Measuring Mortality in Lost Time

July 19, 2020

Previous | Next

This much is obvious: Anyone who dies a preventable death from the pandemic is robbed of a part of their lifetime.  What may not be so obvious is that time and mortality are related in a more collective way. Any time that states lose in getting the virus under control can be measured in additional deaths.

Exhibit A here is the state of Florida, which was making some progress in slowing their covid-19 mortality rate between mid-May and mid-June. But since the recent resurgence, the state’s rate of increase in total deaths is 16% for this past week (4301 to 5002), which is back to about where it was two months ago. Think of that as lost time, time spent spinning one’s wheels getting nowhere, while people die. How many people? 3,038 in those two months alone. To use a military analogy, it’s as if an army lost that many soldiers while it sat in camp, accomplishing nothing.

It’s a similar story in Texas, Arizona, South Carolina and Alabama—brief progress followed by a reversion to a rate of mortality growth not seen in two months. Total lives lost in these four states in the meantime—6,093. And the toll in unnecessary deaths may well rise higher, since death is a lagging indicator of infections—and missed opportunities.

For the sake of comparison, consider the state of Massachusetts, which over the same two-month period brought its weekly increase down from 18% to 1%.  If it had stayed at 18% per week, the state’s total death toll would be 25,304 by now, instead of the actual 8,419. The other states with the worst spring outbreaks also got them under control during that period, saving many thousands of lives. New York and New Jersey moved especially fast, experiencing precipitous drops in new deaths even in early May.

The more effective states have emulated what many countries of the world have accomplished, while the less effective states have contributed to the conspicuously high death rate in the United States as a whole. Also contributing is the lack of federal leadership, which includes the President’s encouragement of states to lift preventive measures prematurely.

Back in March, I titled my earliest post on this topic “No Time to Lose.” That remains as true as ever.

Pandemic Mortality—A Selective Resurgence

July 12, 2020

Previous | Next

As we feared when many states lifted restrictions before seeing their Covid-19 caseloads decline enough, death rates are starting to accelerate in many places. I have been tracking deaths by state, using the weekly percentage increase in cumulative deaths as a measure. For example, total deaths attributed to the virus in Texas were 2,608 one week ago, but they are 3,112 now, an increase of 19%. Since the numbers are cumulative, these percentages never go down, but if they approach zero that means that deaths are coming to a halt and the pandemic is under control for now.

Texas reached a low of only an 8% increase for the week ending June 14, but mortality has accelerated since then. Other states in a similar situation are Arizona, Florida and South Carolina. Here I give the lowest weekly percentage increase and when it was reported, followed by the percentage increase for the most recent week.

Arizona: 13% on June 21, now 19%
Florida: 7% on June 21, now 13%
South Carolina: 8% on June 21, now 17%

Other states to watch:

California: 7% on July 5, now 10%
South Dakota: 5% on June 7, now 12%
Tennessee: 9% on July 5, now 16%
Utah: 8% on July 5, now 17%

Georgia’s increase has only been from a low of 3% for the week ending July 5 to 5% this past week, but the rapidly rising caseload suggests further increases in mortality to come.

The good news is that many states that already had serious outbreaks of the disease have brought their deaths practically to a halt. This week’s increases in cumulative deaths were near 0% in New York and Connecticut. They were around 2% in New Jersey, Massachusetts, Michigan, Pennsylvania and Rhode Island, and 4% in Louisiana.

In just a couple of months, the epicenter of the pandemic has shifted, mainly from the Northeast to the South. No law says it can’t shift again. The main lesson to be learned is that no state can afford to be complacent, and many are paying a steep price for their past complacency. The relentless campaign by the White House to understate the threat and resist national measures to deal with it is not helping.

Living Dangerously (part 4)

May 24, 2020

Previous | Next

As of this Memorial Day weekend, the US death toll from covid-19 continues to rise at a slower rate, both in absolute numbers and as a percentage. This week’s 8,812 new deaths represented an increase of 10%, compared to a increase of 10,239 (13%) the previous week. Slower increase was the story in most states, despite the easing of restrictions on public activity. All ten of the states with the highest percentage increase the previous week (listed in the previous post) had a lower rate of increase this week. A few other states, however, experienced a higher percentage increase this week: Arkansas, Indiana, Kentucky, North Dakota, Rhode Island, Utah and Wyoming. The good news is that no state has been experiencing a sustained acceleration of deaths.

Two words of caution, however. Some states have serious local outbreaks of infection, as in Minneapolis, MN and Montgomery, AL. These have the potential to expand into wider outbreaks. And since death is a lagging indicator of infection, it is too early to conclude that the economy is safely reopening without a resurgence of the disease. Be careful and stay safe this holiday.

Living Dangerously (part 3)

May 17, 2020

Previous | Next

Although total deaths from the coronavirus continue to climb, the rate of increase for the United States as a whole continues to fall. This week’s addition of 10,239 deaths represented a percentage increase of 13%, compared to a 19% increase the week before and a 23% increase the week before that. A month ago, for the week ending on April 18, the weekly increase was 90%.

The Institute for Health Metrics and Evaluation continues to project a gradual decline in deaths over the summer. However, it has once again raised its projection of total deaths from 134,475 to 147,040 by August 4, because mortality is not declining quite as fast as previously projected.

A comparison of state mortality data reveals an interesting pattern. The list of the ten states with the highest total deaths has not changed very much. It now includes New York, New Jersey, Massachusetts, Michigan, Pennsylvania, Illinois, Connecticut, California, Louisiana and Florida. The list was similar a month ago, except that it included Washington and Georgia instead of Pennsylvania and Connecticut.

However, the states with the highest weekly percentage increase in deaths are a completely different group. They include Iowa (37%), New Mexico (36%), New Hampshire (31%), Delaware (29%), South Dakota (29%), Arizona (28%), Nebraska (28%), Minnesota (25%), Missouri (25%), and Texas (tied with Alabama at 24%). This is clear evidence that the epidemic is spreading beyond the coastal cities and posing a greater threat to the interior and more rural regions of the country. Those states need to proceed with caution as they try to promote economic recovery without generating a new spike in mortality.

Living Dangerously (part 2)

May 5, 2020

Previous | Next

Two days ago, I reported that the Institute for Health Metrics and Evaluation at the University of Washington had revised its projection of the ultimate US death toll for this wave of covid-19 from 67,641 to 72,433. Given the number of deaths that had occurred already and the current rate of increase—about 1,700 a day in the preceding week—I didn’t see how that projection could stand up. IHME acknowledged that its model did not yet include increased interactions due to the easing of restrictions on public activity.

Yesterday, the Institute released a more sophisticated model that included greater attention to the mobility of populations as well as some other variables. The result is a dramatic increase in the projected death toll, to 134,475 deaths by August 4. As usual, there is a large margin of error, resulting in a range from 95,092 to 242,890. A death toll of 134,475 would far exceed that of any other country. It is about sixteen times the projection for Germany, although the United States has only four times the population of Germany.

The Institute said that its revision upward “is primarily due to longer peaks and slower declines for locations that have passed their peaks.” In addition, “for a subset of states, the easing of social distancing policies has begun and mobility patterns are on the rise.” Those states include Alabama, Florida, Georgia, Idaho, Louisiana, Minnesota, Montana and Texas.

The Institute did not predict an increase of deaths per day, only a slower decline, from about 1,700 now to about 800 by June 1. There is a much more pessimistic document circulating within the federal government that does project such an increase. Deaths per day would start going up in mid-May, reach 3,000 per day by June 1—that’s like having a 9/11 every day—and continue upward from there. The Centers for Disease Control says that this was only a working model that was not finished or ready for release. But apparently someone is taking seriously the possibility that the relaxation of restrictions will lead to a new surge of exponential growth.

The three scenarios I mentioned remain possible:

  1. The warnings from epidemiologists are exaggerated, and the lifting of restrictions will have little effect on the mortality trend
  2. The lifting of restrictions will raise the infection rate somewhat, but not enough to overwhelm medical systems or increase mortality dramatically
  3. The lifting of restrictions will touch off serious outbreaks with exponential increases in mortality in at least some states

Of the three, #1 looks increasingly unlikely, and the choice now facing the country is between bad and very bad.