Statistical Fragility of Randomized Clinical Trials in Shoulder Arthroplasty.

Abstract Background The p-value is a statistical tool used to assess statistical significance of clinical trial outcomes in orthopedic surgery. However, the p-value does not evaluate research quality or clinical significance. The Fragility Index (FI) is an alternative statistical method that can be used to assess the quality and significance of clinical research and is defined as the number of patients in a study intervention group necessary to convert an outcome from statistically significant to not statistically significant or vice versa. The primary purpose of this study is to evaluate the statistical robustness of clinical trials with regards to shoulder arthroplasty using FI. The secondary goal of this study is to identify trial characteristics associated greater statistical fragility. Methods A systematic review of randomized clinical trials in shoulder arthroplasty was performed. The FI was calculated for all dichotomous, categorical study outcomes discussed in the identified studies. Descriptive statistics and the Pearson correlation coefficient were used to evaluate all studies and characterize associations between study variables. Results A total of 13 randomized controlled trials (RCTs) were identified and evaluated, and were found to have a median sample size of 47 patients (mean 54, range 26-102) and a median number of 7 patients lost to follow-up (mean 5.8, range 0-14). The median FI was 6 (mean 5, range 1-11), a higher FI than what has been observed in other orthopedic subspecialties. However, a majority of outcomes (74.4%) had a Fragility Index that was less than the number of patients lost to follow-up and the majority of outcomes (89.7%) were statistically not significant. Conclusion RCTs in shoulder arthroplasty have comparable statistical robustness to literature in other orthopedic surgical subspecialties. We believe the inclusion of the FI in future comparative studies in the shoulder arthroplasty literature will allow surgeons to better assess the statistical robustness of future research.