ABSTRACT & COMMENTARY
Reduced ICU Bed Availability is Associated with Worse Outcomes on the General Wards
By Betty T. Tran MD, MSc
Assistant Professor of Medicine, Pulmonary and Critical Care Medicine, Rush University Medical Center, Chicago
Dr. Tran reports no financial relationships relevant to this field of study.
This article originally appeared in the September 2014 issue of Critical Care Alert. It was edited by David J. Pierson, MD, and peer reviewed by William Thompson, MD. Dr. Pierson is Professor Emeritus, Pulmonary and Critical Care Medicine, University of Washington, Seattle, and Dr. Thompson is Associate Professor of Medicine, University of Washington, Seattle. Drs. Pierson and Thompson report no financial relationships relevant to this field of study.
SYNOPSIS: This observational cohort study found that reduced ICU bed availability is associated with increased rates of ICU readmission as well as ward cardiac arrest if medical ICU beds were on shortage.
SOURCE: Town JA, et al. Relationship between ICU bed availability, ICU readmission, and cardiac arrest in the general wards.
Crit Care Med 2014; April 25. [Epub ahead of print.]
Prior studies have reported that limited ICU bed availability is associated with various outcomes such as increased severity of illness in patients admitted to the ICU and shorter ICU length of stay.1 Given limited knowledge of other system-related effects of ICU bed availability, Town et al aimed to explore whether ICU bed capacity was associated with rates of ICU readmission within 24 hours and cardiac arrest on the general wards.
The study was conducted at the University of Chicago Medicine, a tertiary academic center that has 63 adult ICU beds and 272 adult general inpatient ward beds. Data on ICU bed availability were collected, and rates of ICU readmission within 24 hours of discharge from the ICU and ward cardiac arrest were calculated per 12-hour shift over a period of 3 years. Care was taken to minimize incomplete data (4.7%) in defining ICU bed availability by availability of nursing rather than number of open physical beds, as well as in excluding comfort care deaths from the analysis.
Over the 3-year period consisting of 8238 discharges from the ICU, there were 245 (3%) readmissions within 24 hours, resulting in an ICU readmission rate of 2.63 per 100 discharges. The overall ward cardiac arrest rate was 2.63 per 10,000 patient-shifts during the study period, with 72% of ward cardiac arrests occurring in patients on a medical service. With each unit decrease in total ICU bed availability, the odds of ICU readmission increased significantly (odds ratio [OR] 1.06; 95% confidence interval [CI], 1.00-1.12; P = 0.03). A similar but non-statistically significant trend was seen between total ICU bed availability and ward cardiac arrest rates; however, MICU bed availability in particular was significantly associated with ward cardiac arrest rates (OR, 1.25; 95%CI, 1.06-1.49; P = 0.01).
ICU beds are at a premium in busy hospitals, and this study highlights important considerations when planning for and evaluating this resource on a wider, administrative scale. The authors found that ICU bed availability had indirect ripple effects throughout their hospital, particularly on the rate at which patients who transferred out of the ICU returned within 24 hours and the overall cardiac arrest rate on the wards. The first association makes intuitive sense. We have all probably received early morning pages from our bed managers requesting expedition of patient transfers out of the unit when ICU beds are full. This push to open up beds may result in moving patients who are "borderline" out of the ICU who otherwise may have spent another night there for closer monitoring.
The relationship between ICU bed availability and ward cardiac arrest rates, however, is more complicated. It is unclear whether the patients who suffered a cardiac arrest on the wards did so because there was a shortage of medical ICU beds. This association would be stronger if, for example, we knew that these patients were recently admitted to the ICU, were being evaluated for ICU transfer prior to their arrest, or had been accepted to the ICU but were waiting for an open bed, but these data are not available in the current study. The authors note that their hospital does have a rapid response team (RRT) that is triggered by general concerns; data on whether those patients who suffered cardiac arrests on the wards had been evaluated by the RRT prior to their events would have strengthened the observed association between ICU bed availability and ward cardiac arrests as it would imply that these patients were sicker but that bed availability may have influenced triage decisions at the time. In addition, institutional policy regarding patient admission to the ICU, specifically who is the designated ICU "gatekeeper" (e.g., resident, fellow, or attending) at each medical center, can also play a role and needs to be considered. Indeed, Town et al dutifully note that their findings are based on shift level rather than patient-specific data, and only associations rather than causal relationships can be gleaned from their observational study.
The effect of reduced ICU bed availability on patient outcomes remains an important topic for continued investigation, not only for patient care but also in ICU organization and management, particularly in times of expected strain such as influenza season or during outbreaks.
1. Sinuff T, et al for the Values Ethics and Rationing in Critical Care (VERICC) Task Force. Rationing critical care beds: A systemic review. Crit Care Med 2004;32:1588-1597.