512324 (1)

Prévia do material em texto

Australian Journal of Physiotherapy 2009 Vol. 55 – © Australian Physiotherapy Association 2009 17
Chen et al: Passive joint mobilisation for painful shoulders
Introduction
Shoulder pain and stiffness is common in the general 
community, with the prevalence up to 33% (Luime et al 
2004, McBeth and Jones 2007, van der Heijden 1999, van 
der Windt et al 1995). The shoulder is the most frequent site 
of musculoskeletal pain after the back and neck (Urwin et 
al 1998), with an incidence of 15 new episodes per 1000 
patients seen in the primary care setting (van der Windt et 
al 1995). Significant disability such as inability to sleep, 
limitations in activities of daily living, and a resulting loss 
of quality of life have been reported, especially in the elderly 
(Ostor et al 2005, Smith et al 2000, Vecchio et al 1995).
Physiotherapy intervention is often the first line of 
management for shoulder problems, with 53% to 79% of 
general practitioners referring to physiotherapists (Glazier 
et al 1998, van der Windt et al 1995). In Australia, shoulder 
pain accounts for up to 10% of all referrals to physiotherapists 
(Peters et al 1994).
Passive joint mobilisation and exercise therapy are 
commonly employed by physiotherapists for shoulder 
problems (Green et al 2003). Systematic reviews have found 
some evidence in support of manual techniques. However, 
these reviews included studies which investigated passive 
joint mobilisation directed at both shoulder region joints 
and vertebral column joints (Desmeules et al 2003, Green 
et al 2003, Michener et al 2004). Only two small, controlled 
clinical trials have specifically investigated the efficacy of 
passive joint mobilisation applied at the shoulder region 
joints in painful, stiff shoulders (Conroy and Hayes 1998, 
Nicholson 1985). Both trials reported no improvement in 
activity or range of shoulder motion in the group receiving 
joint mobilisations compared with the control group. 
However, both of these studies lacked the statistical power 
to detect small but clinically-meaningful effects of passive 
joint mobilisation applied to shoulder region joints.
Whilst there is growing evidence to support the efficacy 
of exercise for shoulder pain (Ainsworth and Lewis 2007, 
Ginn et al 1997, Ginn and Cohen 2005, Grant et al 2004, 
Trampas and Kitsios 2006) evidence for the efficacy of 
passive mobilisation of shoulder region joints to reduce pain 
and disability is lacking. Therefore, the research questions 
for this study were:
Is the addition of passive mobilisation of shoulder 1. 
region joints to advice and exercise for patients with 
shoulder pain and stiffness more effective in reducing 
pain and disability than advice and exercise alone?
Are any gains retained at 6 months?2. 
Method
Design
The study was a single-blinded randomised trial. After 
baseline measurements of pain, disability, and range of 
motion were taken, participants were randomly allocated 
into either the experimental or control group based on 
an assignment schedule that was stored in consecutively 
numbered, sealed opaque envelopes to ensure concealment. 
The assignment schedule was generated by an investigator 
Passive mobilisation of shoulder region joints plus advice 
and exercise does not reduce pain and disability more than 
advice and exercise alone: a randomised trial
Judy F Chen1,2, Karen A Ginn2 and Robert D Herbert3
1The Prince of Wales Hospital, 2The University of Sydney, 3The George Institute for International Health 
Australia
Question: Is the addition of passive mobilisation of shoulder region joints to advice and exercise for patients with shoulder 
pain and stiffness more effective than advice and exercise alone? Design: Randomised trial with concealed allocation, 
assessor blinding, and intention-to-treat analysis. Participants: 90 people who had shoulder pain and stiffness for more than 
one month. Intervention: All participants received advice and exercise. The experimental group also received passive joint 
mobilisation of shoulder region joints. Outcome measures: Primary outcome measures included pain and disability measured 
with the 13-point Shoulder Pain and Disability Index. Secondary outcome measures were self-perceived global improvement 
measured on a 6-point scale and active ranges of motion. Subjects received a maximum of 10 sessions of therapy. Outcome 
measurements were taken at baseline, one month, and six months. Results: The experimental group had 3% (95% CI –5 to 
11) less pain and disability than the control group at one month and 1% (95% CI –13 to 16) less pain at six months, which are 
statistically non-significant. Their global perceived effect was 0.1 out of 5 (95% CI –0.2 to 0.4) worse than the control group at 
one month and 0.1 (95% CI –0.5 to 0.7) better at 6 months, which are also statistically non-significant. Differences between 
groups in all range of motion measures were small and statistically non-significant. Conclusion: The addition of passive joint 
mobilisation of shoulder region joints is not more effective than advice and exercise alone for shoulder pain and stiffness. 
Trial registration: ACTRN 12605000080628. [Chen JF, Ginn KA, Herbert RD (2009) Passive mobilisation of shoulder 
region joints plus advice and exercise does not reduce pain and disability more than advice and exercise alone: a 
randomised trial. Australian Journal of Physiotherapy 55: 17–23]
Key words: Shoulder joint, Shoulder pain, Randomised controlled trial, Exercise, Physiotherapy
Australian Journal of Physiotherapy 2009 Vol. 55 – © Australian Physiotherapy Association 200918
Research
who was not associated with recruitment, intervention, or 
assessment. Disability and range of motion measures were 
obtained at baseline, four weeks, and six months after 
randomisation by a measurer blinded to group allocation. 
Global self-perceived improvement was assessed four weeks 
and six months after randomisation. To maintain assessor 
blinding, participants were specifically requested not to 
discuss any aspects of their intervention with the assessor 
at any stage of re-assessment.
Participants
All patients 18 years and above, presenting at a large 
metropolitan public hospital for physiotherapy intervention 
for shoulder pain and stiffness of more than one month’s 
duration were invited to participate if they understood 
spoken English. Patients were eligible to enrol in the study 
if the shoulder pain was unilateral, over the glenohumeral 
joint or in the proximal upper limb, and reproduced during 
shoulder movements. In addition they needed to have less 
than 140 degrees of active shoulder flexion and abduction 
range of motion or a greater than 10 cm hand-behind-back 
deficit compared to the unaffected side. In order to justify 
the need for passive joint mobilisation, they also had to 
have pain and/or stiffness during accessory movements 
of the shoulder region joints. They were excluded if they 
had trauma within the last month, inflammatory joint 
disease, local neoplastic disorder, a feeling of instability at 
the glenohumeral joint, contraindications to passive joint 
mobilisations at the glenohumeral, acromioclavicular, or 
sternoclavicular joints, or shoulder pain referred from the 
vertebral column structures. Referred pain was defined as 
pain reproduced by active neck movements or by palpation 
of the vertebral column, or if paraesthesia was present in 
the affected limb. Patients with bilateral shoulder pain were 
excluded as one outcome measurement relied on comparison 
with the unaffected side.
Intervention
The experimental group received passive joint mobilisations 
in the form of graded, passive, accessory movements 
directed either at the glenohumeral, acromioclavicular 
and/or sternoclavicular joint in addition to advice and 
exercise. Passive joint mobilisations wereperformed either 
as a passive oscillatory movement or a sustained stretch 
with or without tiny amplitude oscillations at the limit of 
the range. Only low-velocity mobilisations were used; 
manipulations involving high-velocity, low-amplitude 
thrusts were not employed. The movements were aimed at 
restoring structures within the target joints to their normal 
or pain-free positions so as to recover full-range, painless 
movement (Maitland 1991). Progression of these techniques 
was determined by the treating therapist based on the 
participant’s clinical signs and symptoms as per routine 
clinical practice, since the aim of this pragmatic study 
was to investigate the effectiveness of a commonly-used 
physiotherapeutic technique. In addition, participants were 
advised to avoid painful activities involving the shoulder and 
were also advised how to use pain-free methods to perform 
everyday activities such as dressing. They also received 
exercises aimed at restoring neuromuscular control of the 
shoulder muscles in order to restore the dynamic stability 
and muscle force couple co-ordination of the shoulder 
region (Ginn et al 1997, Ginn and Cohen 2005). Exercises 
were conducted in a pain-free manner so as to maximise 
normal muscle function and movement pattern (Ginn and 
Cohen 2005). Exercises were progressed using motor 
learning principles designed to improve shoulder function 
by gradually increasing the complexity of the exercises 
(Stevans and Hall 1998). Exercises which involved a single 
shoulder muscle force couple (such as rotation with the 
humerus supported into abduction range) were progressed 
to exercises which required multiple muscle force couples 
(such as range of motion). Exercises were tailored for the 
individual patient by the therapist and upgraded as muscle 
function improved. Participants were directed to perform 
the exercises at least twice daily.
The control group received advice and exercise only.
Intervention consisted of up to ten 30-minute individual 
sessions delivered within an 8-week period. Participants 
attended up to twice weekly initially, and then once a 
week. Experimental participants were required to receive a 
minimum of six sessions of passive joint mobilisation in the 
course of therapy. Participants were asked not to seek other 
therapy during the trial. Intervention was administered by a 
single experienced physiotherapist who met with one of the 
authors to confirm the intervention protocol and to clarify 
protocol issues as required. A data sheet was kept by the 
treating therapist to record the type of intervention given to 
each participant. The total number of interventions received 
by the participants in both groups, and the exact number 
of sessions of passive joint mobilisation performed in the 
experimental group were recorded.
Outcome measures
The primary outcome measure was the Shoulder Pain and 
Disability Index (SPADI) (Roach et al 1991) which is a self-
administered questionnaire designed to measure the effect 
of pathology of the shoulder on pain and disability. This 
was chosen for its overall validity and patient acceptability 
(Paul et al 2004). It is responsive to change (Bot et al 2004, 
Heald et al 1997, Paul et al 2004), quick to complete (Paul 
et al 2004), and has no floor or ceiling effects (Bot et al 
2004). The SPADI consists of 13 items divided into two 
subcategories addressing pain (five items) and disability 
(eight items). All items were rated with a visual analogue 
scale anchored at ‘No pain’ and ‘Worst pain imaginable’ for 
pain, and ‘No difficulty’ or ‘So difficult it required help’ 
for disability. A numeric value was obtained by dividing 
the scale into 10 segments from 0 to 10 for each item. The 
questionnaire required only 5–10 minutes to complete.
Secondary outcomes included global perceived effect and 
active shoulder range of motion. Global perceived effect 
was measured using a 6-point Likert scale with categories 
‘Significantly deteriorated’, ‘Slightly deteriorated’, ‘No 
change’, ‘Slightly improved’, ‘Significantly improved’ and 
‘Completely recovered’ (Likert 1932). In this study, active 
range of movement in flexion and abduction was measured 
using still photography as described by Ginn et al (1997). 
This method was chosen over goniometry as it was less 
likely to exacerbate shoulder symptoms due to the shorter 
time required to complete measurements. Hand-behind-
back range of movement was measured using a tape measure 
(Ginn and Cohen 2005). Participants were instructed to 
‘take the arm to as far as they could go’ and the distance 
between T1 and the styloid process was measured. The 
difference between the affected and non-affected side was 
reported so that the larger the negative value the worse the 
range of motion. The intra-tester reliability of active range 
of movement measurements taken by the assessor involved 
in the study was determined prior to the commencement of 
Australian Journal of Physiotherapy 2009 Vol. 55 – © Australian Physiotherapy Association 2009 19
the clinical trial. Intra-class correlation coefficients (ICC2,1) 
demonstrated excellent intra-rater reliability (0.92 to 0.98) 
for these measures (Shrout and Fleiss 1979).
Data analysis
The sample size for this study was based on a predetermined 
15-point difference between groups in the reduction on the 
SPADI score. Power calculations indicated that a sample 
of 90 participants (45 per group) would provide an 80% 
probability of detecting a 15-point difference on SPADI 
assuming a standard deviation of 24 points, with an alpha 
of 0.05, and an estimated loss to follow-up of 10%. A 
reduction of 13 points on SPADI has been recommended as 
the minimum clinically-worthwhile improvement (Roach et 
al 1991).
Analyses were conducted on an intention-to-treat principle, 
using all available data from all randomised participants. 
To estimate the effects of intervention, between-group 
differences of primary outcome measures (SPADI) and 
secondary outcomes (patient’s global self-perceived 
improvement scale and range of motion) at one month and 
six months were examined with analysis of covariance using 
a regression approach. Baseline values were included as 
covariates in the regression models to increase the precision 
of estimates.
Results
Flow of participants, therapists and centres 
through the trial
Between November 2004 and May 2007, 224 patients were 
screened for eligibility in this study. Of these, 97 (43%) met 
the inclusion criteria and were invited to participate. Ninety 
patients accepted the invitation and were randomised. The 
reasons for exclusion and flow of participants through the 
trial are illustrated in Figure 1. Three participants allocated 
to the control group withdrew after randomisation and did 
not commence intervention and no data were collected 
from these participants after baseline measurement. Six 
participants were unable to physically attend for range 
of motion measurements at six months but their pain 
and disability as well as the global perceived effect were 
obtained over the phone. Williams et al (1995) demonstrated 
that the SPADI was suitable for telephone administration 
should this be necessary. Therefore, at one month, 87% of 
participants were measured, and at six months 89% of the 
participants were measured for the primary outcome.
Characteristics of participants who completed the study and 
those lost to follow-up are shown in Table 1. Both groups 
were comparable except that there were more participants 
in the experimental group with concomitant neck pain. 
Participants lost to follow-up did not differ significantly from 
those who completed all measures, except that participants 
in the control group lost to follow-up at one month had a 
longer duration of complaint than the other participants.
Asingle senior physiotherapist with 13 years experience 
provided the intervention to both the experimental and 
control groups throughout the study. Although this therapist 
has not been involved in research before, her normal 
workload included over 40% of patients with shoulder 
problems.
This study was carried out at the Physiotherapy 
Musculoskeletal Outpatient Department in a major 
metropolitan public teaching hospital with a throughput 
of 440 patients per year. Referrals for intervention to the 
shoulder represent 51% of the throughput of this department 
per year.
Compliance with trial method
Eighty-seven participants attended for intervention. 
Participants in the experimental group received a mean of 8 
(range 1–10) sessions of intervention and those in the control 
group also received a mean of 8 (range 0–19) sessions. 
One participant in the control group was hospitalised 
due to a medical condition and subsequently developed a 
frozen shoulder, and hence attended a large number (19) of 
intervention sessions. In the experimental group, participants 
received a mean of 8 (range 1–10) sessions of mobilisation 
intervention to the shoulder joints, which was more than the 
minimum requirement of 6 sessions. Fourteen participants 
(16%), seven from each group, terminated intervention 
Chen et al: Passive joint mobilisation for painful shoulders
Table 1. Characteristics of participants.
Characteristic Participants
Randomised 
(n = 90)
Lost to Month 1 follow-up 
(n = 12)
Lost to Month 6 follow-up 
(n = 10)
Exp 
(n = 45)
Con 
(n = 45)
Exp 
(n = 6)
Con 
(n = 6)
Exp 
(n = 4)
Con 
(n = 6)
Gender, n female (%) 35 
(78)
29 
(64)
4 
(66)
4 
(66)
4 
(100)
3 
(59)
Age (yr), mean (SD) 64.7 
(12.5)
65.5 
(12.7)
61 
(16.6)
68.5 
(12.3)
61 
(9.1)
57 
(12.3)
Side of pain, n right (%) 30 
(67)
28 
(62)
5 
(83)
5 
(83)
3 
(75)
4 
(67)
Hand dominance, n right (%) 44 
(98)
42 
(93)
6 
(100)
6 
(100)
4 
(100)
6 
(100)
Duration of complaint (mth), 
mean (SD)
9.3 
(11.1)
11.1 
(13.6)
9.8 
(12.9)
16.5 
(17.7)
4.5 
(2.4)
17.5 
(19.7)
Previous history of shoulder 
problem, n yes (%)
20 
(44)
20 
(44)
2 
(33)
2 
(33)
1 
(25)
1 
(17)
Concomitant neck pain, n yes (%) 31 
(69)
19 
(42)
4 
(67)
1 
(17)
3 
(75)
0 
(0)
Exp= Experimental Group, Con=Control Group
Australian Journal of Physiotherapy 2009 Vol. 55 – © Australian Physiotherapy Association 200920
Research
against the treating physiotherapist’s advice (Figure 1). No 
participants reported having had other intervention during 
the study period.
Of the participants who received passive joint mobilisation, 
70% had mobilisation directed to the glenohumeral 
joint, 10% to the acromioclavicular joint, and 20% to a 
combination of these two joints. No passive mobilisation 
was directed to the sternoclavicular joint as this was not 
found to be symptomatic in this cohort. Anteroposterior 
glides at the end of available range of abduction were the 
most commonly-used technique for the glenohumeral joint, 
and anteroposterior and posteroanterior glides were equally 
Excluded (n = 21)
Cervical involvement (n = 6)• 
No active restriction (n = 8)• 
No passive restriction (n = 7)• 
Lost to Month 1 
follow-up
withdrew after • 
allocation (n = 3)
medical reasons • 
(n = 1)
did not attend • 
(n = 2)
Control Group
– advice
– exercise
– < 10 x 30-min 
sessions in 2 mth
Experimental Group
– passive joint mobilisation
– advice
– exercise
– > 6 and < 10 x 30-min 
sessions in 2 mth
Lost to Month 1 
follow-up
increased pain • 
(n = 1)
medical reasons • 
(n = 3)
did not attend • 
(n = 2)
Time
Month 0
Lost to Month 6 
follow-up
withdrew after • 
allocation (n = 3)
did not attend • 
(n = 3)
ControlGroup
D/C < 2 mth
Experimental Group
D/C < 2 mth
Lost to Month 6 
follow-up
medical reasons • 
(n = 1)
did not attend • 
(n = 3)
Month 1
Month 6
Figure 1. Design and flow of participants through the trial.
Patients with shoulder problems 
screened by telephone (n = 224)
Screened physically (n = 111)
Excluded (n = 113)
Poor English (n = 49)• 
No active restriction (n = 22)• 
Bilateral pain (n = 23)• 
Declined (n = 7)• 
No contact (n = 7)• 
Other reason (n = 5)• 
Measured pain and disability, range of motion
Randomised (n = 90)
(n = 45) (n = 45)
Measured pain and disability, range of motion, global perceived effect
(n = 39) (n = 39)
Measured pain and disability, range of motion, global perceived effect
(n = 41) (n = 39)
Australian Journal of Physiotherapy 2009 Vol. 55 – © Australian Physiotherapy Association 2009 21
commonly applied at the acromioclavicular joint. Only 
Grade II and Grade III movements were used.
Effect of intervention
Group data for all outcomes at baseline, one month and six 
months for control and experimental groups are presented 
in Table 2 while individual data are presented in Table 3 
(see eAddenda for Table 3). One month after randomisation, 
participants in both the experimental and control groups had 
improved in all outcome measures. Further improvements 
were seen at six months.
The experimental group had 3% (95% CI –5 to 11) less pain 
and disability than the control group at one month, and 
1% (95% CI –13 to 16) less at six months which are both 
statistically non-significant. Their global perceived effect 
was 0.1 out of 5 (95% CI –0.2 to 0.4) worse than the control 
group at one month, and 0.1 (95% CI –0.5 to 0.7) better at 
six months which are also both statistically non-significant. 
Likewise, differences between groups in all range of motion 
measures were small and statistically non-significant.
No adverse effects of intervention were reported.
Discussion
This is the first well-powered, randomised trial of passive 
joint mobilisation of the shoulder region joints of painful 
and stiff shoulders. The results of this study demonstrate 
conclusively that the addition of this commonly-used 
technique to advice and exercise is no more effective 
than advice and exercise alone. Pain and disability, global 
self-perceived effect, and range of motion were similar in 
outcome for both groups.
While there exists variable evidence for the efficacy of 
passive joint mobilisation in the intervention of other 
musculoskeletal problems (Green et al 2001, Hoeksma et 
al 2004), our randomised trial demonstrated that passive 
mobilisation of shoulder region joints conferred no added 
benefit to advice and exercise in the management of 
shoulder pain and stiffness. Several authors (Johnson et al 
2007, Vermeulen 2006) have demonstrated improvements in 
shoulder joint range with mobilisation in participants with 
shoulder problems; however, these were not randomised 
controlled trials and hence do not provide high level 
evidence.
Our results confirm those of previous small studies where 
passive mobilisation of shoulder region joints was not 
found to be effective in the management of shoulder pain 
and restriction. Nicholson (1985) carried out a clinical 
trial to determine the effects of passive mobilisation and 
active exercise in patients with painful restricted shoulders. 
Twenty patients with painful glenohumeral restrictions 
were randomly allocated to receive mobilisation and active 
exercise or active exercise alone for four weeks. The mean 
reduction in pain for the experimental group was –5.1 out 
of 10 (SD 4.6) compared with –2.9(SD 4.4) for the control 
group which resulted in a non-significant difference of 
–2.2 (95% CI –6.4 to 2.0). Only passive abduction range 
of motion increased significantly more in the mobilisation 
group than in the control group.
Conroy and Hayes (1998) examined the effect of shoulder 
region joint mobilisation as a component of comprehensive 
intervention for primary shoulder impingement syndrome. 
Fourteen participants with superiolateral shoulder pain, 
Chen et al: Passive joint mobilisation for painful shoulders
Ta
bl
e 
2.
 
M
e
a
n
 (S
D)
 ou
tc
o
m
e
s 
fo
r 
e
a
ch
 g
ro
u
p,
 
m
e
a
n
 (S
D)
 di
ffe
re
n
ce
 w
ith
in
 g
ro
u
ps
,
 
a
n
d 
m
e
a
n
 (9
5%
 C
I) d
iff
e
re
n
ce
 b
e
tw
e
e
n
 g
ro
u
ps
.
O
u
tc
o
m
e
G
ro
u
ps
D
iff
e
re
n
ce
 w
ith
in
 g
ro
u
ps
D
iff
e
re
n
ce
 b
e
tw
e
e
n
 g
ro
u
ps
*
M
o
n
th
 0
M
o
n
th
 1
M
o
n
th
 6
M
o
n
th
 1
 m
in
u
s 
M
o
n
th
 0
M
o
n
th
 6
 m
in
u
s 
 
M
o
n
th
 0
M
o
n
th
 1
 m
in
u
s 
 
M
o
n
th
 0
M
o
n
th
 6
 m
in
u
s 
 
M
o
n
th
 0
Ex
p 
n
 =
 4
5
Co
n 
n
 =
 4
5
Ex
p 
n
 =
 3
9
Co
n 
n
 =
 3
9
Ex
p 
n
 =
 4
1
Co
n 
n
 =
 3
9
Ex
p
Co
n
Ex
p
Co
n
Ex
p 
m
in
u
s 
co
n
Ex
p 
m
in
u
s 
co
n
Im
pa
irm
e
n
t
Sh
o
u
ld
e
r 
AR
O
M
Fl
ex
io
n
 (d
e
g)
97
 
(25
)
10
1 
(26
)
11
6 
(22
)
11
4 
(25
)
11
6 
(26
)
11
8 
(24
)
19
 
(25
)
12
 
(24
)
19
 
(26
)
16
 
(27
)
5 
(–4
 to
 1
4)
0 
(–1
0 
to
 1
1)
Ab
du
ct
io
n
 (d
e
g)
71
 
(23
)
75
 
(23
)
98
 
(30
)
97
 
(32
)
10
6 
(37
)
10
6 
(34
)
27
 
(31
)
21
 
(29
)
35
 
(36
)
30
 
(32
)
4 
(–9
 to
 1
7)
3 
(–1
2 
to
 1
9)
H
BB
 
(m
)
–
0.
13
 
(0.
09
)
–
0.
13
 
(0.
08
)
–
0.
09
 
(0.
09
)
–
0.
09
 
(0.
08
)
–
0.
06
 
(0.
06
)
–
0.
06
 
(0.
06
)
0.
02
 
(0.
09
)
0.
04
 
(0.
07
)
0.
06
 
(0.
09
)
0.
07
 
(0.
08
)
0.
00
 
(–0
.
03
 to
 0
.
04
)
0.
00
 
(–0
.
02
 to
 0
.
03
)
Ac
tiv
ity
 lim
ita
tio
n
SP
AD
I (%
)
65
 
(19
)
60
 
(18
)
46
 
(22
)
47
 
(22
)
47
 
(40
)
43
 
(38
)
–
19
 
(3)
–
13
 
(4)
–
18
 
(21
)
–
17
 
(20
)
–
3 
(–1
1 
to
 5
)
–
1 
(–1
6 
to
 1
3)
G
PE
 
(–2
 to
 3
)
2.
8 
(0.
6)
2.
9 
(0.
8)
2.
7 
(1.
4)
2.
6 
(1.
2)
2.
8 
(0.
6)
2.
9 
(0.
8)
2.
7 
(1.
4)
2.
6 
(1.
2)
–
0.
1 
(–0
.
4 
to
 0
.
2)
0.
1 
(–0
.
5 
to
 0
.
6)
*
 =
 m
e
a
n
 (9
5%
 C
I) d
iff
e
re
n
ce
 a
dju
st
e
d 
fo
r 
ba
se
lin
e
 s
co
re
s,
 
 
Sh
a
de
d 
ro
w
 =
 p
rim
a
ry
 o
u
tc
o
m
e
 m
e
a
su
re
,
 
Ex
p 
= 
ex
pe
rim
e
n
ta
l g
ro
u
p,
 
Co
n
 =
 c
o
n
tro
l g
ro
u
p,
 
AR
O
M
 =
 a
ct
iv
e
 r
a
n
ge
 o
f m
o
tio
n
,
 
H
BB
 =
 H
a
n
ds
 b
e
hi
n
d 
ba
ck
,
 
SP
AD
I =
 S
ho
u
ld
e
r 
Pa
in
 a
n
d 
D
is
a
bi
lit
y 
In
de
x,
 
G
PE
 =
 G
lo
ba
l p
e
rc
e
iv
e
d 
ef
fe
ct
Australian Journal of Physiotherapy 2009 Vol. 55 – © Australian Physiotherapy Association 200922
Research
decreased active humeral elevation, and limited overhead 
movement were randomly assigned to receive joint 
mobilisation and comprehensive intervention (consisting 
of hot packs, active range of motion, stretching, muscle 
strengthening, soft tissue mobilisation and patient 
education), or comprehensive intervention alone. Blinded 
assessors evaluated 24-hour pain, pain with the subacromial 
compression test, active range of motion and activity 
limitations, before and after nine sessions of interventions 
performed within three weeks. The group receiving passive 
joint mobilisations had less 24-hour pain and pain with the 
subacromial compression test than the control group but, 
as in Nicholson’s study, there were no difference in range 
of motion or activity limitations when compared with the 
control group.
Both of these studies demonstrated little or no benefit of 
passive joint mobilisation for painful, restricted shoulders. 
However both studies were small and lacked the power to 
detect clinically-important differences between groups. 
Furthermore, there were no long-term measures collected 
in either study. Our present study was designed to have 
adequate power to detect important intervention effects if 
they had been present, both in the short and medium term.
Michener et al (2004) concluded in their systematic review 
that the addition of shoulder region joint mobilisation in 
combination with exercise should be favoured over exercise 
alone for shoulder pain and restriction in subacromial 
impingement syndrome. This was based on one study by 
Bang and Deyle (2000) where passive joint mobilisation was 
applied to joints and soft tissues of the cervical and thoracic 
spine as well as to shoulder region joints. In our opinion, the 
evidence does not support this recommendation.
The strength of this randomised trial is that it incorporated 
features to minimise bias, thus enhancing the internal 
validity of the study. Participants were assigned to 
experimental and control groups using a concealed random 
allocation procedure. All measurements were performed 
by the same person who was blinded to group allocation. 
The inclusion criteria were general enough to reflect the 
variety of patients commonly presenting to public hospital 
physiotherapy outpatient departments with shoulder pain and 
stiffness. The size of the cohort yielded adequate statistical 
power. The control group received exercise therapy with 
proven efficacy (Ginn et al 1997, Ginn and Cohen 2005). 
Both experimental and control groups received the same 
average number of interventions, and contact time between 
therapist and participant was similar in both groups. Data 
analysis was by intention-to-treat. The analysis was blinded. 
A limitation of the study is that, due to the nature of the 
interventions, it was not possible to blind the participants 
and the treating therapist.
The shoulder region joint mobilisation intervention was 
pragmatic rather than prescriptive. In the trial, as in 
clinical practice, the type of mobilisation was tailored to 
the participant’s specific presentation. For example, the 
different grades and directions of passive joint mobilisation 
and joints treated were determined by the treating therapist 
in response to the presenting signs and symptoms in 
individual participants. This means that the findings of the 
trial can directly inform clinical practice.
The participants in this study represent the cross-section of 
patients referred to hospital physiotherapy departments in 
Australia. More than 50% of participants were referred to 
the physiotherapy department by general practitioners for 
‘painful, restricted shoulders’. Thelack of consensus of 
current diagnostic classification of shoulder problems and 
the lack of accuracy of clinical tests for shoulder conditions 
were the reasons that diagnostic labels were not used to select 
participants in this study (Bamji et al 1996, de Winter et al 
1999, Kuhn et al 2007, Liesdek et al 1997, Norregaard et al 
2002, Park et al 2005, Hughes et al 2008). The presence of 
pain and loss of range were also the criteria used in previous 
shoulder studies (Ginn et al 1997; Ginn and Cohen 2005).
This assessor-blinded randomised trial has shown that the 
addition of passive joint mobilisation to shoulder region 
joints to advice and exercise in patients with shoulder pain 
and stiffness is not more effective than advice and exercise 
alone. Individually-tailored exercises aimed at restoring 
neuromuscular control of the shoulder and advice on how 
to avoid shoulder pain alone are as effective in the short 
and medium term. The present study included participants 
with both shoulder pain and loss of range so it is not known 
whether our results apply to patients who have shoulder 
pain but no loss of range; future studies will be needed to 
address this issue. n
eAddenda: Table 3 available at AJP.physiotherapy.asn.au
Ethics: The South East Sydney Area Health Service Human 
Research Ethics Committee – Eastern Section, and The 
University of Sydney Human Research Ethics Committee 
approved this study. Informed consent was obtained from 
all participants before data collection began.
Competing interests: None declared.
Support: Australian Research Foundation, Musculoskeletal 
Physiotherapy Research Grant 2005. Rob Herbert is 
supported with a fellowship from the Australian NHMRC.
Acknowledgements: We are grateful to Christine Coorey 
from the Prince of Wales Hospital Physiotherapy Department 
who provided all the intervention.
Correspondence: Judy Chen, Physiotherapy Department, 
The Prince of Wales Hospital, Randwick NSW 2031 
Australia. Email: Judy.Chen@sesiahs.health.nsw.gov.au
References
Ainsworth R, Lewis JS (2007) Exercise therapy for the 
conservative management of full thickness tears of the 
rotator cuff: a systematic review. British Journal of Sports 
Medicine 41: 200–210.
Bamji AN, Erhardt CC, Price TR, Williams PL (1996) The 
painful shoulder: can consultants agree? British Journal of 
Rheumatology 35: 172–174.
Bang MD, Deyle GD (2000) Comparison of supervised exercise 
with and without manual physical therapy for patients with 
shoulder impingement syndrome. Journal of Orthopaedics 
Sports and Physical Therapy 30: 126–137.
Bot SD, Terwee CB, van der Windt DA, Bouter LM, Dekker J, 
de Vet HC (2004) Clinimetric evaluation of shoulder disability 
questionnaires: a systematic review of the literature. Annals 
of the Rheumatic Diseases 63: 335–341.
Conroy DE, Hayes KW (1998) The effect of joint mobilisation 
as a component of comprehensive treatment for primary 
shoulder impingement syndrome. Journal of Orthopaedics 
Sports and Physical Therapy 28: 3–14.
De Winter AF, Jans MP, Scholtlen RJ, Deville W, van 
Schaardenburg D, Bouter LM (1999) Diagnostic classification 
of shoulder disorders: interobserver agreement and 
determinants of disagreement. Annals of Rheumatic Disease 
58: 272–277.
Australian Journal of Physiotherapy 2009 Vol. 55 – © Australian Physiotherapy Association 2009 23
Chen et al: Passive joint mobilisation for painful shoulders
Desmeules F, Cote CH, Fremont P (2003) Therapeutic exercise 
and orthopaedic manual therapy for impingement syndrome: 
a systematic review. Clinical Journal of Sport Medicine 13: 
176–182.
Ginn KA, Cohen M (2005) Exercise therapy for shoulder pain 
aimed at restoring neuromuscular control: a randomized 
comparative clinical trial. Journal of Rehabilitation Medicine 
37: 115–122.
Ginn KA, Herbert RD, Khouw W, Lee R (1997) A randomized, 
controlled clinical trial of a treatment for shoulder pain. 
Physical Therapy 77: 802–809.
Glazier, RH, Dalby DM, Badley EM, Hawker GA, Bell MJ, 
Buchbinder R, Lineker SC (1998) Management of common 
musculoskeletal problems: a survey of Ontario primary care 
physicians. Canadian Medical Association Journal 158: 
1037–1040.
Grant HJ, Arthur A, Pichora DR (2004) Evaluation of 
interventions for rotator cuff pathology: a systematic review. 
Journal of Hand Therapy 17: 274.
Green S, Buchbinder R, Hetrick S (2003) Physiotherapy 
interventions for shoulder pain Cochrane Database of 
Systematic Reviews (2): CD004258.
Green T, Refshauge K, Crosbie J, Adams R (2001) A randomized 
controlled trial of a passive accessory joint mobilisation on 
acute ankle inversion sprains. Physical Therapy 81: 984–
994.
Heald SL, Riddle DL, Lamb RL (1997) The shoulder pain and 
disability index: the construct validity and responsiveness of 
a region-specific disability measure. Physical Therapy 77: 
1079–1089.
Hoeksma HL, Dekker J, Ronday HK, Heering A, van der Lubbe 
N, Vel C, Breedveld FC, van den Ende CH (2004) Comparison 
of manual therapy and exercise therapy in osteoarthritis of 
the hip: a randomized clinical trial. Arthritis and Rheumatism 
51: 722–729.
Hughes PC, Taylor NF, Green RA (2008) Most clinical tests 
cannot accurately diagnose rotator cuff pathology: a 
systematic review. Australian Journal of Physiotherapy 54: 
159–170.
Johnson AJ, Godges JJ, Zimmerman GJ, Ounanian LL (2007) 
The effect of anterior versus posterior glide joint mobilization 
on external rotation range of motion in patients with shoulder 
adhesive capsulitis. Journal of Orthopaedic and Sports 
Physical Therapy 37: 88–99.
Kuhn JE, Dunn WR, Ma B, Wright RW, Jones G, Spencer EE, 
Wolf B, Safran M, Spindler KP, McCarty E, Kelly B, Holloway B 
(2007) Multicenter orthopaedic outcomes network–shoulder 
(MOON Shoulder Group). Interobserver agreement in the 
classification of rotator cuff tears. American Journal of Sports 
Medicine 35: 437–441.
Liesdek C, van der Windt DA, Koes BW, Boulter LM (1997) 
Soft tissue disorders of the shoulder: a study of inter-
observer agreement between general practitioners and 
physiotherapists and an overview of physiotherapeutic 
treatment. Physiotherapy 83: 12–17.
Likert R (1932). A technique for the measurement of attitudes. 
Archives of Psychology 140: 1–55.
Luime JJ, Koes BW, Hendriksen IJ, Burdorf A, Verhagen AP, 
Miedma HS, Verhaar JA (2004) Prevalence and incidence of 
shoulder pain in the general population: a systematic review. 
Scandinavian Journal of Rheumatology 33: 73–81.
Maitland GD (1991) Peripheral Mobilisation (3rd edn). Oxford: 
Butterworth-Heinemann, Chs 4–5.
McBeth J, Jones K (2007) Epidemiology of chronic 
musculoskeletal pain. Best Practice and Research in Clinical 
Rheumatology 21: 403–425.
Michener LA, Walsworth MK, Burnet EN (2004) Effectiveness 
of rehabilitation for patients with subacromial impingement 
syndrome: a systematic review. Journal of Hand Therapy 17: 
152–164.
Nicholson GG (1985) The effects of passive joint mobilisation 
on pain and hypomobility associated with adhesive capsulitis 
of the shoulder. Journal of Orthopaedics Sports and Physical 
Therapy 6: 238–246.
Norregaard J, Krogsgaard MR, Lorenzen T, Jensen EM (2002) 
Diagnosing patients with longstanding shoulder joint pain. 
Annals of the Rheumatic Diseases 61: 646–649.
Ostor AJ, Richards CA, Prevost AT, Speed CA, Hazleman BL 
(2005) Diagnosis and relation to general health of shoulder 
disorders presenting to primary care. Rheumatology 44: 
800–805.
Park HB, Yokota A, Gill HS, Rassi GE, McFarland EG (2005) 
Diagnostic accuracy of clinical tests for the different degrees 
of subacromial impingement syndrome. Journal of Bone and 
Joint Surgery 87: 1446–1455.
Paul A, Lewis M, Shadforth MF, Croft PR, van der Windt DA, 
Hay EM (2004) A comparison of four shoulder-specific 
questionnaires in primary care.Annals of the Rheumatic 
Diseases 63: 1293–1299.
Peters D, Davies P, Pietroni P (1994) Musculoskeletal clinic in 
general practice: study of one year’s referrals. British Journal 
of General Practice 44: 25–29.
Roach KE, Budiman-Mak E, Songsiridej N, Lertratanakul Y 
(1991) Development of a shoulder pain and disability index. 
Arthritis Care and Research 4:143–149.
Shrout P, Fleiss J (1979) Intraclass correlations: uses in 
assessing rater reliability. Psychology Bulletin 86: 420–428.
Smith K L, Harryman II DT, Antoniou J Campbell B, Matsen 
III FA (2000) A prospective, multipractice study of shoulder 
function and health status in patients with documented 
rotator cuff tears. Journal of Shoulder and Elbow Surgery 9: 
395–402.
Stevans J, Hall K (1998) Motor skill acquisition strategies for 
rehabilitation of low back pain. Journal of Orthopaedic and 
Sports Physical Therapy 28: 165–167.
Trampas A, Kitsios A (2006) Exercise and manual therapy for 
the treatment of impingement syndrome of the shoulder: a 
systematic review. Physical Therapy Reviews 11: 125–142.
Urwin M, Symmons D, Allison T, Brammah T, Busby H, Rozby 
M, Simmons A, Williams G. E (1998) Estimating the burden of 
musculoskeletal disorders in the community: the comparative 
prevalence of symptoms at different anatomical site, and 
the relation to social deprivation. Annals of the Rheumatic 
Diseases 57: 649–655.
van der Heijden GJ (1999) Shoulder disorders: a state-of-the-art 
review. Best Practice and Research in Clinical Rheumatology 
13: 287–309.
van der Windt DA, Koes BW, de Jong BA, Bouter LM (1995) 
Shoulder disorders in general practice: incidence, patient 
characteristics, and management. Annals of the Rheumatic 
Diseases 54: 959–964.
Vecchio PC, Kavanagh RT, Hazleman BL, King RH (1995) 
Community survey of shoulder disorders in the elderly to 
assess the natural history and effects of treatment. Annals of 
the Rheumatic Diseases 54: 152–154.
Vermeulen HM, Rozing PM, Obermann WR, le Cessie S, Vliet 
V, Thea PM (2006) Comparison of high-grade and low-grade 
mobilization techniques in the management of adhesive 
capsulitis of the shoulder: randomized controlled trial. 
Physical Therapy 86: 355–368.
Williams JW Jr, Hollerman DR Jr, Simel DL (1995) Measuring 
shoulder function with the Shoulder Pain and Disability 
Index. Journal of Rheumatology 22: 727–732.