Researchers created a computer program that can predict which children born with stomach and intestinal problems are at highest risk of serious complications within 30 days of surgery or hospital admission. Using information from nearly 4,000 children across 74 countries, the program correctly identified at-risk patients about 89% of the time. The tool looks at factors like whether complications developed, how long antibiotics were needed, and the child’s weight and health status. This breakthrough could help doctors give extra attention to the most vulnerable patients and potentially save lives.

The Quick Take

  • What they studied: Can a computer program predict which children with birth defects of the stomach and intestines are most likely to have serious problems or die within 30 days of surgery?
  • Who participated: Nearly 4,000 children with gastrointestinal birth defects from hospitals in 74 different countries around the world
  • Key finding: The computer program correctly identified which children were at highest risk about 89 times out of 100, and was especially good at spotting the most vulnerable patients (correct 90 out of 100 times)
  • What it means for you: If your child has a gastrointestinal birth defect, doctors may soon be able to use this tool to identify early which children need the most careful monitoring and aggressive treatment, potentially improving outcomes. However, this tool is still new and needs to be tested more before it’s used in regular hospital care.

The Research Details

Researchers gathered medical information from children with gastrointestinal birth defects who were treated at hospitals in 74 countries. They collected data on nearly 4,000 patients and used this information to teach a computer program (called a random forest classifier) to recognize patterns that predict which children would have serious problems within 30 days. The computer program learned by looking at many different factors about each child, including their diagnosis, weight, health status before surgery, what treatments they received, and whether complications developed. The researchers then tested how well the program could predict outcomes by checking its predictions against what actually happened to the children.

This research approach is important because it uses real-world data from many countries and hospitals, making the results more likely to work in different settings. The computer program can process many pieces of information at once—something that would be difficult for doctors to do manually. By identifying high-risk children early, doctors can provide extra monitoring and treatment, which may prevent serious complications.

This study is strong because it includes a very large number of patients (3,849) from many different countries and hospitals, which means the results are more likely to apply to children everywhere. The researchers used proper statistical methods to make sure their computer program was tested fairly. However, the study is based on data that was already collected for other purposes, so some information might be missing. The program still needs to be tested in real hospitals with new patients before doctors start using it regularly.

What the Results Show

The computer program was very accurate at predicting which children would have serious problems. It correctly identified at-risk children about 89 times out of 100. When the program said a child was at high risk, it was right about 84 times out of 100. Most importantly, when the program identified a child as high-risk, it caught the truly at-risk children about 90 times out of 100, meaning it didn’t miss many vulnerable patients.

The program identified six main factors that were most important for predicting risk: whether the child developed complications after surgery, how long antibiotics were needed after surgery, whether the child needed special nutrition through an IV or a breathing machine, the child’s overall health status before surgery, the child’s weight when admitted to the hospital, and whether doctors used a surgical safety checklist.

About 1 in 5 children in the study (19.5%) had serious problems or died within 30 days. The computer program was especially good at identifying these high-risk children, which is important because these are the patients who need the most attention.

The research also found that other factors mattered, including which continent the child lived on, the family’s income level, and specific details about the type of birth defect. The use of a surgical safety checklist—a simple tool where doctors check off important steps—was one of the most important protective factors, suggesting that following standard safety procedures really does help.

This is the first study to use a computer program to predict outcomes for children with gastrointestinal birth defects using global data. Previous research has used computer programs to predict outcomes in other types of surgery and illness, and those programs have been helpful. This study shows that the same approach works well for children with stomach and intestinal birth defects, which is a new and important finding.

The study has some limitations to keep in mind. Some medical information was missing from some patients’ records, and the researchers had to estimate what that information might have been. The study looked at children who had already been treated, so we don’t know yet if the program will work as well when doctors use it to predict outcomes for new patients in the future. The program was developed using data from hospitals that may have better resources than some hospitals in poorer countries, so it might not work equally well everywhere. Finally, the program is a tool to help doctors make decisions, not a replacement for doctors’ judgment and experience.

The Bottom Line

This research suggests that using a computer program to identify high-risk children with gastrointestinal birth defects is a promising approach (moderate to high confidence). However, the program is not ready for regular use in hospitals yet. Doctors and hospitals should wait for additional testing in real-world settings before adopting this tool. When it is ready, it should be used as a helper tool alongside doctors’ judgment, not as a replacement for it.

This research is most relevant to pediatric surgeons, hospital administrators, and families of children with gastrointestinal birth defects. Parents whose children have these conditions should know that new tools are being developed to help doctors predict and prevent complications. Doctors in hospitals around the world, especially in lower-income countries, should pay attention to this research because it could help them identify which children need the most careful care.

If this tool is adopted by hospitals, doctors would be able to identify high-risk children within hours or days of admission or surgery. The predictions are specifically for the first 30 days after hospital admission or surgery, so benefits or concerns would become apparent within that timeframe.

Want to Apply This Research?

  • For families of children with gastrointestinal birth defects, track daily observations during the first 30 days after surgery: temperature, feeding tolerance, wound appearance, antibiotic administration dates, and any signs of complications (fever, vomiting, swelling, or changes in behavior).
  • Parents and caregivers should ensure all recommended safety protocols are followed, including use of surgical checklists, timely antibiotic administration, and adherence to feeding and nutrition plans. Document these steps in the app to maintain a clear record for the medical team.
  • Create a 30-day post-surgery monitoring checklist in the app that tracks key risk factors: complications, antibiotic duration, nutrition support needs, and overall health status. Share this data with the medical team to support clinical decision-making and ensure high-risk children receive appropriate attention.

This research describes a new computer tool for predicting health risks in children with gastrointestinal birth defects. This tool is not yet approved for regular clinical use and should not be used to make medical decisions without consultation with your child’s doctor. The predictions are based on statistical patterns and cannot guarantee individual outcomes. Always follow your pediatric surgeon’s recommendations for your child’s care. If you have concerns about your child’s recovery after surgery, contact your medical team immediately rather than relying on any prediction tool.