A computerized algorithm to capture patient's past preeclampsia and eclampsia history from prenatal clinical notes

Area of Science:

• Medical Informatics
• Obstetrics and Gynecology
• Clinical Data Mining

Background:

• Electronic medical records (EMRs) contain valuable prenatal data for risk assessment.
• Identifying patients with a history of pregnancy complications like preeclampsia/eclampsia is crucial for timely intervention.

Purpose of the Study:

• To develop and validate a rule-based algorithm, PregHisEx, for extracting past obstetrical history of preeclampsia/eclampsia from prenatal clinical notes.
• To assess the performance of PregHisEx in identifying cases of preeclampsia/eclampsia.

Main Methods:

• A rule-based computerized algorithm, PregHisEx, was developed to mine prenatal clinical notes.
• The algorithm was applied to identify cases of preeclampsia/eclampsia in women who delivered in 2012.
• Validation was performed on a random sample of 200 notes to determine sensitivity, specificity, and predictive values.

Main Results:

• PregHisEx identified a significant number of definite and probable cases of preeclampsia/eclampsia at sentence, note, and pregnancy episode levels.
• Validation demonstrated high performance metrics: 88.0% sensitivity, 98.9% specificity, 91.7% positive predictive value, 98.3% negative predictive value, and an F-score of 0.90.
• The algorithm proved effective in characterizing past obstetrical history.

Conclusions:

• The PregHisEx algorithm is a high-performing tool for identifying past preeclampsia/eclampsia from clinical notes.
• This approach can be extended to characterize other prenatal conditions within EMRs.
• Automated extraction of obstetrical history can enhance risk assessment and patient management.