, physical senses, and psychological
The same blinded assessor (EW) evaluated all partici- wellbeing).38 39 Item responses are all ordinal scales
pants at baseline, at 11 weeks (at the conclusion of the with four levels per item. Scores are scaled from
supervised active or placebo intervention), and at −0.04 (worse than death) to 1.00 (perfect health).
22 weeks after randomisation. Baseline demographic We measured isometric shoulder strength of the
information was collected, and participants rated their symptomatic limb for shoulder abduction and internal
expectation of a beneficial effect of active physio- and external rotation with the Nicholas Manual Mus-
therapy treatment on an ordinal scale from 1 to 5, cle tester (Lafayette, USA). For abduction, participants
with higher scores indicating higher expectations. were in supine position with the shoulder in 90° of
The primary outcomes were the shoulder pain and abduction and the dynamometer positioned on the lat-
disability index (SPADI), average pain on movement eral surface of the distal humerus. Measurements of
assessed by a numerical rating scale, and participants’ external and internal rotation were made in sitting
perceived global rating of change overall. The position with the arm by the side against a folded
shoulder pain and disability index is a self adminis- towel with the elbow flexed to 90° and the dynam-
tered, shoulder specific index consisting of 13 items ometer positioned on the distal forearm. After a
divided into two subscales—pain (five items) and func- demonstration and one warm-up trial, participants
tion (eight items)—with responses to each item were asked to push as hard as they possibly could
recorded on a 10 point scale.31-33 We calculated a against the dynamometer for four seconds while the
total shoulder pain and disability index score by sum- tester provided consistent loud verbal encouragement.
ming the subscales and then averaging for a score out The mean reading of three maximal contractions was
of 100 (higher scores indicate more pain/dysfunction). taken. Reliability is excellent in our laboratory (12 peo-
We measured participants’ overall assessment of ple with rotator cuff disease tested by two examiners
average pain on movement and pain at rest in the pre- twice two to four days apart: intraclass correlation coef-
vious week by separate 11 point numerical rating ficient (2,3) values ≥0.89 for intra-rater reliability of
scales (0 to 10) numbered in 1 cm intervals.34 The mini- each examiner and ≤0.90 for inter-rater reliability27).
mal clinically important difference for shoulder pain We measured participants’ adherence to treatment
on movement measured on this scale is 1.1 units.34 by recording the number of physiotherapy sessions
The amount of weakness, stiffness, and interference attended (out of a maximum of 10). Participants in
with activities of daily living over the previous week the active group also completed a daily log book to
were similarly measured. record the number of home exercise sessions
BMJ | ONLINE FIRST | bmj.com page 3 of 10
 RESEARCH

completed. Adverse events and the use of co-inter- needed 91 participants per group to achieve 80%
ventions in both groups were recorded in a log book. power at a two sided 5% significance level.
To measure the success of blinding, we asked partici- As recruitment was much slower than anticipated, an

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
pants to indicate which treatment they believed they independent statistician not previously associated with
had received at the 22 week assessment. the trial did a blinded assessment of the between parti-
cipant standard deviation and baseline to 11 week cor-
Sample size relation after 46 patients had completed follow-up.
We calculated sample size on the basis of ability to These were more favourable than initially planned
detect a 10 point difference in improvement in total (SD=21, r=0.60), and we revised the total trial sample
shoulder pain and disability index score, previously size downwards to 60 participants per arm to maintain
reported to indicate a clinically important improve- 80% power to detect a difference of 10 units with these
ment (or worsening) of shoulder function.32 Applying revised parameters.
power calculations appropriate for analysis of covar-
iance (adjusting for baseline shoulder pain and disabil- Data analysis
ity index score), to detect a difference in 11 week We did analyses on an intention to treat principle,
using all randomised participants. We replaced miss-
shoulder pain and disability index score of 10 units
ing data by the last score carried forward. For out-
assuming a common between participant standard
comes measured using an essentially continuous
deviation of 27 and a baseline to 11 week correlation
scale, we compared differences in mean change from
of 0.45 (from our pilot study27), we calculated that we
baseline to each time point between groups by using
linear regression modelling with adjustment for base-
line levels of the outcome measure. We checked model
Assessed for eligibility by phone (n=1380)
assumptions by standard diagnostic plots.
Did not meet inclusion criteria or not We dichotomised participants’ measures of per-
interested in participating (n=942) ceived global change after active or placebo treatments
Enrolment: patients
into successful (much better) or unsuccessful (slightly
Assessed for eligibility by physical screening (n=438)
better, no change, slightly worse, and much worse) out-
Excluded (n=318): come. We compared the percentage of successful out-
Did not meet inclusion criteria (n=313) comes between groups by calculating relative risks and
Refused to participate (n=5)
their 95% confidence intervals at each time point with
Randomised (n=120) log binomial regression.40
We calculated an index to assess the success of blind-
Allocation: patients ing after treatment.41 This index takes the value one for
Allocated to active intervention (n=59) Allocated to placebo intervention (n=61) complete blinding and zero for complete lack of blind-
ing.
Allocation:
physiotherapists
Physiotherapists (n=14), centres (n=12) Physiotherapists (n=11), centres (n=9) RESULTS
No of patients treated by each physiotherapist: No of patients treated by each physiotherapist: We recruited 120 participants (59 active, 61 placebo),
median=4, IQR=1.3-5.8, min=1, max=11 median=5 IQR=4.0-7.5, min=1, max=10
No treated at each centre: No treated at each centre:
and 112 (54 active, 58 placebo; 93%) completed the
median=4, IQR=1.0-7.5, min=1, max=16 median=5, IQR=4.0-7.0, min=2, max=19 22 week trial. Figure 1 shows the flow of participants
through the trial. Two participants, both from the
Treatment phase
active group, withdrew before completing the
10 x individual treatment sessions 10 x individual treatment sessions
10 week intervention. A further six participants (three
Week 11 assessment: active, three placebo) withdrew before the 22 week fol-
patients low-up. Demographic and clinical characteristics of
Lost to week 11 assessment: Lost to week 11 assessment (n=0) the eight participants who withdrew from the study
Unable to contact (n=1) Discontinued intervention (n=0)
Discontinued intervention:
did not differ from those of the 112 who remained
Concomitant injury (n=1) (data not shown).
The groups were similar at baseline for demographic
Follow-up phase
and clinical characteristics, although the median dura-
Daily home exercises No exercises
tion of symptoms seemed to be longer in the active
Week 22 assessment: group (table 2). The participants’ expectation of treat-
patients ment outcomes for active physiotherapy was similar in
Lost to week 22 assessment (n=3): Lost to week 22 assessment (n=3): the two groups (P=0.79); 95/105 (90%) participants
Unable to contact (n=2) Unable to contact (n=2)
Moved overseas (n=1) Withdrew (n=1)
who provided this information expected a moderate
or large beneficial effect.
Analysis: patients
Analysed (n=59) Analysed (n=61) Efficacy analysis
Excluded from analysis (n=0) Excluded from analysis (n=0)
Immediately after treatment (11 weeks)
Both groups showed significant improvements imme-
Fig 1 | Flow of participants through study. IQR=interquartile range diately after treatment (11 weeks). However, we found
page 4 of 10 BMJ | ONLINE FIRST | bmj.com
 RESEARCH

Table 2 | Demographic and clinical characteristics of active comparing the active group participants who reported
and placebo groups. Values are numbers (percentages) more than 50% adherence to the home exercise pro-
unless stated otherwise gramme (n=35) with the placebo participants; as a

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
Characteristic Active n=59 Placebo n=61
completers’ analysis without replacing the missing
values; controlling for duration of symptoms by add-
Mean (SD) age (years) 59.3 (10.1) 60.8 (12.4)
ing a term in the regression analysis; and using general-
Median (interquartile range) 24 (6-54) 14 (6-24)
duration of symptoms (months) ised estimating equations to fit population averaged
Mean (SD) height (cm) 169.0 (9.1) 167.5 (10.8) models (data not shown).
Mean (SD) body mass (kg) 79.5 (13.5) 78.9 (15.9)
Mean (SD) body mass index (kg/ 27.8 (4.4) 27.9 (4.8)
Adherence, adverse events, and co-interventions
m2) Fifty-two (91%) of 57 participants in the active group
Male sex 34 (58) 30 (49) and 57/61 (93%) of those in the placebo group attended
Affected shoulder (right:left) 35:24 38:23 all 10 physiotherapy treatment sessions. The number
Dominant side affected 38 (64) 42 (69) of sessions attended by the remainder ranged from
Previous treatment: 29 (49) 35 (57) three to nine with a mean of 4.8 (SD 2.7) in the active
Physiotherapy 20 (34) 26 (43) group and from one to seven with a mean of 6.0 (3.5) in
Corticosteroid injection 10 (17) 15 (25) the placebo group. Of the active group participants
Massage 2 (3) 1 (2) who completed the exercise diaries (52/57 during the
Acupuncture 5 (8) 4 (7)
intervention period and 39/55 in the follow-up period),
Chiropractic/osteopathy 6 (10) 6 (10)
the mean self reported completion of home exercise
sessions was 57.2 (SD 15.7, 82%) during the inter-
vention period and 49.0 (20.3, 70%) during the fol-
no significant between group differences for the pri- low-up period.
mary outcomes of changes in shoulder pain and dis- During the intervention period, 17/55 (31%) partici-
ability index total score (3.6, 95% confidence interval pants in the active group reported adverse events.
−2.1 to 9.4) and pain on movement (0.7, −0.1 to 1.5) These comprised increased short term pain during or
(table 3, fig 2). More participants in the active group after the treatment session (n=3), increased short term
reported an overall successful outcome (defined as pain with the home exercises (12), and mild irritation to
“much better”), although the difference was not statis- the tape used for postural taping (2). In the placebo
tically significant: 42% of active participants and 30% group, 5/61 (8%) reported adverse events comprising
of placebo participants (relative risk 1.43, 95% confi- increased short term pain during or after the treatment
dence interval 0.87 to 2.34) (table 4). session. During the follow-up period, adverse events
With regards to the secondary outcomes, the active were reported only by the active group (7/49, 14%)
group showed significantly greater improvements in and comprised increased short term pain with the
both self reported and objective measures of strength home exercises.
(tables 3 and 4). We found no significant differences Use of analgesics and non-steroidal anti-inflamma-
between treatment groups for other outcomes, includ- tory drugs was similar in the active and placebo groups
ing changes in health related quality of life. over both the intervention period (analgesics: 11/55
(20%) active v 14/61 (23%) placebo; non-steroidal
Twenty-two week follow-up anti-inflammatories: 12/55 (22%) v 13/61 (21%); both
The active group showed a significantly greater P>0.05) and the follow-up period (analgesics: 8/49
improvement in shoulder pain and disability index (16%) v 8/55 (15%); non-steroidal anti-inflammatories
total score at 22 weeks than did the placebo group 6/49 (12%) v 8/55 (15%); both P>0.05).
(mean between group difference 7.1, 0.3 to 13.9) During the intervention period, one (2%) placebo
(table 3, fig 2). However, this was not accompanied participant received a cortisone injection into the
by significant differences between the active and pla- shoulder. During the follow-up period, two (4%) parti-
cebo groups for change in pain on movement (table 3, cipants in the active group received a cortisone injec-
fig 2) or for the percentage of participants reporting an tion into the shoulder and one (2%) participant in the
overall successful treatment outcome (table 4). Within placebo group received acupuncture treatment.
group changes in the primary outcomes remained sig-
nificant at the 22 week follow-up (all P<0.001). Success of blinding
Several secondary outcomes also showed benefits in In the active group, 32/55 (58%) participants correctly
favour of the active group. We saw greater improve- identified their treatment group at 11 weeks compared
ments in shoulder pain and disability index function with 21/61 (34%) participants in the placebo group; 15
score, muscle strength, interference with activity, and (27%) participants in the active group were uncertain
health related quality of life as measured by the assess- which treatment they had received compared with 27
ment of quality of life instrument (tables 3 and 4). (44%) participants in the placebo group; and 8 (15%)
participants in the active group incorrectly identified
Other analyses their treatment group compared with 13 (21%) partici-
The results immediately after treatment and at follow- pants in the placebo group. The blinding index was
up were essentially unaltered when reanalysed 0.70 (bootstrap 95% confidence interval 0.58 to 0.82),
BMJ | ONLINE FIRST | bmj.com page 5 of 10
 RESEARCH

Table 3 | Mean (SD) of groups, mean (SD) difference within groups, and mean (95% CI) difference between groups adjusted for baseline scores for outcomes
with interval data

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
Difference between
Groups Difference within groups* groups†
Week 0 Week 0
Week 0 Week 11 Week 22 to week 11 to week 22 Week 0
Active Placebo Active Placebo Active Placebo to week Week 0 to
Outcome (n=59) (n=61) (n=59) (n=61) (n=59) (n=61) Active Placebo Active Placebo 11 week 22
SPADI total (0-100) 43.3 43.9 27.2 31.2 20.9 28.3 16.1 12.7 22.4 15.6 3.6 (−2.1 7.1 (0.3 to
(18.9) (17.5) (18.9) (21.0) (18.6) (24.5) (17.7) (16.3) (22.0) (17.8) to 9.4) 13.9)
Pain on movement 4.9 (2.2) 4.9 (1.8) 2.9 (2.3) 3.6 (2.3) 2.4 (2.4) 3.3 (2.7) 2.1 (2.6) 1.3 (2.2) 2.6 (2.9) 1.6 (2.4) 0.7 (−0.1 0.9 (−0.03
(0-10) to 1.5) to 1.7)
SPADI pain (0-100) 47.8 48.4 29.8 33.9 23.0 31.0 18.0 14.4 24.8 17.3 3.2 (−3.2 6.8 (−0.7 to
(20.1) (17.5) (20.8) (22.7) (21.0) (26.0) (18.8) (18.5) (23.7) (19.6) to 9.6) 14.3)
SPADI function 33.6 33.8 20.0 25.1 14.1 22.2 13.6 8.7 (13.9) 19.6 11.6 4.7 (−0.1 7.6 (1.8 to
(0-100) (20.0) (20.2) (16.3) (19.3) (14.6) (22.8) (17.3) (20.7) (16.6) to 9.5) 13.4)
Pain at rest (0-10) 2.3 (1.9) 2.1 (1.8) 1.4 (1.9) 1.7 (1.9) 1.0 (2.0) 1.6 (2.1) 1.0 (2.2) 0.4 (2.0) 1.3 (2.5) 0.4 (2.5) 0.4 (−0.2 0.7 (−0.1 to
to 1.1) 1.4)
Weakness on 4.6 (2.4) 4.1 (2.3) 2.6 (2.3) 3.2 (2.4) 2.1 (2.3) 2.9 (2.6) 2.0 (2.6) 0.9 (2.0) 2.5 (3.0) 1.1 (2.7) 0.8 (0.05 0.9 (0.1 to
movement (0-10) to 1.5) 1.8)
Stiffness on 3.3 (2.6) 3.4 (2.4) 2.1 (2.1) 2.6 (2.2) 1.9 (2.3) 2.6 (2.4) 1.2 (2.2) 0.8 (2.2) 1.4 (2.7) 0.8 (2.7) 0.4 (−0.2 0.7 (−0.1 to
movement (0-10) to 1) 1.5)
Interference with 3.9 (2.5) 3.8 (2.3) 2.0 (1.9) 2.6 (2.1) 1.5 (1.9) 2.5 (2.6) 1.9 (2.3) 1.2 (1.9) 2.4 (2.7) 1.3 (2.4) 0.6(−0.04 0.9 (0.1 to
activity (0-10) to 1.2) 1.7)
SF-36 physical 49.3 48.9 61.0 55.0 60.0 53.5 11.7 6.1 (17.4) 10.8 4.7 (22.3) 5.7 (−2.1 6.3 (−2.0 to
(0-100) (23.4) (25.0) (28.1) (27.5) (27.2) (29.1) (26.5) (25.0) to 13.6) 14.5)
SF-36 mental (0-100) 70.2 61.5 69.7 61.9 69.3 63.3 −0.6 0.4 (16.0) −1.0 1.8 (15.8) 2.1 0.6 (−5.2 to
(23.4) (21.4) (22.1) (20.7) (20.4) (21.0) (19.3) (19.7) (−3.8 to 8) 6.4)
AQoL (−0.4–1.0) 0.7 (0.1) 0.7 (0.2) 0.7 (0.2) 0.7 (0.2) 0.7 (0.2) 0.7 (0.2) 0.0 (0.1) 0.0 (0.1) 0.0 (0.2) 0.0 (0.1) 0.0(−0.04 0.0 (0.04to
to 0.03) 0.1)
Abduction strength 7.2 (5.4) 6.2 (3.4) 8.4 (4.7) 6.7 (3.8) 8.3 (3.8) 6.5 (3.9) 1.2 (3.9) 0.5 (2.2) 1.1 (4.4) 0.4 (2.5) 1.0 (−0.1 1.2 (0.1 to
(kg) to 2) 2.3)
External rotation 8.1 (4.5) 7.1 (2.6) 8.3 (3.9) 7.2 (3.0) 8.4 (3.6) 7.0 (3.0) 0.2 (3.8) 0.1 (1.4) 0.3 (4.3) −0.1 (1.9) 0.5 (−0.4 0.9 (−0.1 to
strength (kg) to 1.4) 1.9)
Internal rotation 10.9 (5.5) 10.2 (4.5) 11.7 (5.2) 10.1 (4.3) 12.2 (5.3) 10.2 (4.6) 0.9 (3.1) −0.1 (2.7) 1.3 (3.4) 0.0 (2.7) 1.1 (0.1 to 1.5 (0.4 to
strength (kg) 2.1) 2.5)
AQoL=assessment of quality of life; SF-36=Medical Outcomes Study 36-item short form; SPADI=shoulder pain and disability index.
*Positive change equals improvement, and positive values favour active group.
†Results from regression analyses adjusted for baseline scores.

interpreted as a moderate to high degree of blinded- outcome measures. Changes in overall pain and func-
ness and representing a statistically significant amount tion measured by the shoulder pain and disability
of blinding beyond that expected by chance (the value index favoured the active group, although the mean
of the blinded index is 0.5 for random guessing). between group difference of 7.1 was slightly below
the 8 to 13.2 points reported in the literature as being
DISCUSSION the minimal clinically important difference.42 Several
This randomised, participant and single assessor secondary outcomes also favoured the active group,
blinded, placebo controlled trial evaluated the efficacy including shoulder pain and disability index function
of a 10 week manual therapy and home exercise pro- score, muscle strength, interference with activity, and
gramme delivered by physiotherapists for the treat- quality of life.
ment of chronic rotator cuff disease in middle aged to
older adults. Immediately after the intervention Explanation of results
(11 weeks), the standardised active treatment generally The significant improvements seen in both groups
produced similar beneficial effects on shoulder pain over the 22 weeks may reflect natural recovery of the
and function, the primary endpoints of the trial, com- rotator cuff disease. Although we did not include a
pared with a realistic placebo treatment that controlled third “no treatment” study arm to ascertain this, natural
for therapists’ contact time and the therapeutic envir- recovery is unlikely to explain the whole effect given
onment. Both groups improved by amounts deemed to the long duration of symptoms, particularly in the
be clinically important,42 and more than a third of par- active group, and the moderate baseline disability of
ticipants reported a successful treatment outcome. the cohort—factors that have been associated with a
However, we found significant differences favouring poorer prognosis.43 44 Furthermore, other clinical trials
the active group for objective and subjective measures in patients with chronic rotator cuff disease found mini-
of muscle strength. At follow-up (22 weeks), we saw mal changes in pain or function over similar time-
greater improvements with active treatment for several frames in control groups receiving no treatment.14 19
page 6 of 10 BMJ | ONLINE FIRST | bmj.com
 RESEARCH

Table 4 | Number (percentage) of participants reporting a successful outcome (“much better”) compared with those reporting
an unsuccessful outcome (“slightly better,” “no change,” “slightly worse,” or “much worse”) in both groups, with relative risks

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
11 weeks 22 weeks
Active Placebo Relative risk Active Placebo Relative risk
Outcome (n=57) (n=61) (95% CI) (n=54) (n=58) (95% CI)
Global change overall 24 (42) 18 (30) 1.43 (0.87 to 2.34) 31 (57) 24 (41) 1.39 (0.94 to 2.03)
Global change in pain 22 (39) 20 (33) 1.18 (0.72 to 1.91) 31 (57) 25 (43) 1.33 (0.92 to 1.94)
Global change in strength 19 (33) 7 (11) 2.90 (1.32 to 6.39) 22 (41) 14 (24) 1.69 (0.97 to 2.95)
Global change in stiffness 20 (35) 13 (21) 1.65 (0.91 to 2.99) 25 (46) 18 (31) 1.49 (0.92 to 2.41)

Improvement in our cohort may also have arisen disability index total score. Although the mean
from the statistical phenomenon of regression to the between group difference at the latter time point (7.1
mean.45 This refers to the tendency for extreme symp- units) was slightly below the minimal clinically impor-
toms at baseline to return to a more typical state at final tant difference we used when designing the study, the
assessment. Symptoms associated with chronic rotator 95% confidence interval includes the 10 unit threshold.
cuff disease fluctuate over time, and patients often seek This, together with the fact that the active treatment
medical care or enrol in research when the symptoms also led to significantly greater improvements in
are at their worst. Furthermore, we included patients many secondary outcomes, indicates that manual ther-
only if their pain was worse than a specific threshold apy and home exercise may be beneficial particularly
level. The next change in symptoms is thus more likely over time.
to be an improvement.46 However, we accounted for
regression to the mean in our statistical analyses by Aspects of active and placebo interventions
adjusting for the baseline value of each variable.47 Some debate exists in the literature about whether the
In addition to spontaneous improvement, another use of a placebo treatment as a comparator for complex
factor contributing to the total treatment effect is the interventions such as physiotherapy is appropriate.48
placebo effect.48 A recent meta-analysis showed that The direct and indirect (placebo) effects of the therapy
for active treatment of chronic pain conditions (not spe- have been argued to be unlikely to be distinct and divi-
cifically of the shoulder), spontaneous recovery contri- sible, and elements that may be categorised as indirect
butes around 10% and placebo effects around 30%.46 effects in drug trials may in fact be integral to many
Placebo effects have also been found to be greatest for non-drug interventions. Hence, using a placebo con-
non-drug interventions and for patient reported out- trolled trial design to test an intervention such as
comes, particularly pain.49 Other factors contributing physiotherapy can mean that the differences between
to the placebo effect in our trial include blinding of par- the groups substantially underestimate the total effects
ticipants, a treatment protocol involving considerable
interaction with the therapist (10 individual sessions),
Shoulder pain and disability index

100
and the fact that most (90%) participants expected to Active Placebo
gain a moderate to large benefit from active treatment 80
(given that positive expectations are known to be asso-
60
ciated with improved outcomes50). Interestingly, the
33% reduction in pain reported by the placebo group 40
is consistent with the 38% reduction found in the pla-
cebo group of our previous study in patients with 20
osteoarthritis of the knee, in which we used an identical
placebo treatment and pain measure.28 0

Our primary outcome measures included an assess-

10
Pain on movement

ment of pain but did not include an assessment of func-

tion in isolation. The results of both the primary and 8
secondary outcomes suggest that active treatment did
not substantially affect pain compared with placebo. 6
However, evidence from the secondary outcome of
shoulder pain and disability index function score sug- 4

gests that shoulder function was improved to a signifi-

2
cantly greater extent with active treatment. Given that
our primary outcome of shoulder pain and disability 0
0 11 22
index total score includes both the pain and function
subscales, the benefits of active treatment on function Weeks

could have been masked at the 11 week time point and Fig 2 | Mean (SD) shoulder pain and disability index (SPADI)
attenuated at the 22 week time point, when a significant and pain on movement for active and placebo groups at
treatment group effect was found for shoulder pain and baseline, 11 weeks, and 22 weeks

BMJ | ONLINE FIRST | bmj.com page 7 of 10

 RESEARCH

of treatment. This can lead to false negative results and techniques such as magnetic resonance imaging or ultra-
erroneous conclusions about efficacy. We would have sonography may further improve diagnostic accuracy,53
found significant beneficial effects of treatment if we but these still lack sensitivity for certain pathological

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
had compared the active treatment with a no treatment features,54 55 are costly, and are generally not used in the
control group as other studies in this area have primary care setting. Furthermore, physiotherapy treat-
done.21-23 ment is not directed at the specific pathology but at the
Several aspects of the active intervention warrant movement dysfunction and potential underlying
consideration. Firstly, our standardised programme mechanisms such as altered muscle function, tight struc-
could be argued to have failed to adequately treat the tures, and poor scapular and spinal posture that have
specific physical impairments that patients presented been reported in patients with rotator cuff disease.51
with and that relate to shoulder pain and dysfunction. Whether better outcomes with physiotherapy could be
We noted significant improvements in isometric gained if subgroups of rotator cuff disease could be
strength of 12-15% for the shoulder abductor and inter- defined and specifically treated is not known.
nal rotator muscles, as well as self reported strength
gains, suggesting some effect. Whether the active treat- Comparison with previous studies
ment also successfully tackled other physical factors A limited number of randomised controlled trials of
such as dynamic scapular control and thoracic posture physiotherapy modalities for chronic rotator cuff dis-
is unknown, as these were not measured. Limited ease have been done, and none has tested a combined
research shows that manual therapy techniques and intervention of manual therapy and exercise against a
exercises similar to those used in our study can alter placebo control to allow a direct comparison with our
shoulder and trunk biomechanics.51 results. The only study to use a placebo control treat-
Secondly, several participants failed to complete more ment (detuned laser twice weekly for six weeks) evalu-
than half of the prescribed home exercises, particularly ated a three to six month exercise programme
during the unsupervised follow-up period. Problems supervised by physiotherapists.56 This trial found that
with adherence to exercise programmes are common exercise resulted in a significant 66% reduction in pain
and reinforce the need to better incorporate strategies (measured on the Neer score), which is slightly higher
to enhance adherence, particularly when formal super- than the 43% reduction in pain with our active treat-
vision by therapists ceases. However, our results were ment. However, unlike our study, the researchers
unaltered when we reanalysed the data excluding parti- noted little improvement in their placebo group, ren-
cipants who failed to complete more than 50% of the dering their between group differences significant.
home exercises, suggesting that levels of adherence in Whether blinding of participants was successfully
this study did not unduly influence the outcome. achieved was not stated. Other studies have compared
Thirdly, to ensure a consistent approach and allow exercise with arthroscopic surgery and shown similar
replication, we chose to evaluate a standardised treat- beneficial outcomes in patients with rotator cuff
ment programme based on common elements identi- disease.56-58 The limited studies evaluating exercise
fied from our survey and from the literature. It does not combined with manual therapy have used an exercise
reflect the practice of every clinician involved in the only group as the comparator.59-61 These have found
conservative management of rotator cuff disease, and that the effects of exercise on both pain and function
our results cannot necessarily be generalised to other are augmented with manual therapy, providing a ratio-
manual therapy and exercise programmes given differ- nale for evaluating a combined intervention.
ences in type and dosage. Furthermore, as our treat-
ment was standardised it may have been ineffective Strengths and weaknesses
or even inappropriate for some patients, thus worsen- The strengths of our study include the rigorous study
ing symptoms and attenuating the treatment effects in design, adequate statistical power, excellent retention
the active group. However, a similar proportion of par- of participants, inclusion of a placebo control, and the
ticipants in both groups reported that they were worse use of several therapists and a variety of recruitment
after treatment (9/57 (16%) in the active group; 7/61 sources to increase the external validity of the results.
(12%) in the placebo group), suggesting that this was Our study has some limitations. Therapists were not
not the case. Further research is needed to evaluate blinded to treatment group, which is unavoidable in a
the efficacy of other physiotherapy protocols for trial of this nature. However, their interaction with
chronic rotator cuff disease and to assess treatment patients was standardised and any bias due to non-
that is tailored to individual patients, as occurs in clin- blinding of therapists would probably favour the active
ical practice.52 group, which was not particularly evident in the out-
Like most other trials of rotator cuff disease to date, we comes. Participants were blinded, but given the diffi-
chose to include participants on the basis of clinical fea- culty in designing a credible placebo for physiotherapy
tures alone. Rotator cuff disease is known to most com- interventions, blinding may not have been as complete
monly affect the supraspinatus tendon,10 but we cannot as can be achieved in a drug trial in which an identical
exclude the possibility that differences existed in the placebo pill can be administered. However, formal
underlying structural abnormalities within our study testing of the success of blinding indicated that we
population, although these were likely to be equally dis- achieved a moderate to high degree of blindedness,
tributed between the treatment groups. Imaging representing a statistically significant amount of
page 8 of 10 BMJ | ONLINE FIRST | bmj.com
 RESEARCH

Leaf, Arthur Lee, Stephen Maloney, Christine Roberts, Paul Visentini,

WHAT IS ALREADY KNOWN ON THIS TOPIC Jennifer Maciel, Richard Burton, Jane Burman, Debbie Virtue, and Simon
Wilson. Ben Metcalf also assisted with data analysis and preparation of
Rotator cuff disease is a common shoulder condition causing pain and loss of function
the manuscript.

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
Manual therapy techniques and exercise programmes are often used in the management of Contributors: KB, SC, and RB conceived and designed the trial protocol.
rotator cuff disease, yet little conclusive evidence supports or refutes their efficacy KB, RB, SG, SC, AH, and AF procured the project funding. SC and SG
designed the physiotherapy intervention. AF designed and coordinated
WHAT THIS STUDY ADDS the statistical analysis and generated the randomisation table. MS did the
analyses. EW was the project manager and blinded assessor. KB drafted
Immediate beneficial effects of a standardised manual therapy and home exercise the manuscript, and RB, EW, SC, SG, AH, MS, and AF contributed to the
programme were comparable to those of a realistic placebo treatment in middle aged to older manuscript. All authors read and approved the final manuscript. KB is the
guarantor.
adults with chronic rotator cuff disease
Funding: KB is funded in part by an Australian Research Council future
Benefits of manual therapy and exercise may accrue over time and may be of more value for fellowship. RB is funded in part by an Australian National Health and
improving shoulder function than pain Medical Research Council practitioner fellowship. This work was funded
by the National Health and Medical Research Council (project grant
#299840). Pilot funds were provided by ANZ Trustees, Department of
Physiotherapy and Victor Hurley Grant Royal Melbourne Hospital and the
blinding beyond that expected by chance. Our sample School of Physiotherapy, University of Melbourne. The study sponsors
size was reduced after a blinded interim analysis but had no role in the design; in the collection, analysis, and interpretation of
gave sufficient power (80%) to detect clinically mean- the data; or in the writing of the article and the decision to submit it for
publication.
ingful differences in the primary outcome. We Competing interests: All authors have completed the Unified Competing
replaced missing values by using the last observation Interest form at www.icmje.org/coi_disclosure.pdf (available on request
carried forward method that was commonly used at the from the corresponding author) and declare (1) RB and KB are partly
time the study was being planned but has fallen out of supported by fellowships from the National Health and Medical Research
Council and the Australian Research Council respectively. None of the
favour more recently.62 This method is unlikely to have other authors have financial support for the submitted work from anyone
influenced the results, given the small dropout rate (8/ other than their employer; (2) No financial relationships with commercial
120 participants, 7%) and the fact that we found similar entities that might have an interest in the submitted work; (3) No
outcomes when we used generalised estimating equa- spouses, partners, or children with relationships with commercial entities
that might have an interest in the submitted work; (4) No non-financial
tions to fit population averaged models to the known interests that may be relevant to the submitted work.
scores. Ethical approval: The Royal Melbourne Hospital Human Research Ethics
Committee approved the study (project #2001.115), and all participants
Conclusions and practice implications gave written informed consent.
Data sharing: Physiotherapy protocols, technical appendix, statistical
Our study showed that the particular manual therapy code, and dataset are available from the corresponding author at
and home exercise programme tested conferred no k.bennell@unimelb.edu.au.
additional benefit immediately after treatment com-
pared with a realistic placebo in middle aged to older 1 Chard MD, Hazleman R, Hazleman BL, King RH, Reiss BB. Shoulder
adults with chronic rotator cuff disease. However, disorders in the elderly: a community survey. Arthritis Rheum
1991;34:766-9.
given evidence of significantly greater improvements 2 Roquelaure Y, Ha C, Leclerc A, Touranchet A, Sauteron M, Melchior M,
with active treatment at follow-up in one of the primary et al. Epidemiologic surveillance of upper-extremity musculoskeletal
disorders in the working population. Arthritis Rheum
outcomes and in several secondary outcomes, the ben- 2006;55:765-78.
efits of manual therapy and exercise may accrue over 3 Badley EM, Tennant A. Changing profile of joint disorders with age:
time. Physiotherapy interventions may also be of more findings from a postal survey of the population of Calderdale, West
Yorkshire, United Kingdom. Ann Rheum Dis 1992;51:366-71.
value for improving shoulder function than pain per se 4 Smith KL, Harryman DT 2nd, Antoniou J, Campbell B, Sidles JA,
in this population. Clinicians should thus establish Matsen FA 3rd. A prospective, multipractice study of shoulder
function and health status in patients with documented rotator cuff
whether the patient’s primary presenting problem is tears. J Shoulder Elbow Surg 2000;9:395-402.
pain, impaired function, or both. If pain is a major fac- 5 Bridges-Webb C, Britt H, Miles D, Neary S, Charles J. Morbidity and
treatment in general practice in Australia 1990-91. Med J Aust
tor, then other treatments that reduce pain, such as cor- 1992;157:1-56S.
ticosteroid injections, may be more appropriate. If 6 Peters D, Davies P, Pietroni P. Musculoskeletal clinic in general
both pain and loss of function are factors, then drugs practice: study of one year’s referrals. Br J Gen Pract 1994;44:25-9.
7 Bartolozzi A, Andreychik D, Ahmad S. Determinants of outcome in
or other pain relieving treatments may be needed in the treatment of rotator cuff disease. Clin Orthop Relat Res
combination with manual therapy and exercise to ade- 1994;308:90-7.
8 Gartsman GM, Brinker MR, Khan M, Karahan M. Self-assessment of
quately treat all facets. To facilitate adherence, clini- general health status in patients with five common shoulder
cians should advise patients that the effects of manual conditions. J Shoulder Elbow Surg 1998;7:228-37.
therapy and exercise are not necessarily immediate but 9 Roquelaure Y, Mariel J, Fanello S, Boissiere JC, Chiron H, Dano C, et al.
Active epidemiological surveillance of musculoskeletal disorders in
may take several months before they are evident. a shoe factory. Occup Environ Med 2002;59:452-8.
Further research is needed to evaluate the effectiveness 10 Ostor AJ, Richards CA, Prevost AT, Speed CA, Hazleman BL. Diagnosis
and relation to general health of shoulder disorders presenting to
of different physiotherapy treatment regimens and primary care. Rheumatology (Oxford) 2005;44:800-5.
whether the combination of drug treatment with 11 Buchbinder R, Goel V, Bombardier C, Hogg-Johnson S. Classification
physiotherapy leads to greater benefits in people with systems of soft tissue disorders of the neck and upper limb: do they
satisfy methodological guidelines? J Clin Epidemiol 1996;49:141-9.
mild to moderate chronic rotator cuff disease. 12 Ekeberg OM, Bautz-Holter E, Tveita EK, Juel NG, Kvalheim S, Brox JI.
Subacromial ultrasound guided or systemic steroid injection for
We acknowledge the support and contribution of the Physiotherapy rotator cuff disease: randomised double blind study. BMJ
Department of the Royal Melbourne Hospital, particularly Geraldine 2009;338:a3112.
Millard and Lauren Andrew, as well as the project physiotherapists 13 Green S, Buchbinder R, Hetrick S. Physiotherapy interventions for
Richard Bohen, Tanja Farmer, Marie-Louise Francken, Nigel Friend, Jean shoulder pain. Cochrane Database Syst Rev 2003;2:CD004258.

BMJ | ONLINE FIRST | bmj.com page 9 of 10

 RESEARCH

14 Lombardi I Jr, Magri AG, Fleury AM, Da Silva AC, Natour J. Progressive 39 Whitfield K, Buchbinder R, Segal L, Osborne RH. Parsimonious and
resistance training in patients with shoulder impingement syndrome: efficient assessment of health-related quality of life in osteoarthritis
a randomized controlled trial. Arthritis Rheum 2008;59:615-22. research: validation of the assessment of quality of life (AQoL)
15 Glazier RH, Dalby DM, Badley EM, Hawker GA, Bell MJ, Buchbinder R, instrument. Health Qual Life Outcomes 2006;4:19.

BMJ: first published as 10.1136/bmj.c2756 on 8 June 2010. Downloaded from http://www.bmj.com/ on 19 April 2020 by guest. Protected by copyright.
et al. Management of common musculoskeletal problems: a survey 40 McNutt LA, Wu C, Xue X, Hafner JP. Estimating the relative risk in
of Ontario primary care physicians. CMAJ 1998;158:1037-40. cohort studies and clinical trials of common outcomes. Am J
16 Michener LA, McClure PW, Karduna AR. Anatomical and Epidemiol 2003;157:940-3.
biomechanical mechanisms of subacromial impingement syndrome. 41 James KE, Bloch DA, Lee KK, Kraemer HC, Fuller RK. An index for
Clin Biomech 2003;18:369-79. assessing blindness in a multi-centre clinical trial: disulfiram for
17 Dickens VA. Role of physiotherapy in the treatment of subacromial alcohol cessation—a VA cooperative study. Stat Med
impingement syndrome: a prospective study. Physiotherapy 1996;15:1421-34.
2005;91:159-64. 42 Roy JS, MacDermid JC, Woodhouse LJ. Measuring shoulder function:
18 Haahr JP, Ostergaard S, Dalsgaard J, Norup K, Frost P, Lausen S, et al. a systematic review of four questionnaires. Arthritis Rheum
Exercises versus arthroscopic decompression in patients with 2009;61:623-32.
subacromial impingement: a randomised, controlled study in 90 43 Kuijpers T, van der Windt DA, van der Heijden GJ, Bouter LM.
cases with a one year follow up. Ann Rheum Dis 2005;64:760-4. Systematic review of prognostic cohort studies on shoulder
19 Ludewig PM, Borstad JD. Effects of a home exercise programme on disorders. Pain 2004;109:420-31.
shoulder pain and functional status in construction workers. Occup 44 Thomas E, van der Windt DA, Hay EM, Smidt N, Dziedzic K, Bouter LM,
Environ Med 2003;60:841-9. et al. Two pragmatic trials of treatment for shoulder disorders in
20 Walther M, Werner A, Stahlschmidt T, Woelfel R, Gohlke F. The primary care: generalisability, course, and prognostic indicators. Ann
subacromial impingement syndrome of the shoulder treated by Rheum Dis 2005;64:1056-61.
conventional physiotherapy, self-training, and a shoulder brace: 45 Morton V, Torgerson DJ. Regression to the mean: treatment effect
results of a prospective, randomized study. J Shoulder Elbow Surg without the intervention. J Eval Clin Pract 2005;11:59-65.
2004;13:417-23. 46 Krogsboll LT, Hrobjartsson A, Gotzsche PC. Spontaneous
21 Faber E, Kuiper JI, Burdorf A, Miedema HS, Verhaar JA. Treatment of improvement in randomised clinical trials: meta-analysis of three-
impingement syndrome: a systematic review of the effects on armed trials comparing no treatment, placebo and active
functional limitations and return to work. J Occup Rehabil intervention. BMC Med Res Methodol 2009;9:1.
2006;16:7-25. 47 Barnett AG, van der Pols JC, Dobson AJ. Regression to the mean: what
22 Kuhn JE. Exercise in the treatment of rotator cuff impingement: a it is and how to deal with it. Int J Epidemiol 2005;34:215-20.
systematic review and a synthesized evidence-based rehabilitation 48 Paterson C, Dieppe P. Characteristic and incidental (placebo) effects
protocol. J Shoulder Elbow Surg 2009;18:138-60. in complex interventions such as acupuncture. BMJ
23 Trampas A, Kitisios A. Exercise and manual therapy for the treatment 2005;330:1202-5.
of impingement syndrome of the shoulder: a systematic review. Phys 49 Hrobjartsson A, Gotzsche PC. Is the placebo powerless? An analysis
Ther Rev 2006;11:125-42. of clinical trials comparing placebo with no treatment. N Engl J Med
24 Hawkins RJ, Kennedy JC. Impingement syndrome in athletes. Am J 2001;344:1594-602.
Sports Med 1980;8:151-8. 50 Crow R, Gage H, Hampson S, Hart J, Kimber A, Thomas H. The role of
25 Maitland GD. Vertebral manipulation. Butterworth-Heinemann, expectancies in the placebo effect and their use in the delivery of
2001. health care: a systematic review. Health Technol Assess
26 Bennell K, Coburn S, Wee E, Green S, Harris A, Forbes A, et al. Efficacy 1999;3:1-96.
and cost-effectiveness of a physiotherapy program for chronic rotator 51 Ludewig PM, Reynolds JF. The association of scapular kinematics and
cuff pathology: a protocol for a randomised, double-blind, placebo- glenohumeral joint pathologies. J Orthop Sports Phys Ther
controlled trial. BMC Musculoskelet Disord 2007;8:86. 2009;39:90-104.
27 Coburn S. Efficacy of physiotherapy for chronic rotator cuff 52 May S, Greasley A, Reeve S, Withers S. Expert therapists use specific
pathology: a randomised, double-blind, placebo controlled pilot trial clinical reasoning processes in the assessment and management of
[masters of physiotherapy thesis]. University of Melbourne, 2008. patients with shoulder pain: a qualitative study. Aust J Physiother
28 Bennell K, Hinman R, Metcalf B, Buchbinder R, McConnell J, McColl G, 2008;54:261-6.
et al. Efficacy of physiotherapy management of knee joint 53 Cullen D, Breidahl W, Janes G. Diagnostic accuracy of shoulder
osteoarthritis: a randomised double-blind placebo-controlled trial. ultrasound performed by a single operator. Australas Radiol
Ann Rheum Dis 2005;64:906-12. 2007;51:226-9.
29 Buchbinder R, Youd JM, Green S, Stein A, Forbes A, Harris A, et al. 54 Cothran RL Jr. Imaging in evaluating the rotator cuff. J Surg Orthop Adv
Efficacy and cost-effectiveness of physiotherapy following 2006;15:132-9.
glenohumeral joint distension for adhesive capsulitis: a randomized 55 Moosmayer S, Heir S, Smith HJ. Sonography of the rotator cuff in
trial. Arthritis Rheum 2007;57:1027-37. painful shoulders performed without knowledge of clinical
30 Crossley K, Bennell K, Green S, Cowan S, McConnell J. Physical information: results from 58 sonographic examinations with surgical
therapy for patellofemoral pain: a randomized, double-blinded, correlation. J Clin Ultrasound 2007;35:20-6.
placebo-controlled trial. Am J Sports Med 2002;30:857-65. 56 Brox JI, Staff PH, Ljunggren AE, Brevik JI. Arthroscopic surgery
31 Heald SL, Riddle DL, Lamb RL. The shoulder pain and disability index: compared with supervised exercises in patients with rotator cuff
the construct validity and responsiveness of a region-specific disease (stage II impingement syndrome). BMJ 1993;307:899-903.
disability measure. Phys Ther 1997;77:1079-89. 57 Haahr JP, Ostergaard S, Dalsgaard J, Norup K, Frost P, Lausen S, et al.
32 Williams JW Jr, Holleman DR Jr, Simel DL. Measuring shoulder Exercises versus arthroscopic decompression in patients with
function with the shoulder pain and disability index. J Rheumatol subacromial impingement: a randomised, controlled study in 90
1995;22:727-32. cases with a one year follow up. Ann Rheum Dis 2005;64:760-4.
33 Roach KE, Budiman-Mak E, Songsiridej N, Lertratanakul Y. 58 Ketola S, Lehtinen J, Arnala I, Nissinen M, Westenius H, Sintonen H,
Development of a shoulder pain and disability index. Arthritis Care et al. Does arthroscopic acromioplasty provide any additional value
Res 1991;4:143-9. in the treatment of shoulder impingement syndrome? A two-year
34 Mintken PE, Glynn P, Cleland JA. Psychometric properties of the randomised controlled trial. J Bone Joint Surg Br 2009;91:1326-34.
shortened disabilities of the arm, shoulder, and hand questionnaire 59 Bang MD, Deyle GD. Comparison of supervised exercise with and
(QuickDASH) and numeric pain rating scale in patients with shoulder without manual physical therapy for patients with shoulder
pain. J Shoulder Elbow Surg 2009;18:920-6. impingement syndrome. J Orthopaed Sports Phys Ther
35 Ten Klooster PM, Drossaers-Bakker KW, Taal E, van de Laar MA. 2000;30:126-37.
Patient-perceived satisfactory improvement (PPSI): interpreting 60 Conroy DE, Hayes KW. The effect of joint mobilization as a
meaningful change in pain from the patient’s perspective. Pain component of comprehensive treatment for primary shoulder
2006;121:151-7. impingement syndrome. J Orthopaed Sports Phys Ther
36 Walters SJ, Munro JF, Brazier JE. Using the SF-36 with older adults: a 1998;28:3-14.
cross-sectional community-based survey. Age Ageing 61 Senbursa G, Baltaci G, Atay A. Comparison of conservative treatment
2001;30:337-43. with and without manual physical therapy for patients with shoulder
37 Hawthorne G, Osborne R. Population norms and meaningful impingement syndrome: a prospective, randomized clinical trial.
differences for the assessment of quality of life (AQoL) measure. Aust Knee Surg Sports Traumatol Arthrosc 2007;15:915-21.
N Z J Public Health 2005;29:136-42. 62 Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al.
38 Osborne RH, Hawthorne G, Lew EA, Gray LC. Quality of life Multiple imputation for missing data in epidemiological and clinical
assessment in the community-dwelling elderly: validation of the research: potential and pitfalls. BMJ 2009;338:b2393.
assessment of quality of life (AQoL) instrument and comparison with
the SF-36. J Clin Epidemiol 2003;56:138-47. Accepted: 25 March 2010

page 10 of 10 BMJ | ONLINE FIRST | bmj.com