Vaccine Safety Datalink: A Special Purpose Database

VACCINE SAFETY DATALINK: A SPECIAL PURPOSE DATABASE

In order to identify rare vaccine adverse events, the CDC funded the Vaccine Safety Datalink (VSD), a large database that brings together computerized information on immunizations, medical outcomes and potential confounders. The VSD has been used to evaluate hypotheses from the medical literature, from the VAERS, from changes in immunization schedules, and from the introduction of new vaccines. Beginning in 1991, CDC joined with four HMOs, GHC, KPNW, Kaiser Permanente Northern California and Kaiser Permanente Southern California, all in the western part of the United States. Up to ten HMOs have been utilized for specific studies (DeStefano, 2001), capitalizing on the efficiencies offered by HMOs for population-based health research. Initially focusing on children up to 6 years of age, the database now includes adolescents and adults as well, and totals to approximately 6 million members (DeStefano, 2001). Information on all vaccinations given within the HMO study population, either routinely or for special indications, is computerized, including the vaccine type, date of vaccination, concurrent vaccinations, the manufacturer and lot number and site of vaccination. Outcome data are collected from various sources at each site, such as hospitalizations, emergency depart-ment visits and outpatient clinic visits. To preserve patient confidentiality, each site assigns unique study identifiers to its data before shipping to the CDC annually for merging and analysis (Chen et al., 1997).

Quality control studies have shown high levels of agreement between computerized data and paper medical records. A quality control analysis of three of the HMOs comparing the automated database with paper records for common childhood vaccines showed that from 83% to 99% of the automated records were present in the paper records, and from 82% to 98% of the paper records were present in the automated database (Mullooly et al., 1999).