Estimated wait times at Winnipeg emergency departments ‘quite inaccurate’ at times: memo | CBC News
The system behind the emergency department wait times Winnipeggers see online had multiple flaws that made waits sometimes appear lower than they should, an internal memo shows.
The memo sent out in early February by Shared Health and Winnipeg Regional Health Authority (WRHA) leadership says the information was “quite inaccurate” at times, and that officials had spent months working with the vendor to correct the public boards.
“As many of you know, the publicly facing waiting times boards are not always reflective of the actual number of patients,” the memo began. “Nor have they always given wait times that prepare our patients for the wait that they will experience.”
WRHA posts estimated wait times for emergency departments and urgent care centres on its website, which are updated throughout the day. The boards also include the number of people waiting or being treated at given hospitals.
The February memo said the boards — which were also displayed in waiting rooms — used old data that “no longer represent the reality in 2025-2026” to predict wait times, and that the multi-year dataset had been recently updated to the end of 2024 to make the boards more accurate.
It also said code that resulted in an “unexplained” cap in patients displayed at certain times of the day had been removed.
The code meant that at some points during late evenings and nights, the patient numbers could only go up to a given figure regardless of how many people were actually in an emergency department, the memo said.
‘Unprecedented frequency’
The officials said they were aware showing the numbers on internal waiting room displays may have resulted in “safety events” where people chose to leave an emergency department because they saw wait times were lower at another hospital, even though they had already been triaged.
“As a result, we have removed that information from the internal waiting room displays,” the memo said.
The memo said the system’s predictive model also suffered on days when outlier patients experienced very long wait times.
“We know that this has been occurring with unprecedented frequency and so we are working on shaping the predictions so that they’re more likely to capture that experience,” the memo said, adding that it will continue to monitor the changes.
Manitoba Nurses Union president Darlene Jackson said the information in the memo was not surprising.
“I hear from patients all the time.… They’re told there’s a four-hour wait, and they’ll sometimes wait up to 10 hours, 15 hours,” Jackson said.
A provincial spokesperson said the predictive model showed wait times that may be lower or higher than the time displayed on the emergency department, but did not specify whether it was intended to predict typical waits.
Front-line staff ‘taking the brunt’ of frustrations: union
Jackson said the frustration patients feel when wait times exceed expectations has consequences for staff.
“We’re seeing a lot of escalation and a lot more violence in our health-care facilities, and this is definitely one issue where individuals get very frustrated,” Jackson said.
“Because a nurse or a physician are right there at the front line, they’re the ones that are actually taking the brunt of these patients’ frustrations.”

Health Minister Uzoma Asagwara said the system used to estimate the wait times was introduced about a decade ago, and the province learned over recent months the data was not up to date.
Asagwara said it was doctors and staff who originally flagged the issue.
“It’s really important to really understand that this system was built years ago and was left untouched by the previous [Progressive Conservative] government,” Asagwara said.
“We’re just going to continue to work on this,” they said. “That’s the way, health care goes, you know: you identify a problem … you do the work to fix it.”
Figures may worsen overcrowding: researcher
Dr. Stephen Strobel, a health economist and emergency physician who teaches at McMaster University, has looked into the impact of publicly showing wait time estimates in Hamilton and the Niagara Region in Ontario.
He said that research shows posting the wait times leads to fewer patients — particularly those with minor conditions — at emergency departments, who opt to go to other sites for care instead.
But Strobel said it’s a balancing act.

“As soon as the wait time displayed starts to get very, very high, the effect is probably to dissuade sicker and sicker patients,” he said.
Strobel said in a scenario where the wait times displayed are inaccurately low, there’s reason to expect they would contribute to crowding in waiting rooms, though he added he would be leery to discuss specific impacts on patients’ health.
“There’s probably at least good reason to expect that crowding in an emergency department would make it harder to find sicker patients, sort of the needle-in-the-haystack kind of issue,” Strobel said.
“The result might be … say, missed patients or missed health issues.”
Doctors Manitoba said in a statement physicians strive for accuracy in everything they do, and that they would “expect the information the health system shares with patients to be held to the same standard.”
Dr. Fahed Razak, Canada Research Chair specializing in health-care data analytics, said getting accurate wait-time predictions is hard because there are too many variables at play, with even one abnormal patient being able to “make the entire prediction off.”
But he said the projections have an important role even if they’re not perfect.
“If I was a patient waiting in an emergency room, [having] some information — even if it has problems with accuracy from time to time — is better than not having it,” Razak said.
Razak said institutions must be transparent about how well the system behind the projections works, and which data sets it uses.
“I think they have to be explained with appropriate caution,” he said.
Health-care officials were working for months to fix the system that produces the emergency room wait times Winnipeggers see online because the data was “quite inaccurate,” an internal memo sent out in February says.
