Abstract
An articular index was devised for the sequential assessment of patients with osteoarthritis (OA). Forty-eight joint units, chosen to reflect the characteristic pattern of the disease, were scored for tenderness on pressure or movement on a 4-point scale. Four observers examined patients to assess inter- and intraobserver error. The index was highly reproducible both within and between observers; intraobserver error was, however, significantly smaller. In a double-blind, cross-over trial the index was sufficiently sensitive to detect a statistically significant difference between the responses of patients with OA to an anti-inflammatory agent and to a simple analgesic. It is likely to be a useful addition to current methods of measurement in osteoarthritis.