Clinical and pharmacological factors associated with mortality in patients with COVID-19 in a high complexity hospital in Manaus: A retrospective study

Abstract

COVID-19 is a contagious infection caused by the SARS-CoV-2 virus, responsible for more than 5 million deaths worldwide, and has been a significant challenge for healthcare systems worldwide. Characterized by multiple manifestations, the most common symptoms are fever, cough, anosmia, ageusia, and myalgia. However, several organs can be affected in more severe cases, causing encephalitis, myocarditis, respiratory distress, hypercoagulable state, pulmonary embolism, and stroke. Despite efforts to identify appropriate clinical protocols for its management, there are still no fully effective therapies to prevent patient death. The objective of this study was to describe the demographic, clinical, and pharmacotherapeutic management characteristics employed in patients hospitalized for diagnosis of COVID-19, in addition to identifying predictive factors for mortality. This is a single-center, retrospective cohort study carried out in a reference hospital belonging to the Brazilian public health system, in Manaus, from March 2020 to July 2021. Data were obtained from analyzing medical records, physical and electronic forms, medical prescriptions, and antimicrobial use authorization forms. During the study period, 530 patients were included, 51.70% male, with a mean age of 58.74 ± 15.91 years. The overall mortality rate was 23.58%. The variables age, number of comorbidities, admission to the ICU, length of stay, oxygen saturation, serum aspartate transaminase, and use of mechanical ventilation showed a positive correlation with the mortality rate. Regarding pharmacological management, 88.49% of patients used corticosteroids, 86.79% used antimicrobials, 94.15% used anticoagulant therapy, and 3.77% used immunotherapy. Interestingly, two specific classes of antibiotics showed a positive correlation with the mortality rate: penicillins and glycopeptides. After multivariate logistic regression analysis, age, number of comorbidities, need for mechanical ventilation, length of hospital stay, and penicillin or glycopeptide antibiotics use were associated with mortality (AUC = 0.958).