Predicting COVID-19 mortality with electronic medical records

Abstract

This study aims to predict death after COVID-19 using only the past medical information routinely collected in electronic health records (EHRs) and to understand the differences in risk factors across age groups. Combining computational methods and clinical expertise, we curated clusters that represent 46 clinical conditions as potential risk factors for death after a COVID-19 infection. We trained age-stratified generalized linear models (GLMs) with component-wise gradient boosting to predict the probability of death based on what we know from the patients before they contracted the virus. Despite only relying on previously documented demographics and comorbidities, our models demonstrated similar performance to other prognostic models that require an assortment of symptoms, laboratory values, and images at the time of diagnosis or during the course of the illness. In general, we found age as the most important predictor of mortality in COVID-19 patients. A history of pneumonia, which is rarely asked in typical epidemiology studies, was one of the most important risk factors for predicting COVID-19 mortality. A history of diabetes with complications and cancer (breast and prostate) were notable risk factors for patients between the ages of 45 and 65 years. In patients aged 65-85 years, diseases that affect the pulmonary system, including interstitial lung disease, chronic obstructive pulmonary disease, lung cancer, and a smoking history, were important for predicting mortality. The ability to compute precise individual-level risk scores exclusively based on the EHR is crucial for effectively allocating and distributing resources, such as prioritizing vaccination among the general population.

Fig. 1. Comparing the classification performance between the boosting and standard GLM (binomial logistic) models.

a–d The ROC curves for the overall 45–65, 65–85, and 85+ models, respectively. e–h The area under the ROC curves for the overall, 45–65, 65–85, and 85+ models.

Fig. 2. Calibration curves of models.

a The calibration curve for the overal model. b The calibration curve for the 85+ model.

Fig. 3. Odds ratios for the covariates identified as predictors of mortality in COVID-19 patients.

Risk factor observation represents the number of model iterations that identified a covariate as a predictor of mortality in COVID-19 patients. The total number of model iterations is 10. Median and interquartile ranges (IQR) for odds ratios are available in Supplementary Table 2.

Fig. 4. Odds ratios for the covariates identified as predictors of mortality in COVID-19 patients by age groups.

a Odds ratios for the 45-65 model. b Odds ratios for the 65-85 model. c Odds ratios for the 85+ model. Risk factor observation represents the number of model iterations that identified a covariate as a predictor of mortality in COVID-19 patients. The total number of model iterations is 10.

Fig. 5. Temporal segmentation of clinical records and analysis pipeline.

The analysis pipeline utilizes longitudinal clinical data with a 14-day buffer before COVID-19 infection.