Nursing home estimates of the day

Comments

Massachusetts has really shot up in the case count. They are now the 3rd worst state in the nation both on a per capita basis and a total numbers basis.

You are wrong, as Tyler has proven in his youtube series on the elephant's trunk, hind-legs.

My sister is a school nurse who has stepped in to work at a nursing home in Massachusetts. The virus is everywhere. But the nursing homes haven't stopped using circuit-riding therapists, doctors, and lab techs, who go from home to home, serving patients and spreading disease.
Meanwhile, half the low-paid aides have quit for fear of infecting themselves and their families.

Nursing homes typically have about 100 residents, and are economically marginal operations. Low paid aides and circuit riding techs are intrinsic to the business.

Mostly they provide “mom care”, i.e. unskilled personal care. Most of the residents are in a bad way, and no one gets better. Being an aide is a tough job, physically and mentally.

Your sister is to be applauded, and if a school nurse is at Medical Director level for these facilities.

My sister is a school nurse who has stepped in to work at a nursing home in Massachusetts.

Please thank your sister for me, on behalf of the people she's caring for. She's a better person than I am.

The U.S. has now passed Epstein's predicted total global fatality rate.

I guess putting all the old people in a care home is like putting chickens or pigs in a confinement building.

Except that if you go into a pig barn you leave your clothes, shower and put on clean garments. And if you are sick you can't go in.

There is an economic benefit for the farmer to implement working infection control. In some cases the herd is slaughtered, in other cases they can't sell the product.

There is an economic benefit (pension actuaries love this) in having nursing home infections. Every year.

There is no net economic benefit from others' death. The dollars the dead don't get go to someone else. That's from a utilitarian social welfare function.

Now, if one wanted to go with Man and Superman, by all means, let the weak die! That's not my social welfare function, though.

The dollars do go elsewhere if the resident was on Medicaid.

Just ask the Illinois pension actuaries if they wouldn't mind a few old pensioners bumping off.

I tend to look at revealed preferences. If old people were valued as much as you say why are they being neglected?

The folks putting their mom or dad in a home and then sheltering in their house are some cruel folks. Boomers...

When my parents were dying, it was a combination of their money and ours that let them do so at home. Now financial advisors tell seniors to transfer assets to their children so as to qualify for Medicaid admissions to nursing homes.

Priorities.

So, essentially we're destroying the economy because of nursing homes.

It would be nice if the most eminent economists in the nation, like Tyler, would be screaming this from the rooftops for all to hear.

Something rational along the line of "As the leading economic power in the world, it is IMMORAL and will cause MORE DEATH worldwide to spend Trillions in an effort to extend the lifespans of 100k 80 year olds by an average of 9 months."

Then prove it convincingly with a utility function.

Instead . . . we get what we have :(

Yeah this seems to be getting more and more absurd.

Think of them as
Dorms for the elderly
Without Cheetos
And with bedpans,
With the same amount of
Vomiting in public spaces.

Now, if one didn't know at the outset of the epidemic whether one would be in a nursing home during the epidemic, how would one's interests differ? :-)

Let’s just terminate them like the unborn children we don’t want to take care of... better yet, we could always eat them if there are meat shortages.

This is not a good plan. You can't profitably sell the parts of old people like you can sell the corpses of the aborted.

So costing 100-300 jobs per life saved from coronavirus (according to earlier MR post) is worth it? When half those lives are in nursing homes because the folks are incapable of caring for themselves, and will, on average, die in 1-3 years anyway?

I understand your moral posturing (I used to be on your side), but at some point the evidence has to show you that a cost-benefit analysis will likely not fall on the side of shutdown.

It was sarcasm. Given the state of affairs, we are running out of options... we are about a week away from ending many shutdown orders.

But either way, the virus is causing a deep recession shut down order or no shut down order. It’s a matter of which is worse, a homogenous shutdown order or a heterogeneous shutdown driven by individual action. Either way, not many people are going out to eat or to a movie or sporting event or getting a plane or cruise ship anytime soon.

Spain had a few situations where some caretakers were part timers working at multiple institution. A single worker is careless and catches the virus outside, and then infects elderly in multiple facilities, which then infect each other.

Spain's elderly care facilities also chose to hide the existing cases from residents and family. Phones are 'lost', and by the time the families get a proper hold of the facility, the residents are dead or dying.

It should be noted for scale that about 4% of the elderly in the US live in nursing homes*, so numbers like half the deaths in MA being in long-term care facilities really are staggering.

* "According to the U.S. Bureau of the Census, slightly over 5 percent of the 65+ population occupy nursing homes, congregate care, assisted living, and board-and-care homes, and about 4.2 percent are in nursing homes at any given time. The rate of nursing home use increases with age from 1.4 percent of the young-old to 24.5 percent of the oldest-old. Almost 50 percent of those 95 and older live in nursing homes" https://www.agingoptions.com/blog/2011/10/21/how-many-seniors-really-end-up-in-nursing-homes/

Death doesn't become us: https://www.cdc.gov/nchs/nvss/vsrr/covid19/index.htm

Nursing homes are infected with all kinds of respiratory pathogens. SARS/CoV-2 just adds to this burden. The Incidence of pneumonia in the long-term care setting may be as high as 100-300 cases per 1000 nursing home residents per year.
Residents are the most vulnerable and the most concentrated, and facility management is often sub-par, altogether a deadly combination .
https://www.managedhealthcareconnect.com/article/bacterial-pneumonia-nursing-home

I think another part of it is that so many residents are, if not eager to die, waiting to die.

I base this assertion on the admittedly small sample size of the two people I know who ended up in nursing homes, but neither wanted to live, and that has to hurt your ability to fight off infection.

thank you!

now, what percentage of the remaining deaths from c19 were nosocomial infections?

that number will vary by region and explain much of the remaining heterogeneity.

1. Would the care plan have actually prevented infectious disease, or would it have just been more paper?

2. My mother turned up with the flu in the weeks before she died, in a nursing home in which every member of the staff made a satisfactory impression with one exception (the social worker, natch).

3. What portfolio of measures do you want, can you implement them, and what's the extra cost going to be?

The extra cost is paying workers, including construction workers to build larger facilities, and the high costs of paying workers kills jobs and cuts consumer spending. Only by not paying workers can consumers spend more money - the tax cut from not paying the payroll tax puts more money in the pockets of consumers than the wages the payroll taxes tax.

Clearly, old people need to be forced on their kids and grandchildren for housing, food, and changing bedpans, like in the good old days before the New Deal. Whose bad idea was it to allow kids to leave their home to go to new places with better job opportunity???

Are nursing homes a market for lemons?

They should send the elderly back to their children
Where they can live in the basement
With their grandkids.

Sort of like the Simpsons
When Grandfather visits.

So, NYC and IFRs:

Apparently "Cuomo just announced a little over 21%" seroprevalence in New York City: https://twitter.com/inschool4life/status/1253381963777486850

0.21* 8,398,748 = 1763737.08

11,267 confirmed NYC deaths / 1763737.08 = 0.61% (per Google for NYC coronavirus death)

If excess deaths under-count by 20% you could get up to 0.7%.

OTOH, under a converse assumption if 30% of patients don't develop a strong seralogical signal despite shifting the infection, as indicated by Chinese data, true IFR is probably 0.42%.

If serological signal lags more than deaths, you could also lower that number (since the true infected relative to deaths would be higher).

Maybe higher IFR than I thought (banking on around 0.3%), actually, but sub 1% seems pretty likely. Certainly the ballpark of what the systematically less wrong than others were probably expecting.

New York State is on about 14% vs 15,302 deaths. 19,378,102 *0.14 = 2712934.28. 15302 / 2712934.28 = 0.56% IFR. Similar number.

For NYC
Confirmed deaths:10,290
Probable deaths: 5,121

Probable deaths are "sure, we'll take the Federal Covid Cash"

Confirmed death means anyone who had Covid at the time of death. Imagine 20% of people had it in the last 45 days in NYC. How many people died with Covid but not from it in NYC?

I don't think CFR of 0.1% is unreasonable at all given how they count. But I could see your 0.7% as an upper bound if i squint real hard.

Eh, I do get the skepticism and the perverse incentive at play, but in the conflict between under-counting and over-counting, seems like we have to be willing to count some excess deaths with no explanation beyond the margins of the confidence interval. The alternative being that tons of people died suddenly for no good reason. So long as they're not going beyond that to the extent where they would have to be suggesting that deaths are *lower* than normal net of corona... Whether those are actually corona deaths given high levels of co-infection is a bit more of a technical question, though (if like 20% of corona deaths had flu as well, can we really chalk them up under corona's IFR?).

0.1% seems a little low on this evidence, but perhaps if you put together false negatives (where subjects shifted the infection but don't generate a serological signal, which might be quite common in respiratory tract infection from what I have heard), and maybe some more serological lag than death infection lag, you might be able to get to 0.25%. If those all did apply (it's not certain they do). Which would be closer to "Wu-fu is two-flu", just with a very high susceptible population rate than it is to "Corona is 50x more deadly than the flu" (5% IFR).

Absolutely. Raw deaths vs expected is a pretty good metric for a period. So, too, would raw deaths vs other periods of a high flu.

How many deaths did New York have from all causes in January, February, vs March and April?

I found this helpful for Europe.

https://euromomo.eu/graphs-and-maps/

If you compare total deaths to early 2017, you'll see Europe at 73k vs 70k.

So it is a little bit worse than 2017. Italy and France look about twice as bad as 2017

If I take the Bronx which is probably the worst borough in NYC. It has
205 deaths per 100k population to date. Because aB take about the same time from symptom offset to appear as deaths do , we can use them as a proxy.
Assuming the 21% aB test results are correct. Not guaranteed they often have errors.
In the Bronx IFR = (205/100,000) * (1/0.21) = 976 per 100k = 0.97%. I would just take this as a ballpark not a precise number.
The data is from there
https://gothamist.com/news/coronavirus-statistics-tracking-epidemic-new-york

Tho, why take prevalence estimate from city and then apply to the Bronx alone rather than whole city? Seroprevalence rate for city surely an underestimate for worst hit borough.

False positives could be true, though supposed to be around 1% of results (which was the criticism of the Santa Clare study, within bounds of false positives)? Seems unlikely to greatly change results.

Because I am doing the worst case and the Bronx seroprevalence is not known separately. I assume since the Bronx is the poorest borough, it is likely to have the worst IFR .
For New York City as a whole:
total deaths: 10290 + 5121 ( probable) = 15411
population =8,399M
death per 100 K =183.4
using aB adjustment (0.21) -> IFR =0.87%.
Again these are not precise numbers, only estimates
https://www1.nyc.gov/site/doh/covid/covid-19-data.page

since the Bronx is the poorest borough, it is likely to have the worst IFR .

In what way is it "the poorest borough"? Has there been a study made of the incomes of the infected and deceased in all the boroughs? Where do these people fall in the borough income spectrum? What mechanism would make a poor person, whatever that is, more likely to be infected than anyone else? What are the dividing lines in income between dead, infected, infected but recovered, asymptomatic, uninfected and uninfected but sure to be infected?

The Bronx has the lowest household income of the 5 boroughs. income is negatively correlated with obesity and diabetes, which are risk factors for this disease.
Infection rates I don’t know and number of cases is not reliable data.

70% in Minnesota, as of a couple days ago:

https://www.startribune.com/minnesota-long-term-care-covid-19-deaths-draw-focus/569836902/

Osterholm had a program on this several weeks ago as to where to go first with testing.

Osterholm was wrong a second time a week ago on MSNBC in saying that East Asian countries have lost control of coronavirus. There is zero evidence of that as Singapore has 11 deaths, Hong Kong has 4 deaths, Taiwan has 6 deaths, South Korea has 240 deaths and Japan has 300 deaths with no evidence of cases increasing at a faster rate. I'm not sure why he keeps making up the story that these countries loosened restrictions and then lost control of the situation.

You are absolutely wrong.

In fact, Osterholms prediction stands. It look like you looked at the data the day he made his prediction, and not what he predicted. Google "Singapore covid deaths" You'll see the spike and where they are today.

Here is the Johns Hopkins data and look at the graph on the right. https://coronavirus.jhu.edu/map.html

Next time update your information and check the data against the prediction.

I'm not wrong, Osterholm was and he repeated it again. He wasn't making a prediction either but misstated what was happening in five East Asian countries. Show me any evidence that those countries have lost control of the virus in Korea, Taiwan or Hong Kong. There is no spike in Japan or Singapore either. There was an increase in Singapore's cases for after he said that April 9th.

Singapore cases from Mar 27 to April 8 increased on average 8% a day and there was no change in this when Osterholm said Singapore had lost control of coronavirus. The average has increased to 14% a day on average from April 9 to April 23. That 6 percentage point increase is not a situation out of control with a grand total of 11 deaths over an 11 week period. NYC may have 11 deaths on a ward in a hospital a day.

Todd, Did you look at the charts at Johns Hopkins? Look at the curve.

SINGAPORE: Singapore confirmed 1,037 new cases of COVID-19 as of noon on Thursday (Apr 23), taking the national total to 11,178.

This is the fourth day in a row that Singapore has reported a daily increase of more than 1,000 cases. " https://www.channelnewsasia.com/news/singapore/covid-19-singapore-moh-new-cases-apr-23-workers-dormitory-12669086 This article also shows the rapidly rising curve.

By the way, saying an 8% to 14% is only a 6% increase misleading, as it is compounding and going exponential.

Todd, what is the rate of change from 8% growth to 14% growth in covid cases..

Remember what you learned about acceleration? And, remember each of those persons is in contact with others.

Osterholm is just making shit up...twice.

This is what has floored me. Public health authorities surely have some responsibility here. Why has it taken high level political power to impose simple infection controls to prevent this? It took an order from the Premier of Ontario to have people not working in multiple homes, and the order came 3 weeks or more into the pandemic.

It is almost that the people who need to know early and urgently are the last to know.

These are private nursing homes.
They do not like regulation.

Google what the median time to death is in a nursing home.

One study from a few years ago put it at 5 months. If 25% of nursing home patients become covid infected, that would be a YUGE number of deaths that would have happened anyway, but are now blamed on Covid.

That's why I've been harping on cost benefit analyses that account only for total deaths rather than excess deaths. The net benefits of those analyses thus far are completely overstated because they are adding deaths that would have occurred this year anyway.

That’s so irrational! “ Catinthehat” assured us yesterday that the number was 13.5 years. And “ Catinthehat” was very sure, impossible s/he was wrong.

a). I don't claim infallibility far from it
b). This was a calculation done 4 weeks ago based on the CDC report of deaths demographics available at the time
If the demographics of deaths have now changed because the virus is not infecting all age groups proportionately, but instead infecting older groups at an increasingly higher rate due to the nursing home situation, the calculation may change.

+1. Here is that study: https://www.ncbi.nlm.nih.gov/pubmed/20738438

...here, for non-paywall full version: http://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC2945440&blobtype=pdf

NH care facilities represent half the confirmed COVID-19 cases and deaths. Initially in just two facilities, then in a disabled youth facility, and now in one or two other facillities.

Seems like a third of the staff have confirmed diagnosis of COVID-19 for the earliest homes, more than the number of residents.

All facilities have physical therapy and occupational therapy personnel visiting patients regularly, so that is an obvious vector. PT deals with getting patients moving after injury; OT with figuring out accommodations to allow at least self care, like dressing, eating, combing hair, brushing teeth. The former requires significant physical contact, but even the latter involves.

Others visiting are equipment providers, eg, oxygen generators, bottled oxygen, etc, cleaners, etc. This has been driven by industrial "division of labor", ie, eliminating JOATs from full time staff who get paid higher wages because they can do many tasks like handling equipment and cleaning, or providing all patient hands on care.

Ie, wasteful spending to have ten staff of high skill when 40 staff of low skill can do the work part time at low wage while a single high skill nurse supervises without any patient contact, or observation of care.

Contrast this information with NYC with a much larger swaths of deaths. This explains a lot of the two different sides of the Covid debate.

I think maybe New York (the city in particular) might have transferred a lot of the nursing home illnesses to actual hospitals where the patient then dies and is counted as a hospital death. This might not be a general procedure. It migh explain New York's high hospitalization usage, especially at the start.

How many deaths have occurred in American Nursing homes in the last 45 days that were not blamed on Covid?

Are nursing home facilities now incredibly safe butfor the Covid Grim reaper? What incentives are being setup to encourage reporting deaths as Covid vs actual cause of death at these facilities? How about vice versa?

The median age of covid death is around 80.

Only the tiniest fraction of the many billions of dollars spent/lost locking down a largely invulnerable population of under 50-year-olds would have been sufficient to staff all nursing homes with live-in health care providers that tested negative on entering, perhaps at double or even triple their ordinary pay, and could have been continued for several months, or for as long as we needed until a better way could be found. This was a profound public policy failure and tragedy, one that practically ensured they would be hit first.

I agree completely, and there isn't even the excuse that we didn't know this 5 weeks ago. Italy's doctors said as much in mid March in numerous forums.

More than half the deaths locally, have been connected to nursing homes. They may include in that total a staff member who died, I'm not sure. One of the deceased was profiled today with the arresting headline, Sisters die in pandemics 102 years apart. Before her birth, an older sibling had died in the flu pandemic.

I sense that most people are not really aware of this nursing home skew, and if they were, they would have no real means of making sense of it.

Even though I cannot locate tragedy in their deaths, beyond their essential aloneness, perhaps, not just in death but probably in having outlived nearly all the people they knew, perhaps even some of their children - still it gives one a pang knowing that *very elderly* people had more interesting lives than most of us left behind.

It feels like that being the case - that repeated generational difference - is passing as well, as we are now more homogenized across time as well as space.

I applaud your sensitivity and insight. You must be female. I fear you don't have a place here in the Chicago School Marginal Revolution commentariat.

Looking at the data in Illinois it also seems clear that Covid Death rate is around 70% or so consist of people older than 70, which assume mostly in nursing homes or have other heath issues. So it brings forth a very touchy question, have we just shut down the economy for a real small part of the population. More specifically the whole state is in lockdown mode when most of deaths are in 2 counties. These are questions the media does not want to go into.

I think the deaths in Sweden are like 75+ percent from nursing homes and hospice care facilities. The deaths in Sweden are skewed towards very, very old people.

In Colorado, 63% of COVID deaths have been in nursing homes.

According to https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1532-5415.2010.03005.x, over half of nursing home residents died within six months of admission.

Comments for this post are closed