Joshua F. Baker, Katherine D. Wysham, Mercedes Quiñones, Bryant R. England, Kaitian Jin, Marianna Olave, Sarah Wetzel, Rachel L. Gillcrist, Criswell Lavery, Natalie Keller, Bridget Kramer, Bibianna Ateh, Hannah F. Brubeck, Daniel K. White, Alexis Ogdie, Rui Xiao, Tuhina Neogi, Carla R. Scanzello
OBJECTIVE: The purpose of this study was to determine the benefit of a behaviorally designed incentive program and glucocorticoid injections in reducing symptoms and promoting physical activity in patients with knee osteoarthritis (KOA). METHODS: This was a pragmatic, multisite, factorially designed, blinded, randomized clinical trial. Participants with a clinical diagnosis of KOA and an indication for joint injection (age 40-85 years; Kellgren-Lawrence x-ray grade ≥1) were randomized to receive a behavioral incentive to promote physical activity. They were further randomized to receive glucocorticoid injections or lidocaine-only injections in a crossover design. Primary outcomes were the average daily steps measured biweekly over 28 weeks and change in the Knee Osteoarthritis Outcome Score (KOOS; average of five subdomains) measured biweekly between 2 and 12 weeks after each injection. Mixed-effects models evaluated time-averaged effects. Secondary analyses evaluated within-person changes after crossover injections. RESULTS: Among 221 randomized participants (mean age 62 years; 84% male), 204 (92.3%) completed follow-up. Participants receiving the incentive program walked 771 more steps per day (P = 0.01). KOOS scores were not improved in the incentive arm. Glucocorticoid injections did not result in greater improvement in KOOS scores or subdomains at any time point or when evaluating the time-averaged effect (β 1.72 [95% confidence interval (CI) -1.08 to 4.52]; P = 0.23). A crossover analysis among 196 participants demonstrated no benefit (β 0.36 [95% CI -1.17 to 1.90]; P = 0.29). Participants were 55% accurate in guessing the order of their injections. CONCLUSIONS: Behaviorally designed incentive programs can promote physical activity among patients with symptomatic KOA with no effect on KOA symptoms. Glucocorticoid injections, compared to lidocaine-only injections, did not show a benefit.