emjreviews.com
AI Model Predicts Stoma Risk After Crohn’s Disease Surgery
A predictive model for temporary stoma could influence surgical planning for people with Crohn’s disease, helping tailor decisions and expectations around intestinal resection.
Surgeons and clinicians who perform intestinal resections, researchers working on surgical risk prediction or AI tools, and patients facing Crohn’s surgery who want information about factors that influence stoma formation.
What To Know
What to know This news item reports on a published study that developed machine-learning models to predict the need for a temporary stoma after intestinal resection for Crohn’s disease. Researchers analysed a cohort of 252 patients who had intestinal resection; about 150 required a temporary stoma.
Several algorithms (logistic regression, random forest, XG-Boost) were tested and the random forest model performed best in the training set and moderately in validation. The study used SHAP to rank predictor importance.
Why it matters An accurate predictive tool could help surgeons personalise decisions about whether to form a temporary stoma at the time of surgery, potentially reducing unnecessary stomas or better preparing patients who are likely to need one. More details and caveats The report is based on a single published cohort (BMC Gastroenterology).
Reported model performance was high in training but lower in validation, which is a common sign that more external validation is needed before clinical use. This is an observational/technical study about prediction, not a clinical trial showing improved outcomes from using the model.
Early-stage predictive models often require external validation across other hospitals and populations before being used to change clinical practice. The news summarizes a cohort study using machine learning and SHAP explanations; it does not report prospective testing of the tool in routine care.
