DIRECT VS INDIRECT ADMISSION TO INTENSIVE CARE FROM EMERGENCY CARE CENTER: A QUALITY IMPROVEMENT STUDY OF SEPSIS OUTCOMES IN OLDER ADULTS

Abstract

SESSION TITLE: Clinical Decision Making SESSION TYPE: Original Investigations PRESENTED ON: 10/20/2019 2:15 PM - 3:15 PM PURPOSE: Patients presenting to emergency departments (ED) with suspected infection pose a substantive healthcare burden. We compared outcomes among elderly patients admitted to a community hospital undergoing direct ICU admission for sepsis or indirect ICU admission within 48 hours. METHODS: After IRB approval, an analytic cohort was built with patients at least 65 years old admitted between July 2012 and June 2018 to the ICU with acute physiology and chronic health evaluation (APACHE) IV diagnosis of sepsis, severe sepsis, or septic shock and attendant nidus. Data collection from our institutional APACHE IV Outcomes Database and electronic medical records included demographics, daily APACHE IV score and Acute Physiologic Scale (APS), time to ICU admission, length of stay, ICU and hospital mortality, and readmission within 30 and 90 days of hospital discharge. Continuous and discrete measures underwent apropos parametric and non-parametric univariate and multivariate statistical analyses and when indicated were controlled for age, sex and race to compare measures associated with direct versus indirect ICU admission. Continuous measures are summarized as mean±SD and discrete measures are summarized as proportions. RESULTS: Of 1,321 patients, 970 and 351 respectively underwent direct or indirect admission to ICU with an equivalent length of ICU stay (LOS) (3.3±3.9 vs. 3.3±3.7 days). Age, sex, race, APACHE IV score, APS, and level of sepsis, severe sepsis and septic shock distributions were statistically equivalent (p>.05) across admissions. Intergroup prevalence of sepsis nidus was isomorphic save pulmonary (Direct, 28% vs. 15%, p<.001) and GI (Indirect, 25% vs. 12%, p<.001). Direct (2.8±2.9 hours) versus indirect (17.5±12.8 hours) time to ICU admission was shorter (p<.001), thus hospital LOS for indirect admission (9.3±8.8 days vs. 7.9±6.7 days) was longer (p<.01). All 40 patients readmitted to ICU were from indirect admission trajectory. Direct versus indirect admission did not impact ICU (13.8% vs. 16.8%, p=.10) or hospital (18.6% vs. 20.8%, p=.20) mortality. Independent of hospital trajectory there was an equivalent 30d (7.9%) and 90d (13.8%) readmission rate. CONCLUSIONS: Elderly patients admitted to non-ICU units from the ED and subsequently transferred to the ICU within 48h experience greater hospital LOS versus direct admission to the ICU for treatment of sepsis. ICU and Hospital mortality, as well as 30d and 90d readmission rates, are not affected by the admission strategy. CLINICAL IMPLICATIONS: Appropriate admission triage and clinical unit placement may improve overall hospital LOS. Clinical pathways to identify high risk patients may improve clinical outcomes. DISCLOSURES: No relevant relationships by Kevin Dawkins, source=Web Response No relevant relationships by Mary Geary, source=Web Response No relevant relationships by Justin George, source=Web Response No relevant relationships by Karen Hamad, source=Web Response No relevant relationships by Belissa Ramos-Chaves, source=Web Response No relevant relationships by Yorlenis Rodriguez, source=Web Response No relevant relationships by Joseph Seaman, source=Web Response No relevant relationships by Robert Smith, source=Web Response No relevant relationships by Wilhelmine Wiese-Rometsch, source=Web Response