Case Study: Human Data Science achieves COPD rehospitalization prediction and AFib detection with IQVIA AI

Data can be used to predict COPD rehospitalizations or diagnose AFib in individual patients

Human Data Science, a group driving personalized medicine, faced two clinical challenges: high short-term rehospitalization rates among COPD patients and underdiagnosed atrial fibrillation (AFib) that can lead to stroke. To address these issues they partnered with IQVIA, leveraging data-driven tools—notably an IQVIA-developed machine learning algorithm trained on patient-level electronic health records and registries—to support more precise, individualized care.

IQVIA implemented a machine learning model that used detailed EHR and registry data (consultations, diagnoses, prescriptions, labs, hospitalizations and socioeconomic factors) to predict short-term COPD rehospitalization risk based on drivers like prior exacerbations, outpatient visits and length of stay. Paired with AI-enabled ECG approaches for AFib detection, these analytics demonstrated the ability to predict rehospitalizations and identify previously unrecognized AFib at point-of-care, enabling targeted interventions that improve patient outcomes and help reduce readmissions and downstream healthcare costs.