A team in Australia has developed a digital tool designed to detect serious side effects from cancer immunotherapy faster and more accurately. The system is intended to help clinicians identify patients with major complications without relying on slow manual case reviews.
The innovation comes from Peter MacCallum Cancer Centre, or Peter Mac, in Melbourne. Its main focus is immune-related colitis, an inflammation of the bowel that can appear in patients treated with immune checkpoint inhibitor immunotherapy.
Why immune-related colitis matters
Peter Mac says immune-related colitis can affect up to 50% of patients receiving this type of immunotherapy. That makes early detection especially important as immunotherapy becomes more widely used in modern cancer care.
Fast identification can help medical teams understand who is affected and adjust treatment responses before complications worsen. In this setting, a digital approach can reduce the burden of sorting through large volumes of patient records one by one.
How the algorithm works
The centre describes the tool as a clinically validated “digital phenotype” that uses electronic medical record data to identify patients with immune-related colitis. Because it works with data already stored in hospital systems, the method can streamline case finding without depending entirely on manual chart review.
Lead researcher Jasmine Teng, an infectious diseases physician at Peter Mac, said the tool can replace manual review, which has traditionally taken significant time. Teng said the digital method is faster, scalable, and still able to identify affected patients with high accuracy.
That makes the system useful not only for routine clinical work, but also for larger-scale research. A faster and more consistent way to identify cases can give researchers a stronger foundation for studying treatment complications and patient outcomes.
Clinical and research value
Teng said improved identification at scale could open the door to new clinical insights that were previously difficult to obtain. By gathering a larger and more consistent view of cases, researchers may be able to study patterns in side effects and how patients respond to management more effectively.
The tool could also support biomarker discovery. Teng said identifying biomarkers that predict who will develop immune-related colitis may help patients and treatment teams tailor immunotherapy settings and refine early side-effect management.
That kind of personalized approach is increasingly relevant as immunotherapy continues to play a major role across multiple cancer types. The challenge is not only delivering the treatment, but also spotting serious adverse effects early enough to act on them.
What it could mean for hospitals
Peter Mac positions the tool as a direct use of electronic health records for a real clinical need. As health data volumes continue to grow, algorithms that help doctors understand patient risk are becoming more important.
The ability to duplicate the method is another key advantage. If the same approach can be adopted more widely, other hospitals and cancer centres may be able to use similar systems to detect immunotherapy complications.
For cancer patients receiving immunotherapy, faster recognition of immune-related colitis can be a meaningful step in managing serious side effects. The development shows how existing hospital data can be turned into a practical decision-support tool for care teams and researchers alike.
