Evaluation of Birth Defects Surveillance in Minnesota

BACKGROUND:  Minnesota’s Birth Defects Information System (BDIS) is an active surveillance system that collects information on 45 categories of birth defects for live-born infants from hospitals and outpatient clinics. ICD-9-CM codes from discharge data and birth certificates are utilized to identify possible cases. The goals of BDIS are to monitor birth defects trends; prevent birth defects through targeted education; ensure appropriate services are provided; educate providers and the public regarding birth defects; and stimulate research on risk factors, treatment, prevention, and the cure of birth defects. This evaluation informs recommendations and identifies areas for improvement. 

METHODS: First, the data quality of BDIS was evaluated through assessing death certificates as a data source for case finding. The 2009-2010 death dataset was analyzed to identify deaths among children born in 2009 with cause-of-death as congenital anomalies (ICD10 Q00-Q99). Deaths due to congenital anomalies were linked on birth certificate number with the BDIS dataset to determine the cases that had already been captured through another source (matches).The unmatched deaths were abstracted for medical information to determine if they met the BDIS case definition.  Secondly, BDIS was evaluated through a quality assessment that randomly selected 5 % (29) cases for re-abstraction.

RESULTS: The 2009-2010 death certificates identified 98 children who were born in 2009 with congenital anomaly as their cause-of-death. Fifty-two deaths matched on birth certificate number with a confirmed case in BDIS. There were 46 unmatched deaths that required further investigation to determine if they were missed cases. Death certificates identified four unmatched deaths that met the BDIS case definition of a birth defect. Two cases were diagnosed with anencephaly, one with trisomy 18, and one with multiple birth defects, including hydrocephalus. Abstracted and re-abstracted cases had 69% (20) agreement; the discrepancy between the pairs was 24% (7) due to the re-abstraction missing a diagnosis, 7% (2) due to the abstractor missing a diagnosis.

CONCLUSIONS: Minnesota’s BDIS is a reliable surveillance system, especially when secondary data sources are utilized for case finding. The death certificate analysis is a useful quality assessment tool that serves as a safety net for cases missed either through technical errors or notification issues.  Re-abstraction identified that abstraction practices for patent ductus arteriosus, patent foramen ovale, and microcephaly conditions have changed and might affect rates and counts. Further advancements will improve the data quality and more effectively connect children with services to mitigate the chronic impact of birth defects.