PROLONGED EMERGENCY DEPARTMENT LENGTH OF STAY AND ALTERNATE LEVEL OF CARE STATUS AS MARKERS OF PROLONGED HOSPITALIZATION AMONG OLDER ADULTS: AN EXPLORATORY SYNTHETIC DATA ANALYSIS
Author(s)
Waseem Abu-Ashour, PharmD, PhD1, Marcel Miron-Celis, MSc2, Emelly Rusli, MSc2, Aaron Galaznik, MBA, MD3, Alan Forster, MD, M.Sc1.
1McGill University Health Center, Montreal, QC, Canada, 2MDClone, San Jose, CA, USA, 3MDClone, Belmont, MA, USA.
1McGill University Health Center, Montreal, QC, Canada, 2MDClone, San Jose, CA, USA, 3MDClone, Belmont, MA, USA.
OBJECTIVES: To describe emergency-to-inpatient care pathways among older adults and examine whether prolonged emergency department (ED) length of stay (LOS) and alternate level of care status were associated with longer hospital LOS using MDClone-generated synthetic data to support operational insight into patient flow.
METHODS: We conducted a retrospective encounter-level analysis of adults aged ≥65 years with an ED encounter classified as inpatient/emergency and linked to hospital admission. ED LOS was categorized as ≤12, >12-24, >24-48, and >48 hours. The primary outcome was hospital LOS. Descriptive analyses summarized ED LOS, hospital LOS, stretcher duration, consult count, ICU admission, in-hospital mortality, and Niveau de soins alternatif (NSA; alternate level of care) status. Negative binomial regression estimated incidence rate ratios (IRRs) for hospital LOS across ED LOS categories, adjusting for age, gender, triage priority, and mode of arrival. Sensitivity analysis excluded hospital LOS above the 95th percentile.
RESULTS: The synthetic dataset included 66,625 encounters. Median age was 77.8 years, median ED LOS was 25.0 hours, and median hospital LOS was 7.0 days. Overall, 52.1% of encounters had ED LOS >24 hours and 24.0% had ED LOS >48 hours. Encounters with ED LOS >48 hours had median hospital LOS of 9 days versus 6 days in shorter ED LOS groups. In adjusted models, ED LOS >48 hours was associated with longer hospital LOS versus ≤12 hours (IRR 1.37; 95% CI 1.33-1.41); results persisted after excluding extreme hospital LOS. NSA status occurred in 11.6% of encounters and was associated with longer median hospital LOS (19 vs 6 days).
CONCLUSIONS: In synthetic encounter-level data, very prolonged ED LOS and NSA status identified older-adult encounters with extended hospitalization, suggesting upstream boarding and downstream discharge-transition bottlenecks. Findings warrant validation using original clinical or administrative data.
METHODS: We conducted a retrospective encounter-level analysis of adults aged ≥65 years with an ED encounter classified as inpatient/emergency and linked to hospital admission. ED LOS was categorized as ≤12, >12-24, >24-48, and >48 hours. The primary outcome was hospital LOS. Descriptive analyses summarized ED LOS, hospital LOS, stretcher duration, consult count, ICU admission, in-hospital mortality, and Niveau de soins alternatif (NSA; alternate level of care) status. Negative binomial regression estimated incidence rate ratios (IRRs) for hospital LOS across ED LOS categories, adjusting for age, gender, triage priority, and mode of arrival. Sensitivity analysis excluded hospital LOS above the 95th percentile.
RESULTS: The synthetic dataset included 66,625 encounters. Median age was 77.8 years, median ED LOS was 25.0 hours, and median hospital LOS was 7.0 days. Overall, 52.1% of encounters had ED LOS >24 hours and 24.0% had ED LOS >48 hours. Encounters with ED LOS >48 hours had median hospital LOS of 9 days versus 6 days in shorter ED LOS groups. In adjusted models, ED LOS >48 hours was associated with longer hospital LOS versus ≤12 hours (IRR 1.37; 95% CI 1.33-1.41); results persisted after excluding extreme hospital LOS. NSA status occurred in 11.6% of encounters and was associated with longer median hospital LOS (19 vs 6 days).
CONCLUSIONS: In synthetic encounter-level data, very prolonged ED LOS and NSA status identified older-adult encounters with extended hospitalization, suggesting upstream boarding and downstream discharge-transition bottlenecks. Findings warrant validation using original clinical or administrative data.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
HSD76
Topic
Epidemiology & Public Health, Health Service Delivery & Process of Care, Real World Data & Information Systems
Disease
No Additional Disease & Conditions/Specialized Treatment Areas