This is the new Cancer-Associated Venous Thromboembolism (VTE) Decision Aid website. Development of this Next.js version was started in September 2024 by Omid Jafari.

About this work

This project is the result of an ongoing collaborative effort to conduct patient-centered clinical epidemiology and informatics research using multi-institutional cancer registry–linked electronic health record (EHR) databases. It is co-led by Dr. Ang Li (Baylor College of Medicine) and Dr. Nathanael Fillmore (VA Boston Healthcare System and Harvard Medical School).

We have performed the initial derivation and external validation of a new risk assessment model (RAM) called EHR-CAT for venous thromboembolism (VTE, or blood clots) using retrospective databases of nearly 90,000 patients over 10 years. The model has been tested in highly heterogeneous populations in the U.S., covering nearly all cancer histologies and modern treatment approaches, as well as diverse groups defined by age, sex, race, and ethnicity.

EHR-CAT builds upon the existing Khorana score with additional common data that are readily extractable from most cancer center EHR systems. This added granularity provides a more discriminative risk prediction (c-statistic improved from 0.65 → 0.71) and identifies 20–30% more patients with VTE in the high-risk subgroup.

The VTE outcome was carefully selected and adjudicated via a combination of ICD-9/10 codes, anticoagulant medications, and radiology reports (natural language processing) to achieve high precision and recall.