What’s up with this “programming error” in the Florida data on kids? How does this change things?

Data and Metrics Data Literacy Testing and Contact Tracing

A: TL;DR Whether the positivity rate is 14% or 30% for kids, Florida still has a COVID problem.

And with respect to the data error, we need to have a little understanding. Public health departments are trying to do something very difficult on shoestring budgets. Data collection in a crisis is messy and sometimes we don’t do it right–even when there is funding, and in this case, there isn’t funding.

Florida published numbers for the week of July 10 that showed a (gasp) 30%+ positivity rate among children who were tested for COVID-19 that week. But the next week, the public health department said that the sudden increase was an error. It turned out that a lot of *negative* test results for kids had not been included, and the positivity rate for kids for the next week, July 17th, was down to around 14%.

We posted about how to interpret the Florida positivity rate recently here at Dear Pandemic: https://www.facebook.com/thosenerdygirls/posts/162202478741741
A similar issue played out in Wisconsin. Testing capacity has successfully scaled up, but lab capacity to process tests–including the staff who post the results–is lagging. Negative test results were not being posted as quickly as positive results, which made positivity rates look temporarily high, just like in Florida.

Although it is important to keep an eye on rapid changes in the trends to see if the outbreak is suddenly getting worse, it’s also important to look at the whole picture. These data snafus did not affect the numbers for daily incidence, hospitalizations, and deaths for the same week.

In addition, it’s important to keep one eye on longer-term trends. Positivity among Florida kids (programming error notwithstanding), has not dipped below 10% for 6 weeks. Florida has a lot of COVID-19 circulating right now, among kids and adults. In the case of Wisconsin, daily incidence has been going up steadily since a trough on June 15th, with hospitalizations and deaths holding steady.

This is the latest in a string of data reporting issues during the pandemic. But before you jump to conclusions about a grand public health conspiracy to obscure the facts, consider this cold hard truth: public health just doesn’t have the money to pull off a conspiracy.

Public health departments at the state and local level–including our state laboratories and data analysts–have experienced budget cut after budget cut for decades, and even when budgets aren’t reduced, they are not funded adequately. In the United States, we spend about two cents for every health care dollar on public health efforts broadly. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4695931/

The Trust for America’s Health issued a report *in April 2019* that concluded that in spite of overwhelming, bipartisan public support for what public health provides…
“public health efforts are about $4.5 billion underfunded. That’s led state and local health departments woefully unprepared to address public health emergencies such as infectious disease outbreaks, extreme weather events, and the opioid crisis.”

We should offer a little bit of understanding of our public health staffers. They’re tired. They’re angry. They have no childcare. They’re working incredibly long hours and have been doing it for months. The constant partisan squabbles and refusal by many members of the public to follow the most basic public health suggestions is, frankly, exhausting. Some of us have gotten personal threats to our safety. Some departments have had officials resign amid such threats, and there are new reports about parents–especially moms–quitting their public health jobs just when they are needed most because they’re too tired to go on and have no way to deal with virtual schooling for their children.

Massive, rapid-turnaround data collection efforts in a crisis are going to have some errors, and we need to roll with them. Public health departments are operating far, far beyond their staffing capacities with no end in sight.

Data collection in a crisis is messy and sometimes we don’t do it right–even when there is funding, and in this case, there isn’t funding.

Florida data update from the Sun Sentinel

Link to original FB post