Sample size for prospective studies of hip joint space width narrowing in osteoarthritis by the use of radiographs.

Abstract

To determine the number of participants required in controlled clinical trials investigating the progression of osteoarthritis (OA) of the hip as evaluated by the joint space width (JSW) on radiographs and to evaluate the reproducibility of the JSW measurement methods. Anteroposterior radiographs of hip were taken from 13 healthy volunteers and from 18 subjects with radiographic hip OA. The reproducibility of the JSW was determined from four segments using digital caliper measurements performed on film radiographs and using semiautomatic computerized image analysis of digitized images. Pearson correlation coefficient, coefficient of variability [CV (%)], and sample size values were calculated. It was found that 20 was a typical number of patients for a sufficiently powered study. The highest sample size was found in subjects with OA in the lateral segment. The reproducibility of the semiautomatic computerized method was not significantly better than the digital caliper method. The number of study subjects required to detect a significant joint space narrowing in follow-up studies is influenced by the baseline hip joint OA severity. The JSW measurements with computerized image analysis did not improve the reproducibility and thus performing JSW measurements with a digital caliper is acceptable.