Sample size and follow-up duration predict the robustness of findings in randomized controlled trials comparing anterior cruciate ligament reconstruction with or without lateral extra-articular procedures: A fragility index meta-analysis.

Abstract

Assess the robustness of randomized clinical trial (RCT) findings comparing outcomes after isolated anterior cruciate ligament reconstruction (ACLR) versus ACLR plus lateral extra-articular procedures (LEAP) using both the fragility index (FI) and continuous fragility index (CFI). A systematic search was conducted across PubMed, EMBASE and Cochrane databases from inception to March 2025 in accordance with PRISMA guidelines. Two-arm randomized controlled trials comparing isolated ACLR with ACLR plus LEAPs were included if they reported statistically significant continuous or categorical outcomes. Overall, CFI and FI were calculated using validated methods, with the value demonstrating the minimum number of event reversals needed to make a finding nonsignificant. Outcome-specific indices were also analyzed when the same outcome was reported in ≥3 studies. Multivariable linear regression was used to identify factors associated with higher CFI or FI values. We identified 3819 studies, of which 17 met inclusion criteria and were analyzed, encompassing 2056 knees (30.3% female). The included studies had a mean age and follow-up period range of 18.8-34 years and immediate postoperative to 60 months, respectively. Across 23 significant continuous outcomes, the median CFI was 13.3 (interquartile ranges [IQR] 20.2), and for 21 categorical outcomes, the median FI was 1.6 (IQR 3). For primary outcomes, the median CFI was 3.8 (IQR 60.2), and the median FI was 3.0 (IQR 16). Regression analysis showed no predictors for higher CFI, while higher FI was significantly associated with fewer knees lost to follow-up, larger sample size, higher journal impact factor and longer follow-up. Many significant findings in RCTs comparing ACL reconstruction with or without LEAPs rely on few patient outcome changes to lose significance. Clinically, surgeons should interpret reported benefits cautiously, as small changes in outcomes could alter significance with stronger evidence stemming from studies with larger sample sizes and longer follow-up. Level I-II, meta-analysis.