Associate Name: Andreea Creanga
Funding Source/Period of the Grant: NICHD R03 4/1/2019 - 3/31/2021
Maternal mortality has been increasing in the United States. The extent to which non-medical use of prescription drugs by pregnant and postpartum women has contributed to this trend is unknown since no regional or national database captures this information. We propose a study to demonstrate the feasibility and value of linking Maternal Mortality Review (MMR) and Prescription Drug Monitoring Program (PDMP) data at the state level in Maryland to improve our understanding of the contribution of prescription drug use to maternal mortality. The study will be conducted in close collaboration with the PDMP and MMR Programs in Maryland, both coordinated by the Maryland Department of Health. We will use a fully de-identified, linked MMR-PDMP dataset capturing deaths among Maryland resident women who died in Maryland during 2013-2016 linked to all available prescription drug records available for these women in the Maryland PDMP. Thus, for each death, information from traditional sources of MMR data (e.g. death certificates, medical records) will be supplemented with information on drugs prescribed and dispensed to women. A total of 150 pregnancy-associated deaths occurred in Maryland during 2013- 2016. Based on MMR data alone, 72 of these deaths were among women known to have used street or prescription drugs during pregnancy or in the postpartum period; and 41 women died from unintentional drug overdose. We will use all the information available in MMR and PDMP to develop timelines of known drug use (i.e. street and prescription drugs) for each pregnancy-associated death. We will perform a case-by-case comparison of prescription drug use information captured in MMR versus PDMP data to quantify and classify the information added by the PDMP data. Subsequently, we will examine levels and patterns of prescription drug use separately among drug overdose deaths, pregnancy-related, and pregnancy-associated deaths, by year and overall during the 2013-2016 period. Last, we will examine bivariate associations between each of two outcomes (i.e. prescription drug use and unintentional prescription drug overdose death) and key individual and health systems factors, testing for statistical significance at p<0.05. Results from our study, including guidelines for linking MMR and PDMP data, will be disseminated widely to scientific, professional, and lay audiences. We expect this study to be the first to demonstrate the feasibility and value of linking MMR and PDMP data at the state level, and for these linkages to be used by other states in the future. The PDMP data will provide additional, much needed information on circumstances leading to death among pregnant and postpartum women who abuse or are dependent on prescription drugs, which can help clinicians and public health professionals identify better ways to prevent such deaths in the future.