Why Canada Was So Late to Meet the Pandemic Risk

It’s almost a truism in public health that you need surveillance long before you think you need surveillance. That is, you have to keep watching for early reports of undiagnosed cases of something nasty turning up, somewhere in the world. If you wait until such cases start filling up your own emergency department, you’re already in big trouble.

Surveillance is part of what’s called “epidemic intelligence” — picking up some bit of information, linking it to some other bit and deciding if it means a very unpleasant third bit is on its way. Health agencies around the world rely on epidemic intelligence to alert them to potential trouble. As we’ve seen in the COVID-19 pandemic, those who waste time lose lives.

When the pandemic appeared on Dec. 30, 2019, our own intelligence spotted it — but was slow to tell us how serious it might be. A new report by an independent panel tries to explain that failure and to offer solutions. But the report itself accidentally tells us the failure is far deeper than a single bureaucratic program.

Canada’s surveillance system is based on the Global Public Health Intelligence Network, or GPHIN, originally set up in 1994 after a local plague outbreak in India and greatly expanded in 2004 after SARS. As the early web rapidly grew, GPHIN found it could draw on countless new online sources of information.

GPHIN had successes. It spotted the onset of the H1N1 pandemic in 2009, as well as other outbreaks like Middle East Respiratory Syndrome and Ebola virus. But like all government agencies, GPHIN is a bureaucracy linked to still other bureaucracies, all at the mercy of their current political masters.

Over the years from the founding of the Public Health Agency of Canada by the Paul Martin Liberals in 2004, GPHIN was tweaked (and cut) by the Stephen Harper Conservatives, and cut again in 2019 by the Justin Trudeau Liberals — who cancelled its international surveillance role and limited it to domestic outbreaks. A chief health surveillance officer position went unfilled for years and was scheduled to be dropped.

GPHIN sent its international readers routine daily reports and sporadic alerts about potentially serious outbreaks: almost 900 about H1N1 in 2009, almost 200 about H7N9 avian flu in 2013. In the past ten years, the report says, alerts ranged from 21 to 90 a year. In 2019, just one alert went out, in May.

Who paused the alerts?

“It is clear,” the independent panel said in its interim report, “that some form of direction was given to pause the alert process, and that the level of approval for alerts was elevated and then downgraded at least once. The panel has not seen any written documentation in respect to the timeline of those changes, who requested them and why they might have occurred.”

The final report tells us nothing more about this mystery. Some master bureaucrats seem to have covered their tracks very well.

Much of this 82-page report deals with technical challenges of modern surveillance, using not only humans but artificial intelligence to spot and evaluate key signals in the endless flood of information. (I was relieved to know that AI is smart enough to ignore reports of “Bieber fever.”)

Surveillance comes in two forms. Indicator-based surveillance, or IBS, simply means monitoring formal sources like health agencies and governments for their official pronouncements. They’re very reliable, but can take days or weeks to appear — and they’re not much good when countries lack the resources or desire to monitor themselves. The report says PHAC’s IBS is mostly domestic, checking provincial and municipal health departments.

Events-based surveillance, or EBS, by contrast, tracks media, including social media — and health experts themselves have events-based websites like Pro-MED mail. GPHIN itself spends almost a million dollars a year for access to Factiva, a commercial surveillance service.

EBS is fast but unreliable, and requires prompt, well-informed risk evaluation to spot the real threats amid the rumours and fake news.

The panel says GPHIN picked up the first report of a mysterious pneumonia in Wuhan at 10:30 p.m. EST on Dec. 30, 2019. It came from an Agence France Press news story in the South China Morning Post, a good English-language paper in Hong Kong but not an official source of anything. I woke up the next morning to find my fellow-bloggers and tweeters fully aware of the outbreak as confirmed by another source, the Hong Kong Centre for Health Protection.

The report says that GPHIN issued a daily report on Dec. 31, and that “both PHAC’s president, Tina Namiesnowski, and its chief public health officer, Dr. Theresa Tam, took action upon receipt of the Jan. 1 special report from GPHIN’s management shortly after 9 a.m.”

That suggests someone made a persuasive risk assessment, judging the Wuhan outbreak as a potential threat; yet the report’s key recommendation is that GPHIN develop a real risk assessment capacity itself, instead of just passing along interesting news items.

This in itself is a surprising lapse. Risk assessment is built into most health agencies. The European Centre for Disease Prevention and Control, for example, publishes frequent threat assessments on COVID-19 and many other diseases. The panel notes that PHAC has many kinds of risk assessment: “some program areas have robust risk assessment procedures and investigate public health events regularly, whereas other may not have as much capacity or exposure.” GPHIN, explicitly, “does not undertake risk assessment: it does not develop or recommend specific actions or policy response.”

The panel, after taking a shot at PHAC about “the clear consequences of letting overall co-ordination and governance tools erode and weaken,” recommends “a dedicated risk assessment office,” perhaps under the chief public health officer.

When the diagnosis is hazardous to the doctor’s health

Such an office seems painfully obvious, but on reflection, its absence is understandable. A serious risk assessment is a politically fraught statement, in part because it can be a major economic blow to the country with the outbreak.

According to Alex de Waal in his book New Pandemics, Old Politics, during the 2004 SARS outbreak the then director-general of the World Health Organization, Gro Harlem Brundtland, issued an advisory against travel to China. The Chinese were shocked and furious. Then she shocked and infuriated us by issuing another travel advisory for Toronto. The economic consequences for both countries were unpleasant.

Brundtland could do it only because she was near the end of her term, and since then WHO’s director-generals have been much more tactful in alerting the world.

Risk assessment is also at the mercy of the quality of the information, which early in an outbreak is inevitably poor. ECDC’s first threat assessment of COVID-19 on Jan. 9, 2020, is woefully cautious in hindsight, but the best that could be done at the time:

“... given that there is no indication of human-to-human transmission, the risk to travellers is considered to be low. For the same reason, and since no cases have been detected outside of Wuhan, the likelihood of introduction to the EU is considered to be low, but cannot be excluded. Consequently, the risk of further spread within the EU should a case be identified is considered low to very low.

“Risk of nosocomial [hospital-based] transmission, infection prevention and control: So far, no human-to-human transmission or spread to health-care workers or medical personnel has been reported in China. Therefore, the likelihood of nosocomial transmission is low.”

So a risk assessment office could be a very dangerous place to work. Miss a serious threat, or downplay it, and you’re the scapegoat. Scare the government into action over a false alarm, and you’re the scapegoat. And if you call it right, you put intense pressure on the government to do things it really, really doesn’t want to do.

Communication as ‘product design’

Many of the panel’s recommendations deal with bureaucratic processes for improving and upgrading GPHIN, offering the hope of years of meetings and interim reports before anything actually happens (including the staffing of a risk assessment office). But one suggestion actually amazed me:

“Across government, a broad shift is underway to resituate the consumer, not the producer, at the centre of product design.... In particular, data visualization is a key method of translating data for decision-making, as it makes abstract information easier to understand and shape into a narrative.”

Note that to the panel, readers are consumers and writers are producers — not of documents or presentations, but of product, which is not written but designed.

Note also that this is a new “shift” putting more attention on the reader than before; God knows where attention used to be put. And the single specific method for reaching the reader, who is also a decision-making politician? Data visualization. Infographics. In effect, the panel is telling PHAC to draw a picture for its scientifically illiterate masters.

The panel itself, however, is still mired in an abstract dialect we might call Ottawa Mandarin, understood only by the highest civil servants. Like the Qing dynasty eunuchs of the Forbidden City, they insulate the emperor and his ministers from undesirable information.

I once was a young tech writer at a U.S. high-energy physics laboratory. Assigned to write the lab’s annual research summary, I had one instruction: “Your reader is a first-term congressperson who’s been assigned to the Atomic Energy Committee and doesn’t know anything about it.”

So I wrote it in the clearest English I could, emphasizing the breakthroughs promising practical benefits like cancer therapies. When I moved to Canada and into teaching, I tried for 40 years to tell my students “the writer’s job is to make the reader’s job effortless.” That means using clear, concrete language, thereby respecting the reader — not just using abstract language to make yourself look smart.

Fifty years later, such an idea is apparently only now sinking in with the government of Canada, and only to the extent of dumbing down complex messages to simple infographics. That tells me GPHIN, PHAC, Health Canada and the whole Canadian government are headed for a lot more trouble than a mere pandemic.