Improved Recognition of Maternal Deaths Using Modern Data Analytics

Area of Science:

• Medical Informatics
• Public Health Surveillance
• Reproductive Health

Background:

• Maternal mortality surveillance relies on standard strategies that may miss cases.
• Accurate ascertainment of maternal deaths is crucial for effective public health interventions.

Purpose of the Study:

• To implement a data-fusion approach to improve the detection of maternal deaths.
• To enhance the identification of pregnancy-associated mortality beyond standard surveillance methods.

Main Methods:

• A retrospective cohort study utilized electronic health records from 2011-2018.
• Maternal deaths were identified using standard queries and the TriNetX discovery tool, followed by chart review.
• Pregnancy-associated mortality was compared between the discovery and hospital-identified cohorts.

Main Results:

• The discovery tool identified 18 additional maternal deaths, a 78% increase in ascertainment.
• A higher proportion of newly identified deaths occurred postpartum.
• Cardiac causes and medical comorbidities were the primary drivers of newly ascertained mortalities.

Conclusions:

• Modern data analytics can significantly improve the recognition of pregnancy-associated deaths.
• The data-fusion approach offers a more comprehensive method for maternal mortality surveillance.
• Enhanced identification aids in understanding the full scope of maternal mortality, including deaths related to early pregnancy loss.