2321. FluMex: A New Clinical Severity Index in Mexican Hospitalized Patients with Influenza

BACKGROUND: Influenza virus infection is frequently characterized by a complex clinical behavior and outcomes can be fatal. There are many published scoring methods aimed for pulmonary infections and sepsis severity nevertheless they lack adequate sensitivity and specificity in patients with Influenza. METHODS: From 2013 to 2018, hospitalized patients from five hospitals from the Christus Muguerza health group from Monterrey, Mexico who had a positive rapid influenza-test and/or positive PCR for Influenza virus were enrolled. Risk factors for severity and mortality were evaluated calculating odds ratio with a binary logistic regression model and were adjusted for other factors. The new index was then compared with pneumonia severity scores by assessing area under the curve(AUC), sensitivity and specificity. RESULTS: We analyzed data from 125 patients hospitalized with confirmed Influenza infection. Less than 1% had received the corresponding seasonal influenza vaccine. Type 2 diabetes (T2D) and hypertension (HT) were the most prevalent comorbidities. Odds ratios were significant for age > 65 years, body mass index (BMI) > 30, T2D, HT, pulsoximetry < 90%, respiratory rate > 22 per minute, altered mental status, blood urea nitrogen (BUN) > 19 mg/dL, elevated lactate dehydrogenase (LDH), and an abnormal chest X-ray. The FluMex score was applied to a control group of 125 admitted patients with confirmed Influenza infection. AUC was 0.63 (CI 95%, 0.52–0.74; P < 0.05) for severity and 0.90 (IC 95%, 0.83–0.97; P < 0.05) for mortality, showing better predictive performance than other pneumonia and sepsis scores such as CURB-65, PSI, CROMI, SIRS, SOFA, qSOFA and ILI (Table 1). CONCLUSION: The FluMex scoring system can be a useful tool for patients with suspected Influenza infection in predicting severity and mortality, helping to improve care and resource management. [Image: see text] DISCLOSURES: All authors: No reported disclosures.