Radiography as primary outcome in rheumatoid arthritis:: acceptable sample sizes for trials with 3 months' follow up

Objectives: To investigate whether plain radiographs can show changes in joint damage due to rheumatoid arthritis (RA) within 3 months. Methods: 188 film pairs taken with a 3 month interval were evaluated. They were scored with (chronological) and without (paired) knowledge of the sequence of the films according to the Sharp/van der Heijde method. Changes in joint damage were analysed on a group and an individual level for different subsets of patients. Sample sizes required to detect statistically and clinically significant differences were estimated based on the percentages of patients with progression larger than the smallest detectable change (SDC). Results: Changes in joint damage were seen by both the chronological and the paired scoring method. The percentage of patients with progression of joint damage larger than the corresponding SDCs (1.7 and 2.4) varied in the subsets from 18% to 64% if based on the chronological change-scores and from 9% to 36% using paired change-scores. Acceptable sample size estimates were seen in several subsets, depending on (a) how the investigated drug would reduce the individual risk of progression of joint damage (by an absolute or a relative risk reduction model); (b) how damage was scored (chronological or paired); (c) the baseline risk; and (d) whether a two sided or one sided test would be used. Conclusions: Changes in joint damage due to RA can be detected reliably already within 3 months. This finding can be used to plan short term, randomised controlled trials with radiographic progression as primary outcome.