Abstract
Osteoporotic fractures remain common despite available therapies, highlighting the need for new, safe, and affordable treatment options. Advances in real-world data and causal inference have enabled target trial emulation, with prior frameworks demonstrating drug repurposing using observational data. However, no large-scale trial emulation framework has been applied to systematic drug repurposing for osteoporotic fracture prevention. To address this gap, we developed REACT (Real-world Emulated Clinical Trials), a scalable framework that operationalizes established target trial principles into an automated, comparator-robust pipeline for high-throughput drug screening. Applied to fracture prevention in older adults with osteoporosis, REACT screened 1,222 drugs and identified 18 with statistically significant associations, including 11 with protective effects and 7 with risk-increasing effects. The convergence of these findings with biological and prior observational evidence underscores REACT's ability to surface clinically meaningful patterns and its potential as a generalizable platform for accelerating real-world drug discovery and safety surveillance.