Sensitivity of birth certificate reporting of birth defects in North Carolina, 2011-2017

Statistical Brief No. 49 July 2020 1908 Mail Service Center ♦ Raleigh, NC 27699-1900 919.733.4728 ♦ www.schs.state.nc.us Sensitivity of Birth Certificate Reporting of Birth Defects in North Carolina, 2011-2017 Nina Forestieri, MPH, Birth Defects Monitoring Program Manager Kathleen Jones-Vessey, MS, Women's & Children’s Health Section, Maternal & Child Health Epidemiologist Robert Lee, MA, MS, Statistical Services Manager Introduction In North Carolina, about 4,000 infants (3% of all births) are born with a major birth defect each year. The North Carolina Birth Defects Monitoring Program (NCBDMP) collects information about all medically diagnosed cases of major birth defects among North Carolina residents and maintains a central registry with this information. Select birth defects are also reported on the birth certificate, but research suggests that birth certificates are not a reliable source for identifying birth defects.1'6 The current revision of the US birth certificate, released in 2003 and launched in North Carolina in August 2010, was revised in an effort to improve birth defect reporting and data quality. The revised certificate only includes anomalies diagnosable within the first 24 hours after birth.7'* Despite these changes, a recent study reported poor reliability of birth defect data reported on the revised certificate, and therefore recommended against the use of birth certificates alone as a source of birth defect data.4 This report examines the reliability of birth defects reported on the birth certificate by linking birth certificate data to cases from the NCBDMP, a complete and accurate source of data on birth defects in North Carolina. Methods For births occurring in 201 1-2017, birth certificates were linked to data on birth defects obtained from the NCBDMP; 201 1 was the first full year of revised birth certificate data collected in North Carolina, and 201 7 is the most recent complete year of NCBDMP data available. NCBDMP is an active, population- based surveillance system, in which data are collected by trained field staff who systematically review and abstract medical records in order to verify suspected cases identified by hospitals. Information from administrative health databases such as hospital discharge records and vital records is utilized as well. This method provides the most complete and accurate data on the prevalence of birth defects. The following birth defects captured on the 2003 birth certificate revision were examined: anencephaly, meningomyelocele/ spina bifida, cyanotic congenital heart disease, congenital diaphragmatic hernia, omphalocele, gastroschisis, limb reduction defect, cleft lip with or without cleft palate, cleft palate alone, Down syndrome, and hypospadias. Cyanotic congenital heart disease is a broad grouping, and specific heart defects that this category encompasses are not specified on the birth certificate. For the purposes of this analysis, we defined cyanotic congenital heart disease as any one of the seven critical congenital heart defects most likely to be detected by pulse oximetry screening at birth.9 Sensitivity values were calculated for each birth defect and for all birth defects together (any reported birth defect), along with 95% confidence intervals (CIs). Sensitivity is the proportion of true cases (cases confirmed by the NCBDMP) that were also captured on the birth certificate. For specific birth defects reported on the birth certificate (i.e., each birth defect excluding broader categories of cyanotic congenital heart defects and limb reduction defects), positive predictive values (PPVs) were also calculated. PPV is the proportion of cases captured on the birth certificate that were true cases (as opposed to false positives). For all birth defects together, sensitivity was also calculated by the following characteristics obtained from data recorded on the birth certificate: year of birth, maternal age (<20, 20-24, 25-29, 30-34, or 35+ years), maternal race/ethnicity (non-Hispanic white, non-Hispanic Black, Hispanic, or other), maternal nativity (US-born or foreign-born), maternal education (less than high school, high school, or greater than high school), plurality (singleton or multiple), birthweight category (<2500 grams or 2500+ grams), gestational age category (<37 weeks or 37+ weeks). Adequacy of Prenatal Care Utilization (Kotelchuck) Index (inadequate, intermediate, adequate, adequate plus), Perinatal Care Region (Region 1: western. Region II: northwestern. Region III: southwestern. Region IV: northeastern. Region V: southeastern. Region VI: eastern), and average annual number of births in the infant’s birth hospital (<500, 500-999, 1000-1999,2000+). Statistical Brief No. 49 — July 2020 1