Covichem: A biochemical severity risk score of COVID-19 upon hospital admission

Abstract

Clinical and laboratory predictors of COVID-19 severity are now well described and combined to propose mortality or severity scores. However, they all necessitate saturable equipment such as scanners, or procedures difficult to implement such as blood gas measures. To provide an easy and fast COVID-19 severity risk score upon hospital admission, and keeping in mind the above limits, we sought for a scoring system needing limited invasive data such as a simple blood test and co-morbidity assessment by anamnesis. A retrospective study of 303 patients (203 from Bordeaux University hospital and an external independent cohort of 100 patients from Paris Pitié-Salpêtrière hospital) collected clinical and biochemical parameters at admission. Using stepwise model selection by Akaike Information Criterion (AIC), we built the severity score Covichem. Among 26 tested variables, 7: obesity, cardiovascular conditions, plasma sodium, albumin, ferritin, LDH and CK were the independent predictors of severity used in Covichem (accuracy 0.87, AUROC 0.91). Accuracy was 0.92 in the external validation cohort (89% sensitivity and 95% specificity). Covichem score could be useful as a rapid, costless and easy to implement severity assessment tool during acute COVID-19 pandemic waves.

Risk of Death in Comorbidity Subgroups of Hospitalized COVID-19 Patients Inferred by Routine Laboratory Markers of Systemic Inflammation on Admission: A Retrospective Study

“…Many studies have reported early markers of COVID-19 severity that are capable of predicting clinical evolution towards severe complications, such as demographic predictors, laboratory parameters, chest radiographic abnormalities, and other clinical characteristics such as comorbidities or oxygen dependency [9]. Several studies have integrated clinical and paraclinical variables or signs and symptoms, among others, into a clinical prognostic score for the clinical management of COVID-19 patients, to better establish the prognosis of the disease [10,11]. The utility of multivariable machine learning predictive models has also been explored to stratify the patients at admission into predefined groups of disease severity [12][13][14].…”

Diagnostic utility of the Covichem score in predicting COVID-19 disease

“…It is calculated using a regression function that includes two comorbidities (obesity (defined as a BMI ≥30) and cardiovascular disease (CVD; coronary artery diseases such as angina and myocardial infarction, heart failure, cardiomyopathy, abnormal heart rhythms, valvular heart disease, aortic aneurysms, heart transplant, peripheral artery disease, thromboembolic disease, venous thrombosis and stroke)) and five laboratory test results (Na, albumin, ferritin, LDH, and CK; Supplement 1). It was derived and externally validated by Bats et al, who used the 0.5 (50%) value of this estimate as a threshold to assign patients to positive and negative Covichem groups [ 4 ]. We collected all data and then calculated the estimate and Covichem groups using the regression function described in Bats et al…”

“…Bats et al developed a score, called the Covichem score, for predicting a patient's risk of severe illness with COVID-19. This score can be calculated at hospital admission [ 4 ]. It uses clinical parameters and commonly available laboratory results and does not require imaging results or advanced testing.…”

CT-based severity assessment for COVID-19 using weakly supervised non-local CNN

“…Similarly, Bats et al. designed a framework of clinical characteristics consisting of 26 variables to learn a probability distribution for severe and risk-free patients [20] . Selection of the best predictors was made using the Akaike information criterion and the classifier was learnt as logistic regression.…”