Common intra-articular knee injections demonstrate a similar recovery trajectory over 60 months: A systematic review and meta-analysis of 15,418 participants

Abstract

Purpose: To summarise the recovery trajectories of people undergoing intra-articular knee injections in randomised-controlled trials (RCTs) for osteoarthritis (OA) using patient-reported outcome measures over time. Methods: A systematic review of published RCTs on intra-articular knee injections with at least 100 participants enrolled and a minimum of 6-month follow-up was conducted. MEDLINE, Embase and Cochrane Central Register of Controlled Trials were searched until April 2025. A meta-analysis of within-arm response to treatments was calculated as the standardised mean change and 95% confidence interval (CI) at 1, 3, 6, 9, 12, 24 and 60 months by randomly selecting one arm from each study. Subgroup analyses were performed according to treatment received. The primary outcome measure included the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC). Secondary outcome measures included the visual analogue scale (VAS), Knee Osteoarthritis Outcome Score and International Knee Documentation Centre Score. Results: This study included 73 RCTs (15,418 participants). Pooled SMC for total WOMAC was −2.59 (95% CI: −4.35 to −0.84) at 1 month, −3.3 (95% CI: −5.10 to −1.50) at 3 months, −2.58 (95% CI: −3.95 to −1.20) at 6 months, −3.19 (95% CI: −7.18 to 0.79) at 9 months, −2.09 (95% CI: −3.50 to −0.67) at 12 months, −0.05 (95% CI: −1.43 to 1.34) at 24 months and 0.04 (95% CI: −1.50 to 1.59) at 60 months. Hyaluronic acid (HA), autologous blood product and steroid arms all showed a similar trend. A similar trend was seen for all subscores. Placebo or physiotherapy showed little improvement, except for in VAS pain for 6 months. Conclusion: People undergoing HA, autologous blood product and steroid injections demonstrated a consistent pattern in improvement. This may explain why existing meta-analyses demonstrate inconsistencies in superior treatment. The lack of improvement in placebo or physiotherapy arms suggests regression to the mean is not evident. Clinical Trial: PROSPERO CRD42023445663. Level of Evidence: Level I.