Artificial intelligence (AI) is coming to transform every industry, but advanced data usage in Canada’s health-care system lags well behind other major industries.
However, we still might be sitting on the cusp of a wave of AI-driven health care innovation that could make life simpler for patients and clinicians alike, all while improving overall health care outcomes across the country.
“At a national level, we really have a unique opportunity to apply AI research and solutions to modernize health care, address the challenges we’re facing in our health-care system, and improve the overall quality of care that we provide for patients,” said Azra Dhalla, director of AI implementation at the Vector Institute, a non-profit corporation dedicated to research in the field of AI.
Dhalla works with stakeholders across academia in hospitals and public health agencies on the responsible deployment of AI solutions in clinical environments.
What does AI-driven health care innovation look like?
Dhalla says there are three particular areas worth noting.
- “The first is personalized or precision medicine. With the use of AI, it will enable easier and earlier detection of patient health changes and also prediction of disease. Through predictive analytics, we’re able to speed up the diagnosis and decision making capabilities. And that also increases the amount of time physicians get to spend with patients.”
- “The second is increasing health system efficiencies to target resources more efficiently, which improves both system performance and patient outcomes.”
- “The third is in the area of drug discovery where we can use AI to analyze data and find ways to use existing drugs to treat conditions the drug may not be currently prescribed for, like existing and emerging viruses.”
Here are a few examples of how AI is improving health care:
- IBM’s Watson for Health is being used to significantly cut the time to market and cost of new drugs.
- AI is being used for early detection of diseases like Cancer and Alzheimer’s.
- AI-driven predictive analytics are empowering clinicians’ decision making and prioritization efforts.
AI may not ever replace doctors, but the fact that it’s even a possibility shows how much change might be coming.
“I’m excited about what’s coming with AI,” said Mary Jane Dykeman, managing partner at INQ Law, a Toronto health and data law firm. “But there’s the excitement and then there’s rolling up your sleeves and getting to work and getting it done.”
But what does “getting it done” mean, exactly? What are the issues preventing us from realizing AI’s full potential in our health-care system? Or even the benefits of plain old garden-variety digital transformation?
It’s a long list of issues that need to be sorted through – ranging from education to data privacy, data security, and beyond. But it all starts with actually making our health care data accessible. That’s step one – and it’s a big one.
Before you can leverage AI in health care, you have to make the data available
One of the main reasons that the Canadian health-care system hasn’t been digitally transformed is the complexity and location of its data.
Every health-care organization – from clinic to hospital to regional health authority – has massive volumes of data. And that data may be housed in a variety of systems. Some of it is still on paper, some of it is duplicated between paper and computers. Every region is a bit different. So, while there are other issues with digital transformation in health care, data accessibility is the starting point. And as part of that, we need to think about both the patients and the health-care providers that will be accessing that data. It’s a two-sided equation.
“We need to clean up the data so it’s usable and meaningful,” said Dykeman. “And we need to ground it in the patient experience if we’re going to transform the health system across Canada.”
Will Canadians approve of their health care data being used in AI?
Canadians are increasingly aware of the frequency of cyber security issues. Think of the well-publicized recent data breach that exposed the health data of mothers, newborn children and parents in Ontario, and the cyberattack on national book retailer Indigo.
The use of AI in health care raises questions for Canadians about how their personal data will be used and kept safe. How comfortable will the average Canadian – especially seniors that make more frequent use of the health-care system – be with their data being used by AI?
Privacy and ethics considerations loom large here.
“Privacy rules have been in place in Canada for many years,” said Dykeman. “Now many governments are modernizing their privacy legislation, because suddenly we have shared systems and electronic records and digital opportunities – and legislation needs to reflect that. Privacy is the bedrock though. It’s not a one-and-done thing. It is a constant commitment. And if we get it right, it opens all kinds of doors for us.”
Dhalla expands on that:
“When it comes to diagnostics, for example, the greater the number of X-Ray images we show an AI and the more diverse the data we show it, the more accurate its diagnosis and predictions will be – so the benefits from accessing this vast amount of data for patients and providers can’t be overstated,” she said.
“But there’s a stringent process as it relates to privacy with respect to health data. And rightly so. As AI becomes progressively more ingrained in our health services, there’s a need to build regulatory frameworks across the industry and governments. In fact, regulators and medical organizations are already developing guardrails to address these issues.”
So, privacy and policy are critical pillars.
But they’re not the only ones.
Canada needs to focus on (digitally) improving the patient & clinician experience
The bottom line is that digital transformation – and AI use – is expanding almost daily. And health care for a child born today will look very different when they’re an adult than it does today. In a good way.
That’s why Dykeman believes we all need to rethink the prevailing narrative around AI, data, privacy, and health care. She believes we need to tell a more compelling, positive story. As she noted to Vog App Developers:
“The public deserves transparency about their data – both the negative and positive stories. They don’t know what is possible, because all they hear is the negative, about the last big data breach. Patient-centric design includes them and can bring to them the same excitement we have about the tremendous opportunities to advance our health system with data. Because these transformations will ultimately help them, and if not them personally, others around them.”
There are two key areas where change management will be key:
Health-care workers
Dykeman: “If I’ve been working in health care for years, and I’m quite used to the way I do things, and someone comes along and tells me they’re going to change everything, that’s challenging. But if that change will lead to improvements for patients and clinicians and the family members and caregivers that accompany the patient through the system, that’s different. I can understand that and embrace it.”
Dhalla: “At Vector, we’re helping to work with health-care leaders and clinicians to really change the narrative from fear about AI to how it can really help them augment care. We’re supporting them through this change, providing them with opportunities for learning, knowledge translation, and upskilling.”
Patients
Dykeman: “We need to make it so much easier for patients, because there’s a long legacy of them having to repeat the same information at every step of the way in the health care journey. Or perhaps they show up for an appointment and some aspect of their information is lost, and they have to ask them the same questions again and again. That can be changed. It’s also worth noting that a patient may not even be their own best heath historian – depending on the nature of their situation or condition.”
Dhalla: “It’s important we communicate to patients that it’s de-identified data we’re using, so patients know that we don’t actually have access to, for example, their names. In fact, what we’re doing is building more generalizable models that can be used across patient populations. These types of conversations can help alleviate some of the concerns patients are having.”
Building our digital and AI health care future in clinics and medical schools
There’s a whole new generation of health-care workers who will be the tip of the spear when it comes to digital transformation. Universities are recognizing the potential and need to prepare their students for this data-based future by starting to change curriculums. Ultimately, someone studying to be a doctor or nurse in 2030 could find AI as common a tool as a stethoscope. Such is the rate of change.
But Dykeman also believes that big change must start at a smaller level.
“Organizations need to ask themselves: ‘what can we do with this data? What are some of the pain points that patients and clinicians experience each day? What are the use cases that we can develop?’” she said.
“I think there’s a great opportunity to crowdsource ideas even within individual organizations. Ask the people who are inside them. What are the little problems you’d like to solve? What are the big problems you’d like to solve? How do we get there? We have big audacious goals, but small movements will push them forward piece-by-piece. System transformation is not one big thing. It’s a series of little things.”
