Dave Dixon tells me that the last time he did this series there was a mistake in the cases charts. Dave’s notes below explain exactly where the data comes from and the issues with the state’s backwards reporting. For very bizarre reasons which I am sure they could articulate really well, even though it clearly has the data by date of event, since it on the situation report, the state uses date of report on the weekly report. We are publishing this info about where the data comes from because I get genuine questions, but also because as readership keeps growing, I have attracted a few trolls who accuse me of making this stuff up. You don’t see those in the comments because I trash them and some other lovely notes I get from time-to-time.
Substantively, as I mentioned in another post today, you can see the impact of school testing, which leads to fake cases. I know and I am going to force the state to provide the data to show that the supposed increase in hospitalizations among the young is actually mostly RSV and other illnesses, with incidental CV-19. Otherwise not much going on.
Dave’s Notes for Series Charts with 10 Year Age Groups:
1. Data for tests, ICU hospital admissions, and Non-ICU Hospital Admissions is taken from the Minnesota Weekly Report, published by the Minnesota Department of Health every Thursday, and available here: https://www.health.state.mn.us/diseases/coronavirus/stats/index.html. The data is given for 5 year age groups, which are then combined into 10 year age groups in order for the charts to be more readable.
2. The Minnesota Weekly Report lists cumulative values for each week ending the day before the report is published, so the weeks are defined as running from Thursday to Wednesday. The tests and admissions per week are calculated by subtracting one week’s cumulative tptal from the next.
3. The data in the Weekly Report is presented by date of report, not by date of actual event. We have found through analysis that the reporting for PCR cases takes on average 5.6 days to be 90% complete and 6.6 days to be 95% complete. There are lags for hospital admissions, which have not been computed but appear to be somewhat shorter than the lags for cases.
4. Starting 9/16/21 the data for the covid case charts is taken from a data file from the MN DOH named “wagscd.csv” available from the Weekly Report page https://www.health.state.mn.us/diseases/coronavirus/stats/index.html#wagscd1 . This file contains cases by specimen collection date. In this data file the cases are reported in 5 year age groups per week starting the week ending 3/07/2020, whereas the Minnesota Weekly Reports, although starting on 5/14/2020, have relatively limited information until 8/20/2020. Because the data is presented by specimen collection date this means that the data for past weeks is updated as more cases are reported.
5. The case data published in wagscd.csv uses a different week definition than the Weekly Report weeks. Weeks for cases are defined as ending each Saturday, where the weeks in the Minnesota Weekly Report end on Wednesdays.