A retrospective observational study of daytime and nighttime transfers from the intensive care unit: through the lens of critical care response teams.

Abstract

To evaluate the impact of nighttime compared with daytime transfers from the intensive care unit (ICU) on mortality in a hospital with a critical care response team (CCRT). We performed a retrospective observational study of ICU patients transferred between January 2011 and July 2013 who received CCRT follow-up. The transferred patients were divided into cohorts of daytime and nighttime transfers. A multivariable logistic regression model was used to identify independent predictors of mortality after ICU transfer. There were 1,857 patients included in the study. With the exception of Multiple Organ Dysfunction Score on admission, transfers to a step-down unit, and lower urine output, there were no differences in the baseline characteristics, clinical events identified by CCRTs, and the number of CCRT interventions performed between daytime and nighttime transfers. Patients transferred at night were at higher risk of death in the univariate analysis but not in the multivariate analysis. Independent predictors of mortality included older age (odds ratio [OR], 1.02; 95% confidence interval [CI], 1.002 to 1.04), transfer to a medical service (OR, 1.96; 95% CI, 1.11 to 3.43), CCRT identification of hypoxemic respiratory failure (OR, 5.86; 95% CI, 3.11 to 11.04), decreased level of consciousness (OR, 3.14; 95% CI, 1.23 to 8.02), hypotension (OR, 3.69; 95% CI, 1.36 to 10.01), and longer CCRT duration of follow-up (OR, 1.02; 95% CI, 1.004 to 1.03). Nighttime transfer from the ICU was not an independent predictor of mortality. We identified unique predictors of mortality, including clinical events that CCRTs identified in patients immediately after ICU transfer. Future studies are required to validate these predictors of mortality in transferred ICU patients.