(324) The Association Between Pre-Intensive Care Unit Hospital Length of Stay and Hospital Survival

Monday, February 24, 2025

2:30pm – 3:30pm Eastern Time

Location: Connections Central - RST 02

First Author(s)

Arisheniah Sothilingam, MD (she/her/hers)

Montefiore Medical Center
Yonkers, New York, United States

Co-Author(s)

JG

Jack Galfund, MD

Internal Medicine PGY-2 Resident
Albert Einstein College of Medicine, Montefiore Medical Center, United States
SN

Shivang Narendrakumar Bhakta Bhakta, MD

Internal Medicine PGY-3 Resident
Albert Einstein College of Medicine, Montefiore Medical Center, United States
KP

Kieran Palumbo, MD

Internal Medicine PGY-3 Resident
Albert Einstein College of Medicine, Montefiore Medical Center, United States
AG

Arup Gope, MS

Data Analyst
Montefiore Medical Center, Department of Critical Care Medicine, United States
TB

Thomas Butler, MS

Data Analyst
Montefiore Medical Center, Department of Critical Care Medicine, United States
AM

Ari Moskowitz, MD, MPH

Medical Intensive Care Unit Director and Associate Professor of Medicine
Montefiore Medical Center, Department of Critical Care Medicine, United States
AS

Ariel Shiloh, MD

Director of the Critical Care Medicine Consult Service and Professor of Medicine
Albert Einstein College of Medicine Montefiore Medical Center, United States

Introduction: While physiological predictors for Intensive Care Unit (ICU) survival have been described, the relationship between pre-ICU length-of-stay (LOS) and outcomes remains underexplored. This study assessed the association between pre-ICU LOS and survival, hypothesizing that longer pre-ICU LOS correlates with worse survival.

Methods: This retrospective cohort study, conducted at three hospitals included adults (≥18 years) admitted to any ICU between 2016-2023. Demographics, LOS, and outcomes were abstracted from the Electronic Health Record. Reasons for admission and decompensation for patients with a pre-ICU LOS > 30-days were reviewed. Logistic regression, controlling for confounders including severity-of-illness, assessed the association between pre-ICU LOS and hospital mortality.

Results: Out of 30,417 patients, the median time from hospital to ICU admission was 0.5-days (IQR 0.3, 1.2), and 126 (0.4%) patients had a pre-ICU LOS ≥ 30-days. In the multivariable regression, pre-ICU LOS was associated with higher odds of mortality (aOR 1.09, 95% CI 1.09-1.10), p< 0.01), with significance persisting even after excluding patients who were admitted directly to the ICU without a prior ward stay. Mortality rates increased with longer pre-ICU LOS: 14.5% for < 5-days, 22.0% for 5-10 days, 32.6% for 10-20 days, 39.8% for 20-30 days, and 44.4% for >30-days. Among patients with a pre-ICU LOS ≥30-days, most had an original hospital admission for a medical, non-cardiac problem (67.1%). The reason for ICU admission in patients with a pre-ICU LOS >30-days was acute in 62.1%, subacute in 16.4%, and post-procedural in 21.4%. Of patients who survived after a pre-ICU LOS >30-days, 52.5% had an acute decompensation leading to ICU admission and 31.3% were admitted in the peri-procedural setting.

Conclusions: Our study found an association between longer pre-ICU LOS and hospital mortality. While mortality was high in patients with pre-ICU LOS exceeding 30-days, most of these patients survived to discharge—a finding that should be interpreted in the context of ICU triage decisions which may favor patients with higher survival potential. These results may assist in the triage of patients from the ward to the ICU and provide prognostic information for families.