Survival and Safety Outcomes of ICU Patients Discharged Directly HomeA Direct From ICU Sent Home Study

Objectives: Evaluate outcomes (mortality, morbidity, unplanned return visits) of patients who are discharged directly to home from the ICU. Design: Prospective cohort study. Setting: Two tertiary care medical-surgical-trauma ICUs at Canadian hospitals over 1 year (February 2016-2017). Subjects: All adult patients who were either discharged directly to home (Recruited and Nonrecruited cohorts) from ICU or discharged home within 24 hours after ward transfer (Ward Transfer cohort). Interventions: Direct discharge home from ICU or discharge home within 24 hours of ward transfer from ICU. Measurements and Main Results: One-hundred ninety-eight patients were in the study, 100 patients in the discharged directly to home Recruited arm, 37 patients in the discharged directly to home Nonrecruited arm, and 61 patients in the Ward cohort. All three patient cohorts had 0% mortality at 8 weeks post discharge. The unplanned return visit rate for the Recruited cohort was 24% (emergency department 18%, Ward 4%, ICU 1%), whereas the rate for the Nonrecruited cohort was 52% (emergency department 34%, Ward 14%, ICU 3%) and the Ward Transfer cohort was 46% (emergency department 17%, Ward 26%, ICU 3%) (p = 0.005). No home support was available for 7% of the discharged directly to home Recruited cohort. Twenty-four percent of patients had funded home care nursing, but the majority of patients (81%) relied on help from friends/family. Conclusions: Recruited discharged directly to home patients experienced very good 8-week postdischarge outcomes with 0% mortality and a low rate of ICU readmission (1%) or ward readmission (4%), but not an insignificant rate of emergency department visits (18%). Recruited discharged directly to home patients had better outcomes compared with nonrecruited discharged directly to home patients and patients transferred briefly to the ward prior to discharge home. Future work should include derivation of a clinical prediction tool to identify patient characteristics that make discharged directly to home safe and a randomized control trial to compare discharged directly to home with short stay ward transfers.