Creation and evaluation of a multi-layered maternal and child health database for comparative effectiveness research

Background: Comparative effectiveness research (CER) is a rigorous scientific approach designed to generate valid scientific evidence to make informed decisions that will improve health care quality and patient outcomes. As high-speed computers and sophisticated software packages for data linkage become increasingly available, investigators are creating massive databases for CER. Considering their potential use in informing health care decisions, we must increase transparency of these data so that potential biases can be addressed. Methods: We used an innovative stepwise deterministic record linkage strategy to construct a clinically-enhanced maternal and child health (MCH) database for CER. Mothers and infants that were part of a 12-year cohort were followed up for at least one year (up to 12) after the child's birth or until death, investigating inpatient, outpatient, and emergency department visits. We used multivariable modeling to investigate linkage disparities across a host of maternal and infant demographic, reproductive, and geospatial characteristics. Results: We were able to link 92.1% of the 2,549,738 birth certificate records to an infant birth hospitalization record. The highest crude unlinked rates were seen among infants who died during their first year of life (35.9%), and infants who born to mothers without health insurance (28.1%), with less than a 9th grade education (26.0%), who were foreign-born (12.9%), and who self-identified as Hispanic (12.8%). We discuss potential biases in using these data, particularly for MCH disparities research. Conclusions: Our findings describe a unique algorithm for linking MCH data and demonstrate the importance of evaluating routinely collected and linked health data.