A total of 15 studies was included in the review (n=3,560): five RCTs (n=780, 348 uncemented and 432 cemented) and 10 observational studies (n=2,780, 1,153 uncemented and 1,627 cemented). Follow-up ranged from two to 11 years. There was evidence of publication bias.
There was significantly improved survival of the implant for the cemented fixation compared to uncemented fixation (OR 4.2, 95% CI 2.7 to 6.5; 15 studies); no evidence of heterogeneity. There was no significant difference for Knee Society score between fixation methods (WMD 0.005, 95% CI -0.26 to 0.26; nine studies); significant heterogeneity was present.
Sensitivity analysis using only RCTs showed no differences between the groups for odds of aseptic loosening (OR 1.9, 95% CI 0.55 to 6.40; five studies); there was no evidence of heterogeneity.
Additional complications included patellar subluxation and dislocation (4.3% to 11.8%; four studies), infections and sepsis (0.9% to 2.9%; six studies), subsequent patellar resurfacing (1.1% to 1.2%; two studies) and loosening or failure of the patellar component (5.6% to 39.1%; two studies).