Cohort profile: the Cohort for Risk Prediction Model Evaluation (CORE) for external validation of models identifying high-risk pregnant women in the early second trimester, North India
The Cohort for Risk Prediction Model Evaluation (CORE) is a prospective, single-site Indian cohort of 964 pregnant women established to externally validate early second-trimester risk prediction models for high-risk pregnancies using harmonized clinical and ultrasound data, with the goal of improving maternal and neonatal outcomes in India and comparable settings.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Imagine you are a chef trying to perfect a new recipe for a cake that prevents it from sinking. You've tested your recipe in your own small kitchen with your specific ingredients, and it works perfectly there. But would that same recipe work if you tried it in a different kitchen, with different ovens, different flour, and different humidity? Probably not, unless you test it there first.
This paper is about building a specialized testing kitchen for pregnancy care in India.
Here is the breakdown of the "Cohort for Risk Prediction Model Evaluation" (CORE) study, explained simply:
The Problem: Recipes That Haven't Been Tested
In the world of medicine, scientists are creating "recipes" (called prediction models) using computer algorithms and data. These recipes are designed to look at a pregnant woman early in her pregnancy (around 4 to 5 months) and predict if she or her baby might be at high risk for problems like premature birth or the baby being too small.
The problem is that most of these recipes are tested only in the kitchen where they were invented. They haven't been tried in other places. In fact, the paper notes that only about 6 to 10% of these medical "recipes" are ever tested outside of their original lab. This is risky because a recipe that works in one place might fail in another.
The Solution: The CORE Testing Kitchen
To fix this, the researchers built the CORE cohort. Think of this as a standardized, high-quality testing kitchen in New Delhi, India.
- Who went in? They invited 964 pregnant women (all over 18 years old) who were less than 20 weeks pregnant.
- What did they do? They didn't just ask questions; they took a "snapshot" of the pregnancy.
- They gathered detailed background info (like family size, income, and education).
- They performed high-quality ultrasound scans (like taking a very clear photo of the baby and the mother's cervix).
- Crucially, they saved these photos in two ways: one set with no markings (clean) and one set with measurements drawn on them. This is like saving a photo of a cake both before and after you measure its height, so future chefs can check the measurements themselves.
The Results: What the Kitchen Produced
After following these women until they gave birth, the researchers looked at the "cakes" (the babies) to see how the predictions would have worked.
- The Mix: The group was a mix of first-time moms and moms who had babies before. About half had a normal weight, while others were overweight or underweight.
- The Outcomes: Out of the 724 babies born alive:
- 11% were born too early (preterm).
- 26.5% were smaller than expected for their age (Small for Gestational Age).
- 5.6% were larger than expected.
- 36% fell into a category called "Small Vulnerable Newborns" (meaning they were either born too early, too small, or had a low birth weight).
Why This Matters (According to the Paper)
The paper doesn't claim this study cured anyone or changed medical practice yet. Instead, it claims to have built a reliable reference library.
- A Standardized Ruler: Because they used the same rules and high-quality ultrasound techniques as other major international studies, other scientists can now use this CORE data to test their own "recipes."
- The "Double-Check": By having both the raw photos and the measured photos, they allow experts to verify if a computer model is actually reading the images correctly.
- The Limitation: The paper admits this kitchen is in one specific city (New Delhi). It's like testing a recipe in a city kitchen; it might not work perfectly in a rural village kitchen. So, the authors say this data is most valuable when combined with data from other similar "kitchens" to create a bigger, more reliable picture.
How to Use This Library
The researchers are inviting other scientists to come in and use this data. However, because these are real people's private medical records, you can't just walk in and grab the files. You have to apply for permission through a "Data Access Committee." If your idea is good and ethical, they will let you use the data to see if your prediction model works on this Indian population.
In short: This paper is an invitation to other scientists to stop guessing if their medical prediction tools work everywhere. They can now bring their tools to the CORE kitchen in India, run them against this high-quality data, and see if they actually work before trying to use them on real patients.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.