Be careful with how you respond to COVID-19 statistics | Letter to the editor

I am responding to the article entitled “COVID-19 put a pause on short-term rentals. Manatee is ready to let visitors check in” published May 19. I would like to discuss the phrase in the article: “As the COVID-19 positive rate trends downward in Manatee County…”. While this may technically be true, it is largely a function of the fact that many more tests are being performed, and the criteria for performing tests have been relaxed. Early in the pandemic, the person to be tested had to satisfy stringent criteria of infection and have a doctor’s order for the test. Also, tests and materials to perform the test were in short supply. Now, a doctor’s order is not required, and there is more availability of tests.

In fact, the number of documented cases is the more appropriate criteria. The graph below is from the Florida Department of Health website and is data from Manatee County with the label “New Cases of Residents by Day, accessed 23 May 2020. (https://experience.arcgis.com/experience/96dd742462124fa0b38ddedb9b25e429)

To me, there is no real trend downward in cases in the county but rather appears had a seven day running average been done, to be relatively flat. It certainly is not a graph to celebrate or make an opening-up decision.

How is it true that there is a downward trajectory as stated? The denominator (number of tests) increased, so the proportion of positives decreased. A simple example, if last Monday the number of tests were 20, and 2 were positive, the rate of positive tests was 10%. And if yesterday, the number of tests done was 30, and 2 were positive, the rate would be 6.7%. The number of cases was unchanged, but the rate went down. In the graph above, the number of positive cases reported was 40 on May 20. It was 39 on May 5. Should we feel that there is a downward trajectory based on the real number of cases? I think not.

I am not debating the merits of the decision, but the community should recognize when misleading statistics are being used to justify a decision.

Linda Christmann, MD, MBA

Bradenton