30354 Monitoring the Impact of Human Papillomavirus (HPV) Vaccines On Precancerous Cervical Lesions by Linking Immunization Information System and Cancer Registry Data In Michigan

Monday, March 26, 2012
Poster Hall
Rachel Potter, DVM, MS , Vaccine-Preventable Disease Epidemiologist, Michigan Department of Community Health

Background: The impact of HPV vaccine on cervical cancer may be monitored by linking immunization and cancer registry data.  Michigan is uniquely positioned to examine the most important cancer precursor, cervical intraepithelial neoplasia grade 3 (CIN3), by vaccination status, using two population-based resources: the Michigan Cancer Surveillance Program (MCSP), which has collected cervical carcinoma in situ cases since 1985 and CIN3 since 2009, and its immunization information system, the Michigan Care Improvement Registry (MCIR).

Objectives:  Our objective was to assess the feasibility of identifying a cohort of women from the MCIR who had continuously resided in Michigan and linking their MCIR and MCSP records.

Methods: We identified females with 1976-1996 birthdates in the MCIR and used probabilistic linkage software to match these records with the Michigan birth files; a sample of invalid and possible links was evaluated via manual review.  MCSP cervical carcinoma in situ cases diagnosed in 2006 among women <30 years old were matched with the birth files; MCSP-birth linkages were merged with MCIR-birth linkages using the birth file identifier.  A sample of MCIR-birth linkages was provided to a commercial locator service to identify women with continuous Michigan residence, using child and parent information from the birth file.

Results: Overall, 68% of the 1,274,282 MCIR and 61% of the 1,358 MCSP records matched the birth files.  MCIR-birth linkages increased with birth year, ranging from 24% to 73%; we estimated probable linkages would increase to 79% after manual review.  Over 80% of MCSP-birth linkages merged with IR-birth linkages for cases born after 1984.  Among the sample of MCIR-birth linkages, 86% were continuous residents. 

Conclusions: The study methodology was feasible and used existing data sources.  Its strengths are the high proportion of linkages and population-based data.  Future steps include a study using MCIR-birth linked records for HPV vaccination-eligible women.