J Shoulder Elbow Surg (2014) 23, 500-507

www.elsevier.com/locate/ymse

SHOULDER

A blinded, randomized, controlled trial assessing

conservative management strategies for frozen shoulder
Sarah Russell, MSc, MCSPa, Arpit Jariwala, MChOrth, FRCS(Tr&Orth)b,
Robert Conlon, BSc, MCSPa, James Selfe, PhDc, Jim Richards, PhDc,
Michael Walton, MSc, FRCS(Tr&Orth)a,*

a
Wrightington Upper Limb Unit, Wigan, Lancashire, UK
b
Ninewells Hospital, Dundee, UK
c
University of Central Lancashire, Preston, Lancashire, UK

Background: There is little evidence for the optimal form of nonoperative treatment in the management of
frozen shoulder. This study assesses the efficacy of current physiotherapy strategies.
Methods: All primary care referrals of frozen shoulder to our physiotherapy department were included
during a 12-month period. Of these referrals, 17% met the inclusion criteria for primary idiopathic frozen
shoulder. The 75 patients were randomly assigned to 1 of 3 groups: group exercise class, individual phys-
iotherapy, and home exercises alone. A single independent physiotherapist, who was blinded to the treat-
ment groups, made all assessments. Range of motion, Constant score, Oxford Shoulder Score, Short Form
36, and Hospital Anxiety and Disability Scale (HADS) outcome measures were performed at baseline, 6
weeks, 6 months, and 1 year.
Results: The exercise class group improved from a mean Constant score of 39.8 at baseline to 71.4 at 6
weeks and 88.1 at 1 year. There was a significant improvement in shoulder symptoms on Oxford and Con-
stant scores (P < .001). This improvement was greater than with individual physiotherapy or home exer-
cises alone (P < .001). The improvement in range of motion was significantly greater in both physiotherapy
groups over home exercises (P < .001). HADS scores significantly improved during the course of treatment
(P < .001). The improvement in HADS anxiety score was significantly greater in both physiotherapy inter-
vention groups than in home exercises alone.
Conclusions: A hospital-based exercise class can produce a rapid recovery from a frozen shoulder with a
minimum number of visits to the hospital and is more effective than individual physiotherapy or a home
exercise program.
Level of evidence: Level I, Randomized Controlled Trial, Treatment Study.
! 2014 Journal of Shoulder and Elbow Surgery Board of Trustees.
Keywords: Shoulder; adhesive capsulitis; frozen shoulder; physiotherapy; SF-36; HADS; anxiety

Frozen shoulder, or adhesive capsulitis, is a condition of

Ethical approval: Stockport Local Research Ethics Committee Clinical
Trial Registration Number: 05/Q1401/86.
*Reprint requests: Mr. Michael Walton, MSc, FRCS(Tr&Orth),
Wrightington Upper Limb Unit, Hall Lane, Appley Bridge, WN6 9EP, UK.
E-mail address: mikewalton@shoulderdoc.co.uk (M. Walton).
uncertain etiology characterized by the spontaneous onset
of pain with significant restriction of both active and pas-
sive range of movement of the shoulder.33 It has been
classically divided into phases of freezing (insidious onset

1058-2746/$ - see front matter ! 2014 Journal of Shoulder and Elbow Surgery Board of Trustees.
http://dx.doi.org/10.1016/j.jse.2013.12.026
Physiotherapy randomized controlled trial for frozen shoulder 501

Table I Inclusion and exclusion criteria

Inclusion criteria
Age 40 to 70 years
Patients reported local shoulder pain, frequently present either over the anteromedial aspect of the shoulder extending distally into
the biceps region or over the lateral aspect of the shoulder extending into the lateral deltoid region. Symptoms were present for at
least 3 months.
Spontaneous onset of a painful stiff shoulder
Marked loss of active and passive global shoulder motion, with at least 50% loss of external rotation
Normal findings on anteroposterior and axillary radiographs of the glenohumeral joint
Exclusion criteria
Pathologic findings or glenohumeral osteoarthritis on radiographic evaluation
Clinical evidence of significant cervical spine disease
History of significant trauma to the shoulder
Local corticosteroid injection or any physiotherapy intervention to the affected shoulder within the last 3 months
Cerebrovascular accident affecting the shoulder
Inflammatory joint disease affecting the shoulder
Bilateral frozen shoulder due to possible underlying systemic cause
Thyroid disease
Any coronary event, postcoronary artery bypass, or catheterization before the clinical appearance of frozen shoulder
Prior surgery, dislocation, or fractures on the affected shoulder
Active medicolegal involvement

of diffuse shoulder pain with progressive loss of move-
ment), frozen (gradual subsidence of pain, plateauing of
stiffness with equal active and passive range of motion),
and thawing (gradual improvement of movement and res-
olution of symptoms).33 Differing classifications have also
been described; Neviaser described the process in 4 stages
based on arthroscopic findings,31 and a recent review by
Hanchard20 has suggested that two groups, pain predom-
inant and stiffness predominant, may be simpler and
more appropriate. All of the classifications, however, stress
the importance of recognizing that the disease process is a
continuum rather than having well-defined stages.
A primary or true frozen shoulder occurs when there is
no exogenous cause or preexisting condition. It is manifested
as an idiopathic painful shoulder with a decreased range of
movement in which no systemic diagnosis, precipitating
shoulder condition, or radiographic explanation can be
found.25,29,36 Arthroscopic and histologic studies have shown
that the condition is one of glenohumeral capsular contrac-
tion, particularly of the coracohumeral ligament within the
rotator interval.8 The condition is usually thought to be self-
limited, although Hand et al22 demonstrated that 41% of
patients have mild to moderate symptoms at 7 years and 6%
have severe ongoing symptoms with pain and functional loss.
Despite considerable scientific research, the etiology and
pathology of frozen shoulder remain unknown.23 The preva-
lence has been estimated at approximately 2% to 3% of adults
in the general population.29 However, Bunker8 calculated a
much smaller prevalence of 0.75% of the population on the
basis of clinic attendance in secondary care. It usually de-
velops between the ages of 40 and 70 years5,33,36 and rarely
recurs in the same shoulder unless an injury or disease process
predisposes the joint to repeated episodes of stiffness.13,21
Many studies have attempted to establish the most
effective treatment of frozen shoulder, but there still re-
mains much debate in the literature. Currently, there is no
agreement on the standard management of this condition.14
The controversy is due in part to a failure of many authors
to precisely define and accurately identify frozen shoulder
among other causes of shoulder pain and stiffness.17,36
The Chartered Society of Physiotherapy has completed a
project on the management of frozen shoulder.20 Conclu-
sions drawn from these evidence-based clinical guidelines
suggest that future researchers should report their physio-
therapy interventions in sufficient detail to remove ambi-
guity, consider multicenter trials, and focus on specific
stages of frozen shoulder. In line with these recommenda-
tions, the aim of this study was to investigate the clinical
effectiveness of common physiotherapy interventions in the
treatment of frozen shoulder using validated outcome
measures to determine effectiveness.
Methods
The study used a randomized controlled trial of 3 common
physiotherapy interventions. Patients were randomly allocated to
treatment groups, and the study conformed to the CONSORT
statement.1 All patients gave written informed consent before
participating in the study.
Eligible patients were all new referrals to the physiotherapy
department with a diagnosis of frozen shoulder. Patients were
assessed and inclusion and exclusion criteria verified (Table I).
Inclusion criteria were representative of the typical features of
frozen shoulder: an insidious onset of pain and stiffness with a
clinical reduction in range of motion, principally a >50% reduc-
tion in external rotation, without an underlying radiologic abnor-
mality.30 All patients were required to have had symptoms for a
502
minimum duration of 3 months to minimize the number of patients
in the early pain-predominant phase. The exclusion criteria served
to eliminate patients with an inappropriate diagnosis of idiopathic
frozen shoulder and other medical conditions that may complicate
the pathologic process. All patients were asked not to have adju-
vant therapy for the duration of the study, except oral analgesia.
At baseline, all patients underwent a standardized subjective
and objective examination, as recommended by Wadsworth37 and
Bowling et al.6 Range of motion was measured in a standardized
manner with a universal goniometer from bone landmarks. Most
patients were unable to reach 90! of abduction; therefore, external
rotation was measured at the maximum pain-free angle of
abduction. This point of abduction was recorded at baseline and
used for subsequent measures to ensure comparability of results.
Each measurement was recorded 3 times and a mean taken.
Routine anteroposterior and lateral radiographs were performed to
exclude bone causes of stiffness, such as osteoarthritis. Primary
and secondary outcome measures were taken.
Outcome measures
The primary outcome measure was the Constant-Murley score.11
This has been shown to reflect shoulder function with accuracy,
reliability, and reproducibility.2,16,39 The score combines subjective
and objective measures to produce a 100-point score, comprising 4
parameters: activities of daily living, range of motion, pain, and
strength. The minimal clinically important difference (MCID) is
the smallest change in a score that patients perceive as meaningful,
causing clinicians to consider a change in the patients manage-
ment.15 At the time of writing, there is no data that clearly states
the MCID for the Constant score. However, routine clinical prac-
tice within the organizations involved would normally consider a
change of approximately 15 points to be clinically important.
The secondary outcome measures were the Oxford Shoulder
Score, the Short Form 36 (SF-36) questionnaire, and the Hospital
Anxiety and Disability Scale (HADS). The Oxford Shoulder Score
is a subjective questionnaire that contains 12 questions derived
from 2 parameters, pain and function. Scores from each of the
questions are added to produce a single score ranging from 12
(least difficulties) to 60 (most difficulties).12 Patients complete the
score unaided. The SF-36 is a widely used, self-administered, 36-
item generic indicator used to assess general health.38 It has
recently been applied to the evaluation of shoulder disorders.35
This is a questionnaire designed to assess 8 dimensions of
health status, which include physical functioning (10 items), role
limitations due to physical health problems (4 items), bodily pain
(2 items), social functioning (2 items), mental health (5 items),
role limitations due to emotional problems (3 items), and vitality
and general health perceptions (5 items). HADS consists of 7
depressive items and 7 anxiety-related items40 and has been shown
to be a reliable and valid tool for measuring emotional distress in
medical populations.24
Intervention groups
Patients who met the inclusion criteria and agreed to participate in
the study were then randomly allocated to 1 of the 3 treatment
groups: group 1, exercise class plus home exercises; group 2, in-
dividual multimodal physiotherapy plus home exercises; and
group 3, home exercises alone. These groups were identified to
S. Russell et al.
reflect current national clinical practice. The physiotherapy
treatment period was limited to 6 weeks, after which all patients
continued with the home exercise program.
An independent statistician generated the assignment scheme
by use of computer-generated permuted block randomization. A
random block length (chosen with equal probability from blocks
of length 6, 9, or 12) was used.
The exercise class group treatment consisted of group therapy
scheduled twice per week for the 6-week intervention period. All
patients were given careful instruction and demonstration of each
exercise by a supervising physiotherapist. Patients performed a 50-
minute exercise circuit composed of 12 stations. Each 4-minute
station was designed to facilitate range of motion exercises at the
shoulder and thoracic spine. Stick, pulley, and ball techniques were
used to address forward elevation, abduction, extension, and internal
and external rotation. There was an additional station for scapula
setting exercises and 2 stations addressing trunk rotation and side
flexion. Exercise sheets were given to ensure compliance and to aid
in understanding of the circuit (Appendix 1). The patients were also
instructed on the specific shoulder exercises in the home exercise
program and given the information booklet (Appendix 2).
The individual multimodal physiotherapy group received 2
sessions of individual physiotherapy treatment per week for the 6-
week intervention period. The physiotherapist was a specialist in
musculoskeletal physiotherapy with 11 years of subspecialization
in shoulder therapy. The treatment program was based on local
practice and expert opinion, in the absence of a clear consensus in
the literature.34 There was no attempt to standardize this group,
and management decisions were made on an individual patient
basis as determined by the treating physiotherapist. Treatment
could be adjusted according to the severity of symptoms. It
included Maitland mobilizations that were progressed as the
condition improved, soft tissue massage, myofascial trigger point
release, heat, and stretches. The patients were also instructed on
the specific shoulder exercises in the home exercise program and
given the information booklet.
The home exercise group received instruction on the specific
shoulder exercises in the information booklet. The information
booklet included the home exercises; a description of frozen
shoulder; and advice on sleep, posture, and pain relief.
After baseline evaluation, outcome measures were taken at 6
weeks, 6 months, and 1 year. All outcome measures were per-
formed and recorded by an independent physiotherapist who was
not involved in direct treatment of any patients and was blinded to
the treatment allocation. Patients who expressed a desire to
withdraw from the trial because of inability to cope with ongoing
symptoms were recorded as having failed to respond to treatment
and offered alternative treatment.
Statistical analyses
Data were analyzed within groups to assess the effects of each
intervention on the outcome measures and between groups for
comparison of the effects of the intervention. All data were tested
to determine if they were normally distributed, and where
appropriate, a repeated-measures one-way analysis of variance
(ANOVA) on the outcome data was conducted. All data were
tested by the Mauchly test for sphericity. Pairwise comparisons by
use of the least squares difference were conducted to investigate
the differences between the treatment groups and at the time
Physiotherapy randomized controlled trial for frozen shoulder 503

Table II Mean Constant score domains and range of motion

Baseline 6 weeks 6 months 1 year
Activities of daily living
Exercise class 6.5 12.0 14.0 14.1
Individual physiotherapy 6.7 10.6 12.1 12.7
Home exercises 6.2 8.2 11.0 11.6
Range of movement
Exercise class 9.7 16.8 18.3 19.2
Individual physiotherapy 9.7 14.9 16.1 17.0
Home exercises 10.3 11.4 13.8 15.7
Pain
Exercise class 15.4 33.0 36.1 38.2
Individual physiotherapy 16.6 28.2 32.0 34.0
Home exercises 18.7 25.6 31.4 34.0
Strength
Exercise class 6.1 9.7 13.4 16.6
Individual physiotherapy 7.3 9.2 10.7 13.1
Home exercises 6.4 6.8 8.5 9.8
Total Constant score
Exercise class 37.5 71.5 82.0 88.1
Individual physiotherapy 40.2 62.9 70.8 77.8
Home exercises 41.7 52.0 64.8 72.0
Forward elevation, degrees (range)
Exercise class 95 (85-125) 140 (130-150) 153 (145-160) 166 (155-180)
Individual physiotherapy 96 (85-120) 136 (125-150) 151 (145-160) 165 (145-180)
Home exercises 96 (85-120) 112 (85-135) 129 (95-150) 146 (100-180)
External rotation, degrees (range)
Exercise class 15 (10-20) 39 (30-45) 53 (45-55) 58 (55-60)
Individual physiotherapy 16 (10-25) 37 (25-45) 52 (40-55) 57 (45-65)
Home exercises 16 (10-25) 28 (15-45) 38 (25-50) 49 (35-60)

intervals of outcome measurement after intervention. Adjustments
were made for multiple comparisons by the least significant dif-
ference. The baseline (preintervention) measurement was included
as a covariate as it will be related to the repeated measurements
after introduction of the different interventions rather than being
an outcome of the intervention. The effect of the intervention (the
average effect of the intervention over time) was then tested by the
main effect of intervention group; whether the effect of the
intervention varies over time is represented by the interaction
between the intervention group and the repeated group and the
repeated factor over time.
A power calculation was performed estimating the MCID of 15
points for the Constant score to achieve 80% power and 5% sig-
nificance. A cohort of 117 patients was required, with 39 in each
of the 3 treatment groups.
Statistical analysis was performed by the SPSS PASW Statis-
tics for Windows, Version 18.0 (SPSS Inc, Chicago, IL, USA).
Results
A total of 850 patients were referred to physiotherapy with
a primary care diagnosis of frozen shoulder during a 12-
month period. Of these, 705 did not fit the study inclusion
criteria for primary idiopathic frozen shoulder; 70 patients
declined to participate. Thus, 75 patients entered the study
and were randomly assigned to 1 of 3 groups: exercise class
(n 25), individual multimodal physiotherapy (n 24), or
home exercises (n 26).
The mean age was 51.1 years (40-65 years). The female-
to-male ratio was 1:1.14. The dominant arm was affected in
53% of the study population; 73% of patients were right-
handed. The mean duration of symptoms was 5.79 months
(4-10 months). The primary analysis was intention-to-treat
and involved all patients who were randomly assigned.
There were no statistically significant differences between
the groups in any of the measured baseline characteristics.
The mean Constant score at baseline was 39.8 (18-64), and
the mean Oxford score was 34.4 (20-48). The mean forward
elevation was 95! (85! -120! ), and the mean external rota-
tion was 16! (10! -25! ).
A repeated-measures ANOVA demonstrated a significant
improvement in both Constant and Oxford scores for all
groups between the different time intervals (P < .001).
Further analysis with a pairwise comparison allowed com-
parison between the individual treatment groups. At 6
weeks, the exercise class group demonstrated an improve-
ment in Constant score to a mean of 71.5 (60-89). By 1 year,
the exercise class group had improved Constant score to a
mean of 88.1 (71-96) compared with the home exercise
group score of 72.0 (49-91). This was a significantly greater
improvement in the exercise class group than in either the
504
Figure 1 Mean Constant score for each treatment group over
time.
Figure 2 Mean Oxford Shoulder Score for each treatment group
over time.
individual multimodal physiotherapy group (P < .001) or
the home exercise group (P < .001). The difference between
the exercise class and home exercises exceeded MCID at
each postintervention time point. Although significant, the
difference between the exercise class and individual phys-
iotherapy groups did not meet the MCID of 15 points. This
significant improvement was also demonstrated within each
of the Constant score domains of activities of daily living,
range of motion, pain, and strength (Table II). The exercise
class also showed a greater improvement than the individual
multimodal physiotherapy and home exercises groups on
the Oxford score (P .037; P < .001).
There was a significant improvement from baseline in
forward elevation and external rotation in all 3 groups. The
improvement was significantly greater in both of the phys-
iotherapy intervention groups over the home exercise group at
all time points (P < .001). There were no significant differ-
ences between the exercise class and individual physiotherapy
groups in terms of range of motion at any stage (Table II).
S. Russell et al.
Figure 3 Mean HADS anxiety score for each treatment group
over time.
The individual multimodal physiotherapy group showed
significantly better Constant scores (P .002) and Oxford
scores (P < .001) than the home exercise group at all time
points. A pairwise comparison showed a significant difference
between 6 weeks and 6 months (P < .001), 6 weeks and 1
year (P < .001), and 6 months and 1 year (P < .001) for both
the Constant and Oxford scores for all groups. This demon-
strates a continued improvement over time (Figs. 1 and 2).
All pretreatment and most post-treatment Oxford scores,
Constant scores, and HADS anxiety and depression scores
were strongly correlated (P < .001). HADS anxiety and
depression scores were significantly higher preoperatively
compared with any of the post-treatment time periods (P <
.001). Pairwise comparison between the individual groups
showed no significant difference between the physiotherapy
intervention groups (exercise class and individual multi-
modal), but both groups showed significant improvements
in HADS anxiety score over the home exercise group
(exercise class: mean difference #2.195, P < .001; indi-
vidual multimodal physiotherapy: mean difference #1.509,
P .024) (Fig. 3). There were no significant differences
between the groups on HADS depression score.
Within the domains of the SF-36, there was no signifi-
cant difference demonstrated between the groups in the
general health, physical function, role limitations due to
health or emotional problems, or vitality domains. There
was a significant improvement in bodily pain (P .011),
mental health (P .009), and social function (P < .001)
over time on repeated-measures ANOVA test. Pairwise
comparisons demonstrated a significant improvement in
bodily pain between the exercise class group and home
exercise group (P .032). There were no other significant
differences between the treatment groups for any of the
domains of the SF-36 outcome measure at any time point.
One patient from the exercise class group died and was lost
to follow-up. One patient in the multimodal physiotherapy
group was referred for a glenohumeral local anesthetic and
Physiotherapy randomized controlled trial for frozen shoulder
steroid injection at 6 weeks, and 2 patients from the home
exercise group withdrew at 6 months because of intolerable
pain. All were included in the analysis on an intention-to-treat
basis. No patient underwent surgery or reported any other
interventions during the time frame of the study.
Discussion
Hanchard et al20 reported that the evidence for the management
of frozen shoulder is inconclusive and is generally derived from
studies with weak methodology. Kelly et al26 have suggested
that there is no clear evidence to determine which patients may
need formal supervised therapy as opposed to a home exercise
program. However, the findings of this study support and pro-
vide substantial evidence for the use of physiotherapy, and in
particular a group exercise class, in the treatment of patients
with a diagnosis of idiopathic frozen shoulder.
Clinically, an effective treatment intervention should
result in a significant change in results during the first 6
weeks. With an MCID for the Constant score of 15 as a
reference, an improvement greater than the MCID was
demonstrated by 91% of patients in the exercise group,
68% in the individual multimodal physiotherapy group, and
41% in the home exercise group. The exercise class group
demonstrated a mean Constant score of 72 at 6 weeks
(increasing to 88 at 1 year), which is comparable to the
published outcome of 75.5 after arthroscopic capsular
release despite a significantly more invasive procedure.4
Both physiotherapy intervention groups also demonstrated
significant increase in objective measures, including range
of motion, over the home exercise group.
The results of the study confirmed that patients seen in
an exercise class, supervised by a physiotherapist, had
better patient-reported outcomes and recovered in a shorter
time than those patients in an individual multimodal
physiotherapy or home exercise program. Physiotherapy,
and in particular a group exercise class, provides a clinical
setting in which patients can discuss their condition with
others who are in a similar position. This may reassure
patients and provide them with peer support and the
motivation they need to continue and to progress in their
rehabilitation. Patients were taught self-management of
their condition and how to deal with any increase in pain.
Behavioral changes during the treatment period, relating to
improvement of self-management, could reduce the utili-
zation of health care services during the follow-up period
and reduce sick leave in patients.
We have highlighted that physiotherapy interventions
lead to a significant improvement in patient anxiety
compared with home exercise and that anxiety and
depression are strongly correlated with symptoms in this
frozen shoulder cohort. The impact and importance of
central sensitization aspects of shoulder pain are currently
being highlighted.19 This is the first study to our knowledge
demonstrating that physiotherapy interventions may be
505
particularly beneficial in improving this anxiety aspect of
shoulder pain.
The benefit of exercise classes demonstrated can influ-
ence clinical practice by potentially reducing the number of
individual physiotherapy treatment sessions, improving
cost-effectiveness, as suggested by Carr et al.10 It may also
improve care pathways by initiating effective management
from initial diagnosis. This could standardize treatment
outcomes and have an impact on the need for surgical or
more invasive interventions.
The findings of this study, however, are in contrast to
the work by Levine et al,28 who suggested that patients
prescribed a therapist-directed home exercise program had
the same outcomes at short- and long-term follow-ups as
those of patients treated with other interventions. Kivimaki
et al27 compared patients treated with a home exercise
program to those with manipulation under anesthesia and a
home exercise program. Other than a slight increase in
range of motion, the group performing just a home exercise
program did not differ at any follow-up in pain or working
ability.
Another interesting finding of this study relates to the
diagnosis of frozen shoulder. Only 17% of initial referrals
(145 of 850) met the inclusion criteria for primary idio-
pathic frozen shoulder. Whereas the specific alternative
diagnoses were not recorded, the inappropriate referrals
represented a large number of patients with rotator cuff or
impingement symptoms without stiffness. This suggests
that there is a need to educate primary care physicians and
physiotherapists involved in the diagnosis and management
of frozen shoulder to improve their clinical diagnostic ac-
curacy. This low number of true frozen shoulders in the
population questions the estimations of primary care
prevalence29 and is more in keeping with the lower esti-
mations made by Bunker.8 This difference in true preva-
lence made recruitment much slower than we had
anticipated and led to the key limitation of this study in that
the number of patients recruited did not meet our initial
power calculations. However, despite the smaller numbers,
we have demonstrated statistically significant differences
between the treatment modalities studied.
A further limitation of the study is the absence of a
natural history control group. The constraints of the ethics
process prevented inclusion of a no treatment arm in
the study. However, the home exercises group represents
a control against direct physiotherapy management and
may well represent a close approximation to the natural
history.
This study has also provided information about the
appropriate use of outcome measures. Both the Constant
score2,16,17,39 and Oxford score12,32 have been validated for
the assessment of shoulder conditions. Both of these
scoring systems have shown significant changes in reported
outcomes for the patient cohort in this study. The SF-36,
however, showed few significant differences overall and
only for bodily pain between the groups. This lack of
506
sensitivity of the SF-36 in the assessment of shoulder dis-
ease has also been reported by Carette et al,9 who found no
significant differences between the groups they analyzed in
their study comparing the use of corticosteroid injection, a
supervised exercise program, and a combination of the two
and placebo in the treatment of frozen shoulder. Beaton and
Richards3 concluded that the SF-36 is not sensitive enough
to detect the disability experienced by patients with upper
extremity problems. Griggs et al18 used the SF-36 in a
study evaluating the efficacy of a specific 4-direction
shoulder-stretching exercise program. They concluded
that the SF-36 did not demonstrate significantly lower
scores for the satisfied patients compared with the general
population. Buchbinder et al,7 in a previous trial of oral
steroids for frozen shoulder, discovered that only the bodily
pain subscale of the SF-36 detected a benefit of predniso-
lone over placebo at 3 weeks despite large, clinically sig-
nificant benefits observed for other outcomes, including
pain, function, and range of motion. This lack of sensitivity
to shoulder disease is possibly due to the focus of the SF-36
on lower limb aspects of musculoskeletal disease as
opposed to upper limb, which is a feature of many generic
measures of health.
This study demonstrates that an exercise class, aimed at a
rapid recovery rate with a minimum number of in-
terventions, provides superior patient-reported outcomes in
relieving the signs and symptoms of frozen shoulder
compared with those having individual multimodal phys-
iotherapy or performing home exercises. However, standard
multimodal physiotherapy remains a good alternative and
has been demonstrated to be significantly better than un-
supervised exercise at home. Anxiety and depression appear
to be an important part of frozen shoulder symptoms, and
physiotherapy interventions appear to particularly address
this aspect of the condition. We would caution against the
use of the SF-36 as it does not appear to be a sensitive
reflection of shoulder disease. We have highlighted the poor
level of diagnostic accuracy in primary care referrals and
emphasize the need for better education of primary care
physicians and physiotherapists in the assessment of
shoulder disease. In the current climate of greater emphasis
on referral management, care in the community, and pri-
mary care triage, this has become more important than ever.
Conclusions
A group exercise class provides superior outcomes in
relieving the signs and symptoms of frozen shoulder.
However, standard multimodal physiotherapy remains a
good alternative and has been demonstrated to be
significantly better than unsupervised exercise at home.
We would recommend a trial of physiotherapy for
stiffness- predominant frozen shoulders before more
invasive measures are considered.
S. Russell et al.
Disclaimer
The authors, their immediate families, and any
research foundation with which they are affiliated did
not receive any financial payments or other benefits
from any commercial entity related to the subject of
this article.
Supplementary data
Supplementary data related to this article can be found
online at http://dx.doi.org/10.1016/j.jse.2013.12.026.
References
1. Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D,
et al. The revised CONSORT statement for reporting randomised
trials: explanation and elaboration. Ann Intern Med 2001;134:663-94.
2. Bankes MJK, Crossman JE, Emery RJH. A standard method of
shoulder strength assessment for the Constant Score with a spring
balance. J Shoulder Elbow Surg 1998;7:116-21.
3. Beaton DE, Richards RR. Measuring function of the shoulder. A cross-
sectional comparison of five questionnaires. J Bone Joint Surg Am
1996;78:882-90.
4. Berghs BM, Sole-Molins X, Bunker TD. Arthroscopic release of ad-
hesive capsulitis. J Shoulder Elbow Surg 2004;13:180-5. http://dx.doi.
org/10.1016/j.jse.2003.12.004
5. Binder AI, Bulgen DY, Hazleman BL, Parr G, Roberts S. A controlled
study of oral prednisolone in frozen shoulder. Br J Rheumatol 1986;
25:288-92.
6. Bowling RW, Rochar PA, Erhard R. Examination of the shoulder
complex. Phys Ther 1986;66:1866-78.
7. Buchbinder R, Hoving JL, Green S, Hall S, Forbes A, Nash P. Short course
prednisolone for adhesive capsulitis (frozen shoulder or stiff painful
shoulder): a randomised, double blind, placebo controlled trial. Ann
Rheum Dis 2004;63:1460-9. http://dx.doi.org/10.1136/ard.2003.018218
8. Bunker TD. Time for a new name for frozen shoulderdcontracture of
the shoulder. Shoulder Elbow 2009;1:4-9. http://dx.doi.org/10.1111/j.
1758-5740.2009.00007.x
9. Carette S, Moffett H, Tardif J, Bessette L, Morin F, Fremont P, et al.
Intraarticular corticosteroids, supervised physiotherapy, or a combi-
nation of the two in the treatment of adhesive capsulitis of the
shoulder: a placebo-controlled trial. Arthritis Rheum 2003;48:829-38.
http://dx.doi.org/10.1002/art.10954
10. Carr JL, Klaber Moffett JA, Howarth E, Richmond SJ, Torgerson DJ,
Jackson DA, et al. A randomized trial comparing a group exercise
programme for back pain patients with individual physiotherapy in a
severely deprived area. Disabil Rehabil 2005;27:929-37. http://dx.doi.
org/10.1080/09638280500030639
11. Constant CR, Murley AGH. A clinical method of functional assess-
ment of the shoulder. Clin Orthop Relat Res 1987;(214):160-4.
12. Dawson J, Fitzpatrick R, Carr A. Questionnaire on the perceptions of
patients about shoulder surgery. J Bone Joint Surg Br 1996;78:593-600.
13. Di Fabio RP. Secrets of diagnosis. J Orthop Sports Phys Ther 1998;29:504.
14. Dundar U, Toktas H, Cakir T, Evcik D, Kavuncu V. Continuous pas-
sive motion provides good pain control in patients with adhesive
capsulitis. Int J Rehabil Res 2009;32:193-8. http://dx.doi.org/10.1097/
MRR.0b013e3283103aac
Physiotherapy randomized controlled trial for frozen shoulder
15. Fayers P, Machin D. Quality of life: the assessment, analysis and
reporting of patient-reported outcomes. 2nd ed. Chichester, UK:
Wiley; 2007. p. 441-3. http://dx.doi.org/10.1002/9780470024522
16. Gazielly F, Gleyze P, Matangnon C. Functional and anatomical
results after rotator cuff repair. Clin Orthop Relat Res 1994;(304):
43-53.
17. Green S, Buchbinder R, Hetrick S. Physiotherapy interventions for
shoulder pain. Physiotherapy 2003;89:335-6. http://dx.doi.org/10.
1016/S0031-9406(05)60024-7
18. Griggs SM, Ahn A, Green A. Idiopathic adhesive capsulitis: a pro-
spective functional outcome study of non-operative treatment. J Bone
Joint Surg Am 2000;82:1398-407.
19. Gwilym SE, Oag HC, Tracy I, Carr AJ. Evidence that central
sensitization is present in patients with shoulder impingement
syndrome and influences the outcome after surgery. J Bone Joint
Surg Br 2011;93:498-502. http://dx.doi.org/10.1302/0301-620X.
93B4.25054
20. Hanchard N, Goodchild L, Thompson J, OBrien T, Davison D,
Richardson C. Evidence-based clinical guidelines for the diagnosis,
assessment and physiotherapy management of contracted (frozen)
shoulder: a quick reference summary. Physiotherapy 2012;98:117-20.
http://dx.doi.org/10.1016/j.physio.2012.01.001
21. Hand C, Athanason N, Matthews T, Carr A. Pathology of frozen
shoulder. J Bone Joint Surg Br 2007;89:928-32. http://dx.doi.org/10.
1302/0301.620X.89B7.19097
22. Hand C, Clipsham K, Rees JL, Carr AJ. Long-term outcome of frozen
shoulder. J Shoulder Elbow Surg 2008;17:231-6. http://dx.doi.org/10.
1016/j.jse.2007.05.009
23. Hannafin JA, Chiaia TA. Adhesive capsulitis: a treatment approach.
Clin Orthop Relat Res 2000;(372):95-109.
24. Herrmann C. International experiences with the hospital anxiety and
depression scaleda review of validation data and clinical results.
J Psychosom Res 1997;42:17-41.
25. Kelly IG. Frozen shoulder. In: Kelly IG, editor. The practice of
shoulder surgery. Oxford: Butterworth-Heinnemann; 1993. p. 196-205
(ISBN 0750613831).
26. Kelley MJ, McClure PW, Leggin BG. Frozen shoulder: evidence and a
proposed model guiding rehabilitation. J Orthop Sports Phys Ther
2009;39:135-48. http://dx.doi.org/10.2519/jospt.2009.2916
507
27. Kivimaki J, Pohjolainen T, Malmivaara A, Kannisto M, Guillaume J,
Seitsalo S, et al. Manipulation under anesthesia with home exercises
versus home exercises alone in the treatment of frozen shoulder: a
randomized, controlled trial with 125 patients. J Shoulder Elbow Surg
2007;16:722-6. http://dx.doi.org/10.1016/j.jse.2007.02.125
28. Levine WN, Kashyap CP, Bak SF, Ahmad CS, Blaine TA,
Bigliani LU. Nonoperative management of idiopathic adhesive cap-
sulitis. J Shoulder Elbow Surg 2007;16:569-73. http://dx.doi.org/10.
1016/j.jse.2006.12.007
29. Lubiecki M, Carr A. Frozen shoulder: past, present and future.
J Orthop Surg (Hong Kong) 2007;15:1-3.
30. Miller MD, Wirth MA, Rockwood CA. Thawing the frozen shoulder:
the patient patient. Orthopedics 1996;19:849-53.
31. Neviaser RJ, Neviaser TJ. The frozen shoulder: diagnosis and man-
agement. Clin Orthop Relat Res 1987;(223):59-64.
32. Othman A, Taylor G. Is the Constant Score reliable in assessing pa-
tients with frozen shoulder? 60 shoulders scored 3 years after
manipulation under anaesthetic. Acta Orthop Scand 2004;75:114-6.
http://dx.doi.org/10.1080/00016470410001708230
33. Reeves B. The natural history of the frozen shoulder syndrome. Scand
J Rheumatol 1975;4:193-6.
34. Ryans I, Montgomery A, Galway R, Kernohan WG, McKane R. A
randomised controlled trial of intra-articular triamcinolone and/or
physiotherapy in shoulder capsulitis. Rheumatology (Oxford) 2005;
44:529-35. http://dx.doi.org/10.1093/rheumatology/keh535
35. Smith KL, Harryman DT, Antoniou J, Campbell B, Sidles JA,
Matsen FA. A prospective, multipractice study of shoulder function
and health status in patients with documented rotator cuff tears. J
Shoulder Elbow Surg 2000;9:395-402.
36. Stam H. Frozen shoulder: a review of current concepts. Physiotherapy
1994;80:588-99.
37. Wadsworth CT. Frozen shoulder. Phys Ther 1986;66:1878-83.
38. Ware J, Sherbourne CD. The MOS 36-item Short Form Health Survey
(SF36). Med Care 1992;30:473-81.
39. Yian EH, Ramappa AJ, Arneberg O, Gerber C. The Constant Score in
normal shoulders. J Shoulder Elbow Surg 2005;14:128-33. http://dx.
doi.org/10.1016/j.jse.2004.07.003
40. Zigmond AS, Snaith RP. The Hospital Anxiety and Depression Scale.
Acta Psychiatr Scand 1983;67:361-70.