Evaluation of maternal diseases and medications recorded in Northern Ireland Maternity System (NIMATS) database compared to NI Enhanced Prescribing Database (EPD): a data linkage validation study

The effects of medication exposure during pregnancy on the unborn child are largely unknown, as pregnant women are excluded from pre-marketing clinical trials assessing drug safety. The use of existing administrative datasets, holding routinely collected patient health records, may be feasible for conducting medication safety studies on pregnant women to investigate the risks/consequences of in utero medication exposure on birth outcomes and later child development outcomes. However, these administrative datasets require assessment of data quality to ensure that the data are suitable for research as they were not developed for this purpose. The aim of this study is to evaluate the quality of maternal diseases and medications data recorded in the Northern Ireland Maternity System (NIMATS) database compared to the NI Enhanced Prescribing Database (EPD).NIMATS data for all mothers resident and giving birth in NI will be linked to EPD data using the mother’s Health Care Registration Number (HCN) as a unique patient identifier. Agreement between the two data sources and sensitivity/ specificity estimates will be calculated for each therapeutic class of medications. Self-reported maternal diseases recorded in NIMATS will also be cross-validated with dispensed prescriptions recorded in EPD to assess completeness, consistency and accuracy of information.Awareness of the quality of NIMATS medication and maternal diseases data is essential if these data are to be used in data linkage studies assessing medication safety in pregnancy. These findings will inform a larger ADRN NI/ UK-wide data linkage study exploring child developmental and educational outcomes following maternal medication exposure during pregnancy.