- Analysis shows that causes other than the SARS-CoV-2 pandemic have been more responsible than COVID-19 for the “significant increase in deaths” occurring in NC
- A very troubling implication of the findings is that the potential deadly effects from Gov. Cooper’s restrictions are still ongoing, but to an unknown extent
- This brief gives the data used and explains the analysis further
To recap the series so far: Part 1 on March 4 discussed research findings and experts warning that government lockdowns and severe personal and business restrictions may be deadlier than the virus they’re meant to protect people from, including in North Carolina. Part 2 on March 5 used estimates from the Centers for Disease Control and Prevention (CDC) and data from the North Carolina Department of Health and Human Services (DHHS) to see whether and to what extent North Carolina has been witnessing excess deaths outside of deaths attributed to COVID-19.
The analysis in Part 2 made these significant findings. Without even counting COVID deaths:
- NC has seen at least above-average deaths since the very first week of Gov. Roy Cooper’s shutdown
- NC suffered excess death events in April, May, and steadily since July
- These excess deaths looked to be getting worse by the week ending Sept. 19 (the last week for which NC’s all-cause death data seemed complete)
- Worse, NC was witnessing more excess deaths than COVID deaths
On March 9 WRAL published an investigative report under the headline “‘Jaw dropping’ increase: 17% more died in NC in 2020 over 2019.” The report raps the state for its “Lag in reporting” and used data from Buncombe County’s register of deeds Drew Reisinger, who had obtained death data from all 100 counties. Here is what they found: deaths spiking across the state that could not be blamed on COVID:
Without the more than 7,000 deaths in the state which all noted COVID as a cause, the death rate still would have increased by more than 8 percent.
“There is a death toll associated with this even if it is not COVID itself,” said Reisinger. “The death spike is jaw dropping.”
When responding to inquiries about a year-over-year increase, more than a handful of register of deeds offices told WRAL Data Trackers they saw an increase in what are often called “deaths of despair” – those linked to drug overdoses, suicides or alcohol addiction.
For this series, Part 2 included the following graph. It starts with a baseline represented the average amount of deaths expected in the week. Anything above that baseline would mean above-average deaths. It also has an orange line representing the upper range of average deaths; anything above that line signifies excess deaths (which the CDC calls “a significant increase in deaths“). The bar graphs for that week showed COVID deaths in gray on top of deaths from all other causes in blue, and they are plotted against the baseline and orange line.
Part 2 also explained how to interpret the graph:
If the blue part of a week’s bar is visible at all, it means the week had above-average deaths that can’t be blamed on COVID. If any part of the bar exceeds the orange line, it means the week had excess deaths. If the blue part of a bar exceeds the orange line, it means the week had excess deaths that can’t be blamed on COVID — i.e., that something other than COVID is also responsible for excess deaths in NC that week.
Part 3 of this series will provide the data from the CDC and DHHS used to make the graph. It will use a sample week to demonstrate how to apply the data. An appendix will explain the different categories. The hope here is to bring more clarity to what is a very real threat still facing North Carolinians.
Providing the data and a demonstration
The table below contains the weekly data from the CDC and DHHS and calculations used to produce the graph.
Note: Although the graph begins on the week ending March 28, the table provided here starts in the very first week of 2020 to give an idea of the data pre-Covid. The appendix below explains what each column means.
To understand how to view the data, let us take a look at the week ending Sept. 19. According to the CDC, the average deaths expected from all causes that week was 1,765 deaths. The upper bound threshold for excess deaths that week was 1,871 deaths.
The actual number of deaths for the week, however, was 2,206 deaths. Since that’s well above the threshold for excess deaths, it means North Carolina had experienced “a significant increase in deaths” that week.
According to DHHS, 185 of those deaths were deaths with or from COVID-19. Subtracting out those 185 Covid deaths from the 2,206 observed deaths from all causes, that means North Carolina had 2,021 deaths that week from all causes except COVID-19.
Notice that 2,021 non-Covid deaths is still well above the upper bound threshold of 1,871 for excess deaths. The implication is that North Carolina had experienced “a significant increase in deaths” that week even without counting Covid deaths.
Going further, the total number of excess deaths in North Carolina that week was 335 excess deaths (2,206 deaths minus the 1,871 upper bound threshold). The amount of Covid deaths was 185 deaths. That means there were 150more excess deaths than Covid deaths that week.
Since the amount of excess deaths was greater than the amount of Covid deaths for the week, it warns that causes other than the SARS-CoV-2 pandemic were more responsible for NC’s “significant increase in deaths” occurring that week than was COVID-19.
As stated in Part 2 of this series, these findings suggest the effects of Cooper’s lockdown and severe personal and business restrictions have indeed been deadly, confirming the warnings from scientists, physicians, economists, mental health experts, and others over the past year.
A more troubling implication of these findings is this: while DHHS’s reporting of deaths data to the CDC ended months and months ago, Cooper’s restrictions were worsened, and though they have very recently been relaxed somewhat, they are still ongoing. So their potential deadly effects are still ongoing. We just don’t know to what extent.
But as Reisinger told WRAL, “There are a hell of a lot of people dying right now.”
Appendix: Definitions for the columns presented in the table
Average deaths expected. The average amount of deaths from all causes expected in the state for that particular week of the year, calculated from “historical trends” in actual deaths in the state for that week going back to 2013. (The CDC data are available here; look for “National and State Estimates of Excess Death” under “Download Data.”)
Upper bound (excess deaths threshold). The threshold for excess deaths. It is a “one-sided 95% prediction interval of these expected counts” that is used “to determine whether a significant increase in deaths has occurred” (emphasis added).
Actual number of deaths. The observed number of deaths from all causes reported by state health officials to the CDC for the state for that particular week of the year.
Covid deaths. The number of deaths attributed to COVID-19 (either with or from) for the state in that particular week, according to DHHS.
Non-Covid deaths. The actual number of deaths in the state for that particular week minus the week’s Covid deaths.
Excess deaths. The amount of actual deaths from all causes in excess of the upper bound (excess deaths threshold) in the state for that particular week of the year (quantifying the extent of that “significant increase in deaths”).
Baseline. Set to zero for each week by subtracting the week’s average expected death count from itself. The purpose is to provide a flat baseline and focus attention on any and all deaths occurring in the particular week that would be above average or excess.
Excess deaths threshold. Conforming to the baseline by subtracting each week’s number of average expected deaths from its upper bound. The interpretation remains the same; deaths above this threshold would be considered excess deaths and alert that “a significant increase in deaths has occurred” in the week.
Above-average deaths (non-Covid). Conforming to the baseline by subtracting each week’s number of average expected deaths from its non-Covid deaths. If the number is positive, it means there have been above-average non-Covid deaths for that week. If the number exceeds the excess deaths threshold, it means there have been excess deaths for that week that cannot be attributed to COVID-19. It means that “a significant increase in deaths has occurred” in the state in that week for causes other than the SARS-CoV-2 pandemic.
Excess deaths vs. Covid deaths. Subtracting deaths attributed to COVID-19 in a particular week from excess deaths. If the number is positive, it means that causes other than the SARS-CoV-2 pandemic are more responsible for the “significant increase in deaths” that occurred during that week than is COVID-19.