Kawai AT, Andrade SE, Rosofsky R, Zichittella L, Haffenreffer K, Walraven C, Hayes K, Selvan M, Loughlin AM, Shoaibi A, Anderson S, Lee G. Developing a mother-infant cohort in Sentinel's Prism Program as a resource to monitor the safety of vaccine use during pregnancy. Presented at the 33rd ICPE International Conference on Pharmacoepidemiology & Therapeutic Risk Management; August 26, 2017. Montreal, Canada. [abstract] Pharmacoepidemiol Drug Saf. 2017 Aug; 26(Suppl 2):21.


BACKGROUND: Post-marketing safety data on medication and vaccine use during pregnancy are limited. The Sentinel Initiative, which was created to monitor the safety of FDA-approved medical products and includes electronic claims and/or health record data from 18 partners, could potentially be used to conduct safety surveillance in pregnant women.

OBJECTIVES: To develop a mother–infant cohort within Sentinel’s Post-Market Rapid Immunization Safety Monitoring (PRISM) Program for vaccine safety analyses.

METHODS: In women ages 10–54, we identified live deliveries from 2004 to 2011 in claims data from four Sentinel Data Partners. Live deliveries were linked to infants using subscriber identifiers, last names and addresses. Mother–infant dyads were then linked to birth certificate data from 10 states. Using claims data, we developed an algorithm for gestational age (GA) and validated it using birth certificates.

RESULTS: We identified 651,607 deliveries occurring in women meeting enrollment criteria (180 days prior to pregnancy start through 30 days after delivery). Most deliveries occurred in women ages 25–39 (88%). The prevalence of preterm delivery was 10%, while the prevalence of multiple birth was 5%. We linked 542,278 (83%) of the deliveries identified in claims data to infants. A total of 163,202 (30%) deliveries linked to infants occurred in 10 states with birth certificate linkage; of these, 122,770 (75%) were linked to birth certificates, with 119,856 representing singleton pregnancies. Of the 42,671 singleton mother–infant dyads with birth certificate obstetric estimates (OE) of gestational age available, 40,964 (96%) had claims-based GA within 14 days of the birth certificate OE. Of the 83,292 singleton mother– infant dyads with birth certificate last menstrual period (LMP) date available, 72,670 (87%) had claims-based GA within 14 days of the GA derived from the birth certificate LMP.

CONCLUSIONS: We successfully created a large, mother–infant cohort. Further characterization of other data elements is needed, but the validation of the claims-based GA algorithm supports the feasibility of using Sentinel’s PRISM Program to monitor the safety of vaccine use during pregnancy.

FUNDING:
The Sentinel Coordinating Center is funded by the FDA through the Department of Health and Human Services (HHS) Contract number HHSF223201400030I. This project was funded by the FDA through HHS Mini-Sentinel contract number HHSF223200910006I.

Share on: