View Abstract

A3302 - Characteristic and Outcomes of Delayed Intensive Care Unit Admission in Patients with Infection
Author Block: S. P. Taylor, B. T. Taylor, E. Gunn, A. Heffner, C. Karvetski; Carolinas Medical Center, Charlotte, NC, United States.
Introduction: Delayed admission to the ICU compared to direct admission from the Emergency Department (ED) has been associated with increased in-hospital mortality. One study reported that most early transfers are due to inappropriate admission triage. Interestingly, delayed ICU admission has not been independently linked with increased mortality in patients with sepsis. We hypothesized that examining a broader group of patients identified as infection present on admission might show an association between delayed ICU admission and mortality. Methods: Retrospective analysis of patients admitted from 9 Emergency Departments in our healthcare system with infection at risk for clinical deterioration. This cohort, referred to as Infection Present on Admission (POA), is defined as a final primary or any secondary discharge diagnosis for UTI, bacteremia, pneumonia, cellulitis, C Diff, septicemia, sepsis, severe sepsis, or septic shock where the diagnosis was coded as present on admission. We evaluated patients who required an ICU admission within 120 hours of hospitalization. Patients admitted from the ED to the ICU were direct ICU admit patients. Transfers were early if they occurred within the first 12h, intermediate between 12-72h and late between 72-120h. Results: Of the 13,912 patients with infection POA requiring ICU admission within 120h, 11,605 (84.2%) were directly admitted to the ICU, 638 (4.6%) were transferred in the first 12h, 1198 (8.7%) were transferred between 12-72h, and 337 (2.4%) between 72-120h. APACHE IVa scores at the time of ICU admission were similar for patients requiring direct ICU admission versus transfer at any time. However, APACHE IV predicted mortality at the time of ICU admission differed among patients with direct admission versus transfer (direct admission 17.8, transfer within 12h: 16.1, between 12-72h: 21.5, and between 72-120: 25.6). Looking back at triage clinical markers, triage shock index was higher for direct admissions (0.90) compared to transfers within 12h (0.82), between 12-72h (0.77) and between 72-120h (0.77). Triage lactate level was also higher. Patients with direct ICU admission had similar mortality (13.9%) to patients with delayed transfer within 12h (12.1%), and 12-72h (15.3%), but lower mortality than patients requiring transfer at 72-120h (20.8%).Conclusions: Patients presenting to the ED with infection who require direct admission to the ICU have worse triage markers but no worse mortality than patients with delayed admission. It does not appear that delayed ICU admission is due to failure to appreciate severity of illness at triage. Patients with late admission have the highest predicted and actual mortality.