The Rise of the Win Ratio.

In contemporary clinical trials, investigators often tend to combine multiple events into a composite outcome to enroll less patients and reduce overall costs. Composite outcomes bolster the number of available events and enhance statistical power. For example, if one was conducting a study evaluating the efficacy of a novel heart failure medication, researchers may seek to combine the outcomes of "cardiovascular death" and "hospitalization due to heart failure" into a single outcome termed "cardiovascular death or heart failure hospitalization". Combining these events would successfully increase the number of events, as patients who experience either cardiovascular death or heart failure hospitalization will now contribute to the primary analysis. This makes it easier to show statistical significance.  

While this may appear to be a compelling approach to studying clinical phenomena with low event rates, this traditional approach suffers from the significant limitation of treating all clinical events included within a composite outcome as equally clinically impactful. Going back to our previously stipulated example, it is evident that death carries greater clinical significance than hospitalization, a statement that virtually any clinician would agree with. Some experts would argue that due to its importance, death should never be combined with anything else in a composite outcome. In fact, the traditional approach of compositing these two events into one outcome fails spectacularly in differentiating between these events, assigning them equal clinical weight and impact. Beyond that, another drawback of the traditional approach is the focus on the first event, irrespective of the events that may follow. Referring to our previous example once more, if a patient is hospitalized for heart failure and subsequently dies a month later, their death will be excluded from the primary analysis. This can lead to potentially skewed results since the initial event, even if not the most clinically significant, will be prioritized. (1)

The “win ratio” has emerged as a compelling alternative statistical method that acknowledges the relevance of events within a composite outcome. This approach involves creating a hierarchy of outcomes, with death being of highest importance, followed by hospitalization in our hypothetical example. The win ratio then compares the treatment arms, pairing up patients from each treatment arm, and initially prioritizing the event at the top of the hierarchy. The win ratio itself is the ratio of "wins" to "losses", where a “win” signifies a superior outcome in one treatment arm compared to the other.  Like all ratios, A win ratio of 1 is neutral and a win ratio > 1 is positive. For example, a win ratio of 1.5 means there are 50% more wins in the experimental treatment. A win ratio < 1 would be a "loss ratio" or if inverted it becomes a win for the other arm (control).

To better illustrate, let’s break our previous example into two arms, an experimental (treatment A) arm and a control (treatment B) arm. Once paired up, if a patient in the treatment A arm dies during duration of the trial and a patient on treatment B survives through the trial duration, this would constitute a win for treatment A. Furthermore, if a pair of patients remain alive throughout the study, the subsequent outcome of hospitalizations becomes the next event analyzed. If the patient in treatment A experiences no hospitalizations compared to the patient in group B experiencing hospitalization, this would also be a win for treatment A. The win ratio is then calculated by summing up all the wins across all patient pairs.  

The win ratio presents several advantages over the traditional composite outcome approach. It accounts for the clinical importance of events, ensuring that more clinically relevant events such as death are given priority. It accommodates multiple events, providing a more comprehensive picture of treatment effects. It can also incorporate quantitative outcomes, such as quality of life scores, enriching the analysis with patient-centric data. (2) Furthermore, it is a conceptually straightforward method that is readily grasped by clinicians and researchers.  

There are several ways to calculate them with the preferred approach being matched risk stratified pairs of patients.

While the win ratio has several advantages, it is not without some limitations. Its relative novelty means there is less established guidance on its implementation and less familiarity among trialists. It prioritizes the order of events but does not utilize the “time to events”, potentially discarding some temporal information (1), such as hazard ratios. Another limitation is that there may be subjectivity in ranking certain outcomes where the difference in clinical importance may not be overtly clear, for example, ranking the events of ICU delirium versus ICU length of stay. It is also important to note that while the win ratio can discern which treatment arm demonstrates superior outcomes, it does not provide insight into the specific magnitude of the difference between the two groups, therefore results must be interpreted carefully.  

In conclusion, the win ratio is a statistical method that holds the potential to refine the analysis of composite outcomes in clinical trials. Its application in several contemporary large-scale trials underscores its growing acceptance, and its use is likely to grow as researchers become more cognizant of what it may bring to the table while addressing its current limitations. And as always, authors should give us as much granular detail as possible on the different components of the composite outcome or win ratio. They should not just give us the summary, so we can be more informed consumers of data to better understand the results and be able to provide better recommendations to our patients.

Author: Eric C. Pyles, PharmD

Editor: Martin M. Cearras, MD

References:

(1) Pocock SJ, Ariti CA, Collier TJ, Wang D. The win ratio: a new approach to the analysis of composite endpoints in clinical trials based on clinical priorities. European Heart Journal. 2012;33(2):176-182.

(2) Pocock SJ, Gregson J, Collier TJ, Ferreira JP, Stone GW. The win ratio in cardiology trials: lessons learnt, new developments, and wise future use. European Heart Journal. 2024;45(44):4684-4699.