Documented preoperative statin exposure and short-term mortality after aortic valve replacement: a multicohort observational study

Whether documented preoperative statin exposure is associated with short-term mortality after aortic valve replacement (AVR) is uncertain. This retrospective multicohort study included 5,291 AVR records from SNUH, MIMIC-IV, and eICU. The primary analysis compared no documented versus any documented statin exposure for recorded 30-day mortality in SNUH and MIMIC-IV using overlap weighting and fixed-effect pooling. Exposure-window, recorded 365-day, IPTW, and SNUH available-follow-up analyses were sensitivity analyses; eICU and agent-specific analyses were exploratory. Overlap-weighted 30-day risks with versus without documented exposure were 1.29% versus 2.55% in SNUH and 1.18% versus 2.36% in MIMIC-IV; the pooled HR was 2.01 (95% CI, 0.98 to 4.11; P = 0.056). A 1-day MIMIC-IV window reclassified seven records and, pooled with the SNUH 24-h definition, yielded an HR of 2.29 (95% CI, 1.14 to 4.61; P = 0.020) at 30 days; the recorded 365-day HR was 1.76 (95% CI, 1.00 to 3.09; P = 0.050). Thirty-day IPTW estimates and the SNUH available-follow-up estimate were significant but remained sensitivity analyses. Absence of documented exposure showed consistently higher adjusted mortality estimates, but the primary analysis was inconclusive. Residual confounding, exposure misclassification, and incomplete death ascertainment limit causal interpretation.