Since the publication in this Journal of a description of the effects of placebos in rheumatoid conditions (Traut and Passerelli, 1957), there has been a growing literature on the subject of placebo responses, but little has appeared in print regarding such effects in arthritis.During a recent intra-articular injection trial in which two steroid preparations and a placebo were compared, it became apparent that responses to the latter were frequent and well marked. It was decided, therefore, to make a parallel study of the whole question of placebo responses as these occurred during the trial.The results of injection of the two steroids have already been reported; one paper dealt with the effects on rheumatoid arthritics (Chandler, Wright, and Hartfall, 1958) and the other with the effects on osteo-arthritics (Wright, Chandler, Morison, and Hartfall, 1960).The present paper is concerned chiefly with a detailed examination of the responses to the placebo injections in the same trial. It also reports the results of a subsequent dummy tablet trial on the same group of patients. Finally, some of the implications of these findings as they apply to clinical trials are discussed.Material 49 patients were included in the original trial, and of these, 10 dropped out for various reasons during the 18 months that the trial was in progress. The remaining 39 patients from whom results were complete, comprised 34 females and 5 males; 21 were osteo-arthritics, and the remaining eighteen rheumatoid arthritics, the knees being involved in every case. The age range was 29 to 76 years (mean 58 6). The two steroids used were hydrocortisone acetate, and hydrocortisone tertiary butyl acetate. The placebo consisted of the inert aqueous vehicle.
Design of Injection TrialEvery patient received three courses of injections into the affected knee, each course comprising four injections with an interval of 2 weeks between injections. There was a resting period of 2 months between courses, and the order in which courses were given was arranged so that every patient eventually received a course of each steroid and a course of the placebo.Assessments for local pain and tenderness, range of joint movement, and walking time, were carried out at each visit. After careful consideration it was felt that walking-times would give the most useful single objective measurement by which clinical improvement or deterioration could be assessed. Patients were asked to walk a distance of 75 yards as quickly as possible while being timed by a stopwatch. The value taken was the mean of the four walking times which were measured fortnightly during the 2-month resting period following each complete injection course. These values were then expressed as percentages of the original walking time.Alterations of less than ±10 per cent. were classified as "unchanged", while decreases or increases of more than 10 per cent. were classified as "improvement" or "deterioration" respectively. Throughout this article the values of these walking times are referred to as "find...