Massachusetts Department of Public Health (MDPH) birth certificate data

Information on all live births in Massachusetts is stored in the Registry of Vital Records and Statistics at MDPH. The Massachusetts Standard Certificate of Live Birth, referred to as the ‘birth certificate’, consists of a Parent Worksheet and a Hospital Worksheet. The parent(s) completes the Parent Worksheet, which contains legal and socio-demographic information on the child’s mother and father. While the birth certificate does not confirm that the father is biological, it states that the information provided is about the child’s father regardless of whether the father will appear on the child’s legal birth record. A designated hospital representative (e.g., doctor, nurse, or hospital birth registrar) completes the Hospital Worksheet, which contains information on prenatal care, labor and delivery, neonatal conditions and procedures, and discharge.

The birth certificate contains information on infant’s sex, birth weight, plurality, gestational age based on the last menstrual period and clinical estimates, mode of delivery, and parity.

Mothers self-report the average number of cigarettes they smoked daily before and, separately, during pregnancy. The hospital records the mother’s total weight gain/loss, whether the mother had GDM, whether the mother had hypertension, whether the mother was breastfeeding at the time the birth certificate was completed (referred to as breastfeeding initiation), and month prenatal care began and the number of prenatal care visits.

Mothers and fathers each report their race (white, Black, Asian/Pacific Islander, American Indian, and other), age, place of birth, education, language preference, and marital status (mothers only). The birth certificate in Massachusetts also collects information on each parent’s ancestry or ethnic heritage (referred to as ethnicity) from 39 items, including several write-in options [56]. The hospital records the mothers’ medical insurance status for the delivery.

Mothers report the city and zip code of their residential mailing address on the birth certificate and the Registry reports the census tract. We have the ability to link each child’s census tract with area-level measures of socioeconomic circumstances through the census and the built environment. Information from commercial databases on locations of parks, fast food restaurants and supermarkets can provide indicators of children’s physical activity and food environments.

The birth certificate has undergone multiple revisions since its inception. While a majority of the variables from the birth certificate are available from 1969 onwards, when birth certificate data are first available from MDPH, birth certificates have collected increasingly more information over time. Data for the pregnancy/infant measures as well as maternal race, education, and marital status are available over the entire study period. Data for ethnicity and other socio-demographic characteristics are available primarily from 1987. Similarly, breastfeeding initiation, total weight gain/loss, and pregnancy-related hypertension were included in the birth certificate from 1987. Maternal smoking during pregnancy was collected from 1992 and GDM from 1996.

IRB approval for the Linked CENTURY Study was obtained from Boston College, Harvard Pilgrim Health Care (HPHC), MDPH, and Massachusetts General Hospital. Only approved study personnel at HPHC and MDPH had access to names and dates of birth for data linkage purposes and researchers had access only to a de-identified dataset.

In collaboration with MDPH, we developed a process for transferring the data between institutions (Fig. 1) and linking the datasets (Table 2). The Research Support Data Center at HPHC created a dataset that contained a random ID for each CENTURY Study child, child’s name and date of birth, mother’s date of birth, and all study variables. The Research Support Data Center sent the dataset to MDPH who linked each child with their birth certificate based on a linkage algorithm comparing the child’s name and date of birth and the mother’s date of birth. Table 2 presents the matching phase linkage algorithm and resulting number of matches for the six permutations of the algorithm. The majority of matches occurred only using the child information: 45.2 % of matches were based on the child’s first and last name and date of birth, while a further 33.3 % of matches were based on the child’s first, middle initial, and last name and date of birth. MDPH then removed identifying information and returned the dataset to our study team.

Flow diagram for linking the CENTURY Study data with each child’s Massachusetts birth certificate

Success rate of linkage algorithm by type of match (N = 200,343)