The Difference Between Wisconsin's 'Confirmed' and 'Probable' COVID-19 Data

Efforts to count coronavirus cases and deaths around the state indicate the varying degrees of certainty that can be attributed to individual circumstances as public health agencies keep up with day-to-day rhythms of the pandemic.

By Kristian Knutsen

May 13, 2021

FacebookRedditGoogle ClassroomEmail

Every day of every week of every month since the start of the pandemic — including weekends and holidays, and with the exception of only a single day due to a technical upgrade — the Wisconsin Department of Health Services issues data detailing pandemic metrics for every county in the state. These daily updates include figures about vaccinations, testing, new cases, hospitalizations and deaths, with additional specifics about each in terms of geography, age and other demographics.

The state’s fire hose of COVID-19 data is at the foundation of gauging the shifting status of the pandemic over time and the basis of public health policies seeking to bring it to an end. In some ways, the numbers appear to be simple, but the reality of accurately and comprehensively tracking such information is more complex. Daily discrepancies between state and local figures, a need to revise vast volumes of data gathered in a hurry, and tracking vaccines that require one or multiple doses are just a few of the intricacies that demand close attention by public health authorities.

One data distinction that gets to the heart of these complexities is that between what are labeled “confirmed” versus “probable” cases and “confirmed” versus “probable” deaths. As a standard practice, when distilling the daily datastream into a more digestible update, it’s the confirmed numbers that are highlighted by the state and local health departments, and in turn by local media outlets, including in the WisContext pandemic tracking led by PBS Wisconsin.

The rationale for emphasizing confirmed cases and deaths is simple: These figures represent instances in which the most reliable testing results indicate an individual had been infected with the coronavirus. Yet those numbers are incomplete. Cases and deaths labeled probable reflect, to some degree, the greater breadth of the pandemic about which some degree of information is known.

The difference between the number of confirmed and probable cases and deaths in Wisconsin is not negligible. From the beginning of 2020 up to May 11, 2021, at a statewide level, there have been 604,378 confirmed cases and 63,259 probable cases. Together, these counts add up to 667,637 confirmed and probable cases, a total that is roughly 10% higher than the lower figure. Over the same time, there have been 6,935 confirmed deaths and 752 probable deaths. Combining these two figures results in a total of 7,687 confirmed and probable deaths, a total that is likewise about 10% higher than the lower number.

Due to asymptomatic spread of the coronavirus, it’s thought that the number of overall cases could be considerably higher than any figures tracked by public health authorities. Similarly, the matter of excess deaths — the number of deaths during a given time span that exceeds the total that would be expected given standard health trends — indicates the death toll from the pandemic is some degree higher than official counts, both confirmed and probable.

The phenomena of asymptomatic spread and excess deaths underscore the comparative reliability of probable cases and deaths, but how exactly does the state health department differentiate them from those that are confirmed?

In terms of probable versus confirmed deaths, the difference is simple. Deaths that are labeled probable simply reflect the “number of COVID-associated deaths among probable cases of COVID-19,” according to the state health department’s explanation of its data. (How deaths are determined as associated with COVID-19 is its own process, and reflects individual judgements by physicians and medical examiners or coroners at the local level.)

How the state labels cases as probable versus confirmed cases is a bit more involved. According to a Department of Health Services spokesperson, if a person who may be infected but has not tested positive for the coronavirus via a specific laboratory method (as defined by the Centers for Disease Control and Prevention), they would be counted as a probable case if they meet one or more of three different criteria. One is a positive result via an antigen test method (which is labeled “presumptive” or “supportive” by the CDC based on the type of sample). Another is somebody who exhibits COVID-19 symptoms and is known to have been exposed to the coronavirus — the spokesperson noted “being a close contact of someone who was diagnosed with COVID-19” as an example. The third is if COVID-19 or SARS-CoV-2 (the name of the novel coronavirus) is listed on an individual’s death certificate.

When it comes to case and death revisions, figures can move between the two classifications. Indeed, in its FAQ about data collection and reporting, the state health department notes “a majority of the changes are a result of case status being corrected from ‘confirmed’ to ‘probable.'”

While there is about one probable case and death for every 10 confirmed cases and deaths at the statewide level, the differences between the two can vary widely at the county level. The number of probable cases is comparatively low in Menominee, Jackson, Monroe, Buffalo and Chippewa counties, while it’s comparatively high in Walworth, Green Lake, Pierce, Waupaca and St. Croix counties.

Gauging county-level differences in confirmed versus probable deaths around Wisconsin is a different matter given lower numbers. Nearly a dozen counties haven’t reported any probable deaths, most of which have fewer than 20 confirmed deaths. However, several counties have quite a few probable deaths compared to those that have been confirmed, namely Iron (which has the state’s highest overall death rate), Douglas, Lincoln, Langlade and Adams.

The state health department did not offer an explanation for what may be causing these local differences in confirmed versus probable cases and deaths in response to queries.

When it comes to assessing COVID-19 death counts, though, even adding probable deaths to the confirmed death total may be understating the overall toll.

On May 6, the Institute for Health Metrics and Evaluation at the University of Washington — a group that came to prominence early in the pandemic with its projections and subsequently was criticized for its changing figuresreleased an analysis that found total deaths resulting from COVID-19 could be considerably higher than health authorities have reported.

Based on excess deaths data gathered by the CDC, this analysis indicated the total number of coronavirus deaths in Wisconsin is roughly one-and-a-half to one-and-three-quarters times higher than what the state has reported. The study included both confirmed and probable deaths as reported figures. As of May 2, this analysis projected that Wisconsin had experienced some 11,684 total COVID-related deaths, as opposed to the 6,926 confirmed deaths plus 751 probable deaths — 7,677 together — that the state had reported.

(In its data presentation, figures in the University of Washington analysis lag state figures by more than a week, and its total number of reported deaths is about 2% lower — a difference of 133 deaths — than the state’s own reporting for May 2.)

Overall, the analysis found the global death toll of COVID-19 could be about twice that of official estimates. It sought to sort out direct COVID-19 deaths from increases in mortality related to people delaying health care during the pandemic or suffering higher levels of mental health disorders and substance abuse, as well as decreases in mortality related to decreases in accidental deaths as a result of distancing and lockdowns, a reduction in other transmittable diseases such as influenza, and the effects of the coronavirus on people with serious chronic conditions.

The proportion of higher COVID-19 deaths projected by this analysis for Wisconsin tracked with most other states, though each of its neighbors — Illinois, Iowa, Michigan and Minnesota — had somewhat lower rates of increase.

Determining a single and unequivocal COVID-19 death tally for Wisconsin when the time comes is a goal that’s challenging given the uncertainties inherent to rigorous scientific analysis, but when it comes to the daily numbers reported for the state, those that are known to be confirmed represent a starting point.

Statement to the Communities We Serve

There is no place for racism in our society. We must work together as a community to ensure we no longer teach, or tolerate it.  Read the full statement.