Project #196:
OpenPREGnosis: Developing an Open Algorithm to Identify Pregnancy Episodes and Outcomes in OpenSAFELY

As part of the CORE20PLUS5 initiative, the NHS is focusing on reducing health inequalities in maternity care, especially for the most deprived 20% of the population and those from minority groups. Women of ethnic minority backgrounds face higher risks of pregnancy-related complications, influenced by factors such as age, existing health conditions, and socioeconomic status. The COVID-19 pandemic worsened these disparities, with UK data showing differences in pregnancy outcomes according to ethnic background.

To better understand maternity care nationally, high-quality national data is essential. The OpenSAFELY platform uses two of the largest NHS electronic health record providers, including 58 million records from hospitals and GPs across England. Our initial research was able to identify over 600,000 births (in one of the two providers) and highlighted ongoing disparities in maternal healthcare interactions during the pandemic.

To make better use of the rich data within OpenSAFELY for more in-depth pregnancy our study aims to develop and validate a robust method to accurately identify pregnancy episodes and related events within OpenSAFELY. This will enable analysis of adverse pregnancy outcomes before, during and after the COVID-19 pandemic, considering factors like ethnicity, age, socioeconomic status, existing health conditions, medication use (including vaccinations), and pregnancy history. We aim to understand how these factors influence pregnancy risks and provide evidence for clinicians and policy makers on how risks during pregnancy changed during the pandemic, and the manner in which maternal inequalities were exacerbated. This will provide evidence for strategies to improve preparedness for future public health crises.

Ultimately, this study aims to create a reliable algorithm for identifying pregnancy episodes in OpenSAFELY and to use the pregnancy data to examine health interactions and identify risk factors for adverse outcomes.