Automated Surveillance of Ventilator-Associated Events and Mortality Before and During the COVID-19 Pandemic: A Single-Center Retrospective Study in Japan.

Purpose Using in-house automated surveillance of ventilator-associated events (VAE) from our electronic health record (EHR) system, we investigated clinical and microbiological factors associated with mortality among patients receiving mechanical ventilation (MV). Our primary objective was to identify risk factors for mortality in patients during MV. Our secondary objectives included assessing the impact of COVID-19 and examining any seasonal variations in mortality during MV. Methods All patients who received MV between 2013 and 2025 were included. The internal review board approved the study as a retrospective analysis in 2026. We defined cold seasons as October through March. Using a straightforward VAE calculator in accordance with the Centers for Disease Control and Prevention's guidelines, the primary outcome was all-cause mortality during MV. We excluded deaths after extubation to study patients undergoing MV alone. Patients with missing data due to historical changes in documentation were analyzed collectively in a subgroup analysis. Results A total of 17,881 ventilator days were recorded among 2,003 patients. Multivariate analysis indicated that significant risk factors for mortality included age ≥ 75 (odds ratio [OR] 1.343, 95% confidence interval [95% CI] 1.006-1.792, P = 0.045) and the presence of ventilator-associated conditions (OR 1.926, 95% CI 1.017-3.645, P = 0.044). Cold seasons were associated with lower mortality during MV (OR 0.752, 95% CI 0.572-0.987, P = 0.040). Additionally, there was a marginal association between cold seasons and carbapenem use (P = 0.07), a class of broad-spectrum antimicrobials. Notably, April exhibited the highest mortality during MV, coinciding with a change in physician staffing, with a median turnover of 27.1% among all the house staff. Neither the pandemic era (OR 1.059, 95% CI 0.801-1.400, P = 0.689) nor SARS-CoV-2 positivity (OR 1.211, 95% CI 0.525-2.793, P = 0.654) was found to be associated with mortality.  Conclusions As to the primary objectives, we identified ventilator-associated condition, the first of three tiers of VAE, as a risk factor for mortality during MV. Regarding the secondary objectives, we found that the COVID-19 pandemic and SARS-CoV-2 positivity were not associated with mortality during MV. Meanwhile, mortality during MV was associated with seasonal factors; however, whether antimicrobial prescribing patterns or physician turnover contributed requires further study.