FACTORS ASSOCIATED WITH IN-HOSPITAL DEATH AMONG PATIENTS WITH SEVERE ACUTE RESPIRATORY SYNDROME AND VIRAL CODETECTION IN BRAZIL: A RETROSPECTIVE COHORT STUDY WITH MULTIVARIABLE ANALYSIS, 2022–2023

Abstract

Objective: To identify demographic, clinical, healthcare, and virological factors associated with in-hospital death among patients with severe acute respiratory syndrome (SARS) and respiratory viral codetection in Brazil. Methods: We conducted a retrospective cohort study using 2022–2023 records from the Brazilian Influenza Epidemiological Surveillance Information System. Hospitalised SARS cases with two or more viruses detected by RT-PCR and known outcomes were included. Crude and adjusted odds ratios (ORs) were estimated using logistic regression, with assessment of discrimination, calibration, multicollinearity, and sensitivity to unmeasured confounding using E-values. Results: Among 12,581 hospitalisations, 84.1% occurred in children aged 0–4 years, and 466 patients died (case fatality: 3.7%). After adjustment, age (aOR per year: 1.045; 95%CI: 1.039–1.051), liver disease (aOR: 4.658; 95%CI: 1.868–11.614), immunosuppression (aOR: 2.181; 95%CI: 1.338–3.555), invasive ventilation (aOR: 10.264; 95%CI: 7.202–14.628), and detection of three or more viruses (aOR: 1.665; 95%CI: 1.147–2.418) were associated with higher odds of death. The model had an AUC of 0.901. Conclusion: Death was associated with age, clinical vulnerability, markers of acute severity, and greater codetection complexity. Findings may support risk stratification but do not establish causality and require consideration of selection, measurement, and residual-confounding biases.