Health Canada’s controversial ‘regulatory sandbox’: Enabling innovation or lowering the bar for safety?

The Canadian government is moving forward with a controversial plan to regulate how cutting-edge health products are approved for use. These products are “so novel, complex, and distinct that current regulations are not equipped to handle them,” Health Canada says.

Its solution, a “regulatory sandbox,” could apply to products ranging from 3D-printed organs to algorithms dispensing decision-making support in hospitals and cell therapies tailored to individual patients.

“This is a complete departure from the way regulation typically works,” says Matthew Herder, who examined the policy in his role as director of Dalhousie University’s Health Law Institute. Traditionally, a product moves through the approval process through various phases of a clinical trial, with standards for safety and effectiveness set in advance.

That’s not possible with these products, says Celia Lourenco, director general of Health Canada’s Biologics and Genetic Therapies Directorate, who helped craft the new policy. Like children in a sandbox – improvising and using the tools at hand – a company can work with Health Canada to decide the standards as it goes rather than being restricted by predetermined rules. For example, a biotech company making 3D-printed ligaments would consult with Health Canada to devise a one-of-a-kind plan allowing the company to sell and market the product while gathering data about outcomes in real time.

This “concierge service,” as Health Canada puts it, gives patients faster access to products while giving companies a quicker path to financial return on their investment.

“The idea is, we’re creating a ‘safe environment’” to test a product, says Herder. He pauses before continuing: “The use of that phrase is kind of ironic.”

The potential consequences of getting the regulatory regime wrong are wide-ranging, from mere annoyance to decisions about life and death. One consequence might be purely economic: an algorithm doesn’t offer meaningful insights to improve patient health, but the product stays on the market anyway, with individuals continuing to pay out-of-pocket despite the lack of benefit. But sometimes the consequence can be more dire, such as a recent case in the United States in which an algorithm intended to predict which patients would develop sepsis failed in more than 200 hospitals. 

Lourenco says it’s reasonable to ask about safety and equity, but she’s “very confident” about the system’s ability to protect patient health.

“It’s definitely not a pathway to lower the requirements in any way, or to lower the bar,” she says. “It’s really about tailoring that bar to make sure we have the right evidence on safety, efficacy and quality while not stifling innovation – enabling these important products to enter the market.”

Like children in a sandbox – improvising and using the tools at hand – a company can work with Health Canada to decide the standards as it goes.

By implication, there are risks to maintaining a rigid regulatory system for these products: it could deprive Canadians of the latest therapies while letting the country slip behind as a scientific and economic leader on the global stage.

“Certainly, one of the goals is making sure Canada stays at the forefront of innovation from the get-go,” Lourenco says. “As regulators, we need to have a regulatory framework that enables innovation so these products can be developed in Canada and stay in Canada.”

Too often, Lourenco says, breakthroughs originating in Canada leave the country before being brought to market. The sandbox – created after consultation with biotech start-ups, venture capitalists and academic innovation hubs – is intended to reverse that pattern by creating an attractive environment to keep world-class innovations at home.

The crux of the plan comes down to the prioritization of innovation. For the first time, Health Canada will be working in collaboration with industry to facilitate the approval of products central to patient health. For Herder, this raises alarm bells. For Lourenco, the cause for alarm is the opposite: namely, keeping a more formal system in place while regulators in other countries adjust their practices to the demands of new technology.

Indeed, this past week Health Canada joined with regulators in two other countries – the United States and United Kingdom – to issue 10 “guiding principles” on AI in health care. These principles are intended to “help cultivate future growth in this rapidly progressing field” by signalling the type of information regulators need when assessing products for approval.

The three-country partnership exemplifies the regulatory environment that gave rise to the sandbox, with Canada seeking to attract and retain academics and entrepreneurs who drive innovation forward. Such partnerships are “crucial,” the joint statement says, “if we are to empower stakeholders to advance responsible innovations in this area.

Those involved in the field have been calling for such guidelines. Dr. Andrew Pinto, director of the Upstream Lab at St. Michael’s Hospital, recognizes the promise of AI, but cautions, “There needs to be some standards that guide the ethical application [of algorithms], and also the prevention of harm.” Health Canada’s new guidelines attempt to address those concerns while maintaining a flexible framework. No one yet knows how the process will play out when implemented in a health-care setting.

The first pilot project in the sandbox will be launched in the coming months. Asked whether he has confidence in Health Canada’s ability to strike the right balance between innovation and patient health, Herder laughs, then carefully constructs his response. “That depends on who’s driving this process: Is it going to be Health Canada – and are they going to staff it with scientific expertise? Or is it going to continue to be informed by the industry they’re trying to regulate?

“I think that’s the question we need to keep asking.”