An App to Predict Preeclampsia Risks Is Under Way

Researchers have developed an artificial intelligence model that predicts an individual woman's risk of suffering adverse outcomes of preeclampsia.

Preeclampsia is a pregnancy complication that involves persistent high blood pressure and signs of liver or kidney damage. The condition affects between 5% and 7% of pregnancies and is one of the leading causes of maternal death globally.

Despite its wide prevalence, currently, no tests can predict the future onset of preeclampsia early in pregnancy. The complication is treated once it has been diagnosed, which is insufficient and costly.

This may change soon, thanks to the researchers from the University of Strathclyde in Glasgow and King's College London, who have developed a new machine learning model that can accurately predict the level of risk of severe outcomes in women with preeclampsia. The findings on the model's effectiveness were published in The Lancet Digital Health.

"The model we have developed has been rigorously tested and shown to deliver fast, precise predictions of the risks, in a way which can be adapted to the individual circumstances of women around the world," said Dr. Kimberley Kavanagh, senior lecturer in Strathclyde's Department of Mathematics and Statistics and a co-author of the paper.

Machine learning, a type of artificial intelligence, gives computers the ability to learn without explicitly being programmed. Artificial intelligence is already used in medicine to develop better diagnostic tools, such as for analyzing medical images.

The newly developed machine learning model, named PIERS-ML (Preeclampsia Integrated Estimate of Risk – Machine Learning), consolidates its two previous versions.

To put the model to the test, the researchers recruited 8,843 women from 53 maternity units in 11 countries, including the United States. The model helped to accurately classify pregnant women as very low, low, moderate, high, or very high risk of severe outcomes.

Additionally, the records of a further 2,901 women from South-East England were used to confirm the findings of the main study.

The PIERS-ML model has identified nearly 40% of women with preeclampsia for whom care should be altered. Like similar models, it performed best during the first two to seven days.

The model includes maternal mortality ratios based on their geography, meaning the model is globally relevant.

The researchers now aim to develop an app that could determine a patient's risk of severe outcomes following a preeclampsia diagnosis.

Some risks can be managed

Preeclampsia occurs after week 20 of pregnancy and is often precluded by gestational hypertension. However, high blood pressure during pregnancy does not necessarily indicate preeclampsia.

Some women are at a higher risk of preeclampsia than others. The risk factors for the complication include the following:

• First pregnancy
• History of preeclampsia and hypertension
• Chronic kidney disease
• History of thrombophilia
• Pregnancy from in vitro fertilization
• Type 1 or type 2 diabetes
• Body mass index 35 or more
• Family history of preeclampsia
• Maternal age of 40 years or older
• Prolonged interval since last pregnancy

There are no ways to prevent preeclampsia, but you can reduce its risk by controlling diet and physical activity. The American Pregnancy Association recommends:

• Avoid added salt or reduce its intake
• Drink six to eight glasses of water a day
• Stay away from fried foods and junk food
• Exercise regularly
• Get adequate rest
• Elevate your feet several times during the day
• Quit drinking alcohol
• Avoid beverages that contain caffeine
• Take medications and supplements prescribed by your doctor

While diagnostic tools to predict preeclampsia are improving, it is crucial to take precautions and discuss any symptoms with your healthcare provider.