The statistical fragility of tranexamic acid dosage and route of administration in total hip arthroplasty: A systematic review.

Abstract

Tranexamic acid (TXA) is an established method of reducing blood loss during total hip arthroplasty (THA). Despite many randomized controlled trials (RCTs) examining its use, the optimal dose and route of administration of TXA in THA remain up for debate. Therefore, this study aimed to determine the robustness of RCT findings in this field by applying statistical fragility methodology, including fragility index (FI) and fragility quotient (FQ) calculations. Embase, MEDLINE, and Pubmed were searched for RCTs examining the dose, administration, or the pairing of TXA with another drug in THA. All dichotomous outcomes were extracted, for which the FI and FQ were calculated. The FI was found by determining the number of event reversals required to flip an outcome's statistical significance, and was then divided by sample size to yield the FQ. Literature review yielded 25 RCTs totaling 73 outcomes. Across all outcomes, the median FI was 4.0, with an associated median FQ (mFQ) of 0.047. There were 12 statistically significant outcomes reporting a mFQ of 0.020, while the remaining 61 outcomes were deemed insignificant (mFQ = 0.049). 22 outcomes pertained to blood/platelet transfusion (mFQ = 0.033), 10 outcomes involved thromboembolic events (mFQ = 0.062), 12 outcomes described cases in which additional drugs were required after TXA administration (mFQ = 0.020), and 29 outcomes described other adverse events (mFQ = 0.049). The results of RCTs examining the administration of TXA in THA were statistically fragility, resulting in a median FQ of 0.047. Outcomes reporting significance and those studying the need for additional drugs after TXA administered were particularly fragile. Thus, this study recommends RCTs report fragility statistics in combination with P-values and demonstrates that TXA administration in THA may warrant further level I evidence research.