How complete are immunization registries? The Philadelphia story

To assess accuracy and completeness of Philadelphia, Pa, registry data among children served by providers in areas at risk for underimmunization.

Philadelphia's Department of Public Health selected a simple random sample of 45 children age 19-35 months (or all children age 19-35 months if there were <45 children in the practice) from each of 30 private practices receiving government-funded vaccine and located in zip codes where children are at risk for underimmunization. Chart and registry data were compared with determine the proportion of children missing from the registry and assess differences in immunization coverage.

Of 620 children reviewed, 567 (92%) were in the registry. Significant differences (P < .05) were observed in immunization coverage for 4 diphtheria-tetanus-acellular pertussis vaccinations, 3 polio vaccinations, 1 measles-mumps-rubella vaccination, and 3 Haemophilus influenzae type b vaccinations between the chart (80% coverage) and registry (62% coverage). Providers submitting electronic medical records or directly transferring electronic data to the registry had significantly more children in the registry and higher registry-reported immunization coverage than those whose data were entered from billing records or log forms. All practice types experienced difficulties in transferring complete data to the registry.

Although 92% of study children were in the registry, immunization coverage was significantly lower when registry data were compared with chart data. Because electronic medical records and direct electronic data transfer resulted in more complete registry data, these methods should be encouraged in linking providers with immunization registries.