Does after-hours discharge of ICU patients influence outcome?

Objective: To assess the frequency of after-hours discharges of patients from the intensive care unit and its effect on in-hospital mortality.

Design, Setting And Participants: Observational cohort study conducted in the ICU of Westmead Hospital, a tertiary care teaching hospital in Sydney. All adult patients admitted to the ICU between 1 January 2004 and 31 December 2006 were included in the study cohort. Patients were grouped into two categories based on the time of discharge from the ICU: during work hours (08:00-17:59 hours) or after-hours (18:00-07:59 hours).

Main Outcome Measure: Mortality after discharge from the ICU according to time of discharge.

Measurements And Main Results: 2300 patients accounted for 2451 admission episodes during the study period. There were 151 readmissions, involving 133 patients (5.8%). Excluded from the study cohort were 36 patients (1.6%) who were discharged home, 39 (1.7%) who were transferred to other hospitals, and 354 (15.4%) who died during their first stay in the ICU. Data on the remaining 1871 patients who were discharged alive at the end of their first ICU admission were included in our analysis. Of these patients, 1221 (65.3%) were discharged from the ICU during work-hours and 650 (34.7%) after-hours. Crude mortality for patients discharged after-hours was 13.7%, compared with 10.1% for those discharged during work hours. After adjustment for age, APACHE II score and discharge destination, the risk of mortality among patients discharged after-hours was statistically significant (adjusted odds ratio, 1.38 [95% CI, 1.01-1.88]; P < 0.05).

Conclusions: A high proportion of patients (34.7%) were discharged from the ICU after-hours. Discharge after-hours was associated with a higher risk of in-hospital mortality than discharge during work hours.