Randomized controlled trial outcomes for nonoperative management of lateral epicondylitis of the elbow are statistically fragile: a systematic review.

Background: Nonoperative management remains the first line treatment for lateral epicondylitis of the elbow. In this study, we evaluated the reliability of outcomes reported in randomized controlled trials (RCTs) focusing on nonoperative therapies using fragility analysis techniques including the fragility index (FI), reverse fragility index (rFI), and fragility quotient (FQ). Methods: A comprehensive search across PubMed, Embase, and MEDLINE was conducted to identify relevant RCTs published between January 1, 2010, and November 1, 2023, that investigated nonoperative treatment modalities for lateral epicondylitis. Risk of bias was assessed using the Cochrane risk of bias tool, and data synthesis was conducted using fragility analysis metrics, including the FI, rFI, and FQ. The FI and rFI were used to determine the minimum number of event reversals needed to affect the statistical significance or lack thereof for reported outcomes. The FQ was calculated by dividing the FI by the total number of trial participants. Results: A total of 323 RCTs were screened, and 17 studies were ultimately included. The median FI across 99 total outcomes was 4 (interquartile range [IQR]: 2-6) with an associated FQ of 0.0625 (IQR: 0.0360-0.1042). Of the outcomes extracted from the studies, 29 were statistically significant, with a median FI of 3 (IQR: 1-5) and an FQ of 0.0357 (IQR: 0.0208-0.0714). Conversely, 70 of the outcomes extracted were not statistically significant with a median rFI of 5 (IQR: 3-6) and rFQ of 0.0708 (IQR: 0.0502-0.1042). Importantly, 27.6% (8/29) of statistically significant outcomes had an FI of 1, suggesting extreme fragility in approximately a quarter of all outcomes. The most fragile outcome category was complications and adverse events with a median FI of 4 (IQR: 2.5-4), while pain was the most reported outcome category (median FI: 3.00). Conclusion: These findings point to a lack in robustness for conclusions drawn from RCTs relating to nonoperative treatment for lateral epicondylitis. A reversal of only 3 patient outcomes could considerably influence reported statistically significant findings. Notably, there is increased fragility among FQ and rFQ when sample size is taken into account. We advocate for the standardized inclusion of FI, rFI, and FQ metrics in lateral epicondylitis RCTs to ensure clinicians may more accurately evaluate the reliability of study results.