Total hip arthroplasty (THA) is an increasingly common orthopaedic procedure, with a projected annual rate of 635,000 surgeries per year to be reached by 2030 in the US [1] However, THA often results in perioperative blood loss, potentially requiring blood transfusion and increasing intraoperative risk [2,3] There are a multitude of methods designed to prevent blood loss and the need for transfusion in arthroplasty. Prevalent strategies frequently include the use of tranexamic acid (TXA), an antifibrinolytic agent shown to significantly reduce blood loss in THA [[4], [5], [6]]

While TXA is known to effectively reduce blood loss in THA, it has no gold-standard dose or route of administration. TXA can be administered intravenously, orally, topically, or in combination with other drugs, and debate remains whether any of these methods are superior over alternative TXA regimens [[7], [8], [9]] Despite many randomized controlled trials (RCTs) investigating the ideal route of TXA administration, the literature still does not provide a sufficient foundation to support a superior administration protocol [8,10,11]

RCTs are the highest level of evidence in orthopaedic literature, and are imperative for guiding industry standards. Typically, RCTs report outcomes in terms of a P-value, a metric which can be unreliable and vulnerable to misinterpretation without appropriate scrutiny of sample size or other factors [[12], [13], [14]] To aid in RCT interpretation, fragility metrics including the fragility index (FI) and fragility quotient (FQ) can be applied. FI is defined as the minimum number of outcome event reversals required to flip the statistically significant of an outcome, while FQ is calculated by dividing the FI by the outcome’s sample size. By calculating a FQ, a fragility value representative of total study size is found and yields a more widely comparable statistic.

Given the uncertainty surrounding TXA in THA and the limits of the P-value, this study aimed to apply fragility statistics to analyze the statistical robustness of RCTs studying TXA administration in THA, hypothesizing that conclusions from these RCTs would be statistically fragile, particularly among statistically significant outcomes.