key: cord-0887235-dy02qubd
authors: Shade, Julie K.; Doshi, Ashish N.; Sung, Eric; Popescu, Dan M.; Minhas, Anum S.; Gilotra, Nisha A.; Aronis, Konstantinos N.; Hays, Allison G.; Trayanova, Natalia A.
title: Real-Time Prediction of Mortality, Cardiac Arrest and Thromboembolic Complications in Hospitalized Patients with COVID-19
date: 2022-05-08
journal: JACC: Advances
DOI: 10.1016/j.jacadv.2022.100043
sha: b4c723e11c08b72739484d5add7363bafff5d896
doc_id: 887235
cord_uid: dy02qubd

Background COVID-19 infection carry significant morbidity and mortality. Current risk prediction for complications in COVID-19 is limited and existing approaches fail to account for the dynamic course of the disease. Objectives To develop and validate the COVID-HEART predictor, a novel continuously-updating risk prediction technology to forecast adverse events in hospitalized patients with COVID-19. Methods Retrospective registry data from patients with SARS-CoV-2 infection admitted to 5 hospitals were used to train COVID-HEART to predict all-cause mortality/cardiac arrest (AM/CA) and imaging-confirmed thromboembolic events (TE) (2550 and 1854 patients, respectively). To assess COVID-HEART’s performance in the face of rapidly changing clinical treatment guidelines, an additional 1100 and 796 patients, admitted after the completion of development data collection, were used for testing. Leave-hospital-out validation was performed. Results Over 20 iterations of temporally-divided testing, the mean AUROCs were 0.917 (95% CI: 0.916-0.919) and 0.757 (95% CI: 0.751-0.763) for prediction of AM/CA and TE, respectively. The interquartile ranges of median early warning times were 14-21 hours for AM/CA and 12-60 hours for TE. The mean AUROCs for the left-out hospitals were 0.956 (95% CI: 0.936-0.976) and 0.781 (95% CI: 0.642-0.919) for prediction of AM/CA and TE, respectively. Conclusions The continuously updating, fully interpretable COVID-HEART predictor accurately predicts AM/CA and TE within multiple time windows in hospitalized COVID-19 patients. In its current implementation the predictor can facilitate practical, meaningful change in patient triage and resource allocation by providing real-time risk scores for these outcomes. The potential utility of the predictor extends to COVID-19 patients post-hospitalization and beyond COVID-19.

Patients with COVID-19, the disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), often present with cardiovascular (CV) manifestations such as myocardial infarction, thromboembolism, and heart failure. 1 Clinically overt cardiac injury or cardiomyopathy is reported in 8 to 33% of hospitalized patients 2, 3 and is associated with up to 50% mortality, 4 but imaging studies suggest the true incidence of cardiac involvement in all persons infected with SARS-CoV-2 could be as high as 60%. 5 Thromboembolic events are also frequently reported in severe COVID-19 and are associated with mortality; one study found that 70.1% of non-survivors and 0.6% of survivors met criteria for disseminated intravenous coagulation. 6 Furthermore, thromboembolic complications are more pronounced in acute COVID-19 infection than in other viral illnesses, and include pulmonary embolus and ischemic stroke, which can be fatal and are a significant cause of morbidity even as the infection resolves. 7 Despite the prevalence of thromboembolism and cardiac injury and their associations with poor outcomes, 2, 6, 8 no approach currently exists to forecast adverse these types of events in COVID-19 patients in real time.

Machine learning (ML) techniques are ideal for discovering patterns in high-dimensional biomedical data, especially when little is known about the underlying biophysical processes. ML is thus well-positioned for applications in COVID- 19 and indeed has been employed in screening, contract tracing, drug development, and outbreak forecasting. 9,10 ML approaches have been developed for prognostic assessment of hospitalized patients with COVID-19, including models which predict in-hospital mortality, [11] [12] [13] [14] [15] [16] progression to severe disease, 13, [17] [18] [19] [20] and outcomes related to respiratory function. 9, 14, 21 A continuous remote monitoring system has been developed and validated, 22 but it is designed for outpatient use and does not include laboratory test results. An ML model was also proposed for prediction of thromboembolic events but it required that all J o u r n a l P r e -p r o o f variables be present for all patients; did not provide dynamic risk updates, and was trained with data from only 76 patients. 23 In this study, we develop and validate a prognostic ML model to forecast the real-time risk of all-cause mortality/cardiac arrest (AM/CA) and thromboembolic events (TE) in hospitalized patients with COVID- 19 . We term the model the COVID-HEART predictor. We focus on predicting two clinically important outcomes in COVID-19: in-hospital all-cause mortality/cardiac arrest (AM/CA) and thromboembolic events (TE). In-hospital AM/CA is a clearly identifiable outcome and is often CV-related, thus it was selected for proof-of-concept to demonstrate the potential utility of COVID-HEART. Thromboembolic events are more difficult to identify and require imaging confirmation, thus, this outcome was selected to demonstrate the versatility of COVID-HEART in analyzing real-world clinical data and handling disease-specific outcomes.

Finally, the predictor is tested in two different ways. First, it is tested with data from patients hospitalized after the end of data collection for patients in the development set, to ascertain that COVID-HEART can accurately predict risk in real time for new patients in the face of rapidly changing clinical treatment guidelines. The predictor is next tested with leave-hospital-out nested cross-validation to assess its performance when training and testing are done with data from different populations.

The COVID-HEART predictor was developed and validated in a retrospective cohort study For data from an admission to be included in this study, patients must have had SARS-CoV-2 infection confirmed by polymerase chain reaction (PCR) within 14 days prior to the date of admission or during the admission. The minimum length of time from admission to discharge or death was 4 hours for AM/CA prediction and 72 hours for prediction of thromboembolic events, the difference being necessitated by the time granularity with which each outcome could be identified. Data were censored at the time of outcome or discharge.

Additional exclusion criteria were applied for prediction of each outcome separately.

Patients were excluded from thromboembolic event prediction if they experienced an imagingconfirmed thromboembolic event or were suspected of experiencing a thromboembolic event immediately prior to admission, which was diagnosed on admission or within 24 hours of admission. For prediction of AM/CA, patients were excluded if they experienced cardiac arrest with return of spontaneous circulation immediately prior to admission or if the arrest was precipitated by an event not related to disease severity. These exclusion criteria mean that the scope of the predictor is limited to new, in-hospital events.

For prediction of both outcomes, patients were not excluded based on treatments received, disease severity, need for intensive care, missing clinical variables, or any other reason not listed.

Although excluding patients for these reasons may have improved the ML models' performance, this would have resulted in a "clean" cohort not representative of real clinical data, making the risk predictor less useful in a real-world clinical setting. Outcome definition is discussed in The TRIPOD guidelines for development, validation, and presentation of a multivariable prediction model 24 were followed here (Supplementary Table 1 ). The model uses a selection of features extracted from up to 127 different clinical data inputs (shown schematically in Figure 1A and presented in detail in Supplementary Table 2), some of which are associated with CV complications in COVID-19 and in other severe respiratory illnesses, and others that are associated with general physiologic function. To avoid bias, variables that were directly impacted by a physician's assessment of the patient's condition, such as the fraction of inspired oxygen set on a mechanical ventilator, are excluded. Definition of these predictors, how they were measured, and algorithmic pre-processing steps undertaken prior to dynamic feature extraction are provided in Supplementary Methods. No manual adjustments were made to raw clinical data, and algorithmic pre-processing steps were designed to apply the minimum necessary "corrections" to the raw clinical data inputs to ensure our development and validation data sets were realistic and that our model could be applied in a real-world clinical setting. Pre-processing steps included removal of features which were missing for >60% of time windows, mean-value imputation for missing numerical features, and scaling all numerical features to zero mean and unit variance.

The COVID-HEART predictor was trained to estimate the probability that a patient will experience a particular event within a set number of hours (outcome window) after any point during the patient's hospitalization. It used static variables (demographics and comorbidities) and Figure 1B . Each time-point was assigned a binary outcome label indicating whether the patient experienced the outcome of interest in an "outcome window" following the time-point. Figure 1C schematically shows an array of processed data for a patient who experienced an adverse CV event. The outcome window for prediction of thromboembolic events was 24 hours as this was the minimum interval in which outcomes could be identified. Following training and cross-validation of each classifier configuration for prediction of each outcome with the development set, the optimal classifier configuration was trained on the full development set and used to predict the time-series risk of each event for each patient in the respective temporally divided test set. A binary prediction was also made at each time point using the optimal threshold determined by the development data during training. Model performance was assessed by the following metrics: accuracy, balanced accuracy, sensitivity, specificity, and AUROC. As a secondary analysis, the number of time windows predicted positive for patients who eventually experienced events and for patients who did not were compared. Additional analyses to investigate the effects of missing features and the frequency of new clinical data measurements on testing performance were also performed.

Testing was repeated to obtain a 95% confidence interval for each testing performance metric using the final optimized model from each of the 20 iterations of cross-validation. To maintain the temporal nature of the development-test split, we selected an end cutoff date for the test set such that the development and test sets contained 70% and 30% of patients in the reduced data set, respectively. The earliest train-test cutoff date was June 25, 2020; we did not move the train-test cutoff beyond this date to ensure there was enough data to train the predictor. Since there were few events for each outcome, repeating the train-test split in this way provided an accurate estimate of the models' cross-validation performance and performance on a temporally separate test set. All test patient example predictions and data describing the characteristics of the development and testing sets were generated using the model trained with the full development and testing sets (March 1, 2020 to January 8, 2021).

Finally, to assess the predictor's performance when trained and tested with data from J o u r n a l P r e -p r o o f patients from different populations, we performed leave-hospital-out validation. This is justified by the fact that each of the five hospitals in the study has different characteristics and serves a different patient population (Supplementary Table 3 ). Leave-hospital-out validation was performed by removing all patients admitted to one of the five hospitals in the study, repeating the model training and optimization process using data from patients admitted to the remaining four hospitals, and testing the optimized model with data from patients admitted to the left-out hospital.

If a patient was transferred between hospitals or had multiple admissions to different hospitals, their admission to the left-out hospital was used in testing and the rest of their data were removed from the training data set.

In total, 3650 patients met eligibility criteria for prediction of AM/CA; 1100 (30.1%) were assigned to the test set according to the date cutoff. In addition, 2650 patients met eligibility criteria for prediction of thromboembolic events; 796 (30.0%) were assigned to the test set. Figure 2 shows the flow of patients through the study. Supplementary Table 4 and Supplementary Table 5 For both outcomes, a larger number of time windows in the test set were predicted positive for patients that eventually experienced the outcome compared to those that did not: 38% vs. 10%

for AM/CA, 51% vs. 12% for thromboembolic events. The 95% confidence intervals for these measurements over 20 iterations of temporally divided testing were 36%-41% vs. 9%-11% for AM/CA and 68%-82% vs. 15%-20% for thromboembolic events. This suggests that the ML model is sensitive in identifying warning signs of an impending adverse event earlier than the prespecified outcome window (Supplementary Figure 4) . The interquartile ranges for the median early warning times over 20 iterations of temporally-divided testing were 14-21 hours for AM/CA and 12-60 hours for thromboembolic events, although the classifier was trained to predict outcomes within 2 hours for AM/CA and 24 hours for thromboembolic events. This could represent a clinically useful "early warning" system.

As it is essential for clinical decision-making to identify the features that most contribute to the predicted risk score for a particular outcome, the COVID-HEART predictor was designed to be fully interpretable. Supplementary Table 6 lists up to 20 features with the largest coefficients in the optimal classifier for each of the two outcomes. The final COVID-HEART predictor includes 61 features for prediction of AM/CA, many of which are routinely and continuously acquired vital signs and basic metabolic tests. COVID-HEART includes 9 features for prediction of thromboembolic events. These features are extracted from 39 and 5 unique clinical data inputs, for the two models, respectively. Note that features were normalized prior to classifier training, and that models are not simple logistic regressions, thus interpretation of the coefficients is not 

In this study, we developed and validated the COVID-HEART predictor, a real-time model that can forecast multiple adverse events in hospitalized patients with COVID-19. The COVID-HEART predictor is robust to missing data and can be updated each time new data becomes available, representing a continuously evolving warning system for an impending event. It can also predict the likelihood of an adverse event within multiple timeframes (e.g. 2 hours, 8 hours, 24 hours). Although predictions were made at the same time steps for patients in the test set for consistency with the development set, it is possible to apply the model at any arbitrary time during a patient's hospitalization. We envision that in practice, it could provide the physician with an updated risk score each time any new clinical data input becomes available or only after passing a certain "high risk" threshold, to reduce healthcare provider "alert fatigue". The COVID-HEART predictor is thus anticipated to be of great clinical use in triaging patients and optimizing resource utilization by identifying at-risk patients in real time. Finally, COVID-HEART identifies dynamic predictive features that have not previously been investigated for prediction of these outcomes in patients with COVID-19; these may suggest avenues for future research and personalized targets for clinical intervention.

The COVID-HEART risk prediction approach provides transparency and clinical intepretability, including the ability to determine which features are dominant contributors to a patient's risk level at a particular time, which may suggest potential patient-specific targets for clinical intervention. Prediction models for CV adverse events in patients with COVID-19 have been limited by lack of sufficient data, impractical requirements for use (e.g. that all data be available for all patients or that measurements are taken at the same time relative to time of admission), and overly restrictive inclusion/exclusion criteria that result in idealistic training and testing cohorts not representative of real patient data. 23, 27 Our model is designed to handle realworld data, which may include noise, missing variables, and data collected at different points in a patient's hospitalization. The validation and test results indicate strong generalizability despite statistically significant differences between the temporally-divided development and test sets, and between hospitals in the health system. Finally, the inclusion of multiple time-duration features gives the model the "memory" advantages of a long short-term memory neural network without compromising interpretability or becoming a "black box". It is trained in a manner that achieves high sensitivity and specificity despite severe class imbalance. To our knowledge, these techniques have not previously been combined in real-time predictors for CV events.

Models for risk prediction in hospitalized patients have typically focused on predicting mortality risk or length of stay for patients in the ICU. Traditional models incorporate variables thought to indicate physiologic instability or end-organ injury (e.g. respiratory rate, serum bilirubin level, serum creatinine, etc.). [27] [28] [29] While these models generally have good discriminative power 30 , they fail to provide specific, actionable information and simply notify healthcare teams that particular patients are at increased mortality risk at some point in their ICU stay. In most cases, predictive scores are calculated based on the most extreme variable values during the initial 24 hours of the ICU admission, with repeat calculations every 24-72 hours.

Newer models have higher predictive performance compared to traditional models, they are trained to predict the incidence of a particular outcome (e.g. bleeding, renal failure, mortality, etc.) at an indefinite future time. They are not designed to predict the time periods during which patients are at highest risk. Further, in term of ML for risk prediction in COVID-19, prior studies have focused largely on initial diagnosis, mortality, or severity of illness, but none have specifically focused on cardiovascular events, including in-hospital AM/CA and thromboembolic events, both clinically important complications with implication for cardiac treatment and monitoring. Moreover, to our knowledge, our model is the first to utilize continuous time series physiologic data as well as laboratory and electrocardiographic data to provide a continuouslyupdating risk score for an outcome within a particular future time window (e.g. risk of thromboembolic event in the next 24 hours). By providing a risk score for a specific outcome window, our model provides timely, actionable information, allowing the healthcare team to allocate resources and initiate therapies when they are most needed.

With respect to thromboembolic events, we found that 40 out of 41 events occurred in patients already ordered for high-intensity VTE prophylaxis, suggesting an even more aggressive anticoagulant regimen may be needed for those patients identified by the model. Additionally, VTE prophylaxis is one of the treatments most frequently omitted by nursing staff or declined by

patients. An analysis of VTE events at our institution over a 72-day period during the Spring 2020 J o u r n a l P r e -p r o o f COVID-19 wave demonstrated that 4 out of 11 SARS-CoV-2 positive patients who experienced VTE events had at least one missed dose of VTE prophylaxis. 31 While care providers should ideally strive for 100% compliance with VTE prophylaxis in all eligible patients, the identification of patients at high risk for thromboembolic events may help target these interventions to the patients most in need.

With respect to interventions to address impending AM/CA, we found in our detailed chart review that a number of AM/CA events were not unprovoked but were a consequence of a precipitating event that altered the patients hemodynamics, such as intubation, patient positioning (e.g. supine to prone), or hemodialysis. Therefore, in addition to predicting unprovoked arrest (in approximately half of the cases), our model predicted an unstable physiologic state that resulted in arrest due to otherwise well-tolerated hemodynamic perturbations. Identification of patients as high risk for mortality would aid clinicians by imploring them to defer any treatments that may provoke an arrest until the patient's physiology recovers. For those treatments that cannot be deferred, identification of high-risk patients would prompt the primary team to assemble specialized staff and equipment, given the high risk of arrest (e.g. calling the anesthesia team for intubation in a high-risk patient, having adequate nursing staff for a possible resuscitation, etc.)

A major barrier to clinical adoption of prognostic machine learning models is the lack of appropriate validation on a representative test cohort. The temporally-divided test sets in this study demonstrated the performance of the predictor on a set of patients admitted after the end of data collection for patients in the development set. A prospective cohort would not be expected to have the same composition as the development set; indeed, there were several statistically significant differences in demographics, clinical characteristics, and prevalence of adverse CV events between the development and tests sets in this study. However, the strong test results show that the predictor J o u r n a l P r e -p r o o f is robust to changes in clinical treatment guidelines and evolving demographics. We hypothesize that it maintains its accuracy because it considers data which describe the patient's physiologic state, not variables that are directly influenced by physician input such as ventilator settings or medication use. Further, the predictor maintained strong performance in leave-hospital-out validation, which demonstrated its robustness when trained and tested with data from patients from different populations.

A limitation in this study is the requirement for imaging confirmation of thromboembolic events. All thromboembolic event diagnoses were adjudicated by a clinician to ensure they were clinically relevant. If the radiologist made an incorrect diagnosis and the adjudicating clinician incorrectly agreed that the event was supported by clinical evidence, this would unfortunately constitute an error in our data set. Similarly, it is likely that patients in the study experienced thromboembolic events that were either the precipitating cause of death or that were not identified on imaging and were therefore not counted as events. There were only 35 patients in the development set with imaging-confirmed thromboembolic events and these outcomes could only be identified per-day, not at the exact time they occurred, as with AM/CA. As a result, only a few features could be selected; it is possible that a larger feature set would lead to more accurate prediction of the patients' risk of thromboembolic events since more details of the patients' clinical states could be considered. In addition, complete details about the primary causes of death were not known for all patients and therefore it was not possible to distinguish if all-cause mortality was secondary to cardiovascular, respiratory, or other causes.

Additional limitations stem from the use of the JH-CROWN registry. 32 These include the potential for measurement error, inaccurate patient-reported history (e.g. smoking), and missing data. Another potential limitation is confounding by indication, which means that treatments were selected based on clinical indication. While our model did not include treatments or other variables that were directly influenced by clinical indication, some variables in the model were likely indirectly influenced by clinical indication. For example, the pulse oxygen saturation may have been affected by changes in ventilator settings for patients who were receiving mechanical ventilation. There is also a subgroup of patients who had pre-existing DNR/DNI/comfort care orders. These patients would have received no interventions leading up to an adverse event, which means that the sequalae of physiologic changes for these patients may be different than for patients who received interventions prior to an adverse event. Finally, there is selection bias inherent to including only patients who sought care at a hospital; patients without insurance, undocumented patients, and patients with other barriers to seeking care may be less likely to be included.

In this study we demonstrated highly accurate prediction of AM/CA and TE in hospitalized COVID-19 patients using the continuously-updating COVID-HEART predictor. In its current implementation the predictor can facilitate practical, meaningful change in patient triage and the allocation of resources by providing real-time risk scores for complications occurring commonly in COVID-19 patients. The COVID-HEART can be re-trained to predict additional adverse CV events including myocardial infarction and arrhythmia. The potential utility of the predictor extends well beyond hospitalized COVID-19 patients, as COVID-HEART could be applied to the prediction of CV adverse events post-hospital discharge or in patients with chronic COVID syndrome ("Long COVID"). Additionally, the ML methodology utilized here could be expanded to use in other clinical scenarios that require screening or early detection, such as risk of hospital readmission, with the goal of improved clinical outcomes through early warnings and resultant

Competency in Practice-Based Learning and Improvement: The COVID-HEART predictor can identify patient at-risk all-cause mortality/cardiac arrest and thromboembolic events by (D) Continuously-updating risk score. The COVID-heart predictor provides a risk score (probability) for a given outcome in the K hours following a given time point. Shown is a sample risk score for a patient that experienced an event: green color indicates low risk score; yellow indicates a risk score within a pre-determined range of a threshold value, and the red indicates that the patient is at high risk for an event in the following K hours. Inclusion and exclusion criteria were applied separately for prediction of each outcome. The data were then temporally divided into development and test sets as shown. 

COVID-19 and the cardiovascular system

Association of Cardiac Injury With Mortality in Hospitalized Patients With COVID-19 in Wuhan, China

Environmental Contamination and Viral Shedding in MERS Patients during MERS-CoV Outbreak in South Korea

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Outcomes of Cardiovascular Magnetic Resonance Imaging in Patients Recently Recovered from Coronavirus Disease 2019 (COVID-19)

Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia

Thromboembolism risk of COVID-19 is high and associated with a higher risk of mortality: A systematic review and meta-analysis

Prognostic value of cardiovascular biomarkers in COVID-19: A review

Applications of machine learning and artificial intelligence for Covid-19 (SARS-CoV-2) pandemic: A review

Artificial intelligence and machine learning to fight covid-19

Clinical predictors of mortality due to COVID-19 based on an analysis of data of 150 patients from Wuhan, China

A machine learningbased model for survival prediction in patients with severe COVID-19 infection 2020

Machine Learning to Predict Mortality and Critical Events in COVID-19 Positive New York City Patients: A Cohort Study (Preprint)

Deep learning prediction of likelihood of ICU admission and mortality in COVID-19 patients using clinical variables

Machine Learning and Meta-Analysis Approach to Identify Patient Comorbidities and Symptoms that Increased Risk of Mortality in COVID-19

Common cardiovascular risk factors and in-hospital mortality in 3,894 patients with COVID-19: survival analysis and machine learning-based findings from the multicentre Italian CORIST Study

Unsupervised machine learning for the discovery of latent clusters in COVID-19 patients using electronic health records

Interpretable Machine Learning for COVID-19: An Empirical Study on Severity Prediction Task

A Fully Automatic Deep Learning System for COVID-19 Diagnostic and Prognostic Analysis

Development and Validation of the Quick COVID-19 Severity Index: A Prognostic Tool for Early Clinical Decompensation

Prediction models for diagnosis and prognosis of covid-19: Systematic review and critical appraisal

Wearable sensor derived decompensation index for continuous remote monitoring of COVID-19 diagnosed patients

Predicting thromboembolic complications in COVID-19 ICU patients using machine learning

Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): Explanation and elaboration

Thromboembolic events and Covid-19

COVID-19-associated hyperviscosity : a link between inflammation and thrombophilia ? Caution against COVID-19 treatment

Usefulness of machine learning in COVID-19 for the detection and prognosis of cardiovascular complications

SAPS 3 -From evaluation of the patient to evaluation of the intensive care unit. Part 2: Development of a prognostic model for hospital mortality at ICU admission

Assessing contemporary intensive care unit outcome: An updated Mortality Probability Admission Model (MPM0-III)*

Caution when using prognostic models: A prospective comparison of 3 recent prognostic models

Nonadministration of pharmacologic venous thromboembolism prophylaxis is less common in hospitalized patients with COVID -19

JH-CROWN COVID-19 PMAP Registry

Predictably unequal: understanding and addressing concerns that algorithmic clinical prediction may increase health disparities

COVID-19 and Racial/Ethnic Disparities

Using Social Indicators to Describe Neighborhood-Level Disparities in Adolescent Health in Baltimore City Circa

Comorbidity Measures for Use with Administrative Data

Agency for Healthcare Research and Quality, Rockville M. HCUP Tools and Software. Healthcare Cost and Utilization Project (HCUP)

Lymphopenia in critically ill COVID-19 patients: A predictor factor of severity and mortality

The Science Underlying COVID-19: Implications for the Cardiovascular System

Hospitalized Patients with 2019 Novel Coronavirus-Infected Pneumonia in Wuhan, China

White blood cell count and mortality in patients with acute pulmonary embolism

Deep vein thrombosis and systemic inflammatory response: A pilot trial

tableone: An open source Python package for producing summary statistics for research papers

Abbreviations: white blood cell count (WBC), c-reactive protein (CRP), diastolic blood pressure (DBP), systolic blood pressure (SBP), immature granulocyte (IG) Central Illustration: We developed and validated an interpretable predictor of all-cause mortality/cardiac arrest and thromboembolic events in COVID-19 using retrospective registry data from over 3600 patients admitted to multiple hospitals