Interpret with caution: randomized controlled trials comparing cemented versus press-fit hemiarthroplasty in geriatric patients with intracapsular hip fractures are statistically fragile.

Abstract

Randomized controlled trials (RCTs) investigating cemented (CHA) vs press-fit (PFHA) fixation of femoral implants in hemiarthroplasty have major clinical implications. The purpose of this study is to assess the statistical fragility of reported outcomes in RCTs comparing CHA and PFHA fixation. RCTs published until June 2024 comparing PFHA and CHA fixation were identified systematically from PubMed. Fragility indices were calculated as the minimum number of event reversals needed to alter significance for a variable. Fragility quotients were computed by dividing fragility index (FI) by the sample size. Significance for each of the variables was determined using a two-by-two contingency table and Fisher exact or χ² test as appropriate. A study was defined as statistically fragile if it had an FI less than its described loss-to-follow-up. A total of 12 articles representing 10 RCTs and 115 outcomes were identified, encompassing a total of 2738 patients (1377 CHA and 1361 PFHA). Across all outcomes, median FI was 5 (interquartile range 4-6) with median fragility quotients of 0.019 (interquartile range 0.009-0.032), with 54.8% of outcomes having a FI less than loss-to-follow-up. Mortality by latest follow-up was statistically fragile in 50% of RCTs, with a median of 5.9 event changes per 100 patients sufficient to alter significance. Intraoperative and postoperative periprosthetic fractures, dislocations, and revisions were fragile in at least half of reported RCTs. Results from recent RCTs comparing CHA versus PFHA should be interpreted with caution as differences in mortality, periprosthetic fracture, dislocation, and revision are statistically fragile with high loss to follow-up rates in at least half of RCTs. III.