Inpatient disposition in overcrowded hospitals: is it safe and effective to use reverse triage and readmission screening tools for appropriate discharge? An observational prospective study of an Italian II level hospital

Background: Reverse triage (RT) identifies patients eligible for discharge and have been proposed to cope with daily surge. Nevertheless, early discharge could increase the rate of readmission. Our aim is to test the effectiveness and safety of RT alone and with readmission screening tools (Identification Senior At Risk [ISAR], HOSPITAL, and Groeningen Frailty Index [GFI] scores) to predict appropriate discharge. Material and methods: We prospectively assessed every 4 days (t 0 ) inpatients of medical divisions (High Dependency Unit (HDU), Internal Medicine (IM), and Geriatrics (Ger)) of an Italian Hospital. RT score was calculated for each patient and an RT ≤3 identified those eligible for safe discharge. ISAR, HOSPITAL, and GFI were then applied. We assessed reinstituting of interventions and transferring to an increased level of care unit at 4 days as an ethical proxy of consequential medical events following hypothetical discharge. Date of effective discharge, death, and readmission were measured at 4, 7, 15, and 30 days after the first evaluation. Results: Twenty-five (9.6%) out of 260 patients in our sample had an RT ≤3. Twenty-four (96%) of them compared with 205 (87%) of the RT >3 group (P = NS) were discharged. Patients with RT ≤3 were discharged significantly earlier (3.5 vs 8 days after t 0 [P = 0.0002]). In the RT ≤3 group, all but one patient were alive and healthy at 7, 15, and 30 days. The HOSPITAL score seemed to have the best concordance with RT (84%), in comparison with the ISAR (52%) and the GFI (48%) scores. RT showed a low sensitivity (22%) and high specificity (95%), which was even higher when using RT associated with readmission screening tools. Conclusions: Reverse triage proved to be a safe and conservative tool, with high specificity alone and with readmission screening tools. RT correctly identifies patients that will be discharged earlier.