The Statistical Fragility of Management Options for Acute Achilles Tendon Ruptures - A Systematic Review of Randomized Control Trial with Fragility Analysis.

Abstract

In the treatment of acute Achille's tendon rupture, there is no uniform consensus on which of the many treatment modalities for this common injury is superior with respect to all possible complications. This review is to assess the statistical quality of the available evidence. The P value is the common method to determine the significance of a finding, but it does not convey statistical robustness. The reversal of a small number of outcome events can be enough to change a finding of significance; this is known as statistical fragility, which can be measured with the fragility index (FI) and fragility quotient (FQ). The purpose of this study was to examine the statistical fragility of randomised control trials (RCTs) reporting outcomes of acute Achille's tendon rupture (AATR) management. A systematic search strategy was used to find RCTs published since 1990 investigating AATR management. The FI was calculated using Fisher's exact test by sequentially altering the number of events until there was a reversal of significance. The FQ was calculated by dividing the FI by the sample size. Each trial was assigned an overall FI and FQ calculated as the median result of its reported findings. Overall, 55 RCTs met the inclusion criteria, including 4,205 patients, 82.7% of which were male, there was a mean age of 41 and follow-up of 21 months; 60% of RCTs either did not report a statistical power analysis or were statistically underpowered. The overall FI was 4, indicating the reversal of just four outcomes would change the significance finding. The overall FQ was 0.082, suggesting that reversing eight patients out of every 100 would alter significance. In 22/55 (40%) RCTs, the number of patients lost to follow-up was greater than or equal to the FI of the trial. This review indicates the RCT literature for AATR management may be vulnerable to statistical fragility, with a handful of events required to reverse a finding of significance. We recommend that future trials in this area report the FI, FQ, and P value, to aid readers in assessing the evidence, therefore impacting clinical decision making. I; Systematic Review of Randomised Control Trials.