Pandemic math and the case of the missing cases | The Coast Halifax

Pandemic math and the case of the missing cases

Nova Scotia aims to get its COVID-19 numbers right with the Panorama system.

click to enlarge Pandemic math and the case of the missing cases
Adobe Stock
When Panorama looks at you, this is what it sees.

How many active cases of COVID-19 does Nova Scotia have? This simple question is vital to our physical and economic health. And it’s proven shockingly hard to answer.

Every day the province issues a report about the state of the disease—the number of new cases, how many people are in hospital, how many patients have recovered. There’s been no statistic for active cases, however. And parsing the numbers to figure it out can lead to impossible results that call the province’s competence into question.

Chief medical officer of health Robert "top doc" Strang has said several times over the past months of COVID that disease data is hard to manage because it’s coming from different sources. But today that changed, because all the information got merged onto a system called Panorama.

"It brings public health into the modern age," said Strang, hyping the technology at yesterday’s regular COVID-19 briefing. He said the data merge means there is now a "sole source of truth" about C19.

As Strang describes it, Panorama is an information system used widely across the country in the field of public health (rather than general medicine). Public health pays attention to a host of diseases, and is supposed to be notified when doctors and/or labs make a diagnosis. Unsurprisingly, the current list of "notifiable diseases" includes COVID-19; patient information for every case in Nova Scotia is tracked in Panorama.

For the last couple of years, the system has also been the repository of info about who is getting vaccinated in schools. It’s a far cry from the way those records used to be kept. "When I first came here to Nova Scotia, we had boxes and boxes of index cards in shoeboxes, and that was our immunization record," said Strang.

Over time, as Panorama gets linked to other data systems, all immunization records will become accessible throughout the healthcare system, no matter who administers the shot. "Whether it’s a public health or a family doctor, or pharmacist or somewhere in the hospital—if somebody is vaccinated, all that through data linkages gets in, and we have an immunization registry."

But enough about mumps in 2032. In terms of COVID-19, today’s development is that Panorama got linked to the lab doing the testing. Until now, the lab ran a totally different system for its notifiable disease reporting, with different cutoff times from the public health workers tracking cases in Panorama. That’s lead to all sorts of inconsistencies in the numbers. With the lab at the QEII linked into Panorama—"we are the first province in the country to actually do that full integration," bragged the top doc—Nova Scotia will have what Strang calls "the gold standard" of epidemiological information.

This arrives none too soon. As Nova Scotia’s C19 case numbers have dropped in recent days, the inconsistencies became glaring.

Saturday the province’s report said there had been 1,058 confirmed cases of COVID-19 in total, but 999 people had recovered and 61 had died. In other words, there were 1,060 people who had had the disease (999 + 61 = 1,060), yet somehow only 1,058 people who’d tested positive. As if we imported two people with known cases from Quebec into the province, not bothering to test them although counting them when they recovered.

That math also means the number of active cases wasn’t just zero. Because all of our cases were resolved AND we had the two from Quebec, we were at negative two active cases—we needed to get two new cases just to get to zero active cases. Although if we got two new cases, we’d have at least two active cases, unless we imported two patients from Newfoundland just to test, then sent them home to be active cases there.

A solution would be to give the province the benefit of the doubt and say there were no active cases. But Saturday’s report also said "Northwood in Halifax currently has one resident and one staff active cases," and then in another part of the report: "Three individuals are currently in hospital, two of those in ICU." So there might have been five cases (three in the hospital plus two in Northwood), or three (from the hospital, assuming that number included the Northwood patient(s) in hospital), or two (Northwood), or zero (giving the benefit of the doubt or counting the Newfoundlanders’ tests, whichever strains credulity least) or negative two (counting the recoveries, but not the diagnoses, from Quebec or wherever they came from).

Sunday’s numbers were almost identical to Saturday’s. The same number of recovered patients and deaths, the same number of Northwood cases and people in the hospital. Except there was also a new case diagnosed, making for up to six active cases. Or five. Or three, et cetera.

Monday the report didn’t mention the number of deaths or recoveries or people in the hospital. It did, however, say "Data sources are being reconciled and consolidated to ensure all publicly reported data comes from a single source, Panorama, the province's public health reporting system. Updated data will be reported this week."

And it was a major update. At today's briefing Strang introduced two changes to the way data is reported. 1. Instead of calling cases "recovered" for the patient getting over the disease, a case will be counted as "resolved" when public health completes its investigation and closes it. 2. The province's map of case diagnoses will reflect the NS Health Authority zone where the patient lives, instead of the zone where the diagnosis was made.

He also reset many of our current statistics with the gold-standard numbers. "They're the most up-to-date and comprehensive COVID-19 data," Strang said, "and we've resolved all the discrepancies that people have been pointing out in the last number of days."

There were no new cases or deaths to announce, but the total number of cases rose by one, from yesterday's 1,060 to today's 1,061. Believe it or not, the reason is an imported case—a Nova Scotian who doesn't currently live here and tested positive in another province, but "because of the way we report things between provinces, they are reflected in Nova Scotia numbers."

The 62 deaths reported previously is accurate, but the number of resolved cases dropped by five to 994. Math tells us that leaves five active cases (1,061 - 994 - 62 = 5), and Strang confirmed it. None of those cases are in Northwood (or any other licensed long-term care facility), which is welcome news.

In less welcome news, Strang refused to say where the five active cases are located. "We are very hopeful that soon we are going to have no active cases at all in the province," he said in response to a reporter's direct question about this, "and then it really is a moot point about breaking it down by geography." If hoping away cases was an option, why have we waited 88 days to use it?

This is a question we'll have to wait at least two weeks to get an answer to. At the end of the briefing, Strang announced that he has to go to New Brunswick for a "minor procedure" that isn't available in Nova Scotia, to check the margins from a previous skin cancer operation. And like everyone else, he'll have to go into a 14-day quarantine after returning from travel outside the province. His regular briefings with premier Stephen McNeil will go on hiatus during this time, although he promises to be working from home, focused as usual on the epidemiology.