The trusted source for
healthcare information and
By Elaine Chen, MD
Assistant Professor, Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, Section of Palliative Medicine, Rush University Medical Center, Chicago
Dr. Chen reports no financial relationships relevant to this field of study.
SYNOPSIS: In periods of high medical ICU occupancy, acceptance to the medical ICU may decrease.
SOURCE: Mathews KS, Durst MS, Vargas-Torres C, et al. Effect of emergency department and ICU occupancy on admission decisions and outcomes for critically ill patients. Crit Care Med 2018;46:720-727.
Nationwide, the volume of ICU admissions from the ED has increased significantly over recent years (by 50% from 2001 to 2009). When demand exceeds bed availability, complex decisions regarding ICU must be made. Does bed availability affect triage decisions? If many beds are available, patients who are too ill or too well to benefit from the ICU may be admitted. Conversely, if too few beds are available, ICU admission may be denied to patients who may benefit. In prior studies, ICU denial has been associated with increased hospital mortality.
Matthews et al performed this retrospective cohort study of critically ill ED patients to measure the effect of ED crowding and ICU occupancy on ICU admission decisions and to investigate the potential association of delays in admission with in-hospital morbidity and mortality. The setting was a single-center, urban, academic, tertiary care center with a 14-bed closed medical ICU (MICU) that operates at 91% average occupancy. The institution had four additional specialty ICUs to which patients may be admitted as MICU “overflow.” The ED had a five-bed high-acuity area.
The patient cohort included all adult ED patients for whom medical ICU admission was requested over a 21-month period. ICU admission began with a request from the ED physicians, followed by an in-person evaluation by the MICU team, and concluded with final decision by the ICU attending physician. Patients “boarded” in the ED until a bed in the admitting unit was available, with the ED team as the primary team for ICU admissions. For patients admitted to an acute care unit, the accepting medical team assumed care while patients still were in the ED. A critical care consult service was available to assist those patients not accepted to an ICU or those in another ICU as “overflow.”
The study had two primary objectives: 1) identify whether ED and ICU volume were predictors of ICU admission decisions, and 2) measure whether post-consult ED boarding time affected in-hospital morbidity (defined as persistent organ dysfunction) and mortality at 28 days. Patient-related characteristics collected included: general demographics, severity of illness scores, timing of consult, primary admission diagnosis, and code status/goals of care at various times. Hospital-related predictors included continuous measurement of ED and inpatient census and overall hospital occupancy. Statistical methods applied included T-test, chi-square, analysis of variance, and multivariable logistic regression. Propensity score methods were used to find factors associated with persistent organ dysfunction and/or death (POD+D).
During this period, there were 854 consults from the ED to the MICU, representing 43.7% of all ICU consults. Overall, 455 patients were accepted to the MICU, with 57 requiring overflow admission. Those who were accepted were younger (mean age, 61 vs. 65 years), were not from nursing homes (12.5% vs. 24.8%), and had higher severity of illness (median mortality probability models scores, 0.15 vs. 0.13). Compared to patients denied admission to the MICU, there were more pulmonary system diagnoses in the accepted group (41.5% vs. 30.8%). There was no association between ED census and admission decision. The MICU often was more full at the time of denial (32.8% of time when patients were denied vs. 25.7% of time when accepted).
Regarding patient outcomes, longer ED boarding time after consult, nursing home origin, and higher initial severity of illness were associated with increased POD+D (i.e., worse outcomes). For those accepted to the MICU, outcomes based on location in the primary MICU or overflow in another ICU were not significantly different (P = 0.44).
In this retrospective cohort analysis, MICU bed availability was found to have a significant effect on the decision to admit critically ill patients in the ED to the ICU, even after adjusting for patient characteristics. This is consistent with other studies that have shown that ICU bed availability affects triage. Longer boarding times in the ED were associated with worse outcomes. Census of other ICUs did not affect admission decisions.
This study is an interesting evaluation of a single center with a specific admission system from the ED to the MICU. There exist myriad triage models from the ED to ICU and additional models for management of patients who meet ICU criteria when ICU beds are unavailable.
In the model studied, the MICU team evaluates and decides whether patients are appropriate for MICU level of care; the authors found that when the MICU was full, patients had a lower likelihood of acceptance to the MICU. However, census in the other ICUs and ED had no effect on acceptance decisions. This suggests that physicians accounted for, either consciously or subconsciously, the areas of their own work but not necessarily others. The authors evaluated only those patients for whom ICU admission was requested. Were there patients for whom ICU was considered by the ED physicians, but request was not placed? Does ICU census affect ED physicians’ request rate? Does acute care census affect the request rate?
When there are no beds available, patients cannot go to them. In this study, patients who experienced a prolonged ED boarding time while waiting for available ICU beds had worse outcomes. This highlights the importance of optimizing throughput and improving care for waiting patients. However, patients admitted to other ICUs staffed by non-MICU physicians with critical care consultation had similar outcomes. This is encouraging, and consistent with prior studies.1 In my institution, some overflow patients are staffed by medical intensivists, and others are staffed by intensivists trained in other primary backgrounds (such as neurology, surgery, or anesthesia). Could these staffing differences affect outcomes? As with many retrospective cohort studies, this study does not change practice, but challenges our biases and asks questions that may help improve hospital flow.
Financial Disclosure: Critical Care Alert’s Physician Editor Betty Tran, MD, MSc, Nurse Planner Jane Guttendorf, DNP, RN, CRNP, ACNP-BC, CCRN, Peer Reviewer William Thompson, MD, Executive Editor Leslie Coplin, Editor Jonathan Springston, and Editorial Group Manager Terrey L. Hatcher report no financial relationships relevant to this field of study.