New Brunswick

Province's key COVID decision-making data still not available to public

The province upgraded the COVID-19 dashboard this week, but it still doesn't include a key metric officials use to decide whether to change restriction levels under the winter plan — the seven-day average of new hospital admissions.

7-day average of new hospital admissions isn't part of recent upgrades to website

A nurse tends to a patient suspected of having COVID-19 in an intensive care unit.
The number of people in hospital with COVID-19 remained at a record high 165 on Thursday - 68 of them for COVID, while the other 97 are hospitalized with COVID, meaning they were admitted for something else when they tested positive for the virus. (Evan Mitsui/CBC)

The province upgraded the COVID-19 dashboard this week, but it still doesn't include a key metric officials use to decide whether to change restriction levels under the winter plan  — the seven-day average of new hospital admissions.

Department of Health officials have not responded to questions about whether it will be added.

There is a way to figure it out using the data available on the COVID-19 dashboard, if you know where to look, are willing to do some math and to keep track over time.

You need to calculate the day-over-day difference in total hospitalizations to date and the daily difference in current hospitalizations, add those together to get the new daily admissions, then add the new daily admissions figures for seven days together and divide by seven.

But there's a caveat.

It's unclear whether Public Health considers all new admissions — including people hospitalized "for COVID-19" as well as those hospitalized "with COVID-19," or only those admitted for COVID. Department officials did not respond to a request for information.

If it's the latter, then there isn't enough data available on the dashboard to calculate the figure.

The other problem is the province hasn't said what the trigger number is for assessing whether to move to a more or less restrictive level, only that the seven-day average of new admissions is a key metric.

Hospital admissions for COVID-19 are 'a barometer for the severity of illness in our communities,' Dr. Jennifer Russell, chief medical officer of health, said last Thursday. (Ed Hunter/CBC)

Last week, Dr. Jennifer Russell, chief medical officer of health, released a graph showing the seven-day average is going down, but the graph didn't include exact numbers for each day.

She did say the number had peaked a week prior at 18 per day and had been "on a slow but steady decline since then," which she described as "a very good trend" and what officials were "hoping to see."

New admissions are "an important indicator" of COVID's spread in the province, she said, especially now that it's impossible to get an accurate case count, given the limited use of PCR tests and reliance on people self-reporting positive rapid test results.

Using all hospitalizations to do the calculations, there were 19 new admissions Thursday, surpassing the "peak" of 18 Russell referred to.

Ray Harris, a Fredericton-based data analyst with Data Wazo, who maintains a COVID-19 tracking website, believes the province is using only hospitalizations "for" COVID, and not including people in hospital "with" COVID, who were admitted for other reasons but later tested positive for the virus.

"The only reason I say that is because of the chart that was shown during that news conference, where they did the phase change, didn't quite match my numbers," he said.

The seven-day average of new admissions is trending downward, Dr. Jennifer Russell said last Thursday. (Government of New Brunswick)

"They weren't directionally correct, meaning that it was similar in terms of a trend to what I had. But the numbers themselves weren't exact, so I'm under the impression that what the phase change trigger is is new hospitalizations 'for' COVID, which we don't have information on the dashboard to work that out ourselves."

CBC News asked the Department of Health what the seven-day new admissions were when the province moved to the most restrictive Level 3, as well as in the weeks leading up to the decision to return to Level 2, and as of Wednesday, but did not receive any figures.

"Seven-day average of new hospital admissions was part of the modelling presented at the news conference last week and have never been part of the dashboard," spokesperson Bruce Macfarlane said in an emailed statement.

"The current hospitalizations and ICU numbers are on the dashboard as they have been, along with the recently added age range breakdowns," 

With recent enhancements to the COVID-19 dashboard, daily news releases have become redundant.- Bruce Macfarlane, Department of Health

More people are going to have to turn to the dashboard for information now that the province has stopped issuing daily COVID news releases, effective this week.

"With recent enhancements to the COVID-19 dashboard, daily news releases have become redundant as the information previously contained in them is now available on the dashboard," said Macfarlane.

"A cross-jurisdictional scan also shows New Brunswick provides more data points on our dashboard compared to other Atlantic provinces."

Macfarlane did not respond to questions about what people who don't have computers or aren't computer savvy, and could be among the groups considered most vulnerable to COVID, are expected to do.

But he did say news releases and news conferences will continue to take place when there are policy changes or decisions to make the public aware of. The department also continues to respond to media questions daily, he added.

How to calculate new admissions

To calculate new daily hospital admissions, you need to click on the COVID-19 Case Data tab, the second from the left across the top, if you're using a desktop computer, and scroll down to find hospitalizations. That's the total number of exits from hospital to date, both discharges and deaths. It does not include the people currently in hospital.

On mobile, you can find this figure by clicking on the bottom of the "cases box" until a drop down menu appears and selecting "hospitalizations," or by repeatedly clicking the small arrows pointing to the right or left until the hospitalizations number pops up. On Thursday, the number is 935.

You also need this same figure from the previous day (916) to calculate the difference (19). You will need to have squirrelled the previous day's total away somewhere in order to do this because the number changes daily and historical data isn't available on the dashboard.

Next, you need the current number of people in hospital, which is at the top of the same page (165) and the same figure from the previous day (165) to calculate the difference (zero.)

Then you add those two figures together (19 + zero) to get how many people were admitted to hospital that day (19).

To get the seven-day average, you would add the new admissions figures from each of the seven days together and divide by seven. As of Wednesday, it stood at 17, according to Harris.

But again, this is based on all admissions, both "for" and "with" COVID, and might not line up with the province's figures since the department hasn't revealed what data it's using, or released the numbers themselves.

Still 'good information' on dashboard

Generally speaking, Harris is "pretty happy" with the revised dashboard.

He cites the addition of hospitalizations and ICU admissions by age group, self-reported positive rapid tests, and boosters by age group as examples.

Asked whether there are any other key metrics missing, he replied, "I think there will always be nice-to-haves that we could look for," such as more zone-specific information.

"I always think about this in two different ways. There's what I, an interested citizen, would like to see on the dashboard, just because I'm curious, versus what is really going to affect me in my decision-making day-to-day in terms of how this pandemic … is progressing in the province."

The new landing page (another tab across the top of the website if you're on a desktop computer) has "a lot of good information on it that is in one spot," he said.

The COVID-19 dashboard landing page is a good place for people to start if they're not used to navigating the site, said data analyst Ray Harris. (Government of New Brunswick)

It includes, among other things, the new number of COVID-related deaths, how many people are currently hospitalized and in intensive care, as well as the breakdown of those "for" COVID and "with" COVID, how many of them are on ventilators, and how much those numbers have changed from the previous day.

There's also a graph that shows how many people are in hospital and ICU each day as well as the seven-day averages. (If you hover your cursor along the graph, pop-up boxes with exact numbers and dates appear. You can enlarge the graph by hovering your cursor in the top right corner of the box until a small circle with arrows pointing in four directions appears and then clicking on it. You can minimize it again by repeating these steps.)

The number of health-care workers off sick after testing positive for COVID, hospital and ICU occupancy rates, vaccination rates, and the number of new positive PCR tests and self-reported positive rapid tests are also available.

"Most of what you want for a daily update is just scrolling on that landing page," said Harris.

He recommends it as a good place to start for people who aren't as familiar with navigating the dashboard.

'A lot of numbers'

Isil Flynn, a Fredericton-based IT project manager and user experience architect, thinks there's "too much" information on the landing page.

"That's a lot of numbers," she remarked, while scrolling through.

If people are visiting the site to try to understand the status of COVID in the province or to try to assess their personal risk, Flynn thinks "it's a little more complex than it needs to be."

"Simplifying it is always a good thing."

Isil Flynn, an IT project manager and user experience architect, believes there's a lot of data on the COVID dashboard that isn't clear as to why it would be important to users to have that particular piece of information. (Isil Flynn/Twitter)

Some of the data could also be better explained, said Flynn, citing the number of people in hospital for or with COVID as an example.

"What's not clear is what the difference … is and how that matters to you."

She'd like to see more "trend" data, like the seven-day hospitalization and ICU rates graph.

It's "more impactful to people to just to see — are we flattening the curve? Have we turned the corner? Have we peaked and we're now on the downslide of things? Right? I think that would be more important for people to know."