Evaluation of Data Quality and Utility of the Japan Drug Information Institute in Pregnancy (JDIIP) Consultation Case Database for Pregnancy Pharmacovigilance

Abstract

Introduction: Ensuring medication safety during pregnancy is crucial for protecting maternal and fetal health. However, fragmented data sources and the lack of comprehensive databases present substantial barriers to effective pharmacovigilance. The Japan Drug Information Institute in Pregnancy (JDIIP) database, which contains data on drug treatment counseling for pregnant women, is expected to help address the lack of comprehensive databases for pregnancy pharmacovigilance (PregPV).

Objective: We evaluated the quality and utility of the JDIIP database for PregPV activities, particularly its ability to consolidate and utilize drug-exposure data among pregnant women in Japan.

Methods: To assess the quality and utility of the JDIIP database for PregPV, we examined its alignment with 48 core data elements (CDEs) considered critical for PregPV, as recently proposed by a European Union consortium through the ConcePTION Project. We performed a detailed mapping of each CDE definition-including maternal lifestyle factors, drug exposure, and pregnancy outcomes-against the corresponding data elements captured in the JDIIP database.

Results: The JDIIP database either directly collected or could derive 38 of the 48 specific items (79%) recommended by the ConcePTION Project. At the category level, the JDIIP database aligned closely with the CDE requirements for database management details, pregnancy details, maternal medical history, pregnancy medication exposure, live/stillborn birth outcomes, and malformation details, achieving coverage of over 80% of the necessary variables in each category. Some categories, such as maternal medical conditions arising during pregnancy and infant complications within the first year of life, showed less alignment, with coverage rates below 50%. Although the JDIIP database provides comprehensive coverage of critical pharmacovigilance elements, data collection for specific variables and categories that better align with the CDE framework can be enhanced to improve alignment with the CDE framework and strengthen pharmacovigilance capabilities.

Conclusions: Our findings highlight the potential of the JDIIP database as a valuable resource for advancing PregPV research. Although the collection of certain maternal and infant data elements could be improved, the substantial alignment of the database with established CDEs positions it as a promising tool for advancing PregPV initiatives in Japan.

Fig. 1

Data processing flowchart Figure 1 illustrates the data processing steps of the JDIIP consultation database and describes the filtering steps used to exclude incomplete or unconfirmed cases, resulting in the final dataset used for the study analysis. “Insufficient information” refers to cases in which a response to the follow-up postcard survey was received but the pregnancy outcome could not be determined. This includes responses in which the outcome section was left blank or the information provided was ambiguous or incomplete, making it insufficient to classify the pregnancy outcome. Because the primary analysis required confirmed outcomes, these cases were excluded to avoid potential misclassification

Fig. 2

Number of unique patients for drug classes ATC Anatomical Therapeutic Chemical Figure 2 shows the number and percentage of unique patients for each drug class, based on the ATC classification system. Each patient was counted only once per ATC major category, even if they had been prescribed multiple drugs within the same class. The x-axis represents the number of unique patients who used drugs within each class during pregnancy, and the y-axis represents each drug class.