Skip to content

Projects | Completed

Benefits of ICU Admission for Patients with Acute Respiratory Failure or Sepsis: A Mixed-Methods Study Across 26 Hospitals (ICU Net Benefit)

Research Areas

Principal Investigator

Headshot of Scott Halpern
MD, PhD

Scott Halpern

PAIR Center Research Team

Headshot of George Anesi
MD, MSCE, MBE, FCCM, ATSF

George Anesi

Erich Dress Headshot
MPH, MBE

Erich Dress

MS, MBMI

Brian Bayes

Headshot of Dylan Small
PhD

Dylan Small

Headshot of Kit Delgado
MD, MS

M. Kit Delgado

Topics

Overview

Hospitalizations for patients with acute respiratory failure (ARF) and sepsis are common and costly. Special hospital floors called intensive care units (ICUs) were designed to care for patients who require life support, but they have since expanded to care for patients who do not require these therapies. It is unknown, among patients with sepsis and ARF who do not require life support, who receives a true benefit from receiving care in an ICU. Learning to care for these groups of patients on non-specialized hospital floors (such as a ward) would create healthcare system-wide benefits including better patient outcomes, reduced costs, and enhanced access to scare ICU resources for patients who need them the most. We conducted a retrospective study of over 135,000 patients admitted to 27 U.S. hospitals between 2013 and 2018 with sepsis and/or ARF who did not require life support in the emergency department. We measured the net benefit of ICU admission on patient outcomes including hospital length-of-stay and mortality. We used mixed methods to identify hospital characteristics and care processes that may contribute to observed benefits of ICU- or ward-based care.

Results & Impact

Choices to admit patients with ARF and/or sepsis to the ICU versus ward are associated with substantial differences in outcomes. On average, analyzing data combined across 27 hospitals showed patients with sepsis who do not require life support may have better outcomes, specifically a shorter time in the hospital and lower risk of dying in the hospital, when admitted to a ward for care. On average, analyzing data combined across 27 hospitals showed patients with ARF who do not require life support may have better outcomes (same as above) when admitted to an ICU for care. The size and direction of the difference in these outcomes for ARF and sepsis patients vary among hospitals. Cultural and communication factors were cited as main contributors by clinicians and administrators at hospitals where the highest benefits of ICU admission were observed. More work is needed to explain the observed outcome differences and apply these insights to improve care for patients with ARF and sepsis regardless of where they are treated.

Click here to view the 11 publications from this study as well as editorial and press coverage.

  • A plot association of ICU versus ward admission with hospital length of stay (LOS) among patients with sepsis and ARF.
    Association of ICU versus ward admission with hospital length of stay (LOS) among patients with sepsis and ARF. Among high-acuity patients with sepsis who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with a 1.32-day longer hospital LOS. Among high-acuity patients with ARF who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with a 0.82-day shorter hospital LOS. (Anesi GL, et al. Am J Respir Crit Care Med. 2022. doi:10.1164/rccm.202106-1350OC)
  • A plot of the association of ICU versus ward admission with in-hospital mortality among patients with sepsis and ARF.
    Association of ICU versus ward admission with in-hospital mortality among patients with sepsis and ARF. Among high-acuity patients with sepsis who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with 48% increased odds of in-hospital mortality. Among high-acuity patients with ARF who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with 25% decreased odds of in-hospital mortality. (Anesi GL, et al. Am J Respir Crit Care Med. 2022. doi:10.1164/rccm.202106-1350OC)
  • A conceptual model of ED-ward-ICU patient flow.
    Patients may be admitted through the ED to general medical wards or ICUs, with varying clinician priorities depending on patient and hospital contexts (e.g., ED physicians place high priority on throughput). Not all patients admitted to the ICU require such care. ICU admission in such patients results in unnecessary exposure to ICU harms without ICU benefits. Such harms can prolong hospital stay even after transfer to the ward. Some patients admitted directly to the ward will require later critical care needs either due to later clinical decompensation or mis-triage. For such patients, on-location critical care (e.g., from a rapid response team) can sometimes be provided. Careful conceptual analysis of net benefit and its relationship to contextual factors could lead to stronger strategies for ensuring avoidance of ICU harms and enhancement of ICU benefits. (Anesi GL, et al. Ann Am Thorac Soc. 2025. doi:10.1513/AnnalsATS.202408-806OC)
  • Association of ICU versus ward admission with hospital length of stay (LOS) among patients with sepsis and ARF. Among high-acuity patients with sepsis who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with a 1.32-day longer hospital LOS. Among high-acuity patients with ARF who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with a 0.82-day shorter hospital LOS. (Anesi GL, et al. Am J Respir Crit Care Med. 2022. doi:10.1164/rccm.202106-1350OC)
  • Association of ICU versus ward admission with in-hospital mortality among patients with sepsis and ARF. Among high-acuity patients with sepsis who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with 48% increased odds of in-hospital mortality. Among high-acuity patients with ARF who did not require mechanical ventilation or vasopressors in the emergency department, admission to the ICU was associated with 25% decreased odds of in-hospital mortality. (Anesi GL, et al. Am J Respir Crit Care Med. 2022. doi:10.1164/rccm.202106-1350OC)
  • Patients may be admitted through the ED to general medical wards or ICUs, with varying clinician priorities depending on patient and hospital contexts (e.g., ED physicians place high priority on throughput). Not all patients admitted to the ICU require such care. ICU admission in such patients results in unnecessary exposure to ICU harms without ICU benefits. Such harms can prolong hospital stay even after transfer to the ward. Some patients admitted directly to the ward will require later critical care needs either due to later clinical decompensation or mis-triage. For such patients, on-location critical care (e.g., from a rapid response team) can sometimes be provided. Careful conceptual analysis of net benefit and its relationship to contextual factors could lead to stronger strategies for ensuring avoidance of ICU harms and enhancement of ICU benefits. (Anesi GL, et al. Ann Am Thorac Soc. 2025. doi:10.1513/AnnalsATS.202408-806OC)

Partnering Health Systems

Kaiser Permanente Northern California

Sponsors

National Heart, Lung, and Blood Institute (NHLBI)

Related Publications

View all publications