Document Type


Publication Date


Academic Year



Introduction: Given the current era of decreasing hospital bed availability, there has been a rise in the practice of direct discharge to home (DDH) from ICUs. We evaluated the demographics, clinical characteristics, outcomes and readmission patterns among DDH patients.

Methods: Retrospective review of patients from 2 MICUs from June 2017 to June 2019 at Thomas Jefferson University hospital, an urban tertiary care center. Primary outcome of interest was 30-day hospital readmission. Patients were dichotomized into two groups based on time between ward transfer order and hospital discharge (<24 or ≥24 hours). Risk adjustment performed with Mortality Probability Model (MPM0 -III). ICU workload at admission and discharge was estimated with nine equivalents of nursing manpower use score (NEMS). Patient characteristics compared using t-test and Fisher exact or χ2 test.

Results: 331 DDH patients were analyzed, with the majority (68.3%, 226/331) waiting <24 hours for discharge. Mean LOS significantly longer in patients who had waited ≥24 hours prior to discharge compared to that of patients who waited <24 hours (4.63 vs 2.65 days, p<0.001). 10.3% (45/331) presented to TJU for evaluation within 30 days of discharge. Of these patients, 75.6% (34/45) were readmitted. No significant difference in severity-of-illness, admission NEMS, or 30-day readmission between the 2 groups (p=0.70).

Discussion: Shorter wait-times for ICU patients after being determined ready for DDH were associated with shorter hospital and ICU LOS but not with an increase in 30-day readmissions. Further examining pre-discharge and post-discharge data could better identify those at risk of readmission.