Eighty-nine trials (involving 12 667 participants) were included. Thirteen trials reported adequate concealment of allocation (15%), 68 trials used a sham intervention in the control group (76%), 16 studies were judged to have adequately blinded patients (18%), and 48 had blinded outcome assessment (54%). Seventeen trials analysed data according to the intention-to-treat principle (19%), and 23 trials had sample sizes of 100 patients or more per trial group (26%). Forty studies had a follow-up duration greater than three months.
Viscosupplementation moderately reduced pain (SMD −0.37, 95% CI −0.46 to −0.28; 71 studies). There was moderate between-trial heterogeneity and a significantly asymmetrical funnel plot. Smaller reductions in pain were significantly associated with larger sample size, unpublished studies and blinding of outcome assessment, with results below the threshold for a minimal clinically important difference (for all individual time points).
Viscosupplementation also moderately improved physical function (SMD −0.33 95% CI −0.43 to −0.22; 52 studies) but was also accompanied by moderate between-trial heterogeneity and a significantly asymmetrical funnel plot. The estimate for the 15 large studies with blinded outcome assessment was -0.09 (95% CI -0.17 to 0.00); results did not reach the minimal clinically important difference for any individual time point.
There was no statistically significant difference in the risk for flare-ups, although viscosupplementation was associated with an increased risk for serious adverse events (RR 1.41, 95% CI, 1.02 to 1.97; 11 studies), drop-outs due to adverse events (RR 1.33, 95% CI 1.01 to 1.74; 23 studies), and local adverse events (RR 1.34, 95% CI 1.13 to 1.60; 31 studies). The authors reported that when they restricted the analyses to large trials with blinded outcome assessment similar results were found (no data were presented, except for serious adverse events).