Tracking Coronavirus data might seem as easy as going to the state’s website or following your local health department's web page, but as you dig into the data things start to get a bit more complicated.
“We’ve been expecting worse than we’ve actually seen so far,” said Dr. Michael Collins, medical director of the Grand Traverse County Health Department, referencing to data on the COVID infection rate in the county. According to the data, Grand Traverse County's infection rate is low-- much lower than other counties of it's size.
Grand Traverse County, like many counties, had a shortage of testing early on. Collins said early on they had to prioritize testing.
He said,“originally the only people outside of hospital workers who were being tested were those who were pretty seriously ill because they were at least requiring hospitalization.”
Collins said there were probably many more cases than the official numbers showed, but they weren’t able to be identified due to lack of testing. Grand Traverse is just one example where the data doesn’t tell the whole story. In many areas of the state and country there are testing limitations and procedures that complicate analysis.
“That is probably the number one challenge of anyone tasked with analyzing data, is verification, do you have a clean data set?" said Professor John Daniels.
Daniels is a member of the American Statistical Association, and teaches statistics at CMU. He said it can be difficult to compare infection data. Different areas have different testing capacities and procedures, he said that's a problem when you try to compare numbers.
“When you’re collecting data and analyzing data, there needs to be consistencies in the process. If the processes being done by the different counties are not consistent then you run the risk of introducing bias into the data, and that’s the biggest problem we face," explained Daniels.
Daniels said researchers can try to fill in the blanks by looking at other metrics, one example Daniels said could work would be looking at a county's death rates during the earlier stages of the pandemic, where testing was limited to the worst cases. Grand Traverse County death rates are still lower than other similar counties but they are closer to average than the case count.
Of course a lack of testing isn’t the only factor that can cause statistical anomalies for Coronavirus. Having a surge in testing availability or changing how testing is distributed can have a huge impact on numbers
Currently Oceana county’s new cases are on the rise, or so it seems.
Jeannine Taylor is with the 10th District Health department, which includes Oceana County. She said case numbers are up, but there’s a reason.
“Oceana County has a drive through clinic a few weeks ago, and it’s available to anyone that wants to get tested, so I think that’s why the numbers are going up in Oceana County because more people are able to get tested and have been tested there,” said Taylor
Increased testing and a more open testing procedure began last week, and coincided with a sudden rise of cases. Taylor said it's not surprising to see cases rise since more people than ever are being tested in the county.
“They will test anybody that feels may have it or if they have mild symptoms, and they actually don’t even need a doctor's order at this point,” she explained.
Oceana is one of few places in the state with testing like this, meaning now the area will be able to better track its cases of the virus. And the more testing becomes widespread, the closer Michigan will be to the consistency that statistician John Daniels said is so critical. and maybe it will become easier to analyze and compare data on coronavirus
