Machine Learning Predicts Early Mortality in IBD Patients

A groundbreaking study published in the Canadian Medical Association Journal (CMAJ) has found that nearly half of individuals with inflammatory bowel disease (IBD) who died between 2010 and 2020 passed away prematurely. Researchers applied machine learning models to healthcare data to predict early mortality in IBD patients, offering insights that could help enhance patient care and prevent avoidable deaths.

High IBD Rates and Increased Mortality Risk

Canada has one of the highest IBD rates worldwide, including Crohn’s disease and ulcerative colitis. Patients with IBD tend to have a shorter life expectancy than those without the disease, often developing other chronic conditions that further increase their risk of premature death—defined as passing away before the age of 75.

The study found that individuals diagnosed with chronic conditions before age 60 were at an even higher risk. The most common conditions at the time of death included:

• Arthritis (77%)
• Hypertension (73%)
• Mood disorders (69%)
• Kidney failure (50%)
• Cancer (46%)

Machine Learning Enhancing Mortality Predictions

Researchers from ICES, SickKids, and the University of Toronto explored whether machine learning—which has successfully predicted early death in the general population—could also be used to assess mortality risk in IBD patients. The study confirmed that including chronic conditions diagnosed before age 60 improved predictive accuracy, allowing for earlier intervention.

“The clinical implication is that chronic conditions developed early in life may be more important in determining a patient’s health trajectory,” said Dr. Eric Benchimol, senior scientist and professor at the University of Toronto.

A Call for Integrated Healthcare Strategies

With 47% of IBD-related deaths classified as premature, experts stress the importance of multidisciplinary care involving gastroenterologists, mental health professionals, dietitians, and other specialists. The study’s findings emphasize the need for early and effective treatment strategies to prevent avoidable deaths.

By leveraging machine learning and predictive analytics, healthcare providers can better identify high-risk patients and develop targeted interventions, improving long-term outcomes for those living with IBD.

For more information: Canadian Medical Association Journal