HealthPals collaborates with the FDA Reagan-Udall Foundation to accelerate the response to COVID-19

HealthPals joins forces with the Reagan-Udall Foundation for the FDA and the Friends of Cancer Research to contribute to the COVID-19 Evidence Accelerator. Read more about the collaboration here.

AUC-Stroke.png
Stroke prediction with AUC of 90%
AUC-MI.png
MI prediction with AUC of 89%

Leveraging the Veradigm linked EHR and Claims data covering 35 Million patients, HealthPals tracked the natural history of cardiovascular events in over 500,000 COVID-19 patients. Using several machine learning algorithms, we developed predictive models to identify patients who were at high risk of strokes and heart attacks within 90 days of their COVID-19 diagnosis.

 

Our models achieved promising results in predicting the occurrence of cardiovascular events with an AUC prediction of 90% for stroke and 89% for heart attacks in these COVID-19 patients, creating a huge potential to significantly and positively impact clinical outcomes in the COVID-19 population.

COVID-19 - Event-rate-curve.png
COVID-19 patients are at higher risk of cardiovascular events than influenza patients

HealthPals also explored the differences in cardiovascular events among patients who contracted COVID-19 versus those who contracted influenza. Using our CLINT platform, we performed a propensity score-weighted analysis to ensure a fair comparison between nearly one million COVID-19 and influenza patients. We found that those with COVID-19 were significantly more likely to have strokes, heart attacks, deep vein thrombosis, and pulmonary embolisms after their diagnoses than those with influenza, demonstrating a novel comparison between these two diseases at an unprecedented scale.