Duplicate patient records pose a major problem for many immunization registries, as well as for many electronic patient record systems. This paper reports two complementary studies exploring the deduplication of immunization registry records. One study explores the utility of different demographic data elements, singly and in combination, to assist in the deduplication process. The second study explores how clinical patient data (vaccination history data) might assist in this process. To assess the utility of demographic data elements, data were used from three registries after duplicates had been identified. A computer program, IMM/Scan, was written to count the number of true-positive (TP) matches and false-positive (FP) matches found when using different Boolean combinations of demographic data elements. In this study, a strategy of "ORing high value ANDed pairs of data elements" appeared to be most powerful. To assess the utility of vaccination history data, record pairs were drawn from 440,000 patient records. Two metrics on patient history were tested: (1) the number of identical doses shared by two records, and (2) the number of "extra" doses in the combined history of two records. In this study, sample findings include: (1) for pairs of nonduplicate records, 93% had no identical doses and 90.6% had "extra" doses, and (2) for pairs of duplicate records, 83.8% had one or more identical doses and 82% contained no "extra" doses. These studies demonstrate potentially useful approaches to using demographic data and patient history data to assist the automated deduplication of immunization patient records.
[1]
D. Clark,et al.
Comparison of probabilistic and deterministic record linkage in the development of a statewide trauma registry.
,
1995,
Proceedings. Symposium on Computer Applications in Medical Care.
[2]
R. Thompson,et al.
Development and validation of an immunization tracking system in a large health maintenance organization.
,
1993,
American journal of preventive medicine.
[3]
Matthew A. Jaro,et al.
Probabilistic linkage of large public health data files.
,
1995,
Statistics in medicine.
[4]
A. Pennisi,et al.
Evaluating the accuracy of transcribed clinical data.
,
1993,
Proceedings. Symposium on Computer Applications in Medical Care.
[5]
Thomas B. Newman,et al.
Implementation Brief: Use of Commercial Record Linkage Software and Vital Statistics to Identify Patient Deaths
,
1997,
J. Am. Medical Informatics Assoc..
[6]
Perry L. Miller,et al.
IMM/Scrub: A Domain-Specific Tool for the Deduplication of Vaccination History Records in Childhood Immunization Registries
,
2000,
Comput. Biomed. Res..
[7]
Sue Stokes Du Bose,et al.
VERIFICATION AND VALIDATION PROCEDURES FOR IMMUNIZATION REGISTRIES
,
1997
.