To assess the efficacy of steroid injections for shoulder disorders.

MEDLINE was searched from 1966 to 1995, and EMBASE from 1984 to 1995; relevant citations of retrieved papers were obtained.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) were included.

Specific interventions included in the review

Steroid injections for shoulder disorders; other interventions included physiotherapy, exercise and oral analgesic, and local anaesthesia medication.

Participants included in the review

Patients with shoulder disorders (n=993) were included.

Outcomes assessed in the review

The outcomes assessesd were differences in proportions of treatment success (pain relief, improvement or maintenance of joint function, diminishing of disability).

How were decisions on the relevance of primary studies made?

One reviewer selected the study reports based on the following criteria: patients had shoulder pain at the moment of inclusion, at least one of the treatment regimens included steroid injections, and clinically-relevant outcomes were included.

The assessment was based on 4 categories: study population (selection, randomisation, size, comparability at baseline, drop-outs described, loss to follow-up described), interventions (description of treatments, co-interventions avoided), measurement of effect (blinding, relevant outcome measures, duration of follow-up) and data presentation (adequate analysis and presentation). The likelihood of potential bias was evaluated and reported as either satisfactory or unsatisfactory. The studies were blinded for author, journal, results and conclusions. Two reviewers independently assessed the reports with respect to the quality of study methods, and any disagreements were resolved by discussion. A weight was assigned for validity, precision or clinical relevance. For each study, a method score was calculated by summing the weights for all criteria satisfied, and studies were ranked according to their score.

Two reviewers independently extracted the data.

How were the studies combined?

The success rates were determined by dividing the number of documented successes by the number of patients randomly allocated to the intervention, based on the intention to treat principle. The differences between success rates were calculated for all relevant comparisons, along with 95% confidence intervals (CIs). Results of individual studies are presented in table format.

How were differences between studies investigated?

The studies are ranked according to their methodological quality and discussed narratively.

Sixteen studies: 7 compared steroid injections with placebo; 7 compared steroid injections with currently applied treatment modalities for shoulder disorders (5 studies provided insufficient data for the calculation of confidence intervals); 2 did not have a reference treatment.

Only 3 out of 16 studies scored more than 50 points for methodological quality. Most studies reported small sample sizes.

Studies comparing steroid injections with placebo: the 2 studies with the highest method scores reported a significantly higher success rate for steroid injection; differences in success rates were 14% (95% CI: 1, 27) and 70% (95% CI: 50, 90).

Studies comparing steroid injections with competing treatment modalities: in 2 out of the 3 studies with method scores above 50 points, the lower limit of the 95% CI exceeded a 10% difference in success rate, in favour of steroid injection.

The evidence in favour of the efficacy of steroid injections for shoulder disorders is scarce. The methods employed in most studies appear to be of poor quality. The few studies that appear to be credible do not provide conclusive evidence about which patients, and at what time in the course of shoulder disorders, benefit most from steroid injections.

No information is presented on the search strategy used, so it is difficult to know whether all relevant studies were retrieved.

This additional published commentary may also be of interest. Shoulders to the wheel. Bandolier 1996;32:2-3.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.