Sask. Health Authority says it sent bad data showing half of Saskatoon hospital beds were empty

SHA: hospital occupancy study was not accurate WATCH: SHA says it sent bad data to an independent agency studying hospital capacity which was widely publicized last week.

The Saskatchewan Health Authority says it sent the wrong information to an independent agency studying hospital capacity that incorrectly showed nearly half of Saskatoon hospital beds were unoccupied.

On Friday multiple media outlets, including CTV News, published data from a 2022 – 23 study from the Canadian Institute for Health Information (CIHI) that showed nearly half of the bed’s at Saskatoon’s three hospitals were unoccupied.

But on Tuesday morning, the Saskatchewan Health Authority sent a statement disputing the study’s findings — because it sent the wrong data.

"The data submitted by the Saskatchewan Health Authority (SHA) to CIHI was incorrect, resulting in significant discrepancies with the occupancy level information reported by CIHI in their latest report," the statement read.

"The SHA is working to provide the correct numbers to CIHI."

CIHI, a national non-profit organization that provides general information about Canadian healthcare, uses information from all territories and provinces as part of a national agreement to help produce a variety of reports and publicize health data.

"The ministry and the SHA have been meeting to address the discrepancies in the reported numbers and develop a plan to ensure the numbers are more accurately reflected in the 2023-24 report," the statement said.

Steven Lewis, an adjunct professor of health policy at Simon Fraser University said he realized the information was incorrect almost as soon as he read it over the weekend and says the SHA, CIHI and media who published the data are all to blame, but not equally.

"I read it and said, 'this is wildly implausible and cannot be true,'" Lewis said. "So I think it reflects a bit of a cultural problem."

Lewis said the province is mostly at fault for sending incorrect information in the first place. He said numerous errors had to have taken place for the data to be this far off the actual numbers.

The occupancy rate was calculated by taking the total number of inpatient days, dividing that by the number of calendar days, dividing the number of acute care beds and multiplying the figure by 100.

With a basic calculation at play, Lewis speculates the number of beds was the issue.

"They obviously made some big error in transmitting how many beds there were," he said.

The report said Saskatoon City Hospital's occupancy rate was 36 per cent, but Lewis said "everyone dropped the ball" since the province or CIHI didn't double check the number before publication when the occupancy rate for the year prior was 93 per cent.

"If a reporter read a report that said Saskatoon has a population of 750,000, just about every reporter and editor would say that's just ridiculous," Lewis said. "This is the same order of magnitude error."

Lewis said the report and the absence of people calling out or asking if it was full of errors points to data illiteracy prevalent in society. He said in a world inundated with up to the minute data and information, people are less likely to ask questions or spot errors.

Once you combine two trustworthy institutions in an industry where information isn't widely accessible, people tend to trust the information they receive.

"In the grand scheme of things, this is this is one of the biggest and most obvious errors I've ever seen," he said.

"There's a difference between being a good technician and a good statistician."

Lewis has no doubt both CIHI and the SHA are re-examining their internal data quality assurance systems after publishing a report with such errors.

However, he says the SHA is unlikely relying on faulty data to make policy decisions.

"It's a unicorn as a type of error, and without question, it has not guided any policy decisions,” said Lewis.

“The policy decision it should be guiding is do we need to look a little bit harder at our data quality processes and at least at the very least, find out what happened,” he said.

"Don't be pointing fingers and blaming anybody."