To determine the long-term effects (a minimum follow-up time 8.8 years) of cemented and cementless fixations used for total knee arthroplasty (TKA).
PubMed, EMBASE, Ovid, Cochrane Library, CINAHL, China National Knowledge Infrastructure and China Wangfang database were interrogated for appropriate randomized controlled trials (RCTs) through July 2020. Data were extracted and assessed for accuracy by 2 of the authors acting independently. Any controversial discrepancies were resolved after discussion with a third author.
Eight RCTs were included with low to moderate bias risks. The cemented fixation of TKA was comparable to cementless fixation in terms of implant survival (relative risk, 1.016; 95% CI 0.978 to 1.056; P = 0.417), Knee Society (KS) knee score (standardized mean difference (SMD), − 0.107; 95% CI − 0.259 to 0.045; P = 0.168), KS function score (SMD − 0.065; 95% CI − 0.238 to 0.109; P = 0.463), KS pain score (SMD − 0.300; 95% CI − 0.641 to 0.042; P = 0.085), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score (SMD − 0.117; 95% CI − 0.307 to 0.073; P = 0.227), HSS score (SMD − 0.027; 95% CI − 0.270 to 0.217; P = 0.829), range of motion (SMD 0.061; 95% CI − 0.205 to 0.327; P = 0.652) at ≥ 8.8 years of follow-up. In terms of radiographic outcomes at ≥ 8.8 years of follow-up, the incidence of a radiolucent line in the cementless group was lower than for the cemented group (SMD 3.828; 95% CI 2.228 to 6.576; P < 0.001). However, the maximum total point motion (MTPM) of the cementless group was greater than for the cemented group (SMD − 0.739; 95% CI − 1.474 to − 0.005; P = 0.048).