Background
Conditions causing shoulder pain are common and contribute substantially to the musculoskeletal morbidity of the community (Bjelle 1989). The prevalence of shoulder disorders has been reported to range from seven to 36% of the general population (Lundberg 1969). Shoulder disorders account for 1.2% of all general practice encounters in Australia, being third only to back and neck complaints as musculoskeletal reasons for primary care consultation (Bridges Webb 1992). In Dutch general practice the incidence of shoulder disorders has been estimated to be 11.2 per 1000 registered patients per year (van der Windt 1995). The shoulder is frequently injured, particularly in competitive sports. Eight to 13% of athletic injuries involve the shoulder (Hill 1983).
Prevalence of shoulder disorders has been shown to increase with age (Badley 1992). This finding has implications for the provision of health care in view of the aging of the population as a whole. In contrast, others (Allander 1974, Ingemar 1993) have demonstrated a decline in both the prevalence and incidence of shoulder pain with age, the peak prevalence occurring in the 56 - 60 year age group.
Substantial disability may result from shoulder disorders. Moving the shoulder allows placement of the hand, hence compromised shoulder mobility impacts substantially on the performance of tasks essential to daily living (e.g. dressing, personal hygiene, eating and work). In addition, shoulder pain is often associated with impaired ability to sleep, so affecting mood and concentration. People with shoulder pain have been shown to score substantially less than normal values on the SF-36 (a standardised measure of general health) for physical function, social function, physical role function, emotional role function and pain (Beaton 1996;Gartsman 1998). Shoulder disorders are often recalcitrant with some studies demonstrating persisting pain and disability from 12 months (van der Windt 1995) to 18 months (Chard 1991) in up to 50% of cases.
There are many commonly employed forms of treatment for shoulder disorders, including, non-steroidal anti-inflammatory drugs, glucocorticosteroid injections, oral glucocorticosteroid medication, manipulation under anaesthesia, physical therapy, hydrodilatation (distension arthrography) and surgery. A previous version of this systematic review of randomised controlled trials investigated all these treatments and concluded that there was very little evidence to either support or refute the efficacy of interventions commonly used to treat shoulder pain. Furthermore, the interpretation of results of studies that have been performed is hampered by the fact that these disorders are labelled and defined in diverse and often conflicting ways. (Green 1998) In a review of the diagnostic labels and/or definitions of the study populations, we concluded that most trials can be broadly categorised as studying adhesive capsulitis (specific diagnoses also including periarthritis and frozen shoulder) and/or rotator cuff tendonitis disease Green 1998. Shoulder pain and disorder may be caused by varying underlying pathologies, and the diagnostic criteria for defining these disorders are not consistently nor reliably applied. No standardised definitions are used and often there are conflicting criteria defining the same condition in different trials.
Since our original review (Green 1999), many new clinical trials, studying a diverse range of interventions, have been performed. In order to update and simplify the review, it has been subdivided into a series of reviews investigating the evidence for efficacy of single interventions. The review has also been broadened by including all randomised or pseudo-randomised clinical trials regardless of whether outcome assessment was blinded.
This review examines the evidence for efficacy and safety of physiotherapy for the treatment of adults with shoulder pain. Physiotherapy encompasses a broad range of interventions. This group of interventions are often the first line of management for shoulder pain. The aim is to relieve pain, promote healing, reduce muscle spasms, increase joint range and strengthen weakened muscles and ultimately to prevent and treat functional impairment (Lee 1973). Physiotherapy interventions include manual physical therapy where passive joint mobilisation is employed to mobilise and stretch the soft tissue. Supervised and prescribed exercises aim to improve range of movement and muscle function by restoring shoulder mobility and stability. Physiotherapy interventions also include a number of electrotherapeutic modalities including Laser Therapy, Ultrasound, Bipolar Interferential Current, Transcutaneous Electromagnetic Stimulation, and Pulsed Electromagnetic Field Therapy. Laser therapy is light amplification by stimulated emission of radiation. This results in a beam of light of a single frequency with little divergence, thought to reduce inflammation and improve circulation (England 1989). Ultrasound is used as a physiotheray intervention for its physiological effects which include argumentation of blood flow, increased capillary permeability and tissue metabolism, enhancement of tissue extensibility, elevation of pain threshold, and alteration of neuromuscular activity leading to muscle relaxation (Downing 1986). Bipolar Interferential Current is believed to promote recovery by elevation of the pain threshold and promotion of muscle relaxation (van der Heijden 1996).Transcutaneous Electromagnetic Stimulation (TENS) uses analgesic currents and while its mechanism of action is not completely understood it is thought that it serves to release endogenous opiates in specific areas of the Central Nervous System (Herrera-Lasso 1993). Pulsed Electromagnetic Field Therapy is thought to improve vascularisation, so promoting healing (Binder 1984). In practice, patients with shoulder pain seldom receive a single treatment intervention in isolation.
This review will specifically address the effectiveness of physiotherapy interventions alone or in combination for relief of pain and dysfunction of the shoulder.
Objectives
To determine the efficacy of physiotherapy interventions for shoulder pain and dysfunction.
Methods
Criteria for considering studies for this review
Types of studies
a) Randomised or pseudo-randomised controlled trials. Studies where participants were not randomised into intervention groups were excluded from the review.
b) Trials in which allocation to treatment or control group was not concealed from the outcome assessor were included but recorded as such in the table of included studies.
c) Studies in all languages were translated into English and considered for inclusion in the review.
Types of participants
Inclusion in this review was restricted to trials with participants meeting the following criteria:
a) Adults >16 years of age.
b) Shoulder pain or disorder for greater than 3 weeks. Studies that included various soft tissue disorders were considered if the results for shoulder pain were presented separately or if 90% or more of participants in the study had shoulder pain.
c) Studies of participants suffering a history of significant trauma or systemic inflammatory conditions such as rheumatoid arthritis, hemiplegic shoulders, post-operative and peri-operative shoulder pain and pain in the shoulder region as part of a complex myofacial neck/shoulder arm pain were excluded.
Trials were sub grouped into type of shoulder disorder for analysis (see methods section).
Types of interventions
All randomised controlled comparisons of a physiotherapist delivered intervention versus placebo, no treatment, another intervention, or of varying physiotherapy interventions compared to each other were included.
Types of outcome measures
No studies were excluded on the basis of outcome measure used. The clinically relevant outcomes of interest in shoulder disorder are pain, range of motion (active and passive), function/ disability and quality of life, strength, return to work, participants' perception of overall effect, global preference, physicians' preference and adverse effects.
Search methods for identification of studies
MEDLINE, EMBASE, CINAHL (includes all major physiotherapy and occupational therapy journals from U.S.A., Canada, England, Australia and New Zealand), and Science Citation Index (SCISEARCH) were searched 1966 to June 2002 .
1 Shoulder Pain/
2 Shoulder Impingement Syndrome/
3 Rotator Cuff/
4 exp Bursitis/
5 ((shoulder$ or rotator cuff) adj5 (bursitis or frozen or impinge$ or tendinitis or tendonitis or pain$)).mp.
6 rotator cuff.mp.
7 adhesive capulitis.mp.
8 or/1-7
9 exp Rehabilitation/
10 exp Physical Therapy Techniques/
11 exp Musculoskeletal Manipulations/
12 exp Exercise Movement Techniques/
13 exp Ultrasonography, Interventional/
14 (rehabilitat$ or physiotherap$ or physical therap$ or manual therap$ or exercis$ or ultrasound or ultrasonograph$ or TNS or TENS or shockwave or electrotherap$ or mobili$). mp.
15 or/9-14
16 Clinical trial.pt
17 random$.mp.
18 ((single or double) adj (blind$ or mask$)).mp.
19 placebo$.mp.
20 or/16-19
21 8 and 15 an 20
In addition, the Cochrane Controlled Trials Register (CCTR) Issue 2, 2002 was searched.
Data collection and analysis
Following identification of potential trials for inclusion by the previously outlined search strategy, the methods sections of all identified trials were reviewed independently according to predetermined criteria (see selection criteria), by two reviewers. All articles were coded and details of source, intervention, population and funding recorded. Where the two reviewers disagreed, discussion was facilitated in order to reach consensus. If this was to fail, the trial was sent to a third reviewer for arbitration.
Trials meeting inclusion criteria were collated, and the methods and results sections re-sent to the same two reviewers for assessment of trial validity and data extraction.
ASSESSMENT OF VALIDITY
Validity of included trials was assessed by comment on whether they met key criteria (including appropriate randomisation, allocation concealment, blinding, number lost to follow up and intention to treat analysis). These criteria were based on the PEDro scale specifically designed and validated for the assessment of validity for trials of physiotherapy interventions (http://ptwww.cchs.usyd.edu.au/pedro/). Trials were not scored numerically. The only quantitative scoring was given for allocation concealment, ranked as:
A: adequate
B: unclear, or
C: inadequate.
Whether or not trials met the key methodological criteria was recorded on a pre-piloted data extraction sheet and later transposed into the "Characteristics of Included Studies" table.
Validity of trials was assessed in this qualitative way as opposed to using a numerical or summary scale. There are concerns regarding the validity of such scales and a lack of information about whether all the criteria included in such scales impact on the overall outcome of the trial (Juni 1999).
DATA EXTRACTION AND ANALYSIS
In order to assess efficacy, raw data for outcomes of interest, specifically means and standard deviations for continuous outcomes and number of events for binary outcomes were extracted where available from the published reports. All standard errors of the mean were converted to standard deviation. Wherever reported data was converted or imputed, this was recorded in the notes section of the included studies table. For trials where the required data was not reported or not able to be calculated, further details were requested from first authors. If no further details were provided, the trial was included in the review and fully described, but not included in the meta-analysis. An entry to that effect was made in the notes section of the included studies table.
When trial results were not normally distributed and so reported as median and range, the trial was not included in the meta-analysis but results presented in Additional Tables.
Meta-analysis was facilitated by RevMan 4.1. The following choices of statistic and 95% confidence intervals were presented for all outcomes.
CONTINUOUS OUTCOMES:
Weighted mean difference using a fixed effect model was selected when outcomes were measured on standard scales. When outcomes were reported on non standard scales, using differing units and methods of assessment (for example disability scales), a standardised mean difference was selected. Possible clinical reasons for heterogeneity were explored, and in the presence of significant heterogeneity, trial results were not combined.
DICHOTOMOUS OUTCOMES:
Relative risk using a fixed effects model was selected for interpretation of dichotomous outcome measures in this review as we believe that this is the most appropriate statistic for interpretation when the event is common . Reasons for heterogeneity were evaluated and in the event of significant heterogeneity trial results were not pooled.
SUBGROUP ANALYSIS
Shoulder pain and disorder may be caused by varying underlying pathologies, and the diagnostic criteria for defining these disorders are not consistent nor reliably applied. In general, adhesive capsulitis was defined as the presence of pain with restriction of active and passive glenohumeral joint movements, and rotator cuff tendonitis was defined by the presence of painful arc and pain with resisted movements, and/or normal passive range of motion. However there were no standardised definitions used in the included trials and often there were conflicting criteria defining the same condition in different trials. For example, "pain with resisted movements of the shoulder and loss of passive abduction" was used to define rotator cuff tendonitis in one trial, whereas another trial used "pain on resisted abduction and full passive range of motion". Based upon review of the diagnostic labels and/or definitions of the study populations (Green 1998), most trials could be broadly categorised as studying adhesive capsulitis (specific diagnoses also including periarthritis and frozen shoulder) and/or rotator cuff tendonitis disease. For the purposes of subgroup analysis rotator cuff disease was categorised as tendinitis (specific diagnoses also including supraspinatus, infraspinatus and subscapularis tendonitis) or full rotator cuff tear. Results for each intervention were analysed within each diagnostic subgroup where described. If not described by the trialist, the population was labelled as general shoulder pain. This was planned a priori.
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies.
Individual studies are fully described in the table of included studies. Sixty seven potentially eligible trials were identified by the search strategy. There was agreement in all cases between reviewers on inclusion of studies. A total of twenty-six trials fulfilled the inclusion criteria. The reasons for exclusion of the other 41 trials are listed in the table of excluded studies. The 26 included studies are described below.
PHYSIOTHERAPY MODALITIES COMPARED WITH PLACEBO OR NO TREATMENT
Fourteen studies compared a physiotherapy modality to placebo (one study included two different modalities and placebo). The physiotherapy modalities studied were bipolar interferential current (one trial) (van der Heijden 1999), ultrasound (five trials) (van der Heijden 1999; Ebenbichler 1999; Berry 1980; Downing 1986; Nykanen 1995), laser (three trials) (Taverna 1990; Saunders 1995; England 1989) pulsed electromagnetic field (two trials) (Dal Conte 1990; Binder 1984), combined iontophoresis of acetic acid plus ultrasound (one trial) (Perron 1997), supervised exercises (two trials) (Brox 1993/7; Ginn 1997) and mobilisation (one trial) (Bulgen 1984).
COMPARISONS OF ONE TYPE OF PHYSIOTHERAPY MODALITY TO ANOTHER
Eight trials compared one type of physiotherapy modality to another.
Electrotherapeutic agents compared to non electrotherapeutic interventions:
One trial compared electrotherapy plus exercise to mobilisation and manipulation (Winters 1997/9) and one trial compared physiotherapy including electromagnetic therapy to an identical intervention without electromagnetic therapy (Leclaire 1991).
One type of electrotherapeutical agent to another:
One trial compared bipolar interferential current to ultrasound (van der Heijden 1999) and two compared transcutaneous nerve stimulation (TENS) to ultrasound (Herrera-Lasso 1993; Shehab 2000)
Manual interventions:
Three trials compared mobilisation plus exercise to exercise alone (Nicholson 1985; Conroy 1998; Bang 2000)
Exercise interventions:
One trial investigated the effect of isokinetic resisted exercise compared to biofeedback (Reid 1996).
PHYSIOTHERAPY MODALITIES COMPARED TO OTHER, NON PHYSIOTHERAPY TREATMENT INTERVENTIONS
Seven trials compared injection to physiotherapy. These comprised two trials comparing intra-articular corticosteroid injection with a combined physiotherapy intervention (van der Windt 1998; Berry 1980), two trials comparing intra-articular and subacromial corticosteroid injection to electrotherapy and exercises (Winters 1997/9; Lee 1973), and three comparing injections to mobilisation and manipulation (Winters 1997/9; Bulgen 1984; Dacre1989).
One trial compared laser to non-steroidal anti-inflammatory medication (England 1989).
Risk of bias in included studies
Included studies were of varying methodological quality. A full description of whether or not the methods of each trial met the predetermined quality assessment criteria can be found in the table of included studies. Trial populations were generally small (median sample size = 48, range 14 to 180) with many trials underpowered to demonstrate a difference between groups if one was present. Six of the 26 trials (23%) had adequate allocation concealment (Downing 1986; Ebenbichler 1999; Ginn 1997; Lee 1973; van der Heijden 1999; van der Windt 1998) (1 unclear and 19 inadequate), 19/26 (73%) trials had a blinded outcome assessor (Bang 2000; Berry 1980; Binder 1984; Brox 1993/7; Conroy 1998; Dacre1989; Ebenbichler 1999; England 1989; Ginn 1997; Leclaire 1991; Nicholson 1985; Nykanen 1995; Perron 1997; Saunders 1995; Shehab 2000; Taverna 1990; van der Heijden 1999; van der Windt 1998; Vecchio 1993) and 10/26 (38%) trials blinded the participants ( Binder 1984; Dal Conte 1990; Downing 1986; Ebenbichler 1999; England 1989; Leclaire 1991; Nykanen 1995; Saunders 1995; Taverna 1990; Vecchio 1993). In eight trials (8/26; 31%) there was greater than 20% loss to follow-up (Brox 1993/7; Downing 1986; Lee 1973; Nicholson 1985; Reid 1996; Saunders 1995; Shehab 2000; Winters 1997/9) and in 6/26 (23%) the trialists described intention to treat analysis (Brox 1993/7; Saunders 1995; Shehab 2000; Taverna 1990; van der Heijden 1999; van der Windt 1998).
Twenty trials (20/26; 77%) presented sufficient data to be included in meta-analysis, two presented data in a form which could not be included in meta-analysis so these results are included as additional tables, and four trials did not present any data that could be included in the review.
Effects of interventions
PHYSIOTHERAPY MODALITIES COMPARED WITH PLACEBO OR NO TREATMENT
- ELECTROTHERAPY INTERVENTIONS
One trial of 145 participants demonstrated bipolar interferential current to be no more beneficial than placebo in general (mixed population) shoulder disorders for the recovery or substantial improvement in pain in the short or long term (6 weeks to 12 month follow-up) (van der Heijden 1999).
Based on the results of one trial (Ebenbichler 1999), ultrasound appears to have some significant benefit over placebo in calcific tendinitis (RR for recovery or substantial improvement in the short term (end of treatment) 1.81 (1.26, 2.60). In addition, the same trial demonstrated a significant effect in terms of improvement in radiological appearance of calcific tendinitis in the short term (end of treatment) (RR 4.53 (1.46, 14.07)) and long term (nine month follow-up) (RR 3.74 (1.62, 8.66)). However, an additional trial investigating the effect of iontophoresis of acetic acid plus ultrasound (Perron 1997) found no significant benefit in calcific tendinitis. There is no evidence of effect of ultrasound in general shoulder pain or rotator cuff tendinitis. A pooled analysis of three trials assessing the effect of ultrasound on short term recovery or substantial improvement in three varying clinical conditions (van der Heijden 1999; Ebenbichler 1999; Berry 1980) demonstrated a very small but significant benefit over placebo (RR 1.41 (1.04,1.90)). This benefit was attributable to the trial in calcific tendinitis and was not supported by two additional trials, not included in the meta-analysis, measuring the effect of ultrasound on pain (as opposed to recovery/ improvement). These trials demonstrated no benefit of ultrasound over placebo (Downing 1986; Nykanen 1995). Ultrasound had no significant effect demonstrated from pooled analysis of three trials on range of motion (WMD -2.89 (-10.43, 4.66) (Downing 1986; Berry 1980; Nykanen 1995). No trial included in this review assessed adverse effects of ultrasound.
The effect of laser compared to placebo has been assessed by four trials, three included in meta-analysis (Taverna 1990; Saunders 1995; Vecchio 1993) and one with results presented as an additional table (England 1989). The pooled analysis demonstrates laser to be significantly more effective than placebo in bringing about a good or excellent result in the short term (RR 3.71 (1.89, 7.28) for adhesive capsulitis but not for rotator cuff tendinitis. This finding is supported by the results of the third trial demonstrating the difference in medians of pain measured on a 10 cm VAS at 2 weeks to be 2.5cm (2,3)Table 01. No trial has assessed adverse effect of laser.
Pulsed electromagnetic field (PEMF) has been shown in one trial to have a significantly beneficial effect on calcific tendinitis in both the short (RR 19 (1.16, 12.43) and medium (RR 39 (2.46, 617.84) term (Dal Conte 1990). A second trial assessing this intervention in general shoulder pain did not present quantitative analysis but concluded significant short term benefit (Binder 1984). PEMF resulted in more post treatment pain than placebo, but was not associated with increased adverse effect.
- EXERCISES
A supervised exercise regime has been demonstrated to be of significant benefit in both the short and longer term. One trial of 56 participants with mixed shoulder disorders demonstrated significantly greater recovery (RR 7.74 (1.97, 30.32), function (RR 1.53 (0.98, 2.39)and range of abduction (RR for worsening range 0.33 (0.11, 0.96) than placebo at one month (Ginn 1997). A second trial, with a two and a half year follow up demonstrated sustained significant benefit with respect to function for exercise over placebo in rotator cuff disease (RR for good or excellent function 2.45 (1.24, 4.86) (Brox 1993/7).
- MOBILISATION
Only one small trial of 42 participants with adhesive capsulitis divided into four groups assessed the effect of mobilisation compared to no treatment (and to ice and to intra-articular corticosteroid injection) (Bulgen 1984). The data from this trial was not presented in a form allowing inclusion either in meta-anlyses or additional tables (presented graphically), but the authors concluded no significant differences between groups with respect to pain or range of motion.
COMPARISONS OF ONE TYPE OF PHYSIOTHERAPY MODALITY TO ANOTHER
- ELECTROTHERAPY INTERVENTIONS COMPARED TO NON ELECTROTHERAPY INTERVENTIONS.
One trial compared exercises and electrotherapy to mobilisation and manipulation and demonstrated no significant difference between the two groups in both the short or long term (Winters 1997/9). This trial contains no information about the benefits of either intervention over nothing. One trial demonstrated no additional benefit of electromagnetic therapy over physiotherapy alone (Leclaire 1991).
- ONE TYPE OF ELECTROTHERAPEUTIC MODALITY COMPARED TO ANOTHER:
The effect of ultrasound was not significantly different to bipolar interferential current in the short or long term (van der Heijden 1999), however one trial showed significantly greater improvement with ultrasound than TENS (Shehab 2000). This was not supported by the results of a second trial (Herrera-Lasso 1993).
- MANUAL INTERVENTIONS COMPARED TO EXERCISE
Based on three small trials it appears that mobilisation plus exercise is of greater benefit than exercise alone in rotator cuff disease, but not in adhesive capsulitis (Bang 2000; Conroy 1998; Nicholson 1985). Due to differences in scale and use of a combination of change scores and final value, results for pain could not be combined, however the two trials conducted in participants with rotator cuff disease both demonstrate a significant difference in reduction in pain at 3-4 weeks for the exercise plus mobilisation group over the group performing exercise alone (WMD -186.23 (-319.34, -53.12 (Bang 2000) and WMD -32.07 (-58.04, -6.10)) (Conroy 1998) . The difference between groups with respect to range of motion, strength and function are based on only one of the three trials, but demonstrate benefit of adding mobilisation to exercise.
- EXERCISE INTERVENTIONS COMPARED TO EACH OTHER
One trial with only 20 participants (Reid 1996) has shown no significant difference between a routine of isokinetic resistance exercises and use of electromyographic biofeedback for anterior instability. This was the only trial identified in a population with gleno-humeral instability and tells us nothing about the benefit of exercise over no treatment.
PHYSIOTHERAPY MODALITIES COMPARED TO OTHER, NON PHYSIOTHERAPY,TREATMENT INTERVENTIONS
- INJECTION
Seven trials compared injection to physiotherapy. These comprised two trials comparing intra-articular corticosteroid injection with a combined physiotherapy intervention van der Windt 1998; Fernandes 1980, two trials comparing intra-articular and subacromial corticosteroid injection to electrotherapy and exercises Winters 1997/9 Lee 1973, and three comparing injections to mobilisation and manipulation Winters 1997/9, Bulgen 1984, Dacre1989.
Several trials have compared the effect of physiotherapy to injection, however most have used differing physiotherapy modalities and injection sites making it not clinically sensible to combine the results of these trials in a meta-analysis. One study with multiple outcomes assessed at many time points (van der Windt 1998) has demonstrated intra-articular corticosteroid injection to be significantly more beneficial than a combination physiotherapy approach (mobilisation, exercise and electrotherapy) with respect to improvement in main complaint at 3 weeks, 7 weeks and 13 weeks, but not beyond. This benefit was maintained when combined with a second study assessing short term pain and demonstrating no significant difference between groups (Berry 1980). With respect to adverse effect, injection was associated with an increased risk of facial flushing (RR 9 (1.18, 68.74).
These findings are supported by another trial comparing intra-articular and subacromial cortico-steroid injection to exercises and electrotherapy (Winters 1997/9) and demonstrating significant benefit of injection over physiotherapy in the short term (RR for 'cured' 3.72 (1.88, 7.37)), however in the longer term there was no difference between groups (RR for 'cured' 1.23 (0.47, 3.26). These results are consistent when injection is compared both to physiotherapy comprising mobilisation and manipulation (short term cure RR 1.83 (1.17, 2.88), long term cure RR 0.88 (0.36, 2.06).
A further study of injection compared to mobilisation presented results as a graph without numerical data but support the findings of the above trials, concluding no significant long term difference between injection and mobilisation, although some short term benefit with respect to pain in favour of injection (Bulgen 1984). In an old trial presenting their results graphically, no significant between group difference was reported for injection and exercise compared to heat and exercise at six weeks, but range of motion was the only outcome assessed (Lee 1973).
The only trial concluding no difference in short term benefit between physiotherapy and cortico-steroid injection did not present results in a manner that could be included in the meta-analysis, but was performed in a population with adhesive capsulitis (Dacre1989). All other trials were in a population with mixed shoulder disorders and rotator cuff disease.
- MEDICATION
One trial compared laser to non-steroidal anti-inflammatory medication (England 1989) and demonstrated significant short term benefit in favour of laser with respect to pain, function and range of motion. The follow up for this trial however did not extend beyond the treatment period and hence little can be concluded about sustained effect ( Table 1).
Discussion
While 26 trials are included in this review, there is substantial clinical heterogeneity with respect to the interventions tested and hence few trials could be combined in meta-analysis to reach an overall conclusion about the effect of physiotherapy interventions for shoulder disorders. In addition, the results generated by this review are based on trials of very small numbers of participants and hence may be biased by Type II error (the failure to demonstrate a difference which is in truth present, or false negatives). Findings of no significant benefit are therefore consistent with no evidence to support or refute the use of the intervention.
Many of the trials included in this review were of poor methodological quality, with few concealing allocation or analysing results using intention to treat principles. Where possible, data was entered into the analysis section of the review as intention to treat, however the bias introduced by failure to conceal allocation, blind outcome assessors or obtain adequate follow up cannot be corrected in the review analysis and is likely to result in an overestimation of treatment effect.
Of major clinical concern is that few of the identified trials tested combinations of interventions, either more than one physiotherapy intervention or physiotherapy interventions combined with another intervetnion, despite this being the most common way in which shoulder disorders are treated by physiotherapists in practice. Shoulder disorders are difficult to diagnose and classify due to a common overlap of symptoms. Similarly, patients with shoulder disorders rarely receive a single treatment modality in isolation. Current clinical practice for shoulder disorders is likely to comprise not only a combination of physiotherapy intervetions, but also physiotherapy interventions combined with other medical treatments such as corticosteroid injection or medications. Very few trials have assessed combinations of treatments, with some directly comparing two interventions generally given together. This failure of trials to reflect actual practice in their tested interventions needs to be considered not only in interpreting the evidence available for the management of shoulder disorders using physiotherapy interventions, but also in planning future research. Trials should consider testing standardised methods of delivery of combination of physiotherapy interventions reflective of actual practice.
In many cases, included studies tested interventions in an ill-defined or mixed population. It can be assumed from the selection criteria that these populations included a mix of diagnostic categories, for example adhesive capsulitis and rotator cuff disease. The same treatment modalities are rarely employed in clinical practice to treat people with varying shoulder disorders and it is likely that these trials were not able to accurately assess the effects of the intervention due to different subgroups of their populations responding in different ways. While the diagnosing of varying shoulder disorders is difficult and potentially unreliable , future trials would be of greater clinical benefit if performed in defined diagnostic categories.
Authors' conclusions
Implications for practice Further research, in particular larger trials of higher methodological quality, of well defined interventions and in specific populations need to be conducted. Furthermore, high quality trials more reflective of the current clinical practice of combined interventions using standardised methods of delivery need to be conducted before we can draw conclusions regarding the benefits and optimal use of physiotherapy interventions in the treatment of shoulder disorders. The evidence to date can be summarised as follows: There is weak evidence from few, methodologically compromised trials to indicate:
|
Implications for research There is a clear need for trials of physiotherapy interventions, including trials of combinations of modalities, in the treatment of shoulder disorders. There is a need for validation studies of the inclusion and exclusion criteria used to define specific conditions which result in painful shoulder and trials should aim to use properly defined interventions. Trials should be adequately powered and address key methodological criteria (allocation concealment, blinding of participants and outcome assessors, adequate follow up and appropriate statistical reporting). Specifically, further research is needed before we can draw conclusions about:
|
Acknowledgements
We are grateful to Steve McDonald from the Australasian Cochrane Centre for his help with the search strategy and to the Centre of Evidence-based Physiotherapy at the University of Sydney for the use of the PEDro trial quality assessment method and database.
Data and analyses
- Top of page
- Background
- Objectives
- Methods
- Results
- Discussion
- Authors' conclusions
- Acknowledgements
- Data and analyses
- What's new
- History
- Contributions of authors
- Declarations of interest
- Sources of support
- Notes
- Index terms
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
What's new
Last assessed as up-to-date: 23 February 2003.
| |||||||||
History
Review first published: Issue 2, 2003
| |||||||||||||||
Contributions of authors
Sally Green and Rachelle Buchbinder modified the updated protocol. Sally Green and Rachelle Buchbinder identified trials and extracted study results. Sally Green and Sarah Hetrick entered study details and results. All reviewers wrote the review.
Declarations of interest
None known
Sources of support
Internal sources
- Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia.
- Australasian Cochrane Centre, Australia.
External sources
- No sources of support supplied
Notes
Since the original review which included all interventions for shoulder pain, many new clinical trials, studying a diverse range of intervetions, have been performed. In order to update the review, it has been subdivided into a series of reviews investigating the evidence for efficacy of single interventions. The review has also been broadened by including all randomised or pseudo-randomised clinical trials regardless of whether outcome assessment was blinded.
This review will be split into two separate reviews on updating: Physiotherapy interventions for rotator cuff disease, and Physiotherapy interventions for adhesive capsulitis. Physiotherapy interventions for shoulder pain will be withdrawn from publication in The Cochrane Library, once the two new reviews are published.
Index terms
Medical Subject Headings (MeSH)
*Physical Therapy Modalities; Randomized Controlled Trials as Topic; Shoulder Pain [*therapy]
MeSH check words
Humans
* Indicates the major publication for the study