KEYWORDS:
Duplicate Records, Data Quality, Database Integration, Connect
BACKGROUND:
One challenging problem facing health databases is duplication of records. To address this problem, the New York City Department of Health and Mental Hygiene (DOHMH) is creating a centralized de-duplication service. The Master Client Index (MCI), which integrates data from the Citywide Immunization Registry (CIR), the Lead Quest Registry (LQ), and Vital birth records, will provide this service.
OBJECTIVE:
To describe the DOHMH centralized de-duplication service.
METHOD:
Since the MCI’s main function is to facilitate matching and be extensible to all DOHMH databases, the data model has been designed with attributes common to all systems. In addition, information specific to a particular system may also be stored in the MCI to improve matching. The MCI has been developed to be "person-centric" and to accomplish this goal, it employs a sophisticated artificial intelligence program to identify duplicates entering the system. This matching interface is “trained” by program-specific data.
RESULT:
In the summer of 2003, the CIR and LQ will integrate through the MCI. After deployment, the duplication rate for the CIR is expected to decline from its current rate of 30% to a negligible level and LQ’s low de-duplication rate will also be improved. Currently, 9 FTEs are allocated to manual de-duplication for the CIR, but after MCI deployment this number will be reduced to 1 FTE. Further, the Bureau of Communicable Disease currently has a 30% duplication rate, and their new surveillance system is being developed to communicate with the MCI for de-duplication purposes. DOHMH is also planning to leverage the MCI for other units in the future.
CONCLUSION:
Multiple programs benefit from a centralized de-duplication service.
LEARNING OBJECTIVES:
To describe a centralized de-duplication service.
Handout (.ppt format, 166.0 kb)
Back to Quality Improvement -- The Key to a Successful Registry
Back to The 2003 Immunization Registry Conference (October 27-29, 2003)