10 June 2021: How Big Is Michigan's Undercount of COVID Nursing Home Deaths?

Problems with state stats on nursing home deaths due to COVID-19

This article is by Stephen Delie of the Mackinac Center for Public Policy, whose investigation with Charlie LeDuff turned up a very large gap in Michigan's official count of COVID deaths among residents of its long term care facilities.

In this first excerpt, he describes the kind of long term care facilities where Michigan Department of Health and Human Services officials have exempted from reporting COVID death data:

Adult Foster Care and Homes for the Aged are also required to report data, but only to MDHHS. Additionally, these facilities are only required to report death and case information if they were licensed by the state Department of Licensing and Regulatory Affairs to house 13 or more residents.

Smaller facilities are not required to report, despite accounting for 76% of these facilities in Michigan. In other words, 3,480 smaller long-term care facilities, serving more than 22,000 elderly Michiganders, are not being tracked.

This is the first part of the problem. Because these facilities are not required to report COVID-19 deaths, there is no way to know how many people have died at these facilities from this disease. This is important information, especially valuable to inform policymakers regulating these facilities.

Delie then describes how death certificate data may be used to improve the accuracy of the state's accounting of COVID nursing home deaths:

... death certificate records could be used to determine how many residents of these facilities died from COVID-19. The state regularly reviews these records and relies on them as a source for the state’s total tally of COVID-19 deaths. But MDHHS has not tried to link these death certificate records to long-term care facilities since last June, citing time and cost constraints. This is the second part of the problem.

The state is not counting any of the COVID-19 deaths it finds by reviewing death certificates to the tally for long-term care facilities. When the state had done this previously, it found that 44% of death certificates were for residents of long-term care facilities. Because of this, it is very likely that the state’s overall tally of deaths of residents of long-term care facilities is an undercount.

Creating more accurate data is simple. The health department could simply check if the address of the deceased listed on each death certificate matches an address of a licensed long-term care facility.

If it does, and it was not already reported directly by a long-term care facility, the death would be added to the state’s total long-term care facility count. Again, the state used to conduct this type of analysis, so it is obviously possible in practice.

Of course, as Director Hertel pointed out in the committee hearing, this will not provide a full accounting. If, for instance, the address on a death certificate is listed as the deceased’s residential home, rather than their long-term care facility, it would not get classified as a death at a long-term care facility. But that’s not a reason not to do it. The resulting data would still be more accurate than the state’s current figures.

Michigan's official tally of COVID deaths among residents of nursing home and other long term care facilities in the state is 5,900. Delie's analysis suggests a more accurate count could add as many as 3,000 deaths to that total if Michigan's public health officials were to ever complete the death certificate review practices they abandoned.