Semi‐invasive therapies for pain in knee osteoarthritis: A systematic review and network meta‐analysis

Abstract Background The increasing number of semi‐invasive pain therapies in knee osteoarthritis poses challenges in decision‐making. This review aimed to simultaneously compare established intra‐articular therapies with newer peri‐articular therapies and explore effect modifiers. Methods Randomized controlled trials were searched from five electronic databases without date or language restrictions. Study selection and data extraction of reports, retrieved up to May 2024, were performed independently by paired assessors. The primary outcome was 6‐month pain score. Nine treatments were included. The effect size (ES) for each treatment, relative to placebo, was estimated using standardized means difference and expressed with 95% confidence intervals (CI). The rigor of results was evaluated with subgroup/sensitivity analyses. Results A total of 111 studies (14,695 participants) were included, with intra‐articular hyaluronic acid having the greatest number of participants. Neuroablation demonstrated the greatest ES (1.08, 95% CI: 0.07, 2.10). While platelet‐rich plasma (PRP) ranked second (ES: 0.75, 95% CI: 0.28, 1.22), it was the only intervention demonstrating statistically significant effect at 3, 6, and 12 months. However, this statistical significance was lost in some sensitivity analyses. Larger estimates for biologics and PRP compared with prolotherapy, steroid, and hyaluronic acid injections were consistently observed across different timepoints and in multiple sensitivity analyses. Generally, no statistically significant difference was found between the nine types of therapies. Conclusion Although there is robust evidence suggesting greater efficacy of PRP, potentially including biologics, over other interventions, future research is needed to identify the phenotype or patient subgroup that would benefit most from PRP.