Response to Xu et al: Time-to-event analysis, data variability, and consideration of model selection: Considerations in relation to stroke-heart syndrome