Direct versus indirect admission of COVID-19 patients to intensive care from emergency department

ABSTRACT

INTRODUCTION: Approximately 20% of COVID-19 patients require hospitalization including 5% needing intensive care. We compared outcomes among patients admitted to an academic community hospital undergoing ICU admission for COVID either directly from emergency department or from general medical units. METHODS: Data were collected after IRB exemption from APACHE IV database and electronic medical records including demographics;administrative data;baseline inflammatory, metabolic and coagulation biomarkers significantly associated with primary outcome;ICU day 1 APACHE IV score and Acute Physiology Score (APS);length of stay (LOS);ICU readmission within same hospital visit;and ICU and hospital mortality. Primary outcome was hospital mortality. Continuous data were summarized with mean [?SD] or median [IQR] respectively compared using t-test or Kruskal-Wallis Test. Discrete data were summarized as counts (proportions) compared with chi-square. Two-tailed p<.05 was significant. RESULTS: Of 244 COVID patients discharged between March 23 and July 28, 2020, respectively 26 and 40 underwent direct (2.6 [1.0-3.4] hrs), or indirect admission (49.8 [19.2 - 136.3] hrs) to ICU. Intergroup similar baseline data were pooled (63±2 years, 58% male, 61% White, 14% Black, 25% Other);temperature 99.0±1.2 °F;SpO2 95±3%;C-reactive protein 14.7±8.5 mg/dL;ferritin 1292±1741 ng/mL;lactate dehydrogenase 455±195 U/L;D-dimer 3.9±6.9 mg/L;prothrombin time 13.4±5.1 seconds;hemoglobin 13.0±2.3 gm/dL;APACHE IV score (66±29);APS (52±27);mechanical ventilation days 14.7±10.9 days;ICU LOS 11.6±10.7 days;and hospital LOS 20.9±14.6 days. There was no difference in number of comorbidities (3.6±1.6) or Elixhauser comorbidity score (9.9±9.8) across ICU admission pathways. Chronic condition prevalence included hypertension (56%), diabetes (28%), obesity (28%) and hypothyroidism (16%). COVID directed drug class treatment was similar. Six patients readmitted to ICU were indirect admissions. ICU (39.4%) and hospital (42.4%) mortality were independent of ICU trajectory. CONCLUSIONS: Outcomes were similar in patients admitted to the ICU directly versus those who decompensated and required transfer from medical floors. Given equivalent biomarker risk at presentation, this finding evinces care quality depth of our safety net.