Potential biomarkers for fatal outcome prognosis in a cohort of hospitalized COVID-19 patients with pre-existing comorbidities.

The difficulty in predicting fatal outcomes in patients with coronavirus disease 2019 (COVID-19) impacts the general morbidity and mortality due to severe acute respiratory syndrome-coronavirus 2 infection, as it wears out the hospital services that care for these patients. Unfortunately, in several of the candidates for prognostic biomarkers proposed, the predictive power is compromised when patients have pre-existing comorbidities. A cohort of 147 patients hospitalized for severe COVID-19 was included in a descriptive, observational, single-center, and prospective study. Patients were recruited during the first COVID-19 pandemic wave (April-November 2020). Data were collected from the clinical history whereas immunophenotyping by multiparameter flow cytometry analysis allowed us to assess the expression of surface markers on peripheral leucocyte. Patients were grouped according to the outcome in survivors or non-survivors. The prognostic value of leucocyte, cytokines or HLA-DR, CD39, and CD73 was calculated. Hypertension and chronic renal failure but not obesity and diabetes were conditions more frequent among the deceased patient group. Mixed hypercytokinemia, including inflammatory (IL-6) and anti-inflammatory (IL-10) cytokines, was more evident in deceased patients. In the deceased patient group, lymphopenia with a higher neutrophil-lymphocyte ratio (NLR) value was present. HLA-DR expression and the percentage of CD39+ cells were higher than non-COVID-19 patients but remained similar despite the outcome. Receiver operating characteristic analysis and cutoff value of NLR (69.6%, 9.4), percentage NLR (pNLR; 71.1%, 13.6), and IL-6 (79.7%, 135.2 pg/mL). The expression of HLA-DR, CD39, and CD73, as many serum cytokines (other than IL-6) and chemokines levels do not show prognostic potential, were compared to NLR and pNLR values.

Accurate diagnosis of sepsis using a neural network: Pilot study using routine clinical variables.

BACKGROUND AND OBJECTIVES: Sepsis is a severe infection that increases mortality risk and is one if the main causes of death in intensive care units. Accurate detection is key to successful interventions, but diagnosis of sepsis is complicated because the initial signs and symptoms are not specific. Biomarkers that have been proposed have low specificity and sensitivity, are expensive, and not available in every hospital. In this study, we propose the use of artificial intelligence in the form of a neural network to diagnose sepsis using only common laboratory tests and vital signs that are routine and widely available. METHODS: A retrospective, cross sectional cohort of 113 patients from an intensive care unit, each with 48 routinely evaluated vital signs and biochemical parameters was used to train, validate and test a neural network with 48 inputs, 10 neurons in a single hidden layer and one output. The sensitivity and specificity of the neural network as a point sampled diagnostic test was calculated. RESULTS: All but one case were correctly diagnosed by the neural network, with 91% sensitivity and 100% specificity in the validation data set, and 100% sensitivity and specificity in the test data set. CONCLUSIONS: The designed neural network system can identify patients with sepsis, with minimal resources using standard laboratory tests widely available in most health care facilities. This should reduce the burden on the medical staff of a difficult diagnosis and should improve outcomes for patients with sepsis.