Sean G Kelly1, Miles Hawley, James O'Brien. 1. Department of Internal Medicine, University of Wisconsin Hospitals and Clinics, 4240-01A MCFB, Madison, Wisconsin 53705, USA. SKelly@uwhealth.org
Abstract
PURPOSE: To evaluate whether bed availability affects a physician's decision to request or offer an intensive care unit (ICU) transfer. MATERIALS AND METHODS: We administered mail-based surveys to determine the respondents' probability of either requesting an ICU transfer (generalist respondents) or offering an ICU transfer (intensivist respondents). Respondents randomly received clinical vignettes that were identical except for the number of available ICU beds (one or seven available ICU beds). Respondents also made predictions about the patient's outcomes. RESULTS: Among generalists and intensivists, there were wide ranges in decisions about ICU transfer. In the Generalist ICU request study, the average probability of transfer with one versus seven available ICU beds was 52.2% and 58.5% (P = .41), respectively. In the Intensivist ICU offer study, the average probability of transfer with one versus seven available ICU beds was 62.5% and 57.4% (P = .24), respectively. The most consistent association with decisions about ICU transfer was the predicted probability that a patient would require an ICU bed in the future if not transferred currently. CONCLUSIONS: There is high variability in the decision to request or offer ICU beds. There was not a significant association between bed availability and ICU transfer decisions.
PURPOSE: To evaluate whether bed availability affects a physician's decision to request or offer an intensive care unit (ICU) transfer. MATERIALS AND METHODS: We administered mail-based surveys to determine the respondents' probability of either requesting an ICU transfer (generalist respondents) or offering an ICU transfer (intensivist respondents). Respondents randomly received clinical vignettes that were identical except for the number of available ICU beds (one or seven available ICU beds). Respondents also made predictions about the patient's outcomes. RESULTS: Among generalists and intensivists, there were wide ranges in decisions about ICU transfer. In the Generalist ICU request study, the average probability of transfer with one versus seven available ICU beds was 52.2% and 58.5% (P = .41), respectively. In the Intensivist ICU offer study, the average probability of transfer with one versus seven available ICU beds was 62.5% and 57.4% (P = .24), respectively. The most consistent association with decisions about ICU transfer was the predicted probability that a patient would require an ICU bed in the future if not transferred currently. CONCLUSIONS: There is high variability in the decision to request or offer ICU beds. There was not a significant association between bed availability and ICU transfer decisions.