Statistical fragility of reporting hemidiaphragmatic paralysis after brachial plexus blocks in randomized controlled trials: a systematic review

Abstract

To characterize the fragility index (FI) of statistically significant results reported in randomized controlled trials (RCTs) investigating the incidence of hemidiaphragmatic paralysis (HDP) after brachial plexus blocks. A systematic review of RCTs retrieved from the PubMed-Medline, Embase, and Web of Science electronic databases was conducted. All alternative RCTs published between January 2012 and October 2022 were identified. Only RCTs with two parallel arms designs, and reporting HDP as the primary outcome, statistical significance, and superiority results were selected. The FI was calculated according to Fisher’s exact test using previously described methods. In addition, the risk of bias was evaluated using the Cochrane Risk-of-Bias tool for randomized trials. The 23 RCTs that fulfilled the inclusion criteria had a median FI of 4 (interquartile range [IQR]2–8) and a median Fragility Quotient of 0.077 (IQR 0.038- 0.129). However, in 13 (56.5%) trials, the calculated FI value was ≤ 4. In 3/23 (13.0%) trials, the number of patients who dropped-out exceeded the FI value. Most trials (91.3%) had an overall low risk of bias. This systematic review revealed that the statistical results of RCTs investing HDP after brachial plexus blocks have tended to be fragile in the past decade. The FI should be an important aid in the interpretation of clinical results in combination with the P-value, particularly when statistically significant results are dependent on a small number of events. Future RCTs with larger sample sizes are needed to obtain more robust results in this field.