We performed a prospective, randomised trial on 106 patients to compare the effects of local corticosteroid injections with physiotherapy as advocated by Cyriax in the treatment of tennis elbow. The main outcome measures were the severity of pain, pain provoked by resisted dorsiflexion of the wrist, and patient satisfaction.At six weeks 22 of 53 patients in the injection group were free from pain compared with only three in the physiotherapy group. In the corticosteroid-treated group 26 patients had no pain on resisted dorsiflexion of the wrist compared with only three in the physiotherapy group. Thirty-five patients who had injections and 14 who had physiotherapy were satisfied with the outcome of treatment at six weeks. At the final assessment there were 18 excellent and 18 good results in the corticosteroid group and one excellent and 12 good results in the physiotherapy group. There was a significant increase in grip strength in both groups but those with injections had a significantly better result.After one year there were no significant differences between the two groups. Half of the patients, however, had received only the initial treatment, 20% had had combined therapy and 30% had had surgery.We conclude that at six weeks, treatment with corticosteroid injections was more effective than Cyriax physiotherapy and we recommend it because of its rapid action, reduction of pain and absence of side-effects. Cyriax (1936, 1982) claimed substantial success for using local friction in combination with Mills' manipulations. There have been no comparative studies of these forms of treatment.Our aim was to compare the outcome and side-effects of treatment with local corticosteroids with physiotherapy as described by Cyriax.
PATIENTS AND METHODSFor one year all patients with tennis elbow referred to the University Hospital, Maastricht were included in our study. The criteria for entry were pain on the lateral side of the elbow, tenderness over the forearm extensor origin, and pain on the lateral epicondyle during resisted dorsiflexion of the wrist with the elbow in full extension. Patients were excluded if they had had a previous operation on the lateral side of the elbow, or had arthritis or allied conditions, neurological disorders of the painful extremity, more than three local corticosteroid injections during the six months before consultation, and if the same elbow had been treated before by Cyriax's methods.In those with bilateral symptoms only the most painful arm was included. The patients were randomised into two groups using sealed numbered envelopes without strata or blocks. For ethical reasons we did not continue the randomised treatment if the results were unsuccessful.There were 106 patients (59 men and 47 women) with a mean age of 43 years (SD 9). Physical examination consisted of inspection, measurement of range of movement of the elbow, and assessment of the pain provoked by resisted movement and palpation. The mean duration of the symp-