The win ratio (WR) approach is used to assess composite endpoints in a hierarchical fashion. This novel method offers an excellent opportunity to assess the robustness of the findings yielded by landmark trials, such as the DapaTAVI trial.
MethodsWe applied the WR method to evaluate the treatment effect of dapagliflozin in hierarchically ordered clinical outcomes. Several combinations of outcomes were tested, including time-to-event, binary, and continuous endpoints.
ResultsThe WR of the original primary endpoint was 1.36 (95%CI, 1.03-1.78; P=.028), comparable to the reciprocal of the original hazard ratio (1/HR, 1.38; 95%CI, 1.06-1.81). The win difference was 4.84% (95%CI, 0.55-9.12), confirming consistent findings in terms of absolute effect. Alternative combinations of the primary outcome with different prioritization of its components yielded similar treatment effects and statistical significance. Ignoring a time-to-event approach and including recurrent events did not substantially affect treatment efficacy and its statistical significance. In contrast, the inclusion of the total length of stay for heart failure hospitalizations in the hierarchy shifted the point estimate toward the null. Including New York Heart Association functional class improved the precision of the estimate (WR=1.31; 95%CI, 1.09-1.56; P=.003). Conversely, including quality of life through Kansas City Cardiomyopathy Questionnaire comparisons shifted the overall estimate toward the null (WR=1.10; 95%CI, 0.94-1.30; P=.236).
ConclusionsThe WR approach is a solid method to assess treatment efficacy. We observed consistent findings using this approach in the DapaTAVI trial.
Keywords
Identify yourself
Not yet a subscriber to the journal?
Purchase access to the article
By purchasing the article, the PDF of the same can be downloaded
Price: 19,34 €
Phone for incidents
Monday to Friday from 9am to 6pm (GMT+1) except for the months of July and August, which will be from 9am to 3pm