Exploring inaccuracies in fetal and infant death certificates

Background. In 2011, Baltimore City had a fetal-infant mortality rate of 21.3 per 1000 live births, which was 1.5 times higher than Maryland's rate. Fetal–Infant Mortality Review (FIMR) programs in the US rely on birth and death records to identify trends in fetal/infant mortality and make recommendations to improve birth outcomes. Baltimore City FIMR routinely encounters records with missing data or data that are discrepant with medical records; poor data impairs the FIMR team's ability to analyze fetal/infant death trends. This study aimed to identify the prevalence of inaccuracies in fetal/infant death certificates. Methods. A convenience sample of fetal and infant death certificates (N=29) from Baltimore City birthing hospitals (2011-2012) were compared to medical records to identify inconsistencies and missing data. Six key informant interviews were conducted to identify contributors to inaccurate death certificate data. Thirty additional cases were abstracted and will be added to the sample. Results. Twenty-five (86%) death certificates had at least one inaccuracy. Inaccuracies included missing/unknown data (60%) or medical record discrepancies (60%). Eleven (37%) death certificates had inaccurate data related to prenatal care and/or previous birth outcomes. Key informant interviews revealed contributing factors to be inadequate training in certificate completion and absence of monitoring. Conclusions. Most certificates had inaccurate data, including important variables. Since death certificate data drive processes such as FIMR, it is vital to further explore, and prevent, poor certificate completion. Improving death certificate data accuracy is critical for developing targeted recommendations to further reduce fetal/infant mortality in cities such as Baltimore.