Danielsen BH, Carmichael SL, Gould JB, Lee HC. Linked birth cohort files for perinatal health research: California as a model for methodology and implementation.
Ann Epidemiol 2023;
79:10-18. [PMID:
36603709 PMCID:
PMC9957937 DOI:
10.1016/j.annepidem.2022.12.014]
[Citation(s) in RCA: 14] [Impact Index Per Article: 7.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/28/2022] [Revised: 12/20/2022] [Accepted: 12/29/2022] [Indexed: 01/04/2023]
Abstract
PURPOSE
Rigorous perinatal epidemiologic research depends on population-based parental and neonatal sociodemographic and clinical data. Here we describe the creation of linked birth cohort files, an enriched data source that combines information from vital records with maternal delivery and infant hospital encounter records.
METHODS
Probabilistic linkage techniques were used to link vital records (i.e., birth and fetal death certificates) from the California Department of Public Health with hospital inpatient, ambulatory surgery and emergency department encounter data for mothers and infants from the California Department of Health Care Access and Information.
RESULTS
From 2012 to 2018, 95% of live birth records were successfully linked to maternal and newborn hospital records while 85% of fetal death records were linked to a maternal delivery record. Overall, 93% of postnatal hospital encounters of infants (i.e., <1 year old) were matched to a linked record.
CONCLUSIONS
The linked birth cohort files is a rich resource opening many possibilities for understanding perinatal health outcomes and opportunities for linkage to longitudinal, social determinant, and environmental data. To optimally use this file for research, analysts should evaluate possible shortcomings or biases of the data sources being linked.
Collapse