Predicting mortality of individual patients with COVID-19: a multicentre Dutch cohort

Abstract

ObjectiveDevelop and validate models that predict mortality of patients diagnosed with COVID-19 admitted to the hospital.DesignRetrospective cohort study.SettingA multicentre cohort across 10 Dutch hospitals including patients from 27 February to 8 June 2020.ParticipantsSARS-CoV-2 positive patients (age ≥18) admitted to the hospital.Main outcome measures21-day all-cause mortality evaluated by the area under the receiver operator curve (AUC), sensitivity, specificity, positive predictive value and negative predictive value. The predictive value of age was explored by comparison with age-based rules used in practice and by excluding age from the analysis.Results2273 patients were included, of whom 516 had died or discharged to palliative care within 21 days after admission. Five feature sets, including premorbid, clinical presentation and laboratory and radiology values, were derived from 80 features. Additionally, an Analysis of Variance (ANOVA)-based data-driven feature selection selected the 10 features with the highest F values: age, number of home medications, urea nitrogen, lactate dehydrogenase, albumin, oxygen saturation (%), oxygen saturation is measured on room air, oxygen saturation is measured on oxygen therapy, blood gas pH and history of chronic cardiac disease. A linear logistic regression and non-linear tree-based gradient boosting algorithm fitted the data with an AUC of 0.81 (95% CI 0.77 to 0.85) and 0.82 (0.79 to 0.85), respectively, using the 10 selected features. Both models outperformed age-based decision rules used in practice (AUC of 0.69, 0.65 to 0.74 for age >70). Furthermore, performance remained stable when excluding age as predictor (AUC of 0.78, 0.75 to 0.81).ConclusionBoth models showed good performance and had better test characteristics than age-based decision rules, using 10 admission features readily available in Dutch hospitals. The models hold promise to aid decision-making during a hospital bed shortage.

Age is the main determinant of COVID-19 related in-hospital mortality with minimal impact of pre-existing comorbidities, a retrospective cohort study

“…Over the last months, studies revealed multiple risk factors independently associated with COVID-19 related (in-hospital) mortality, including higher age, pre-existing co-morbidities, and male sex, which is in line with current findings [ 9 , 10 , 22 – 25 ]. A recent study of 2273 COVID-19 hospitalized Dutch patients showed that a mortality prediction model using ten clinical features including age, number of home medications, admission blood values urea nitrogen/LDH/albumin, oxygen saturation, blood gas pH and history of chronic cardiac disease, improved discrimination over age-based decision rules only [ 26 ]. Nonetheless, no data regarding to which extent age and cardiovascular comorbidities contributed to the mortality risks were provided.…”

Mortality prediction of COVID-19 patients using soft voting classifier

“…With data obtained from five hospitals in New York City for 4098 Covid-19 patients, Extreme Gradient Boosting (XGBoost) was utilized to predict in-hospital mortality at time intervals of 3,5,7, and 10 days from admission ( Vaid et al, 2020 ). On a dataset containing 2273 patients from the Netherlands, linear logistic regression and non-linear gradient boosting were used to outperform other age-based models at the time ( Ottenhoff et al, 2021 ). A combination of longitudinal chest X-Rays and clinical data of 654 patients was utilized.…”

“…In this study, it was also found that those with age > 80 and 70 < age < 80 were at more risk as compared to others. The patients are more prone to mortality after 21 days if they are already on medication for some other disease ( Ottenhoff et al, 2021 ). In studies ( An et al, 2020 ) and ( Ottenhoff et al, 2021 ), it can be observed that feature selection is necessary and it can reflect the real-life variables that are important before and during the treatment of the patient.…”

“…The patients are more prone to mortality after 21 days if they are already on medication for some other disease ( Ottenhoff et al, 2021 ). In studies ( An et al, 2020 ) and ( Ottenhoff et al, 2021 ), it can be observed that feature selection is necessary and it can reflect the real-life variables that are important before and during the treatment of the patient. In the studies, age, medication, and data known for other diseases increased the AUC.…”

Mortality and readmission rates among hospitalized COVID-19 patients with varying stages of chronic kidney disease: a multicenter retrospective cohort

“…We performed a multicenter cohort study among COVID-19 patients included in the Dutch COVIDPredict cohort 6 – 8 . The COVIDPredict is a consortium of eleven hospitals in the Netherlands that aim to understand better and predict which COVID-19 patients should receive which treatments and which type of care.…”