Understanding the profile of errors that cause duplicate entries in a patient registry

Duplicate records are detrimental to the cost-effective and efficient delivery of health care. Manually identifying and resolving duplicates can cost $60 per case. Patterns have been found in the types of errors that occur in patient registries, suggesting that undetected duplicate records may be similar to those already identified. At the University of Utah, records from all community clinics are merged with hospital records in the Enterprise Data Warehouse (EDW). The Pedigree and Population Resource group at Huntsman Cancer Institute links demographic records from the EDW to the Utah Population Database (UPDB). In last year's linkage, 76,922 duplicate records were identified. The purpose of this study was to compare the differences between clinic and hospital records in the EDW with existing literature.