Real-world data (RWD) in healthcare comprise data routinely collected in electronic health records, registries, administrative claims, and more. RWD offer tremendous research opportunities due to their size, population coverage, and resource efficiency. However, since RWD are not collected as part of a rigorously designed epidemiologic study, research with RWD is challenging owing to its vulnerability to multiple sources of bias. In this talk, Dr. Shardell will focus on two case studies involving RWD collected from Medicare fee-for-service beneficiaries who sustained a hip fracture. In the first study, Dr. Shardell will describe a strategy to overcome informative observation time bias when estimating associations of patient-level characteristics with physical recovery using data from the Medicare Minimum Dataset, a federally mandated standardized clinical assessment tool administered by U.S. Center for Medicare and Medicaid Services to inform care management for residents in Medicare and Medicaid certified nursing facilities. In the second study, Dr. Shardell will describe multiple strategies across the artificial intelligence pipeline to enhance algorithmic fairness when predicting number of days at home, a patient-centered outcome operationalized using Medicare claims data, during the first six months after discharge from hospitalization due to hip fracture. Dr. Shardell will conclude with a discussion on RWD resources and how the UMB ICTR Biostatistics Core can support your research with RWD.