The fragility of statistically significant findings from randomized controlled trials in pediatric appendicitis: A systematic review

PURPOSE: Randomized controlled trials (RCT) in pediatric appendicitis remain limited, and the robustness of available evidence is unknown. The aim of this study was to determine the fragility of results in pediatric appendicitis RCTs. METHODS: A systematic search of Embase and MEDLINE was performed. Eligible studies were two-armed RCTs that included at least one statistically significant dichotomous outcome, had parallel-group allocation, and assessed pediatric patients (0-17) with a primary diagnosis of appendicitis. The Fragility Index (FI) for one statistically significant outcome per trial was calculated using a Fisher's exact test, with statistical significance set at p < 0.05. RESULTS: Six studies were identified for inclusion. Studies included a median of 103 patients (interquartile range [IQR] 86-127), with a median of 18 (IQR 4.5-41.25) events for analyzed outcomes. The primary outcome variable was included in analysis for 4(67%) studies. The median FI across studies was 3 (IQR 0.75-4.25), with results ranging from 0 to 5. Results indicate that overall, converting 3 patients from non-events to events in a single trial arm would change the significant dichotomous outcome to nonsignificant. CONCLUSION: The fragility of results in RCTs in pediatric appendicitis should be considered before clinical practice is changed. Investigators should consider reporting the FI alongside study results, as p-values alone may be misleading. TYPE OF STUDY: Randomized Controlled Trial. LEVEL OF EVIDENCE: Level I.