STEROID INJECTIONS FOR SHOULDER DISORDERS - A SYSTEMATIC REVIEW OF RANDOMIZED CLINICAL-TRIALS

Background. Patients with shoulder disorders are believed to benefit c onsiderably from steroid injections. However, the controversy about th eir efficacy persists. Aim. The study was designed to assess the effic acy of steroid injections for shoulder disorders. Method. A systematic computerized literature search in Medline (Index Medicos 1/1966-10/19 95) and Embase (Excerpta Medica 1/1984-10/1995) was conducted, supplem ented with citation tracking of all relevant publications. Studies pub lished before November 1995 were selected if steroid injections were r andomly allocated to patients with shoulder disorders and when clinica lly relevant outcome measures were reported. Because the validity of s tudy outcomes depends heavily on the strength of methodological qualit y, the methods were assessed systematically by two 'blinded' independe nt reviewers. This resulted in a method score (maximum 100 points) tha t was based on four categories: study population, interventions, measu rement of effect, and data presentation and analysis. Confidence inter vals for the differences between groups in success rates were calculat ed in order to summarize the efficacy of steroid injections. Results. Only three out of the 16 studies scored more than 50 points, indicatin g a generally poor quality of methods. Most studies reported small sam ple sizes. The flaws most often found were incomparability of co-inter ventions and poor blinding of therapist. The methods assessment was fr equently hampered by incomplete information about randomization, progn ostic comparability, compliance, outcome measures included, blinding o f patients and blinding of outcome measurement. Conclusions. The evide nce in favour of the efficacy of steroid injections for shoulder disor ders is scarce. The methods of most studies appear to be of poor quali ty. The few studies that appear to be credible do not provide conclusi ve evidence about which patients at what time in the course of shoulde r disorders benefit most from steroid injections.