Rheumatology Advance Access published online on September 27, 2008
Rheumatology, doi:10.1093/rheumatology/ken360
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Shoulder acute pain in primary healthcare: is retraining effective for GP principals? SAPPHIRE—a randomized controlled trial
1York Trials Unit, Department of Health Sciences, University of York, York, 2Rheumatology and Rehabilitation Research Unit, University of Leeds, Leeds, 3Department of Health Sciences, University of York, York, 4Department of Rheumatology, Barnsley District General Hospital, Barnsley, 5Primary Care Rheumatology Society, Northallerton and 6Institute of Medical and Social Care Research, University of Wales, Bangor, UK.
Correspondence to:
P. Helliwell, Rheumatology and Rehabilitation Research Unit, University of Leeds, 36 Clarendon Road, Leeds LS2 9NZ, UK. E-mail: p.helliwell{at}leeds.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Objective. To address the question whether general practitioners (GPs) should receive practical training in shoulder problems and to test whether cortisone injections are better than anaesthetic injections for rotator cuff problems.
Methods. A pragmatic split-plot, randomized trial with a cluster factorial design, conducted in general practices across five centres across the United Kingdom. Ninety-one practices were randomized to receive additional training in diagnosing and injecting rotator cuff problems or no additional training. Two hundred patients consulting their general practices with shoulder pain were then randomized to receive either a corticosteroid or lignocaine injection. The main outcome was score on the British Shoulder Disability Questionnaire (BSDQ). The Short-Form 36-item Health Survey and EuroQol at 12 months from entry to the trial were also scored.
Results. Over the course of the trial there was a mean difference of 0.94 (S.E. = 1.01) on the BSDQ score between the groups, with patients treated by the untrained group having a mean of 9.46 (S.E. = 0.82) and those by the trained group having a mean of 8.51 (S.E. = 0.60). There were no statistically significant differences between the groups. Analysing by substance injected, there was a mean difference of 0.15 (S.E. = 0.48) throughout the trial between the groups, with patients given the cortisone having a mean BSDQ of 9.67 (S.E. = 0.39) and those given lignocaine, 9.82 (S.E. = 0.39). This was not statistically significantly different.
Conclusions. Training GPs in the diagnosis and treatment of shoulder disorders does not make any difference to the outcome, in terms of pain and disability, 1 yr later. Further, there is no advantage to injecting steroid in a group with predominant rotator cuff disorder.
Trial registration. International Standard Randomized Controlled Trial Number 58537244. Trial steering committee comprised Prof. Paul Dieppe, Prof. Elaine Hay, Dr Brian Hazleman and Dr Kerenza Hood.
KEY WORDS: Randomized trial, Cluster, Shoulder, General practice, Rotator cuff, Injection, Diagnosis, Training
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Shoulder problems are the third most frequent disorder of the locomotor system after back and neck problems. In the general population, estimates of prevalence range from 7 to 25% [1]. On the basis of one comprehensive review of shoulder disorders [2] it is estimated that the cost of shoulder pain to society is in the order of £100 million. Five percent of general practice consultations relate to shoulder complaints [3], and a postal survey in West Yorkshire showed that the incidence of painful joints increases with age [4]. As a result, as the elderly population increases, the number presenting with shoulder complaints is likely to increase.
Current management of this disabling condition includes steroid injections, NSAIDs, physiotherapy and manipulation. Earlier trials [5, 6] and systematic reviews [1, 7] suggest that steroid injections are more effective than physical therapy, with one of the trials [5] reporting that injections combining cortisone with lignocaine were much more effective than manipulation or physiotherapy in treating disorders of the synovial structures. However, one trial in nine English practices by the same team found no significant difference in effectiveness for shoulder pain between cortisone injections and community-based physiotherapy [8].
Nevertheless, concern has been expressed about the size, quality and quantity of trials in this field [1, 7, 9, 10]. In summary, the relative effectiveness of steroid injections and physiotherapy is still not clear, especially in the United Kingdom. To throw more light on this issue SAPPHIRE focused on the related, but even less clear, question whether training for shoulder problems, including diagnosis and treatment (and injection techniques) is effective. As a starting point we believed the evidence that shoulder injections confer at least as much benefit as physiotherapy justifies further evaluation of relevant training.
Aware that the trial by Winters et al. [5] combined cortisone with lignocaine, SAPPHIRE was also designed to test whether cortisone injections alone were more effective than lignocaine alone. Thus, it will test whether cortisone or lignocaine is the more effective of the two drugs combined in the Dutch trial [5].
Our main aim was to address the question whether general practitioners (GPs) should receive practical training in shoulder problems. We also aimed to establish whether cortisone injections are better than anaesthetic injections.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Recruitment and assignment of participants
General practitioners
GPs across five centres in the United Kingdom (North and East Yorkshire/Northern Lincolnshire, South Yorkshire/North Derbyshire, West Yorkshire, Teesside and North Wales) were invited by letter to take part in the study. If interested, a simple form was completed outlining practice list size, number of partners, details of any previous training in shoulder injecting and membership of relevant specialist societies. For example, Members of the Primary Care Rheumatology Society, the British Institute of Musculoskeletal Medicine or the Society of Orthopaedic Medicine were excluded. Interested GPs were invited to attend a seminar to discuss the trial in more detail. As a result of these seminars and as part of an iterative process, participating GPs contributed to the practical methodology of the trial. We randomized participating general practices between attending a shoulder training day before the trial or after the trial using a computer-generated sequence. Practices were stratified by area and the number of principals in their practices. The participating GPs in these practices had not received practical training in shoulder disorders but currently treated such problems including administering injections.
Patients
Patients consulting with pain in one or both shoulders were identified from the participating practices. Practices were also asked to display posters and leaflets telling patients that the practice was conducting research into shoulder disorders. GPs completed one side of a form summarizing history, examination, diagnosis, treatment and referral about each patient identified with shoulder pain. This was done for all shoulder pain patients for the duration of the trial. Eligible patients (Fig. 1) received an explanation of the trial and an invitation to take part. GPs then had the option of inviting all patients with pain lasting for <12 months to whom they give a clinical diagnosis of rotator cuff tendonitis to see the practice research nurse with a view to joining the trial. Some, however, chose to conduct the whole recruitment process themselves. Eligibility was confirmed, written informed consent was gained, further baseline data were collected and a future appointment was made. Patients were randomized by a computer-generated sequence to injection substance by the GP or practice nurse telephoning a remote randomization service run by the York Trials Unit, usually after the patient had left the surgery. Patients were not informed of the allocation.
|
Subjects’ written consent was obtained according to the Declaration of Helsinki (Br Med J 1996; 31:1448–9) and the design of the work was approved by Northern and Yorkshire Multi-centred Research Ethics Committee (MREC). Local research ethics committees approval was obtained from: Barnsley, Doncaster, Rotherham, South Sheffield, North Sheffield, North Derbyshire, Hull and East Riding, Northallerton, South Humber, York, Airedale, Bradford, Calderdale, Dewsbury, Huddersfield, Leeds St James, Leeds West, Wakefield, County Durham and Darlington, Hartlepool and North Tees, South Tees, North Central Wales, North East Wales and North West Wales. Research Governance approval was obtained from Sheffield (West, South East, South West and North), Barnsley, Rotherham, Doncaster (West, East and Central), Chesterfield, Hambleton and Richmond, Craven Harrogate and Rural, Selby and York, Scarborough Whitby and Ryedale, East Yorkshire and Wolds, Eastern Hull, West Hull, North East Lincolnshire, North Lincolnshire, Bradford South and West, Central Huddersfield, Calderdale, North Kirklees, Middlesbrough, Hartlepool, Langbraugh, North Tees, North West Wales, North East Wales, Conwy and Denbighshire.
Interventions
Training GPs in shoulder disorders
Members of the trial team had developed and piloted the programme for the training day. Before the start of patient recruitment, we trained half of the participating GPs at a residential venue over 24 h, with training within practices available as an alternative. After the trial, the same trainers devoted another weekend to training the rest of the participating GPs or again, within practices if required.
The training programme consisted of a 60-min lecture on shoulder disorders. The approach was based on previous experience of the clinicians and simplified shoulder disorders into four groups: rotator cuff tendonitis, capsulitis, acromioclavicular disease and referred pain. Diagnosis was based on the technique of selective tissue tensioning. Under each heading (anatomy, history, examination, red flags, injections) simplified ‘rules of 4’ were given and these were summarized in hand outs. Following a question and answer session participants were split into small groups of five or less for individual training in examination and diagnosis and a model was used for practical training in injection techniques.
Although one study has shown that injections are more effective at the site of the tendon insertion than at the sub-acromial bursa [11], most GPs (and rheumatologists) inject the subacromial space for rotator cuff tendonitis. Furthermore, there is evidence that the accurate placement of injections improves outcome [12, 13]. As evidence for changing current practice is weak, we judge that our training programme is more likely to be effective if it aims to improve current practice.
Drug therapy
Randomized patients received 2 weeks of oral therapy (400 mg of ibuprofen, three times a day) with their appointment for the shoulder injection. This is an accepted practice in the treatment of acute painful shoulders in general practice [14]. However, we recommended paracetamol for patients with contra-indications to NSAIDs, COX-2 inhibitors for patients at high risk of gastrointestinal complications, and ibuprofen for the rest.
Specifically for the trial, we provided numbered vials containing 40 mg/ml triamcinolone acetonide or 1% lignocaine. Injections were given in 1 ml quantities.
On their return for injection, patients who no longer complained of symptoms did not receive an injection, which had the advantage of avoiding unnecessary treatment. Analysis was nevertheless by intention to treat.
Masking
Specifically for the trial, we provided numbered vials containing 40 mg/ml triamcinolone acetonide or 1% lignocaine. As GPs were able to distinguish visually between the steroid and the anaesthetic, to further ensure patient blinding they were requested not to reveal this to patients.
Exclusion and inclusion criteria
Follow-up and adverse events
Patients completed questionnaires on shoulder pain and general health before randomization and at 1 month, 3 months and a final follow-up of 6–12 months thereafter. Adverse events were monitored by the trial team and forwarded to the Data Monitoring Ethics Committee.
Patient-assessed outcomes
The questionnaires included a combination of generic and specific instruments, which is considered best for measuring health outcomes of shoulder problems [15].
Questionnaires included two generic instruments—the Short-Form 36-item (SF-36) Health Survey and the EuroQol (EQ-5D) [16, 17]. The SF-36 is particularly valid in patients with shoulder problems [15]. Two specific instruments were also included based on different approaches to measuring the effect of shoulder problems on health—the British Shoulder Disability Questionnaire (BSDQ) [18], and visual analogue scales, which are a useful complement to instruments based on functional limitations [19, 20]. Three scales were used to assess the amount of shoulder pain that patients experience—at night, in daytime at rest and in daytime on movement [21].
Each of these four instruments has an important role to play in the evaluation of practical training for injecting shoulders. However, the BSDQ most resembles the instruments used in the two pivotal randomized trials of steroid injections vs physiotherapy [5, 6], hence it was designated as the primary outcome.
Sample size
As SAPPHIRE took place in primary care, small differences are clinically more important than in secondary care, because the large denominator implies more impairment and disability in need of effective treatment. To aim for an effect size of, for example, 0.5 would carry a major risk of type II error and deny general practice many opportunities to add to its own evidence base. Our re-analysis of van der Windt et al. [6] suggested that, by recruiting only 109 patients, they may have missed detecting effect sizes of 0.4 at 13 weeks. Although SAPPHIRE's main comparison (trained GPs with untrained GPs) differs from theirs (injections with physiotherapy), the target population for SAPPHIRE (patients with a clinical diagnosis of rotator cuff tendonitis) is more precise than theirs. This increased the chance that, when expressed as standardized differences, our effect sizes would be larger than theirs, notably at 26 and 52 weeks. For all these reasons, we judged that SAPPHIRE should have the power to detect a standardized difference of 0.3 and that there would be little clinical relevance in detecting standardized differences smaller than this.
Our primary outcome—the BSDQ—has a mean of about 10 and an S.D. of about 4.5 in general practice patients [18]. The validity of the BSDQ has evolved from the Functional Limitations Profile [22]. Two complementary surveys showed that it had criterion and construct validity [18]. Furthermore, the BSDQ closely resembles the Dutch Shoulder Disability Questionnaire (DSDQ) used by van der Windt et al. [6]. Moreover, the paper that validated the DSDQ recognized the similarity between these two measures [23]. Thus, we believe that the BSDQ provides a robust basis for power calculations. Additionally, although there may be a theoretical case for basing power calculations on either the EQ-5D or costs [24], the EQ-5D is much less responsive than specific outcome measures [25]. Furthermore, much less is known about the distribution of costs arising from shoulder pain than about the distribution of BSDQ scores [18]. In practice, therefore, we believed it was prudent to base power calculations on the BSDQ.
We know of no data on intra-practice correlation coefficients (CCs) for the BSDQ. However, both the North of England Study of Standards & Performance in General Practice [26] and the York Primary Care Back Pain Trial [27] included patient outcome measures analogous to the BSDQ. Since neither intra-practice CC exceeded 0.01, we base our preliminary estimate of the power of the SAPPHIRE trial on an intra-practice CC of 0.01. Since each practice will recruit an average of only five patients, this is equivalent to an inflation factor of only 1.04.
In order for the trial to have 80% power, using a significance level of 5% to detect a standardized difference of little more than 0.30, i.e. about 1.38 BSDQ points, would require 200 patients in each arm. Allowing for an attrition rate of 20%, 500 patients were needed—250 in each arm.
Analysis
We used two-sided significance tests at the 5% level to analyse the primary outcome, BSDQ score, by intention to treat. The analyses were performed using PROC Mixed in SAS v9 (SAS software). As the data are nested, observations within patients within practice, practice and id were included in the model as random effects.
The data from all three follow-up points were used as the dependent variable in a mixed model. This produces an average score throughout the whole trial, similar to that of area under the curve. The independent variables were month since randomization [1 month, 3 months, and a final time-point (6–12 months)], number of GPs in each practice, baseline score of the dependent variable, practice and patient identification number. Practice and patient id were included as random effects in order to take account of the repeated measures over time, month, number of GPs, baseline score and treatment were all included as fixed effects.
In the original analysis plan, it was decided to include centre (geographical area, for example, North Yorkshire) as a covariate in the model but problems with convergence of the model occurred when this variable was included and the estimates were not reliable. Centre was not included in any of the final models.
The first analysis was tested for an interaction between substance and training. An interaction term between substance and training was added to the model. This interaction was not significant, hence the analysis focused on two separable trials, training vs no training and lignocaine vs cortisone.
For the training vs no training comparison, those participants in practices that were not included in the randomized training element of the trial were excluded from the analysis. All randomized participants were included for the steroid vs anaesthetic comparison. Intra-class CCs were calculated for the cluster training vs no training comparison. Model checking was performed to ensure that the model was a good fit.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Despite an extension to the recruitment period, this trial failed to reach its original target of 500, due to far fewer that expected eligible shoulder pain patients being identified. In addition, less than half of the practices on board recruited at least one patient. As the extension granted was unfunded it was not possible to continue until the original target was achieved.
Participant flow and follow-up
Ninety-one practices from five centres were randomized to training before (46 practices—labelled ‘trained’) or after (45 practices—labelled ‘untrained’) the trial's patient recruitment period, with an extra 15 ‘expert’ practices recruited to boost patient recruitment. Forty practices (12 untrained, 24 trained and four expert practices) recruited 200 patients (Table 1) in the 18 months between February 2005 and July 2006. Nineteen of the original 91 randomized practices withdrew at some point during the trial. Patients were followed up for a variable length of time ranging from 12 months to a minimum of 6 months.
|
Baseline data
There were no significant differences between the three groups at baseline for age, gender, employment status and baseline BSDQ score. The mean age for the trained group was 58 (S.D. = 11.5), for the untrained group 59 (S.D. = 14.2) and the expert group 55 (S.D. = 15.8). The percentage of females in each group was 52, 55 and 51% for the trained, untrained and expert groups, respectively. The BSDQ score at randomization was 12.2 (S.D. = 4.2) in the trained group, 13.1 (S.D. = 4.4) in the untrained group and 12.1 (S.D. = 4.4) in the expert group (Table 1).
Based on the information provided on the Process Forms (n = 199 returned), the duration of shoulder pain was seen to range from 1 to 52 weeks in the untrained and expert groups and from 1 to 40 weeks in the trained group. The median in the trained group was 6.5 weeks, in the untrained group 8 weeks and the expert group 7 weeks. In all the groups, <50% had previously consulted for shoulder pain (Table 2).
|
Test for interaction
There were no statistically significant interactions between injection type and training for BSDQ (interaction estimate –1.12; S.E. 1.77: Sig. 0.53), or the SF-36 physical component score (interaction estimate 5.10; S.E. 3.26: Sig. 0.12), or mental component score (interaction estimate 0.12; S.E. 3.96: Sig. 0.98); therefore, the analysis was treated as two separable trials.
Training
For the training analysis, those 45 participants from practices that were not involved in the training randomization part of the trial were excluded.
Over the course of the trial there was a mean difference of 0.943 (S.E. = 1.01) on the BSDQ score between the groups, with the untrained group having a mean of 9.46 (S.E. = 0.82) and the trained group having a mean of 8.51 (S.E. = 0.60). The trained group had better outcomes on physical component score (PCS) and BSDQ. However, there was no statistically significant difference between the two groups (Table 3; Fig. 3).
|
|
Substance
All patients were included in these analyses.
Over the course of the trial there was a mean difference of 0.15 (S.E. = 0.48) between the groups, with the cortisone group having a mean BSDQ of 9.67 (S.E. = 0.39) and the lignocaine group 9.82 (S.E. = 0.39). There was no statistically significant difference between the two groups (Table 4; Fig. 4).
|
|
Unequal recruitment rates
One anomaly seen was the differential recruitment of patients between the ‘trained’ and the ‘untrained’ practices (106 vs 47), similar to that reported elsewhere [28]. However, as shown in Table 1, there were no differences in either severity or age between the patients recruited by each group. These were not statistically different.
Follow-up
The overall follow-up rate at the final follow-up point was 89.5% (179/200). The follow-up rate for the trained group was 91%, the untrained group 88% and the expert group 89%. There were no significant differences in follow-up rate between the groups.
There were no significant differences in BDSQ, SF-36 scores or gender between those patients that were followed up and those who were not. There was a significant difference in age, with those followed up having a higher mean age (59; S.D. = 14.0) than those that did not return the final questionnaire (52; S.D. = 12.9) (Table 5).
|
Adverse events
Only one adverse event was reported during the trial, which was facial flushing due to a reaction to the corticosteroid.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Principal findings of the study
This trial sought to address the question of whether GPs should receive practical training in shoulder problems. We also aimed to establish whether cortisone injections are better than anaesthetic injections. In order to do so, we randomized GPs without practical training, but who do currently give such injections, to receive training before or after the trial. We further randomized 200 of their patients with shoulder pain of >12 weeks duration, but <12 months to receive either lignocaine or cortisone injections.
In this study, the 12-month outcome was not dependent on the substance injected, both groups improving at 12 months. However, and in common with other studies of this type, there appeared to be an early (1 month) advantage to both training and cortisone (Figs 3 and 4), but this difference was not maintained at 12 months. Furthermore, we did not observe a statistically significant difference between the trained and untrained groups—although we did observe a clinically significant difference favouring the trained group. There are a number of explanations for these observations. First, the trial was originally powered to observe an effect size of 0.3 S.D. between the intervention and control group patients in terms of the BDSQ scores. We did observe such a difference. However, due to problems with recruitment, our sample size was not sufficient for any difference of this magnitude to be statistically significant. Whether there is a ‘true’ difference between the groups and we missed this difference (type II error) or our observed difference is simply due to chance, is unknown.
Secondly, it is possible that the training was ineffective, perhaps due to the interval between training and patient recruitment. In this respect, it is important to note that there was an average interval of 
5 months between training and first recruitment of patients (due to delays in clinical governance permission) and this may have caused some loss of effectiveness of the training package. It is also possible that the theoretical basis of the training was flawed. The diagnosis of rotator cuff disease was based on the concept of selective tissue tensioning [29] and this may be a false diagnostic algorithm. However, many of the patients included had pain on passive movement of the affected shoulder (Table 2), thus contradicting the algorithm used in this study (where pain on passive movement was taught as a feature of capsulitis). Although some authors believe that making a clinical distinction between, for example, rotator cuff disease and capsulitis is neither possible clinically nor useful therapeutically [30], including a mixed population of diagnoses in this cohort may have masked any specific therapeutic effect of cortisone in rotator cuff tendonitis.
Finally, it may be that the natural history of shoulder disorders (where the clinician makes a diagnosis of rotator cuff disease) is good whether treatment is given or not, rather like with low back pain. In this case, providing the clinician does nothing to retard recovery, the patient is going to improve with or without any intervention. However, it is obviously of importance to the patient if that recovery can be achieved earlier, and such interventions as tested in this trial may be important in achieving this early improvement.
Strength and weaknesses of the study
This study is, as far as the authors are aware, the largest study of its type in the world and as a result is a major contribution to the area of shoulder pain management.
This is in spite of the fact that recruitment was below expected and a number of issues are thought to have contributed to this. First, although 91 practices were taking part, only 40 actually recruited any patients (44%). When GPs were sent a questionnaire enquiring about any recruitment problems, those that replied cited lack of time, lack of eligible patients and excessive paperwork as the main reasons affecting recruitment of patients. Although seminars had been held prior to trial commencement and paperwork modified according to GPs’ suggestions, there still appeared to be a problem combining research and usual working schedules. There is also the possibility that, despite pilot work being done to establish the prevalence of rotator cuff tendonitis, there was an overestimation of the number of eligible patients within general practice.
Secondly, delays due to research governance approval hindered the start of many practices. Of the 31 research governance committees applied to, only 14 gave approval within 60 days. Approval times ranged from 3 days to 49 weeks. Submitted amendments to the inclusion criteria were given approval between 3 days and 17 weeks. Changes with research governance organization created some additional problems. For example, six practices previously given approval by one large committee were no longer able to commence recruitment because their Primary Care Trusts decided to create their own research governance committees. In doing so, these new committees would not honour the original approval given and fresh applications had to be made.
Strength and weaknesses in relation to other studies
The strength of this study is the numbers of patients recruited and the high follow-up rate. In addition, the recruitment of a relatively homogenous clinical group allows us to make some meaningful observations on the natural history of shoulder pain in this group, irrespective of treatment given. This study has also demonstrated that it is important to look at early outcomes in this type of trial as differences may be apparent at this time that are subsequently less obvious with the diminution of treatment effect and the passage of time. Clearly, it is of little comfort to the patient to be told that pain and disability 1 yr later will not be influenced by any treatment given now when a reduction of symptoms could be achieved sooner.
Unanswered questions and future research
In order to determine which patients are most likely to benefit from such injections, further studies are required to look at various clinical subgroups.
|
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Study was the original idea of P.H., I.R. (guarantor), J.D. and A.A., who with J.W. and D.T. carried out the design. P.H. and J.D. developed the training package. V.M. conducted the analyses. J.W. drafted the manuscript which was revised by P.H., I.R., A.A., D.T. and V.M.
The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit the publication.
The trial team would like to thank the following for their input and assistance with this trial.
Contributing practices and GPs
North Wales. Strathmore Medical Practice (Dr Baker), Bodnant (Dr Thompson), The Surgery Llanrwst (Drs Johl, Ramsey and Britt-Compton), Beaumaris Health Centre (Drs Vousden and MacVicar), Rysseldene Surgery (Drs Murphy, Weis and Owen) and Beechley Medical Centre Wrexham (Drs Thomas and Wilkinson).
North Yorkshire. The Surgery Brigg (Drs Whitaker and Willis), Churchfield Surgery Sleights (Drs Lomas and Jackson), Pilgrim Primary Care Centre Immingham (Dr Singh), Albert Road Surgery Cleethorpes (Dr Hurst), University of York Health Centre (Drs Price and Wallam), The Marshes Snaithe (Drs Tinker and Summers), Eastfield Surgery (Dr Walker), Cornlands Road Surgery York (Drs Fisher, Muhammed and Barry) and Leeds Road Practice Harrogate (Dr Banks).
South Yorkshire. Heeley Green Surgery Sheffield (Dr Driscoll), Grimethorpe Surgery (Dr Sriramulu), The Brimington Surgery (Dr Livings), Upwell Street Surgery Sheffield (Dr Key), The Crookes Practice Sheffield (Dr Mascott), Leyfield Surgery Staveley (Dr McConnell), The Royston Surgery (Drs Littlewood and Naish), Wincobank Medical Centre Sheffield (Dr Maher), Totley Rise Medical Centre Sheffield (Drs Pressley and Shawcross), The Goldthorpe Centre (Dr Sen), Brinsworth Medical Centre (Drs Singh and Venkatraman) and Broom Lane Medical Centre Rotherham (Drs Kacker and Sanders).
West Yorkshire. Wibsey and Queensbury Medical Centre Bradford (Drs Harding and Richardson), Rydings Hall Surgery Brighouse (Dr Reed), Oakwood Surgery Leeds (Dr Singh), Saltaire Medical Centre Shipley (Drs Faruque, Dean, Millns Sizer and Livingstone), The Surgery South Milford (Dr Hirst), Windmill Health Centre Leeds (Dr Gerrard), Station Road Surgery Sowerby Bridge (Dr Catlow), Rockwell Medical Centre Bradford (Dr Qazi), Kilmeny Surgery Ingrow (Drs Hodgson and McGill), Cleakheaton Health Centre (Dr Fox), Brig Royd Surgery Sowerby Bridge (Drs Pickles, Knowles, Littlewood, Wyatt and Pool),The Lindley Village Surgery (Dr Spencer), Windhill Green Medical Centre Shipley (Drs Passant and Chadwick) and Netherton Surgery Huddersfield (Dr Boulton).
Teesside. Lawson Street Surgery Stockton (Dr Williams), Coatham Surgery Redcar (Drs Stocking and Lyle), North Ormesby Health Centre (Drs Marshall, Nadah, Adebayo and Lakin), Springwood Surgery Guisborough (Dr Halloway), Gladstone House Surgery Hartlepool (Drs Ray and Kalia), Borough Road and Nunthorpe Medical Group Middlesbrough (Drs Selby and Lakeman).
The trial team would also like to thank all the nurses, practice managers and receptionists from the above practices who were involved.
Research networks
Wolds Primary Care Research Network, Trent Focus, Northern Primary Care Research Network, Yorkshire Primary Care Research Network and Capricorn.
Funding: This trial was funded by the Medical Research Council (grant number G0001147) and received support for the education seminars and training events from Merck, Sharp and Dohme. The MRC established a trial steering committee to advise the grant holders and trial team on trial design, the collection, analysis, interpretation and writing up of data and publication policy.
Disclosure statement: V.M. and J.W. received salary from the MRC research grant. J.D. has received travel grants from Pfizer, Wyeth, Novartis and Napp and honoraria for tutorials from Pfizer and Novartis. He has served on advisory boards for pharmaceutical companies including GlaxoSmithKline, Wyeth, Novartis and IDEA. All other authors have declared no conflicts of interest.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
- van der Heijden GJMG, van der Windt DAWM, de Winter AF. Physiotherapy for patients with soft tissue shoulder disorders: a systematic review of randomised clinical trials. Br Med J (1997) 315:25–30.
[Abstract/Free Full Text] - van der Heijden GJMG. Shoulder disorders: a review of the state of the art. BailliereClin Rheum (1999) 13:287–309.
- Peters D, Davies P, Pietroni P. Musculoskeletal clinic in general practice: study of one year' referrals. Br J Gen Pract (1994) 44:25–9.[Web of Science][Medline]
- Badley EM, Tennant A. Changing profile of joint disorders with age: findings from a postal survey of the population of Calderdale, West Yorkshire, United Kingdom. Ann Rheum Dis (1992) 51:366–71.
[Abstract/Free Full Text] - Winters JC, Sobel JS, Groenier KH, et al. Comparison of physiotherapy, manipulation and corticosteroid injection for treating shoulder complaints in general practice: randomised study. Br Med J (1997) 314:1320–5.
[Abstract/Free Full Text] - van der Windt DAWM, Koes BW, Deville W, et al. Effectiveness of corticosteroid injections versus physiotherapy for treatment of painful stiff shoulder in primary care: randomised trial. Br Med J (1998) 317:1292–6.
[Abstract/Free Full Text] - van der Heijden GJMG, van der Windt D, Kleijnen J, et al. Steroid injections for shoulder disorders: a systematic review of randomised trials. Br J Gen Pract (1996) 46:309–16.[Web of Science][Medline]
- Hay EM, Thomas E, Paterson SM, Dziedzic K, Croft PR. A pragmatic randomised trial of local cortisone and physiotherapy for the treatment of new episodes of unilateral shoulder pain in primary care. Ann Rheum Dis (2003) 62:394–9.
[Abstract/Free Full Text] - Green S, Buchbinder R, Glazier R, Forbes A. Systematic review of randomised controlled trials of interventions for painful shoulder: selection criteria, outcome assessment and efficacy. Br Med J (1998) 316:354–60.
[Abstract/Free Full Text] - Green S, Buchbinder R, Glazier R, Forbes A. Interventions for shoulder pain (Cochrane review). In: The Cochrane library (2000) (3). Oxford: Update Software. 2000.
- Hollingworth GR, Ellis RM, Hattersley TS. Comparison of injection techniques for shoulder pain: results of a double blind randomised study. Br Med J (1983) 287:1339–41.
[Abstract/Free Full Text] - Berry H, Fernandes L, Bloom B, et al. Clinical study comparing acupuncture, physiotherapy, injection and oral anti-inflammatory therapy in the shoulder. Curr Med Res Opin (1980) 7:121–6.[Web of Science][Medline]
- Eustace JA, Brophy DP, Gibney RP, et al. Comparison of the accuracy of steroid placement with clinical outcome in patients with shoulder symptoms. Ann Rheum Dis (1997) 56:59–63.
[Abstract/Free Full Text] - van der Heijden GJMG. Non-steroidal anti-inflammatory drugs (NSAIDs) for shoulder complaints. J Clin Epidemiol (1995) 48:691–704.[CrossRef][Web of Science][Medline]
- Beaton DE, Richards RR. Measuring function of the shoulder: a cross-sectional comparison of five questionnaires. J Bone Joint Surg (1996) 78A:882–90.
[Abstract/Free Full Text] - Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36): conceptual framework and item selection. Med Care (1992) 30:473–81.[Web of Science][Medline]
- Brooks R. EuroQol: the current state of play. Health Policy (1996) 37:53–72.[CrossRef][Web of Science][Medline]
- Croft P, Pope D, Zonca M, et al. Measurement of shoulder-related disability: results of a validation study. Ann Rheum Dis (1994) 53:525–8.
[Abstract/Free Full Text] - Dixon JS, Bird HA. Reproducibility along a 10 centimetre vertical visual analogue scale. Ann Rheum Dis (1981) 40:87–9.
[Abstract/Free Full Text] - Felson DT, Anderson JJ, Boers M, et al. American College of Rheumatology preliminary definition of improvement in rheumatoid arthritis. Arthritis Rheum (1995) 38:727–35.[Web of Science][Medline]
- Rizk T, Pinals T, Talaiver A. Corticosteroid injections in adhesive capulitis: investigation of their value and site. Arch Phys Med (1991) 72:70–2.
- Charlton JRH, Patrick DL, Peach H. Use of multivariate measures of disability in health surveys. J Epidemiol Community Health (1983) 37:299–304.
- van der Windt DAWM, van der Heijden GJMG, de Winter AF. et al. The responsiveness of the (Dutch) shoulder disability questionnaire. Ann Rheum Dis (1998) 57:82–7.
[Abstract/Free Full Text] - Briggs A, Gray A. The distribution of health care costs and their statistical analysis for economic evaluation. J Health Serv Res Policy (1998) 3:233–45.[Medline]
- Hurst NP, Kind P, Ruta D, et al. Measuring health-related quality of life in rheumatoid arthritis: validity, responsiveness and reliability of EuroQoL (EQ-5D). Br J Rheumatol (1997) 36:551–9.
[Abstract/Free Full Text] - North of England Study of Standards & Performance in General Practice. Medical audit in practice: effects on the health of children with common conditions. Br Med J (1992) 304:1484–8.
[Abstract/Free Full Text] - Moffett JAK, Torgerson DJ, Bell-Syer SEM, et al. Randomised trial of exercise for low back pain: clinical outcomes, costs and preferences. Br Med J (1999) 319:279–83.
[Abstract/Free Full Text] - Farrin A, Russell I, Torgerson D, Underwood M. Differential recruitment in a cluster randomized trial in primary care: the experience of the UK Back pain, Exercise, Active management and Manipulation (UK BEAM) feasibility study. Clin Trials (2005) 2:119–24.
[Abstract/Free Full Text] - Cyriax J. Textbook of orthopaedic medicine (1978) 7th. London: Bailliere Tindall.
- Kuijpers T, van der Heijden GJ, Vergouwe Y, et al. Good generalizability of a prediction rule for prediction of persistent shoulder pain in the short term. J Clin Epidemiol (2007) 60:947–53.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. McKenna, L. Bojke, A. Manca, A. Adebajo, J. Dickson, P. Helliwell, V. Morton, I. Russell, D. Torgerson, and J. Watson Shoulder acute pain in primary health care: is retraining GPs effective? The SAPPHIRE randomized trial: a cost-effectiveness analysis Rheumatology, May 1, 2009; 48(5): 558 - 563. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||