Cost of Illness in Patients With COVID-19 Admitted in Three Brazilian Public Hospitals.

OBJECTIVES: The severity and transmissibility of COVID-19 justifies the need to identify the factors associated with its cost of illness (CoI). This study aimed to identify CoI, cost predictors, and cost drivers in the management of patients with COVID-19 from hospital and Brazil's Public Health System (SUS) perspectives. METHODS: This is a multicenter study that evaluated the CoI in patients diagnosed of COVID-19 who reached hospital discharge or died before being discharged between March and September 2020. Sociodemographic, clinical, and hospitalization data were collected to characterize and identify predictors of costs per patients and cost drivers per admission. RESULTS: A total of 1084 patients were included in the study. For hospital perspective, being overweight or obese, being between 65 and 74 years old, or being male showed an increased cost of 58.4%, 42.9%, and 42.5%, respectively. From SUS perspective, the same predictors of cost per patient increase were identified. The median cost per admission was estimated at US$359.78 and US$1385.80 for the SUS and hospital perspectives, respectively. In addition, patients who stayed between 1 and 4 days in the intensive care unit (ICU) had 60.9% higher costs than non-ICU patients; these costs significantly increased with the length of stay (LoS). The main cost driver was the ICU-LoS and COVID-19 ICU daily for hospital and SUS perspectives, respectively. CONCLUSIONS: The predictors of increased cost per patient at admission identified were overweight or obesity, advanced age, and male sex, and the main cost driver identified was the ICU-LoS. Time-driven activity-based costing studies, considering outpatient, inpatient, and long COVID-19, are needed to optimize our understanding about cost of COVID-19.

Diagnosis and prediction of COVID-19 severity: can biochemical tests and machine learning be used as prognostic indicators?

OBJECTIVE: This study aimed to implement and evaluate machine learning based-models to predict COVID-19' diagnosis and disease severity. METHODS: COVID-19 test samples (positive or negative results) from patients who attended a single hospital were evaluated. Patients diagnosed with COVID-19 were categorised according to the severity of the disease. Data were submitted to exploratory analysis (principal component analysis, PCA) to detect outlier samples, recognise patterns, and identify important variables. Based on patients' laboratory tests results, machine learning models were implemented to predict disease positivity and severity. Artificial neural networks (ANN), decision trees (DT), partial least squares discriminant analysis (PLS-DA), and K nearest neighbour algorithm (KNN) models were used. The four models were validated based on the accuracy (area under the ROC curve). RESULTS: The first subset of data had 5,643 patient samples (5,086 negatives and 557 positives for COVID-19). The second subset included 557 COVID-19 positive patients. The ANN, DT, PLS-DA, and KNN models allowed the classification of negative and positive samples with >84% accuracy. It was also possible to classify patients with severe and non-severe disease with an accuracy >86%. The following were associated with the prediction of COVID-19 diagnosis and severity: hyperferritinaemia, hypocalcaemia, pulmonary hypoxia, hypoxemia, metabolic and respiratory acidosis, low urinary pH, and high levels of lactate dehydrogenase. CONCLUSION: Our analysis shows that all the models could assist in the diagnosis and prediction of COVID-19 severity.