Effect of non-patient-level factors on mortality in intensive care units

Several studies have suggested that mortality of intensive care unit(ICU) is influenced by various factors. But most of them were focused on patient-level variables. This study was performed to determine whether staffing, organization structure and admission time have a significant effect on ICU mortality. A total of 751 patients were enrolled in this multicenter, retrospective cohort study. Patients were admitted to ICUs of four tertiary care medical centers from January 2013 through December 2014 in South Korea. We evaluated mortality difference between ICU with intensivist and ICU without intensivist. We also compared mortality between high and low patient-to-nurse ratio group(6:1 versus 3:1). ICU type was divided into two groups (divided model includes medical (MICU)/surgical (SICU) and mixed model MICU+SICU). Mortality of two groups was compared. The relationship between admission time(weekday, weekend) and mortality was evaluated. The overall mortality was 23.3%. ICU with intensivist is associated with reduced mortality(21.3% vs. 24.8%; OR, 0.57; 95% CI, 0.38-0.87; p =0.008). The Mortality of divided ICU was lower than that of mixed ICU(19.5% vs. 25.3; OR, 0.49; 95% CI, 0.28-0.83; p =0.034). However, there was no significant relationship between patient-to-nurse ratio and mortality(6:1 vs. 3:1; 25.7% vs. 21.6%; OR, 0.88; 95% CI, 0.50-1.54; p =0.801). No mortality difference was found between weekday and weekend admission(22.6% vs. 23.6%; OR, 0.65; 95% CI, 0.38-1.11; p =0.765). This study suggests that intensivist staffing and divided ICU model are significantly associated with reduced ICU mortality. However, patient-to-nurse ratio and admission time are not associated with ICU mortality.