Hospital factors that influence ICU admission decision-making: An ethnographic study of six hospitals

Document Type

Conference Proceeding

Publication Date


Publication Title

American Journal of Respiratory and Critical Care Medicine


Rationale: Intensive care unit (ICU) triage-the decision whether to admit a patient to an ICU- should be guided by a patient's severity of illness; yet, hospital characteristics are known to be a key driver of ICU use. We sought to understand how factors at the hospital-level (e.g., capabilities, staffing, structure) influence triage.

Methods: We performed an ethnographic study of six hospitals across the state of Michigan. Acute care hospitals were selected based on urban/rural location, academic/community status, number of ICU beds, and all-cause ICU admission rates (based on 2015 Medicare data). We conducted hospital observations, surveyed hospital leadership, and interviewed ICU, emergency medicine, and hospitalist physicians as well as key, non-physician informants [(administrators, charge nurses, rapid response nurses, or advanced practice professionals)]. Interviews sought to understand how hospital characteristics influenced ICU triage and were semi-structured, audio-recorded, transcribed, and coded. Content was analyzed to identify prominent themes using an inductive, interpretive description approach. Hospital observations and surveys provided additional, detailed hospital-level data.

Results: Among the six hospitals, 66 participants [15 ICU physicians, 14 emergency medicine physicians, 15 hospitalist physicians, 22 key informants] were interviewed. The six hospitals differed by key characteristics (Table). Three key themes related to ICU triage were identified from interviews, surveys, and observations. First, in smaller, rural hospitals, ICU triage was defined, not by whether a patient should receive ICU care, but rather by whether a patient should be transferred to a larger hospital for a higher-level of care. These patients were not admitted to the ICU in order to facilitate more rapid transfer by remaining in the emergency department. Second, ICU nurse staffing played an indirect, yet critical, role in triage. For example, some hospitals frequently experienced transient closure of ICU beds due to ICU nursing shortages-resulting from a difficulty in recruiting and retaining nurses with critical care training. Third, all hospitals noted that the severity of illness threshold for ICU admission often fluctuated based on ICU bed availability. The presence of intermediate care mitigated some, but not all, of this variability by providing an alternate care location for patients with moderate acuity.

Conclusions: In a large, ethnographic study of six hospitals, we identified that physicians viewed the concept of ICU triage differently depending on the characteristics of the hospital in which they practiced. Yet, some of these characteristics cannot currently be measured on a large-scale, which may hinder efforts to change broad patterns of ICU use.





This document is currently not available here.