Article Review

J Clin Epidemiol Vol. 48, No. 5, 691-704, 1995 pp. Corwizht 0 1995 Elsevier Science Ltd

Printed’& &ea;Britain. All rights reserved 0895-4356/95 $29.00+0.00

o8!x-43!%(!400170-7

Review

THE EFFICACY OF NON-STEROIDAL ANTI-INFLAMMATORY DRUGS (NSAIDS) FOR

SHOULDER COMPLAINTS. A SYSTEMATIC REVIEW

DANIBLLE A. W. M. VAN DER WINDT*‘, GEERT J. M. G. VAN DER HEIJDEN,’ ROB J. P. M. SCHOLTEN,’

BART W. KOES and LEX M. BOUTER12 ‘Institute for Research in Extramural Medicine, Faculty of Medicine and *Department of Epidemiology and Biostatistics, Vrije Universiteit, Amsterdam, The Netherlands and ‘Department of Epidemiology, Faculty of Medicine and Health Sciences, University of Limburg, Maastricht,

The Netherlands

(Received 27 October 1994)

Abstract-The medical literature was evaluated with respect to the efficacy of NSAIDs for shoulder complaints. Nineteen randomized clinical trials met the selection criteria and were included in this review. Each publication was independently scored by two blinded reviewers, according to a standardized set of 17 methodological criteria. The studies were ranked according to their total methods score (maximum 100 points). The methodological quality of the majority of the studies was rather disappointing; only 5 trials received a methods score exceeding 50 points. Furthermore, methodological criteria were often inadequately described, particularly those regarding the randomization procedure, co-interventions and control of compliance. The three trials with the highest methods scores demonstrated superior short-term efficacy of NSAIDs in comparison with placebo intervention. Fourteen trials comparing two types of NSAIDs showed no conclusive evidence in favour of a particular NSAID with respect to efficacy or tolerability. Future studies should compare the benefit-risk ratios of NSAIDs and analgesics for shoulder complaints in order to establish whether the use of NSAIDs is more favourable than analgesics, despite the higher risk of adverse reactions from NSAIDs.

Shoulder Non-steroidal anti-inflammatory drugs Blinded review Efficacy Adverse reactions Randomized controlled trials NSAIDs

INTRODUCTION

A painful or stiff shoulder may be the result of a variety of diseases and conditions. Shoulder complaints occur frequently: epidemiological studies have reported point-prevalences ranging from 7% to 25% in western general populations [l-3]. The annual incidence is estimated at 10 per 1000, peaking at 25 per 1000 in the

*All correspondence should be addressed to: MS D.A.W.M. van der Windt, Institute for Research in Extramural Medicine, Vrije Universiteit, Van der Boechorststraat 7, 1080 BT Amsterdam, The Netherlands.

age-category of 42-46 years [l]. Approximately 50% of these patients consult a physician for their shoulder complaints. The incidence of shoulder complaints in Dutch general practice is estimated at 13 per 1000 registered patients per year [4].

Shoulder complaints are usually a self-limiting condition of relatively short duration. In 48% of the patients, pain or stiffness will disappear or substantially improve within 1 month after consulting their general practitioner (GP) [unpublished observations]. Unfortunately, the

691

692 Danielle A. W. M. van der Windt

long-term outcome is not always this favourable. Persisting pain or restriction of movement, due to more chronic conditions, may last for several years [5-71.

A painful shoulder may be caused by articular or periarticular rheumatic conditions, or may be the result of neurological or vascular disorders, neoplasms, or referred pain from internal organs. Because of the complex functional anatomy of the shoulder girdle, even intrinsic complaints which emanate from the shoulder joint may constitute a diagnostic challenge. To date, there is no conclusive evidence on either the pathogenesis of shoulder disorders or the accuracy of diagnostic procedures such as X-ray, magnetic resonance imaging, computed tom- ography or ultrasonography [S, 91.

A wide variety of treatment modalities has been introduced, including medication, physical therapy modalities, exercise therapy, corticos- teroid injections and surgery. In primary health care, therapy is frequently initiated with the prescription of a non-steroidal anti-inflamma- tory drug (NSAID) [4]. A variety of NSAIDs is available to the GP, and new drugs are being introduced on a regular basis. NSAIDs are assumed to act by inhibiting prostaglandin synthesis, resulting in relief of pain and suppression of inflammatory processes in articular or peri-articular structures [lo]. Ad- verse reactions include mild gastrointestinal symptoms which occur frequently, as well as more serious complications, such as gastro- intestinal bleeding, renal insuffiency, hepatitis and bronchospasm. The efficacy of NSAIDs should be studied in terms of benefit-risk ratios, in order to determine whether the advantages of NSAIDs outweigh the risk of adverse reactions.

The efficacy of NSAIDs has been studied extensively, but most randomized clinical trials have included rheumatic conditions affecting various joints. The purpose of this paper was to evaluate the medical literature with respect to the efficacy of NSAIDs for patients with shoulder complaints only. There is still much confusion and corresponding lack of consensus regarding the classification of shoulder disorders. Peri- arthritis humeroscapularis or a painful stiff shoulder do not constitute clearly defined clinical entities. Diagnostic criteria may even differ between studies for disorders seemingly straight- forward labelled as tendinitis or bursitis. We therefore prefer to present one overview of the use of NSAIDs for all indications. In this systematic review, conclusions with respect to the

efficacy of NSAIDs were made after blinded assessment of the methodological quality of the studies according to standardized methodologi- cal criteria.

METHODS

Search strategy and screening

A Medline literature search of the period (19661993) was conducted, using the keywords shoulder, analgesics, anti-inflammatory drugs, bursitis, tendinitis, periarthritis, capsulitis, com- parative study and evaluation study. All Medline citations referring to review articles or random- ized controlled trials were retrieved and exam- ined. The references of all retrieved articles were screened for additional relevant publications.

Selection

Studies had to meet the following criteria to be included in this review:

1. The study population included patients with intrinsic shoulder complaints.

2. At least one of the interventions under study included NSAIDs.

3. Patients were randomly allocated to the interventions under study.

4. The publication was a full report. Letters and abstracts were excluded.

5. In publications concerning various rheu- matic disorders, the results of patients with shoulder complaints were presented separately, or at least 90% of the study population had to be composed of patients with shoulder com- plaints.

6. The publication was written in English, German, French or Dutch.

Methodological quality assessment

The selected publications were blinded for author(s), source and results by removing sections of the text, tables and figures. Subsequently, two blinded reviewers (GJMGH and RJPMS) independently scored each publi- cation according to a standardized set of 17 methodological criteria (A-Q), referring to the selection of the study population, description of interventions, outcome measurements, analysis and presentation of data (Table 1).

The reviewers rated each methodological criterion according to the following rules:

+ : Informative description of the criterion at issue; adequate design and conduct (prevention of bias).

The Efficacy of NSAIDs for Shoulder Complaints 693

– : Informative description, but inadequate design or conduct.

?: Lacking or insufficient information, assign- ing either ‘ + ’ or ‘ – ’ was not possible.

Disagreements between the reviewers were identified and subsequently discussed during a consensus meeting. If disagreements were not resolved during this meeting a third blinded reviewer (BWK) was consulted for a final judgement. Next, all methodological criteria rated ‘plus’ were scored using the weighting factors listed in Table 1. Each study was assigned a total methods score by summing up all weighted criteria. This evaluation finally resulted in a hierarchical order, ranking the studies according to methodological quality. The methodological criteria and matching weights are based on the criteria used by Koes et al. [ 11, 121, but were adjusted to fit the disorders and interventions included in this review.

Data abstraction

The success rate in each intervention group was determined by using information from text and tables of the original publications. Whenever possible, success was defined as recovery or substantial improvement from baseline, accord- ing to the patient. The exact definition of success

may somewhat differ between studies, and will be presented in tables summarizing the details and results of the studies. Success rates were calculated according to the intention-to-treat principle, dividing the number of documented successes at the end of the intervention period by the number of patients allocated to the intervention by randomization (irrespective of drop-outs and loss to follow-up during the study). There will probably be a number of successfully recovered patients among the drop-outs, but the reasons for drop-out are not always clearly described. Consequently, for some studies our calculations will be an underestima- tion of the real success rate and produce somewhat lower success rates than the results reported in the original publications. The differences between the success rates in groups treated with NSAIDs and reference groups were calculated, along with the corresponding 95% confidence intervals (95% CI).

In order to estimate the ability to detect a clinically relevant difference, the power of each study was determined [ 131. For these calculations a difference in success rate of 25% between intervention groups was considered to be clinically relevant.

The conclusions of the studies regarding the

Table 1. Criteria list for a methodological assessment of randomized clinical trials for shoulder complaints

Methodological aspect Weighting

Study population 35 A. Selection and restriction: homogenous study population 4 B. Adequate randomization procedure 3 C. Prognostic comparability at baseline I D. Drop-outs 3 E. Loss to follow-up 120% or 1

loss to follow-up < 10% 3 F. Smallest study group > 25 subjects or 6

smallest study group 3 50 subjects or 9 smallest study group 3 75 subjects 15

Interventions 25 G. Informative description of each intervention 10 H. Placebo-controlled study 5 I. Pragmatic study 5 J. Co-interventions avoided 5 Measurement of outcome 25 K. Blinding of patient and physician I L. Relevant outcome measures M. Blinded assessment of outcome measures i N. Adequate follow-up period 5

Analysis and presentation of data 15 0. Analysis according to intention-to-treat principle 5 P. Adequate presentation of data 5 Q. Adjustment for confounding variables and/or differences of 5

prognostic indicators at baseline.

A more detailed description of the methodological criteria is given in the Appendix.

Our website has a team of professional writers who can help you write any of your homework. They will write your papers from scratch. We also have a team of editors just to make sure all papers are of HIGH QUALITY & PLAGIARISM FREE. To make an Order you only need to click Ask A Question and we will direct you to our Order Page at WriteDemy. Then fill Our Order Form with all your assignment instructions. Select your deadline and pay for your paper. You will get it few hours before your set deadline.