COMPARISON OF SIRS CRITERIA, QSOFA SCORE, AND A COMBINATION OF BOTH FOR THE DIAGNOSIS OF SEPSIS AND PREDICTION OF MORTALITY IN HOSPITALIZED ADULT PATIENTS WITH COVID-19

ABSTRACT

TOPIC Chest Infections TYPE Original Investigations PURPOSE: Sepsis is common in critically ill COVID-19 patients and is associated with high mortality. However, the diagnosis of sepsis and the prediction of mortality are challenging in COVID-19 patients, especially in the early stage of sepsis. Research about the performance of SIRS criteria and qSOFA score in patients with COVID sepsis is limited. Therefore, this study aimed to compare the performance of SIRS, qSOFA, and a combination of both in hospitalized COVID-19 adult patients regarding the diagnosis of sepsis and the prediction of in-hospital mortality. METHODS: Retrospective data analysis included 341 patients who were hospitalized with laboratory-confirmed COVID-19 infection during April 2020 and January 2021 (the peaks of the first and second waves in New York state) at a 242-bed university-affiliated community-based hospital. Data from the initial patient encounter at the emergency room was collected by reviewing the electronic health records. SIRS criteria, qSOFA score, and a combination of both (SIRS+qSOFA) were analyzed by the MATLAB software. SIRS+qSOFA includes 5 parameters tachypnea (respiratory rate ≥ 22), tachycardia (heart rate >90/min), fever or hypothermia (temperature >38°C (100.4°F) or < 36°C (96.8°F)), leukocytosis, leukopenia or bandemia (WBC > 12000/mm³, < 4000/mm³, or > 10% bands), altered mentation, and hypotension (systolic blood pressure ≤100 mmHg). The area under the receiver operating characteristic curves (AUC) was calculated to evaluate the performance of SIRS, qSOFA, and SIRS+qSOFA. RESULTS: In 341 COVID-19 patients, 128 patients were diagnosed with sepsis (37.5%), which included 77 expired patients (60.2%);213 patients were not diagnosed with sepsis (62.5%), which included 50 expired patients (23.5%). The AUCs for the diagnosis of sepsis with SIRS, qSOFA, and SIRS+qSOFA were 0.739, 0.731, and 0.793, respectively. The AUCs for the prediction of mortality with SIRS, qSOFA, and SIRS+qSOFA were 0.593, 0.665, and 0.646, respectively. For diagnosing sepsis, the sensitivity of SIRS (≥ 2), qSOFA (≥ 2), and SIRS+qSOFA (≥ 3) were 82.0%, 48.4%, and 68.8%, respectively;the specificity of SIRS (≥ 2), qSOFA (≥ 2), and SIRS+qSOFA (≥ 3) were 54.9%, 85.4%, and 76.5%, respectively. For predicting mortality, the sensitivity of SIRS (≥ 2), qSOFA (≥ 2), and SIRS+qSOFA (≥ 3) were 68.5%, 42.5%, and 55.9%, respectively;the specificity of SIRS (≥ 2), qSOFA (≥ 2), and SIRS+qSOFA (≥ 3) were 46.7%, 81.8%, and 68.7%, respectively. CONCLUSIONS: For diagnosing sepsis in COVID-19 patients, the performance of SIRS+qSOFA with 5 parameters (AUC=0.793) is better than either SIRS (AUC=0.739) or qSOFA (AUC=0.731). For predicting mortality, the performance of qSOFA (AUC=0.665) is superior to either SIRS (AUC=0.593) or SIRS+qSOFA (AUC=0.646). CLINICAL IMPLICATIONS This study will facilitate the diagnosis of sepsis and the prediction of mortality in the hospitalized COVID-19 adult patients. DISCLOSURES No relevant relationships by Shraddha Acharya, source=Web Response No relevant relationships by Inigo Atienza, source=Web Response No relevant relationships by Stephen Jesmajian, source=Web Response No relevant relationships by Jeffrey Lederman, source=Web Response No relevant relationships by Liyun Liu, source=Web Response No relevant relationships by Junfeng Xue, source=Web Response