Introduction: The win ratio approach to analysing composite endpoints gives higher priority to events that are clinically more important.

Aim: To assess the effect of nintedanib on a hierarchical composite endpoint in patients with progressive pulmonary fibrosis (PPF) using the win ratio approach.

Methods: In the INBUILD trial, patients with PPF other than idiopathic pulmonary fibrosis were randomised to receive nintedanib or placebo. We analysed a composite endpoint comprising (in decreasing order of importance) death, acute exacerbation, respiratory hospitalisation, hospitalisation, absolute decline in forced vital capacity (FVC) ?10% predicted, absolute decline in FVC ?5% predicted. Every patient in the nintedanib group was compared with every patient in the placebo group. Components of the composite endpoint were evaluated in decreasing order of importance until one of the pair had a better outcome. If the patient on nintedanib had the better outcome, it was a ?win?. If the patient on placebo had the better outcome, it was a ?loss?. Otherwise, it was a tie. The win ratio was calculated as the number of wins divided by the numbers of losses.

Results: There were 332 patients in the nintedanib group and 331 in the placebo group. A total of 109,892 pairs were analysed. There were 59,992 wins, 43,834 losses and 6066 ties. The win ratio for nintedanib vs placebo was 1.37 (95% CI 1.14, 1.64; p<0.001).

Conclusions: In the INBUILD trial, nintedanib was associated with a higher probability of better outcomes than placebo based on a win ratio approach to analysing a composite endpoint that accounted for the differing clinical importance of events.