Randomized Controlled Trials Evaluating Treatments for Carpometacarpal Arthritis Are Statistically Fragile: A Systematic Review.

Abstract

Randomized controlled trials (RCTs) on carpometacarpal (CMC) arthritis have a significant impact on treatment decisions, yet the robustness of their findings is seldom evaluated. This study aims to investigate the statistical fragility of RCTs evaluating treatments for CMC arthritis. PubMed, Embase, and MEDLINE were systematically searched for recent RCTs (2010-2023) evaluating treatment variations for CMC arthritis. The fragility index (FI) or reverse FI (rFI) was calculated for each dichotomous outcome, representing the number of event reversals required to alter statistical significance for significant and nonsignificant outcomes, respectively. The fragility quotient (FQ) was determined by dividing the FI or rFI by the total sample size. Of the 46 RCTs screened for inclusion, 15 were included for analysis, resulting in 46 total outcomes. The median FI across the 46 outcomes was 4.5, with an associated median FQ (mFQ) of 0.075. Statistically significant outcomes were more fragile (mFQ = 0.026) than nonsignificant outcomes (mFQ = 0.079). In 66.7% of significant outcomes, losses to follow-up were greater than that outcome's FI. Postoperative functional tests, adverse events, and patient satisfaction were the most common outcome categories reported, with mFQ's of 0.170, 0.066, and 0.068, respectively. Outcomes reported in RCTs studying CMC arthritis are statistically fragile, particularly among significant outcomes where losses to follow-up may have also affected outcome significance. For future studies, we recommend reporting fragility statistics along with <i>P</i> values to provide hand surgeons another metric by which to gauge the results of CMC RCTs.