Patient matching: cap overlays before chasing duplicate rates

Hospitals are pushed to cut duplicate patient records, and the fastest route is more aggressive auto-merging. This episode asks what that does to overlay risk. Using OneFlorida's 99.3% matching precision (7 false merges per 1,000), the BMJ Quality and Safety duplicate mortality study, an ECRI overlay case, the MATCH IT Act and TEFCA, the guest argues for capping overlays at one false merge per 1,000 automated merges, while the host pushes back on duplicate deaths. Listeners leave able to set an overlay ceiling, split matches into auto-merge, human review and link-only bands, set per-source thresholds as Intermountain did, sample past merges, and weigh AHIMA cost figures and equity risks for clinical AI. 0:00 Duplicate records vs overlays, BMJ mortality study 1:41 OneFlorida's 7 false merges per 1,000 3:55 MATCH IT Act and match rate standards 4:30 Capping overlays versus clearing duplicate backlogs 6:26 Overlay costs and linking instead of merging 8:02 Clinical AI and uneven matching errors