How many Canadians die each year from medical mistakes? 

In about 1 in 17 hospital stays, the patient experiences something unintended, such as a medical mistake.¹ 

Most of these mistakes happen in the gaps: a warning sign buried in a chart, a lab result no one connected, a handoff at shift change.

That's where AI is starting to help.

What AI is already preventing

Fewer unexpected deaths. At St. Michael's Hospital in Toronto, an AI early warning system called CHARTWatch alerts care teams when a patient is quietly getting worse. Unexpected deaths dropped by 26%.³

Faster sepsis care. At five U.S. hospitals, an AI sepsis alert reduced deaths by 18.7% when clinicians acted on it within three hours.⁴

More cancers caught. In a Swedish trial of more than 105,000 women, AI-supported mammography found 29% more breast cancers without more false alarms.⁵

How many mistakes can AI prevent?

There's no single number yet. Early results show real reductions, but only when someone acts on the alert – and not every tool works. One widely used sepsis system caught only about one in three cases.⁴ AI can spot the warning sign. People still have to respond.

Still, the direction is clear: AI will form the basis of the next-generation operating system for Canada's advanced healthcare system.  

What to do next

AI isn't available everywhere, and it can't follow you between providers. If you or a family member are facing a complex diagnosis, a hospital stay or a treatment plan that doesn't feel right, RCM Health can review your care and make sure nothing is missed. A medical second opinion can confirm whether the plan is the right one.

Or call 647-350-5500 to speak with our team

Raymond Rupert | CEO

RCM Health Consultancy Inc.

647-350-5500 | info@rcmhealth.ca | rcmhealth.ca

References

1. RiskAnalytica. The Case for Investing in Patient Safety in Canada. Canadian Patient Safety Institute, 2017.
2. Patient harm in Canadian hospitals? It does happen. Canadian Institute for Health Information and Healthcare Excellence Canada, 2024.
3. Verma AA, et al. Clinical evaluation of a machine learning–based early warning system for patient deterioration. CMAJ, 2024.
4. Adams R, Henry KE, Sridharan A, et al. Prospective, multi-site study of patient outcomes after implementation of the TREWS machine learning-based early warning system for sepsis. Nature Medicine, 28, 1455–1460, 2022.
5. Screening performance and characteristics of breast cancer detected in the Mammography Screening with Artificial Intelligence trial (MASAI). The Lancet Digital Health, 2025.

This newsletter is for general information only and is not a substitute for individualized medical advice, diagnosis or treatment. Decisions about your care should be made in consultation with your treating physician.