External Validation

Fillmore has focused on the external validation of predictive models to ensure that machine learning and electronic health record tools maintain accuracy across diverse clinical environments. His work demonstrates that novel risk scores for cancer-associated thrombosis and cardiovascular disease often outperform traditional benchmarks when tested in independent cohorts, including safety-net hospitals and veteran healthcare systems. Through these studies, he has established the necessity of rigorous validation to confirm that algorithmic tools are scalable and reliable for personalized patient management.