Contrasting factors associated with COVID-19-related ICU and death outcomes: interpretable multivariable analyses of the UK CHESS dataset.

摘要: Abstract Background Identifying factors associated with severe COVID-19 is a priority to guide clinical care and resource use in this pandemic. Methods This cohort comprised 13954 in-patients confirmed COVID-19. Study outcomes were death intensive unit admission (ICUA). Multivariable logistic regression estimated odd ratios adjusted for 37 covariates (comorbidities, demographic, others). Gradient boosted decision tree (GBDT) classification generated Shapley values evaluating the impact of each patient. Findings Deaths due immunosuppression disease (Odds Ratio 1.39, 95%CI [1.10-1.76]), type-2 diabetes (1.31, [1.17-1.46]), chronic respiratory (1.19, [1.05-1.35]), obesity (1.16, [1.01-1.33], age (1.56/10-year increment, [1.52-1.61]), male sex (1.54, [1.42-1.68]). Associations ICUA differed direction (e.g., age, disease) scale, e.g., (3.37, [2.90-3.92]) some factors. Ethnicity was strongly but variably both outcomes, example Irish ethnicity negatively not ICUA. GBDTs had similar performance (ROC-AUC, 0.83, 0.68 GBDT; 0.80 regression). explanations overall consistent odds ratios. Chronic heart disease, hypertension, other comorbidities, ethnicities impacts on ranging from positive negative among different patients, although consistently all. Immunosuppressive diabetes, liver diseases either or Interpretation Very association factors, obesity, may review practice. explanation identified varying effects patients emphasising importance individual patient assessment.