A Model for Identifying Patients Who May Not Need Neurologic Intensive Care Unit Admission: Resource Utilization...

Limited resources, neurointensivists, and neurologic intensive care unit (neuro-ICU) beds warrant investigating models for predicting who will benefit from admission to neuro-ICU. This study presents a possible model for identifying patients who might be too well to benefit from admission to a neuro-ICU.

We retrospectively identified all patients admitted to our 16-bed neuro-ICU between November 2009 and February 2013. We used the Acute Physiology and Chronic Health Evaluation (APACHE) outcomes database to identify patients who on day 1 of neuro-ICU admission received 1 or more of 30 subsequent active life-supporting treatments. We compared 2 groups of patients: low-risk monitor (LRM; patients who did not receive active treatment [AT] on the first day and whose risk of ever receiving AT was ≤ 10%) and AT (patients who received at least 1 of the 30 ICU treatments on any day of their ICU admission).

There were 873 (46%) admissions in the LRM group and 1006 (54%) admissions in the AT group. The ICU length of stay in days was 1.7 (± 1.9) for the LRM group versus 4.5 (± 5.5) for the AT group. The ICU mortality was 0.8% for the LRM group compared to 14% for the AT group (odds ratio [OR] = 17.6; 95% confidence interval [CI], 8.2-37.8, P < .0001). Hospital mortality was 1.9% for the LRM group compared to 19% for the AT group (OR = 9.7; 95% CI, 5.8-16.1, P < .0001).

The outcome for LRM patients in our neuro-ICU suggests they may not require admission to neurologic intensive care. This may provide a measure of neuro-ICU resource use. Improved resource use and reduced costs might be achieved by strategies to provide care for these patients on floors or intermediate care units. This model will need to be validated in other neuro-ICUs and prospectively studied before it can be adopted for triaging admissions to neuro-ICUs.