External Validation of the Novel ABCD Scoring System for Prediction of In-Hospital Mortality among Severe and Critically Ill COVID-19 Patients: A Prospective Cohort Study

COVID-19 is a worldwide public health emergency. The availability of a brief and objective tool predicting the mortality among hospitalized COVID-19 patients might be helpful to direct limited medical resources to patients at higher risk of mortality. The aim of this work was to assess the novel ABCD scoring system as a tool for the prediction of in-hospital mortality among severe and critically ill COVID-19 patients admitted to the quarantine ICU. A prospective cohort study was conducted at Zagazig University Hospitals and comprised one hundred and seventy-nine COVID-19 patients admitted to the ICU. The ABCD score was calculated for all the patients. The primary outcome was the in-hospital mortality among the study participants.The median age of the participants was 63 (range: 18-95). About fifty-one percent (n= 92, 51.4%) were males. Of the total 179 patients, (n= 91, 50.8% survived, and n= 88, 49.2% died). A statistically significant association between the ABCD grading and the mortality outcome was detected (p-value = 0.029). By multiple logistic regression, the ABCD grading was a predictor of mortality in COVID-19 patients where the OR was 2.66 (95% CI: 1.284-5.517) (p-value = 0.008). The AUROC of the ABCD score was 0.608 (95% CI: 0.526-0.691), sensitivity was 61.4%, specificity was 50.5%, PPV was 54.5%, and NPP was 57.5%. To conclude: the ABCD grading system showed a discriminative ability to predict death in critically ill COVID-19 patients. The ABCD grading might be a quick approach for bedside assessment of severe COVID-19 patients, and it could also be used to rule out the probability of additional deterioration in patients in the non-critical zone.