Designed in partnership with Hamilton Health Sciences catheterization lab and using data from tens of thousands of patients seen in Hamilton and Niagara over twelve years, CREATE (CentRe for digital hEalth and dATa sciEnce) has developed a highly accurate AI algorithm that can save health care system millions of dollars every year while improving outcomes for patients.
Cardiovascular disease is the leading cause of death worldwide. Invasive coronary angiography has long been the gold-standard test used to accurately diagnose blockages in the arteries supplying the heart muscle. However, more than 50 percent of patients referred for this test are being exposed to risk of stroke or death without getting the full benefit.
As a first step to addressing this challenge, the CarDIA trial was created by a team led by Dr. Jon-David Schwalm, an interventional cardiologist at Hamilton Health Sciences. The CarDIA study evaluated a clinical pathway that diverts some patients referred to invasive angiogram to a lower cost, lower risk, and highly accurate cardiac CT test. The CarDIA trial showed that they could reduce the number of unnecessary invasive angiograms by 76 percent, which both save money and decrease risk.
“The hardest part of the study is that we used an archaic system to pull out the right patients for CCTA. It was a labour-intensive process that involved manual screening of referrals by the triage staff and cardiologists. That is not ideal in the real world.” – Dr. JD Schwalm
To address these challenges, Dr. Schwalm partnered with Dr. Jeremy Petch and his CREATE team to develop an AI model using more than 10 years of data on patients undergoing elective invasive angiogram. The AI model was better able to predict which patients would most benefit from invasive angiogram than any existing clinical prediction algorithms. Detailed results were recently published in the peer-reviewed literature.
This research program has received the attention of the Auditor General’s report released in December 2021: Value for Money Audit on Cardiac Disease and Stroke Treatment. The report specifically calls out the CarDIA study as Ontario Health’s primary strategy for optimizing the selection of patients for invasive angiogram.
We are now working with all of Ontario’s data to develop a generalizable and externally validated model. We are simultaneously partnering with Ontario Health to co-design a clinical decision support tool for use at the point of care to support triage at all of Ontario’s cardiac catheterization labs.
“A decision support tool based on AI machine learning will help reduce the number of unnecessary angiograms, reduce health system costs and help ensure that patients receive the best test based on their risk for coronary artery disease.” – Dr. JD Schwalm
Students opportunities include prospective evaluation of the AI model, as well as interoperability and human factors work related to the clinical decision support tool.