Using active birth defects surveillance programs to supplement data on fetal death reports: Improving surveillance data on stillbirths

Abstract

Surveillance of stillbirths using fetal death reports (FDRs) has been challenging because of under-reporting of fetal deaths and missing data on the FDRs. Using active case finding and chart abstraction within the infrastructure of established birth defect surveillance programs could potentially enhance the data from FDRs. The data collection form for the Metropolitan Atlanta Congenital Defects Program, an active, population-based birth defects surveillance system, was modified to collect additional information on stillbirths from medical records. The study population was a 25% simple random sample of stillbirths recorded on FDRs in 2004 (n = 125) by residents in the five central counties of metropolitan Atlanta. Stillbirth was defined as a fetal death at > or =20 weeks gestation or > or =350 g if age was unknown. Data on demographic characteristics and risk factors collected from the two sources were compared for completeness and agreement, as well as causes of and conditions associated with the fetal death. Combining data sources provided more information. Demographic and risk factor variables in the two data sources showed strong agreement (categorical variable, kappa range = 0.79-1.00; continuous variable, correlation coefficient range = 0.61-1.00). The actively ascertained data provided more complete information for causes and conditions of fetal death. Data from the FDRs yielded 42% of cases with no listed cause of death or associated condition compared with 10% using Metropolitan Atlanta Congenital Defects Program data. Expanding the potential of existing active birth defects surveillance programs to include stillbirth surveillance could potentially improve the quantity and quality of available data on fetal deaths. Ongoing studies are needed to corroborate these findings and to assess completeness of case ascertainment.