To the Editor:

We wish to thank Dr. Jiang et al. for their thoughtful comments [1] on our recent work [2].

Several reasons may explain the difference in findings on mortality between our study and the meta-analysis by Fraser et al. [3]. One possibility is the fundamental difference in data source: while the meta-analysis involved pooled data from six papers, our study used a multi-center database. Next, the meta-analysis utilized Mantel–Haenszel risk ratios for short-term mortality, whereas our study employed a Cox proportional hazards analysis for in-hospital mortality. Moreover, the meta-analysis compared mortality between patients who had been administered benzodiazepines and nonbenzodiazepines, while our study focused on those who had been administered only benzodiazepines or propofol. Any or all of these factors may have contributed to the conflicting results.

We agree that sedation depth is a potentially important element in the relationship between sedative drugs and patient outcome, but were unable to incorporate this factor due to a lack of relevant information, such as Richmond Agitation-Sedation Scale. Nevertheless, it may be possible to collect additional information on drug usage for integration into current data. The inclusion of sedation depth is a worthwhile consideration for future studies.