Viscosupplementation for knee osteoarthritis: systematic review and meta-analysis

Pereira TV, Jüni P, Saadat P, et al. · BMJ

Pain reduction vs placebo of SMD -0.08 (95% CI -0.15 to -0.02), equal to -2.0 mm (95% CI -3.8 to -0.5) on a 100 mm VAS — the upper bound of the interval sits below the minimal clinically important difference the review set at 9 mm. Function moved as little (SMD -0.11 across 19 large trials). Serious adverse events ran 3.7% on hyaluronic acid against 2.5% on placebo (RR 1.49, 95% CI 1.12 to 1.98). Trial sequential analysis puts the crossing into conclusive equivalence at 2009.

Population

169 trials, 21,163 randomized participants; main pain analysis from 24 large placebo-controlled trials (8,997 participants)

Intervention

Intra-articular hyaluronic acid

Comparator

Placebo or no intervention

Limitations

Funded by the Arthritis Society and St Michael's Hospital Foundation, with no hyaluronic acid manufacturer among the funders and no author reporting a personal payment from a drug or device company. Heterogeneity was low across all three analyses (tau-squared 0.02 for pain, 0.01 for function, 0 for serious adverse events). The authors' own caveats are that the summary estimates cannot exclude benefit in some selected population, and that the serious-adverse-event risk may run higher outside trial populations than within them.