Abstract:
|
Since proposed in 2012, the win ratio analysis methodology has been well recognized in pharmaceutical research, especially in clinical studies with composite time to event endpoint. This endpoint is usually made complicated by the existence of competing risks in which patients are likely to experience competing events, such as dose change, rescue therapy or death before observing the primary endpoint events. While methodology on win ratio exists, research on the impact of competing risks on the win ratio estimation lacking. In this research work, we first use a simulation study to evaluate the impact of competing risk event on win ratio estimation under various scenarios. Specifically, two situations are investigated: (i) the competing risk event is treatment dependent; (ii) the competing risk event is correlated with primary endpoint event. Based on these numerical explorations, an analysis strategy will be proposed in handling competing risk in win ratio analysis.
|