Doyle Index is a valuable additional pain measure in osteoarthritis

Abstract

To determine reliability, feasibility, and validity of the Doyle Index (DI), a pain score proposed for osteoarthritis (OA). The DI was performed in 260 patients with OA at multiple sites (mean age 64.9 years, 84% women) by grading pain (0-3) in 48 joints and joint groups by pressure or passive movement. Reliability and feasibility were determined in a random sample of 18 patients, by examining them twice using four raters. Intraclass correlation coefficients (ICCs) for intra- and interrater reliability were calculated, as well as the mean time to perform the DI. Validity was assessed in 260 patients, by correlating DI total scores and DI scores for the hand and knee/hip joints separately, to the pain and function subscales of the Australian/Canadian Osteoarthritis Hand Index (AUSCAN) and Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), using Spearman's rank coefficient (r). In the total population the median (interquartile range) DI score was 11.0 (5.0-19.0). Intraobserver ICCs [95% confidence interval (CI)] ranged from 0.94 (0.84, 0.98) to 0.97 (0.93, 0.99). Interobserver ICC was 0.88 (0.77, 0.94). The mean time to perform the total DI was 5.1min (range 2.4-7.8). DI total scores as well as scores for the hand and knee/hip joints separately were related to AUSCAN (r range 0.61-0.65) and WOMAC (r range 0.43-0.51), although the level of correlation was moderate. The DI is a reliable, easy to perform, and valid measure for OA pain during physical examination and therefore a promising additional outcome measure not only for OA research but also for clinical practice.