Re: Liu et al., “On the reported methodology in published TriNetX-based studies: an analysis of impossible index event designs”