Win-Ratio for Composite Endpoints

Composite endpoints are a well-established method for evaluating multiple treatment effects simultaneously. This increases statistical power 💪 and can reduce the required sample size 🔢. However, traditional time-to-event analyses consider only the first event for each patient - regardless of its clinical relevance.

💡The Solution: The Win Ratio by Pocock et al. (2011)
The Win Ratio is an alternative statistical method for analyzing hierarchically structured endpoints. Instead of only considering the first event, it compares patient pairs step by step according to predefined priority levels:
 1️⃣ More severe events are prioritized – e.g., mortality before hospitalization
 2️⃣ If tied on the first level, the next event category is considered

 There are two versions of the method: matched (paired analysis) and unmatched (unpaired analysis).

👉 Real-World Applications
 🔹 The EMPHASIS-HF trial investigated Eplerenone in heart failure patients. In this study, cardiovascular death and heart failure hospitalizations were defined as hierarchical composite endpoints and analyzed using the Win Ratio (publication in NEJM referenced by IQWiG: https://www.nejm.org/doi/full/10.1056/NEJMoa1009492).
 🔹 In the benefit assessment dossier for Tafamidis in the indication of amyloid cardiomyopathy, the Finkelstein-Schoenfeld test was used for a similar hierarchical evaluation of all-cause mortality combined with cardiovascular hospitalizations (G-BA resolution: https://www.g-ba.de/beschluesse/4832/).

➡️ The Win Ratio allows for a more nuanced assessment of clinical endpoints and is expected to play an even greater role in the future. The IQWiG mentions this method in the draft of their General Methods 8.0, signaling broader adoption in future clinical trial analyses.