Health Information Solutions

Home Data Linkage Website Construction Graphical User Interfaces Statistical Analyses SAS Support Services Publications Contact Links and Client Support E-Mail Last Updated: 01/11/2007 The present situation of data and information technology in the health care industry is characterized by the development of large independent data collection systems developed by different entities for different purposes using non-standard data elements. The expense of data collection contributes to the pressure of improving presently available data by simply linking it to other complementary data sources. Health Information Solutions has developed customized linkage algorithms for several linkage situations. For instance, we have developed probabilistic techniques for linking the infant hospital discharge record, maternal hospital discharge record, and vital statistics birth records with linkage percentages exceeding 98%. We have also developed a probabilistic linkage technique for matching re-hospitalizations and transfers of under-one-year olds to their birth hospitalization as well as for matching maternal prenatal and postnatal hospitalizations. The probabilistc linkage replaces the deterministic linkage technologies that are described in a manuscript that was published in 1997. For more information on obtaining access to these linked data files, contact the Health Care Information Division at the California Office of Statewide Health Planning and Development. The platform we primarily use to develop our probabilistic and deterministic record linkage algorithms is the Statistical Analysis System, (SAS™). We have developed a strategy for linking California death records for persons who died during a year to all hospitalizations experienced by these persons during the year preceding their deaths. This information is currently used to determine differential health care access patterns during the last year of life by demographic characteristics. In collaboration with the California Health Information for Policy Project (CHIPP) and the Family Health Outcomes Project (FHOP), Health Information Solutions has devised probabilistic techniques for the unduplication of vital records based on core data elements. These methods have been used to unduplicate the California State Client Index database, Marin County, San Diego County, and Orange County Public Health Data. More information about these linkage efforts can be found at the Office of Statewide Health Planning and Development. Past linkage projects include matching Medicare beneficiary data to California hospital discharge data and matching WIC data to vital statistics birth, maternal and infant hospital discharge data. We are also currently involved in linkage activities for the California CABG Outcome Reporting Program (CCORP).