Document Type


Publication Date


Academic Year



Introduction: Discharging patients directly home from the ICU is becoming increasingly common, largely driven by decreased ward bed availability. We evaluated readmission patterns of ICU patients discharged directly home.

Methods: Retrospective review was conducted of direct discharges from the ICU to home between June 2017 and June 2019. The primary outcome of interest was 30-day hospital readmission. Patients were dichotomized by “wait-time” between transfer order and hospital discharge (<24 hours or ≥24 hours). Outcomes were compared using t-test, Fisher exact, and chi-squared. Risk-adjustment was performed using the Mortality Probability Model (MPM0-III). ICU workload was estimated using the nine equivalents of nursing manpower use score (NEMS).

Results: 331 patients were identified, with a mean time of 0.72 [0 - 5.84] days between ICU transfer order and discharge to home. 68.3% (226/331) of patients waited <24 hours for discharge. There was no difference in severity-of-illness or admission NEMS between the groups. 10.3% (45/331) of patients presented for evaluation within 30 days of discharge. 10.3% (34/331) of patients were readmitted. There was no significant difference in 30-day readmission between patients who were discharged after waiting <24 hours vs. waiting ≥24 hours (p=0.70).

Discussion: Patients returning directly home from the ICU without discharge delay were not readmitted more frequently within 30 days than those discharged after a delay exceeding 24 hours. Further investigation into identifying patients eligible for safe, early discharge may reduce unnecessary critical care resource utilization.