Driving in the Wrong Direction: Exploring the Unintended Consequence of an Early Discharge Program on Length of Stay in Hospital Setting

Abstract

Importance: Early discharge of patients has become standard work in acute care settings to reduce inpatient length of stay (LOS), improve patient flow, and reduce boarding in the emergency department (ED). Objective: Retrospective analysis of outcomes from a discharge by 11 am program at an academic medical center from January 1, 2020, to June 30, 2022. The analysis addresses the effects of a discharge by 11 am goal on time from discharge order release to patient discharge, ED boarding, LOS, and observed-to-expected LOS. Design, Setting, and Participants: Patient-level electronic health record data included discharge order entry time, discharge time, LOS, and diagnosis-related group geometric LOS (GMLOS). Additional unit-level data for ED boarding volumes and hours were included. Analyses were conducted at the hospital and unit levels where indicated. Results: Patients with a discharge order by 9 am have longer mean hours from order to discharge than patients without a discharge order by 9 am (9.04 vs 2.48 hours, P < .001) ED boarding total (R 2 = 46.2%, P ≤ .001), percentage (R 2 = 50.4%, P ≤ .001), median minutes (R 2 = 24.6%, P = .005), and total minutes (R 2 = 40.8%, P ≤ .001) all increased as discharge by 11 am performance improved. The mean LOS is longer for the discharge by 11 am group than the non–discharge by 11 am group -1.67; 95% CI, -2.03 to −1.28, P < .001). Discharge by 11 am patients had a LOS/GMLOS ratio 21.9% higher than the non–discharge by 11 am cohort (difference -0.31; 95% CI, -0.36 to -0.26, P < .001). Conclusions: Discharge order entry and release by 9 am and patient physically discharged by 11 am initiatives demonstrate a statistical increase in time from discharge order to discharge time, ED boarding, LOS, and observed-to-expected LOS.