Discrepancies in ICU use identified

By Sarah Guy

An analysis of intensive care unit (ICU) use in the USA reveals that many patients admitted to this specialist department are at low risk for mortality, while more than half of patients with a high mortality risk are admitted to other departments.

This finding suggests that severity of illness is only one of several factors contributing to the ICU triage decision, and that other contributors are likely to include goals of care reflecting patient comfort rather than prolongation of life, and available alternatives such as hospice care, report the researchers.

The results also indicate wide variation in ICU use between hospitals, suggesting "a considerable lack of consensus about when to use the ICU," say Lena Chen (University of Michigan, Ann Arbor, USA) and co-workers.

"This work could form the basis for validating standards for ICU admission that are tailored to different risk groups and that are sensitive to hospital-specific environments," the team writes in the Archives of Internal Medicine.

Using data for all nonsurgical admissions to Veterans Affairs acute care hospitals between 2009 and 2010, Chen and colleagues assessed the severity of the health of the 31,555 (10.9%) ICU admissions. Overall, these patients had higher mean predicted (and observed) 30-day mortality rates than did patients admitted to non-ICU wards, they report, at 7.5% versus 3.5%.

However, both ICU and non-ICU patients had relatively low overall predicted mortality rates at admission, at 1.7% and 1.0%, respectively, and over half (53.2%) of patients with a 2.0% or lower predicted 30-day mortality were admitted to an ICU.

In an accompanying commentary, Jeremy Kahn and Christopher Seymour (University of Pittsburg, Philadelphia, USA) warn against filling ICU beds with low-acuity patients, saying that "low-risk patients in an ICU not only occupy a bed that could be used for someone at higher risk but also can subject patients to a less restful environment."

By contrast, Chen et al found that the majority (69.6%) of patients at high risk for death (ie, with a 30-100% predicted mortality at 30 days) were not admitted to an ICU.

"This could be due to poor triage decisions, or, it is hoped, successful solicitation of end-of-life treatment preferences among patients likely to die," remark Kahn and Seymour.

Finally, for high-risk patients with predicted mortality above 30%, Chen et al found significant variation in ICU admission rates between hospitals. For each one-standard deviation increase in patient severity, the odds of ICU admittance ranged from a 15% decrease to a 122% increase.

Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.