An Election Is Coming Up - So Let's Change Covid Reporting Methods

In essence, the purpose of statistics is to accurately measure things. 

It becomes rather hard to accurately measure a thing when you change the reporting methods about the thing, making comparisons difficult if not impossible, likely due to an upcoming election.

The Detroit Free Press: Michigan health department to change how it reports COVID data

Starting next week, the Michigan Department of Health and Human Services is changing the way it reports coronavirus data.

It will update Michigan cases and COVID-19 deaths weekly, rather than three times a week, starting April 4 on its website

       . . .

The change in the way the state will report cases and deaths going forward adheres to a national surveillance strategy created by the U.S. Centers for Disease Control and Prevention, said Lynn Sutfin, spokesperson for the state health department.

"At this stage in the pandemic, daily updates do not inform policy or contribute to the national surveillance program," Sutfin said. "Many states have also begun to move to less frequent public reporting of data.

Thus causing it to reflect a lagging rather than leading trend and likely end up reporting more cases but less deaths as a result. 

Most  interestingly, they're now going to reduce the number of deaths publicly attributed to Covid as well:

Another change is that the health department will no longer report how many COVID-19 deaths were discovered during a review of vital records.

That's quite a change, as that retrospective review was previously used to pump the death-from-Covid numbers up.

So the numbers will now be adjusted to reduce the numbers of deaths from/with Covid in the run up to the mid-terms, and also deliberately lag in case of an increase in cases.  Our Gov can now announce there are fewer deaths based on these new statistical methods as the midterms approach. 

Now isn't that wonderfully convenient?