Corrective advice: How to read town Covid numbers

Over the past 10 months, I have compared the current total to the day before. Then, when numbers started rising more quickly, about mid-October, I should have paid closer attention to what was happening.

I have been reporting Arlington's Covid-19 numbers since last March 6, a day after my wife and I returned from vacation, just as the pandemic's curtain fell.

Updates occur every weekday, Monday through Friday, between 4 and 5 p.m. They are based on numbers that the town reports. See the town Covid link here >> 

That information is summarized in a box showing numbers of cases and age ranges of those affected. Accompanying the box is a brief account of the changing numbers. The only names reported are town residents whose obituaries are published in local media. Otherwise, for privacy reasons, the public sees numbers. See the latest summary here >> 

Over the past 10 months, I have compared the current total to the day before. Then, when numbers started rising more quickly, about mid-October, I should have paid closer attention to what was happening.

If I had, I would have avoided a mistake in early December. My error is instructive, and we all can learn from it.

From Dec. 2 to 3, the number of Covid cases in town rose unlike anything previously reported -- by 35. That was news, and I highlighted it. The YourArlington link to this news flew to varied social-media sites.

Number questioned

Following were the usual wow! emojis, but so was a sharp-eyed reader, who questioned the number.

I stuck to my guns: Yes, on Dec. 3 the total cases was indeed 35 higher than the day before.

But it wasn't. A poster who seemed to be supporting my case noted the nine cases reported at St. Agnes.

That should have been a tip-off. That cluster at the Medford Street parochial School occurred *before* the 35 surge, on Dec. 1.

Numbers explained

I asked Christine Bongiorno, director of town Health and Human Services, Dec. 4 to explain the numbers. She is understandably busy these days, and on Dec. 8, she got back to me:

"Here is the answer to your question straight from the person who manages the daily data updates:
 
"The data reported is based on the positive specimen date as to be consistent with MDPH [Mass. Department of Public Health] reporting -- not the date that the lab is entered into MAVEN or the date BOH [Board of Health] is notified of the case.

"As a result, there is often a lag in data reporting where I have to retroactively fill in case numbers for days that have passed once the labs have been entered into MAVEN. As a result, I could see how it looks like there is a periodic update of cases over certain days.
 
"Basically, depending on the lab processing the results, we could see some positives come in up to a week after the specimen is collected, that can really be confusing when we go back to update the system."

Translation: Sometimes the numbers have to be updated over a range of days, or even a week, to catch up with what is reported.

Understandoing MAVEN

MAVEN stands for the Massachusetts Virtual Epidemiologic Network, which is explained here >> 

You can see how MAVEN provides Arlington numbers here >> If you navigate its chart into the past, you see over the months the falling and rising toll of illness.

Let's return to the original example of "35 in one day." Those 35 cases from Oct. 2 to 3 were recorded over a much longer period than one day and certainly included the nine cases reported at St. Agnes.

Because of the dynamic nature of the reporting process, you cannot return to that time frame and review a breakdown of specifics numbers at the time. What you can do is become more familiar with the basic reporting sources, including MAVEN.

Scroll down at the Arlington MAVEN site, and you can see dynamic bar charts showing cumulative cases and a seven-day moving average from March 9 to December.

Note to the troll on Patch comments: I have no interest in St. Agnes kids other than wanting to document Covid cases in town.

Dec. 10, 2020: Vaccine shipments to Mass. to begin

This explanation, which includes opinion, was published Saturday, Dec. 12, 2020.