Canada’s approach to productivity: longevity and patient care over innovation

Economist Impact, the business intelligence arm of the Economist magazine, recently conducted an international study of productivity in the public sector, including healthcare. Canadian organizations are generously represented, with half of the Canadian respondents coming from the healthcare sector.

By a significant margin over their worldwide peers, Canadian organizations see budget constraints, lack of public trust, and labour shortages as barriers to productivity.

Fifty-six percent of Canadian respondents felt budget constraints were a concern, compared to 48 percent globally and 47 percent in the U.S. Lack of public trust was cited by 45 percent of Canadian respondents (versus 38 percent globally and in the U.S.), while labour shortages were named by 23 percent of Canadians (compared with 17 percent globally and 12 percent in the U.S.).

Peeling the layers of productivity
The Canadian federal government has weighed in on the healthcare labour shortage, allotting $14.3 million in March to help healthcare workers with foreign accreditation enter the workforce, with some of that funding aimed specifically at women with international training.

The funding will “help grow capacity in our healthcare system and give relief to workers already providing care to Canadians,” former minister of health Mark Holland said earlier this year.

“We’re short of staff in all levels in healthcare,” said Dr. Patrick Rogalla, radiologist at Toronto’s University Health Network in the Joint Department of Medical Imaging, where he also serves as the division head of cardiovascular and thoracic imaging and site director of medical imaging at the Toronto General Hospital.

Operationally, he said, imaging productivity tends to be narrowed to throughput — how many patients can we scan per time slot? — but it can look very different depending on the strategies and goals of the institution.

Through the lens of quality, the picture might look different because less interaction with the patient might improve throughput, but it can also increase the need for recalls and additional testing. And at the strategic level, there’s “a conglomerate of resources” — human, environmental, technical, financial — that are interwoven to maximize big picture efficiency, he said.

At the operational level, technology offers new opportunities.

“It can open up new treatment plans, new possibilities of treating disease that we never had before,” Dr. Rogalla said. The automation that replaces strenuous, repetitive tasks doesn’t just improve throughput. It’s also a safety net.

“We eliminate downtime. We eliminate problems from fatigue. We may eliminate unexplained variability — a very important term in high-reliability organizations.”

Unexplained variabilities can lead to waste and even “catastrophic events,” he said.

Augmented intelligence
Lorne Rothman, industry specialist and national healthcare lead at data and AI firm SAS Canada, said there is a preference for “augmented intelligence” — the use of artificial intelligence to support, not replace, decision-making — among healthcare users. Large language models (LLMs) and natural language processing (NLP) are used to retrieve and standardize information in medical records and laboratory results.

“Relying solely on humans to pore over large volumes of textual information is incredibly time consuming. The problem is compounded by variation in how medical conditions, and drug treatments, dosages, and measurement units are presented,” he said. “Applications that employ powerful and very specific and contextual rules sets are used to automatically retrieve information and standardize document content.”

Resource demand forecasting, for example, can help predict hospital admissions, hospital and ICU bed usage, and impacts of infectious disease outbreaks on resource usage.

“During the Covid-19 pandemic we developed and applied infectious disease models to help forecast the local progression of infections, and to estimate virus growth rates as well as the impact of pandemic policy changes on disease spread,” Rothman said.

Deep learning computer vision applications are used to speed diagnostic imaging and help prioritize patients for care. While there remains a cautious approach towards innovation in Canada, Rothman said it’s often in service of prioritizing patients.

“Proactive improvements require time, planning and resources, but budget limitations in our publicly funded system push priorities toward patient care over technological innovation.”

Gaining trust in AI
When it comes to the pursuit of increased productivity, there remains the thorny issue of public trust in technology, systems, and the government itself.

“In Europe, the trust is more in the government and less into private sector,” said Dr. Rogalla. “In the U.S., the trust is more in the private sector, and not at all the government. I think in Canada, we’re in between. And that is a typical Canadian position.

“We don’t want to the be first movers. We want to be the first survivors.”

AI applications are particularly suspect in the public eye, said Rothman.

“Lately, many people assume that AI is synonymous with LLMs and generative AI. Many are particularly wary of its use in health,” he said.

“Let’s prioritize low-risk AI applications, boost public and front-line understanding of AI, and promote augmented rather than artificial intelligence,” Rothman suggests.

In terms of infrastructure, survey respondents from Canada said they feel like they’re ahead in digitizing data and having a comprehensive data governance framework. They’re also confident in the quality and relevance of their data, and on the human resources front, Canada is more generous with training opportunities and encourage upskilling existing employees.