Experts say Sask. could be more transparent with COVID-19 data without sacrificing privacy

Eleanor Widdowson made a pointed call for transparency from the Saskatchewan government after her sister died from the novel coronavirus in March.

"Don't hide shit."

Saskatchewan refuses to reveal communities where COVID-19 has been detected unless there is an outbreak declared. Widdowson said that more knowledge might have protected her sister, who died after testing positive.

The province continues to gives case numbers broken up into six regions for the whole province.

There are discrepancies in how provincial governments are handling data release across Canada. There is no law that dictates how and when it should be released.

Last week, Toronto Public Health released COVID-19 case numbers for all of the city's postal codes after public pressure. Public health officials in Alberta have identified cases in cities like Calgary by neighborhood.

"It's actually a judgment, more of an ethical call, in determination," Dr. Anne Huang, a former deputy medical health officer for Saskatchewan and Health Canada, said. She agreed individual privacy was of concern with a low case count, but said that argument no longer stands with Saskatchewan's numbers.

Huang said the province should re-evaluate what information it shares. She said releasing case numbers by postal code, as was done in Toronto, wouldn't work for Saskatchewan's rural population, but suggested releasing the data in map form, showing the 38 Saskatchewan Health Authority (SHA) health network boundaries.

Huang said the boundaries were created "based on the pattern of how people access health services."

"To reopen the economy safely and gradually, it means that we need to become better at identifying where the hot spots are," she said. "The best way of illustrating where the hotspots of cases and transmission are, is by showing this data in the map format."

'There's inequality in our society'

Some provinces, like Ontario, have held back data, citing fear of stigmatizing people living in communities with higher rates of COVID-19. Stigma has also come up in Saskatchewan.

"I think it's an omission, an implicit omission by the government that they don't want to face public criticisms of the underlying social and economic factors that contribute to the higher disease burden in these communities," Huang said.

Huang said governments can use data to reduce stigma by educating the public and policy-makers. She said scientific evidence shows that social determinants of health — things like housing issues, low income and education — are linked to an increased risk of communicable or chronic disease like sexually transmitted infections or diabetes.

"You're starting out at a disadvantage," Huang said

She said stigma only happens if provincial leaders don't address the link between social determinants and disease outcomes.

"We pull back the curtain of this inconvenient truth for the government that there's inequality in our society," she said. "And that disparity needs to be addressed."

Saskatchewan's far north region has had far more cases — more than 250 — than any other in the province. Many of them stemmed from an outbreak in La Loche. Five of the 11 people who have died after a positive diagnosis are from this area. Four others are from the north, but the province refuses to say where they lived.

Current approach 'being reviewed'

Huang said community data can be released without infringing on privacy and "will force a necessary public discussion about what we need to do better."

A spokesperson said Monday afternoon that the Ministry of Health "is committed to continuously improving its reporting processes in line with privacy legislation." However, the spokesperson would not answer at the time whether the province was considering releasing more specific data, like using postal codes or the health network boundaries, only saying the approach to reporting is "currently being reviewed."

Dr. Saqib Shahab, the chief medical health officer for Saskatchewan, elaborated at a news conference Tuesday.

He said a new, updated map would be broken down into 13 to 15 sub-provincial areas and include for each area the number of cases in the last two weeks, the overall testing rate and the rate of positive tests.

"We did want to provide more geographical granularity just to make sure that everyone felt comfortable because some of the zones that we have been using are huge and this will provide a bit more granular detail and we will be obviously monitoring that that is meeting everyone's needs as well."

No value to releasing death locations: chief medical health officer

Dr. Shahab has maintained that releasing the specific location of deaths is a privacy issue.

In its written response this past weekend, the Ministry of Health pointed to the confidentiality section of The Public Health Act, which would prohibit the release of identifying information from being shared. However, the section doesn't specify that revealing a person's community would identify them.

It also states that information could be released if the minister orders it for purpose of protecting public health.

"The virus is the same wherever it is in Saskatchewan and when you get an unfortunate outcome like a hospitalization or death it's very tragic," said Shahab last Thursday.

"There's no value from a public health perspective or a public risk assessment perspective on knowing where a person lived who unfortunately passed away."

He said opening up testing criteria is one way the province is trying to react rapidly as cases are identified in new communities. The SHA has, on occasion, released warnings about exposures at individual locations, such as a Walmart in North Battleford.

Commissioner says transparency means trust

Saskatchewan Information and Privacy Commissioner Ron Kruzeniski said as much transparency as possible is needed for the province to maintain the trust of the community.

He said the opportunity exists for provincial officials to break down the information further, but how much further is up to health officials. He said consistency is important.

"I would certainly encourage decision makers to go as granular as they can go with the caveat that if they provide information that would identify you or me that that would be going too far," said Kruzeniski.

"But they are always able to release de-identified information, in other words it doesn't give my name or or or to any specifics as to where I live, and they can do that to the extent that people in their community can't figure out who exactly they're talking about."

He said linking one case to one small community would be a problem, but doing the same for cities like Regina, Saskatoon or Prince Albert would likely still protect the person's privacy.

Kruzeniski said the issue is where to draw the line.

"I certainly think that in regions where there's five or more people who have contracted COVID-19 or five or more deaths they certainly can go to the point of [saying there were] five people in a community."

Dr. Cory Neudorf, a professor in the department of community health and epidemiology at the University of Saskatchewan, said authorities should be revisiting their approach to releasing information as numbers increase.

He said identifying patterns and trends — such as differences in health outcomes in different regions — is difficult with limited data.

"The challenge in doing that when you have a very small number of deaths, or a very small number of cases, is you can't subdivide that very much before you start getting very unstable numbers," said Neudorf.

Calls to collect more data as well as share it

"To say, 'Is this a meaningful difference or not?' is really difficult. But as you get enough cases out there and you can actually start evaluating that then it becomes more important."

He said there is an increasing desire across Canada for more data on COVID-19 to be collected, as well as shared.

"Not just geographically, but are there differences in terms of ethnicity of individuals who are either infected or dying, differences in their socioeconomic status, or are they a vulnerable population of some kind?"

"The age breakdowns, males versus females, all of those kinds of differences. And that's needed to see — as we try to limit the spread of the disease — are the broad-based interventions we're putting in place equally effective in different parts of the province and in different subgroups?"

He said once those differences are identified through statistics they should be shared publicly.

In the meantime, Neudorf said more clarity from officials about why some information cannot be shared would also be helpful.

"Even to just give the public assurance that 'Hey, you know, we're looking at this and we don't see a difference yet,' or 'There's something worrying here, we're looking into it,'" he said.

"Just that assurance that it's even being looked at would be valuable even before information is available."