Tests vs. people tested: the numbers don't add up

Posted

HURON — From the point of the first confirmed case of COVID-19 in the state of South Dakota on March 10, the residents of the state have relied on the state’s Department of Health (DOH) to provide timely and accurate numbers to reflect the status of the virus within the state. When winter sports championships and eventually schools were canceled for the year, the numbers showing the significant spread of the virus in the state allowed for an explanation of why such measures were needed.

The state is now four months into the battle with COVID-19, and recent revelations from the DOH itself bring to question the accuracy of the department’s numbers throughout the pandemic.

On July 2, DOH officials announced a new window on the dashboard display of the state’s COVID-19 website, which displayed the state’s “total tests” that had been taken. Over the past week, multiple follow-up emails and questions in subsequent press briefings have attempted to clarify the difference between the counting of a test and what the state has released as testing numbers.

Those queries generated the description from DOH officials of previously-released numbers as persons tested, which would differentiate tests as individual tests versus the individuals who have been tested in the state.

That would be all well and good, especially when calculating the attempted goal of reaching 5% of residents tested per month.

The United States Census Bureau estimates the state’s population in 2019 to be 884,659. To get 5% of the population tested each month, the state would need to test 44,233 people, or roughly four times the size of Huron every single month.

On July 2, state epidemiologist Dr. Josh Clayton declared that the state had met its threshold for 5% testing by conducting “over 48,000 tests in the month of June.” The DOH dashboard on the COVID-19 website actually listed 48,687 tests performed, which would be 5.5%.

However, the DOH COVID-19 website only listed 36,694 people who had been tested in the month of June, which works out to 4.15% of the state’s population, according to census estimation.

When asked about the discrepancy on Thursday, DOH Secretary Kim Malsam-Rysdon attempted to explain that individuals could be tested in multiple months and be counted in multiple months, which she said explains the difference. To make up that difference to achieve just 5% in June, 7,539 people in the state would have needed to be people who were retesting. However, the state’s difference was 11,993 tests, which, when you do the math, meant more than 24% of the total tests run were retested.

Wanting to get an idea of the reporting, the Centers for Disease Control (CDC) website was contacted. The CDC shows South Dakota with more than 75,000 tests as of July 9, with 7,242 positive cases. The case number was accurate for the state, but the test number was short by roughly 12,000 from the people tested, according to the state and short by nearly 50,000 tests from what the state had on its test total dashboard window, which was removed from the DOH dashboard on July 6, after multiple requests for further explanation led the department to better explain its numbers.

In an email to the contact link at the CDC, it was confirmed that the DOH directly reports numbers to the CDC and that the understanding they presented to reporting agencies was that the “test data” was to be the tests performed on individuals, not another number. That number was significantly lower on the CDC website for South Dakota than either number reported to its own people.

What numbers are real at this point? When the DOH announces the number of positive cases, deaths, and/or recovered cases on a day, is full and complete information being given?

Multiple attempts to contact DOH generated no response to the questions on these concerns.