5 Levels of Risk Analysis: We’re All Going to Die

We all know people in their 50s who, having either come down with diseases, abused their bodies, or both, are very unhealthy, and people in their 70s or 80s who are vigorous. A healthy 80 year old might be more likely to live longer than a 50 year old with cancer and heart trouble -no surprise there, I should hope. And either of them might be hit by lightening or fall in the shower or experience any one of the myriad accidents that put many of us on the slip n slide to the grave.

Here’s an about-to-turn-eighty year old who probably is about as likely to die this year as I am, a just turned 63 year old. I doubt anyone would be surprised if she made to 100. But she probably won’t.

It would seem best, then, to assess our risk of death over the next year as follows:

  • Am I healthy? If so, my risk of death in the next year is very low, lower than the average represented in actuarial tables. Those tables include both healthy and unhealthy people, and the unhealthy drag the average down. So, I need to go live! Safety 4th!!
  • Am I over 50? If so, I need to knock off the unhealthy behaviors that weaken me – really, good advice at any age, just starting to get serious around 50. Get some exercise, eat better, get those 8 hours of sleep. AND devote more time to relationships that are fulfilling and comforting and stress reducing. Church going people with loving families live longer – go figure.
  • Am I seriously ill/unhealthy? If so, I need to do what I can to improve, if possible. Provided I’m doing what I can to get healthier/mitigate the damage, I need to live now. If I become obsessed with my health, my life will become merely a living death of fear. I am going to die – am I going to live?
  • Am I over 70? If so, my affairs need to be in order, because my body will start shutting down eventually, probably sooner rather than later, and diseases will likely get me. Smile more. Hug somebody. Hang out with friends and family. Live! Pray! Eat!

And – that’s about it. Health is for living. If we’re not living well, what does it matter how healthy we are?

BONUS: Level VI – Use Your Brain: There is no such thing as a statistical analysis of a single person’s risk. All that can be said is that, for groups of people more or less like you, certain things did and did not happen. Whether those things will or won’t happen to you is simply unknown until after they do or don’t happen. But the experience of others can help point you in the right direction. To a very large extent, what happens to you is within your control, excepting, of course, that you will die eventually. Do things that are good for you; avoid things that are bad for you. Figure out the hand God dealt you – if you have shaky health, the things that are good or bad for you might need modification. Live, love, pray, be prudent, and don’t worry.

Now the present pertinent application of this wisdom: Does COVID change any of this? In other words, is COVID a significant additional risk, or just something that adds a little to the already existing risk?

Using the publicly available data from the CDC, the Imperial College report, the news and taking it at face value: from the beginning, the majority of deaths form COVID have been people already very sick with other problems. How do we know this? Two ways: 1) nursing homes make up something like 2/3rds of all deaths in countries where the infirm elderly are sent to nursing homes to die. Nursing homes are full of very sick people, so weak and ill that they can no longer take care of themselves and need expert help just for daily survival. 2) Two outbreaks took place under contained conditions on board ships. On both the Diamond Princess cruise ship and the aircraft carrier Theodore Roosevelt, people were often gathered together in closed areas, eating together, passing each other in narrow halls. These provide near-perfect conditions for the spread of an airborne. Further, these ships, with their known and largely isolated sample populations, take a lot of uncertainty out of the analysis.

On the Roosevelt, crewed by about 5,000 people each of whom was, presumably, healthy enough to crew a ship, 25% caught the bug; 55% who tested positive developed any symptoms; 3 got seriously ill; one died.

So, among the presumably reasonably healthy population of 5,000 Rooselvelt crew, the case and infection fatality rate was 0.079%; the population fatality rate was 0.02%. To put it another way, based on these numbers, about 8 out of every 10,000 people who caught the virus would die; about 2 out of every 10,000 people exposed to the virus would die. Note that this was very early on, when all treatments were necessarily unproven and experimental. For reference, the Wuhan numbers were used to suggest about 250 out of 10,000 people who caught the virus would die from it.

On the Diamond Princess were 2,666 passengers and 1,045 crew, including many elderly people. 14 people died. Here’s the data:

#DateAgeSexFromNotes
12020.02.2087MJapan (Kanagawa)Hospitalized on 2020.02.11[c]
22020.02.2084FJapan (Tokyo)Hospitalized on 2020.02.12
32020.02.2380sMJapanHospitalized on 2020.02.05
42020.02.2580sMJapan (Tokyo)Hospitalized on 2020.02.09
52020.02.2870sFJapan (Tokyo)Hospitalized on 2020.02.07
62020.02.28MUnited KingdomFirst death of a UK citizen from COVID-19
72020.03.0178MAustraliaEvacuated to a hospital in Perth, Australia[e]
82020.03.06MHong Kong
92020.03.1970sMCanada
102020.03.2270sMJapan
112020.03.2270sMJapan
122020.03.2860sFHong Kong
132020.04.09Japan
142020.04.1470sMJapan (Chiba)Hospitalized on 2020.02.07[f]

The families of three of the dead declined to give ages, several others declined to be precise; the average age was roughly 78. None of the crew, with an average age of 36, died. All 14 deaths were passengers. Again, crew is presumably reasonably fit, while you don’t need to be very healthy to go on a cruise. The infection and case fatality rate on the Diamond Princess was 1.97%; the population fatality rate was 0.38%.

1.97% death rate is pretty scary – that’s about 1 out of every 50 people in the sample. I could do the math for each person, but this is already a long post, and the purpose here is to get a feel for how to think rationally about risk, so I’m going with generalizations. To get an idea of how many people in a representative sample of people around the average age of the passengers would be expected to die in the normal course of things, we might look at their underlying actuarial fatality rate. To get a rough idea, using the US actuarial table (which is not materially different from any other first world nation’s for our purposes), for 78 year old men and women, it’s a 4.78% and 3.43% annual fatality rate, respectively. That means that, out of 1o,000 78 year old men. about 478 are expected to die within the year; and out of 10,000 78 year old women, about 343 would similarly die.

So, based on age and sex alone, the people who died of COVID on the Diamond Princess were about 1.5 to 2.5 times as likely to die of of something else over the next year than to die of COVID during the about 3 months over which the passengers got infected, came down with the disease, and died. Look at it this way: if about 40 out of 1,000 78 year olds are going to die over the next year, then 10 would be expected to die over a 3-month period, on average. 14 died on the Diamond Princess, only 4 more than would have been expected to die anyway. All things being equal, which of course they never are.

We know, from the discussion above, that age alone is not definitive as to one’s risk of death. So: were these 14 any sicker than average? The case fatality rate over 3 months is remarkably similar to 25% (3 months out of 12) of the annual actuarial rate. In other words, were some or all of these 14 unfortunate people among those with a high risk of dying in 2020 anyway, regardless of catching the virus or not? In a better world, that would have been the first thing investigated and published.

Another point: the average age of the passengers was 69. Even living in close quarters on a cruise ship, most did not catch the virus; over 98% of those who did recovered.

Recap: We’re all going to die. Actuarial tables give us a useful starting point for how many people of a particular age and sex are likely to die over the next year of their lives. Actuarial data lumps healthy and unhealthy people together. At any age, a healthy person’s chance of dying will be lower, possibly much lower, than the applicable actuarial death rate; a unhealthy person’s will be higher, possibly much higher. Ultimately, your individual risks are merely hinted at by the statistics, which should be prudently considered in light of our individual lives. All the data, taken at face value, (2) from Wuhan out, says: COVID adds a small additional risk to the preexisting underlying risk of death for anyone.

Finally, why would anybody sane decline to be vaccinated against COVID? Setting aside any possible mid-term and long-term side affects, which are by definition unknown – the drugs have only been around for a short while, so what happens in a longer while simply cannot be known – and ignoring reports of short-term deleterious affects, let’s talk simply about effectiveness. People hear that a drug will lower their chances of catching, and therefore dying of, COVID, and that’s enough for them.

But is even a 100% reduction in risk always a good thing that compels action? Let’s assume there exists a 100% effective shark repellent. Rub this stuff on, and you’ve completely eliminated your risk of getting eaten by a shark. Would you feel morally obliged to rub it on?

Human beings run about a 1-in-2 billion chance per year of getting eaten by a shark. In numbers:

  • chances of getting eaten by a shark, no repellent: 0.0000000005.
  • chances of getting eaten by a shark, w/ repellent: 0.0000000000.

A difference of 5 parts per billion is meaningless. Those are the same number for any meaningful assessment of risk.

Of course, many of us rarely if ever swim in the ocean, so our risk is even much smaller (but never zero!) than that. Maybe, just maybe, if you plan to surf in the relatively shark-infested water of Australia, for example, you might take the time to rub on the repellent. Otherwise? You’re wasting 2 minutes of your life you’ll never get back.

Back to COVID: The CDC says the population fatality rate for COVID in the 55-64 age range I fall in is 0.0012, or 0.12% Twelve out of every 10,000 Americans age 55-64 died with COVID in 2020. (Sloppy numbers, grouped by large 10-year bands. But that’s how the CDC reports them.) Applying the 2/3rds heuristic – that 2/3rds of the deaths have been in nursing homes, and I am not in a nursing home – we can, for a first pass, reduce that rate by 2/3rds, to 0.08% The actuarial risk of death as a 63 year old American male is about 1.4% This rate is pre-COVID, so, assuming (and it is quite an assumption!) that the rates are additive, then, for starters, the risks are:

  • base actuarial likelihood of death: 1.40%
  • base rate plus adjusted COVID risk: 1.48%

Brief aside on various fatality rates: all the fatality rates – case, infection, population – are backward-looking. At this point, the virus has been ‘raging’ through the population for well over a year and completely petered out over the last few months. These rates don’t reflect what’s happening now, which is pretty much nothing, but include what used to be happening. But of the 3 rates, perhaps it is most meaningful to look at population fatality rates. If you or I haven’t already had it – remember, 99%+ of all COVID infections are asymptomatic or generate only mild symptoms – it is getting more and more likely we won’t catch it, what with huge numbers already having had it or getting vaccinated. The population fatality rate reflects the reality that a huge number of people simply won’t get the disease. But all these numbers are based on activity that is no longer happening. The real risk, today, is effectively zero – effectively, because no logically possible cause of death ever has a zero possibility of killing you.

Anyway, back to the quantified risk. Those are not very different numbers. On the one hand, since I’m fairly healthy (I need drop quite a few pounds, which is serious, but am otherwise good), my underlying risk is probably a lot lower; on the other hand, the risk of COVID to reasonably healthy people is also much, much lower than for already sick people. Do these factors offset each other? Who knows. So let’s just go with this.

Let’s say I take the most effective vaccine by the current, still very preliminary, and constantly updated evidence: Pfizer. Doing so will reduce my risk – on average, based on preliminary numbers – by 95%. The theoretical 0.08% risk is reduced to a 0.004% risk. Thus:

  • base rate plus adjusted COVID risk: 1.480%
  • absolute risk reduced after Pfizer: 1.404%

This means – and remember, all these numbers are just estimates, overestimates for healthy people, underestimates for seriously sick people, and based on what happened in the past, not the nothing happening now – that, for 63 year old American men, on average:

  • without the Pfizer vaccine: about 1,480 per 100,000 will die per year;
  • with the Pfizer vaccine: about 1,404 per 100,000 will die per year.

While a 95% reduction in relative risk sounds good, what it really means is (1404/1480) about a 5% absolute reduction in risk.

So, I’m looking at a possible 5% reduction in my absolute risk of death, under a set of outdated, overly pessimistic assumptions. I could get that much reduction by skipping desserts and getting in some more walking.

Could I die of COVID? Sure! I could also get eaten by a bear. Here’s the thing: even if I caught COVID and died, it doesn’t change the reality. I’m a reality guy. I am going to die of something. The microscopic risk of COVID being the Grim Reaper’s weapon of choice doesn’t worry me at all.

What does worry me: the rabbit-like surrender of our rights and our country because we were told to be very, very scared.

  1. Thus, the spectacle of children riding scooters without a helmet, but wearing a mask; chain smokers pulling down their masks to take one drag after another; people pulling a mask out of their pockets as required, then stuffing it back in their pockets when done, all the while maintaining proper social distance. From a tribal perspective, these behaviors are understandable; looked at from the point of view of overall risk, they are insane.
  2. By ‘taken at face value’ I mean ‘ignoring the glaring problems in and egregious misuse of’ the data, leading to an inescapable overstatement of risk. In reality, once the numbers are properly understood, vetted, and cleaned up, the added risk of COVID is much lower for 99.5%+ of us than even the low risk explained above.
  3. Note that all these statistics from the CDC use deaths ‘involving’ COVID as the numerator. As discussed elsewhere, all this means is that COVID showed up on a death certificate in either section A – immediate sequence of events that lead to death – or section B – conditions that may have contributed to death. So a person who had a massive stroke and then died later in a nursing home, but had two or more symptoms of COVID (all that’s required for a COVID diagnosis – no positive test results needed) is all but guaranteed to have COVID show up, at least in section B. Since those symptoms include fever, aches, breathing troubles – symptoms shared with everything from colds and flues to working a little too hard in the sun and seasonal allergies – it’s hard, if the impossible can be called difficult, to believe that an awful lot of people are classified as deaths ‘involving’ COVID where the virus played little or no part in their demise.

Actuarial Table: (This one is from 2017 – the numbers do change over time, but very slowly.) You can look up how likely, historically, you are to die this year at your age – if age were really all that mattered. If you’re basically healthy, then you’re likelihood is less, probably a lot less, than the averages represented by these numbers.

Tip: it might help to think in terms of thousands or 10s of thousands: where the table says a 40 year old woman has a 0.143% chance of dying this year, that would mean about 14 out of every 10,000 40 year old women are likely to die on average. Then remember that those 14 women are very ill (cancer, heart disease, etc.), otherwise unlucky (unavoidable accidents), or unwise (doing dangerous, stupid things.) If you are a normally careful and otherwise healthy 40 year old woman, your chances are lower.

Exact
age
MaleFemale
Death
probability
(%)
Death
probability
(%)
00.630%0.523%
10.043%0.034%
20.029%0.021%
30.023%0.016%
40.016%0.014%
50.015%0.013%
60.014%0.011%
70.013%0.010%
80.012%0.010%
90.010%0.009%
100.010%0.009%
110.011%0.010%
120.015%0.011%
130.023%0.014%
140.034%0.017%
150.047%0.021%
160.059%0.025%
170.072%0.029%
180.086%0.034%
190.100%0.038%
200.115%0.043%
210.129%0.047%
220.141%0.052%
230.149%0.055%
240.156%0.058%
250.161%0.061%
260.167%0.065%
270.172%0.068%
280.177%0.073%
290.182%0.078%
300.187%0.083%
310.191%0.089%
320.196%0.094%
330.201%0.099%
340.207%0.103%
350.214%0.109%
360.221%0.114%
370.228%0.121%
380.234%0.127%
390.241%0.135%
400.248%0.143%
410.258%0.152%
420.271%0.163%
430.287%0.175%
440.306%0.188%
450.329%0.203%
460.354%0.220%
470.383%0.239%
480.418%0.261%
490.457%0.285%
500.500%0.312%
510.546%0.340%
520.597%0.371%
530.653%0.405%
540.713%0.442%
550.777%0.481%
560.845%0.523%
570.916%0.565%
580.990%0.604%
591.067%0.644%
601.152%0.689%
611.242%0.739%
621.331%0.793%
631.416%0.851%
641.503%0.914%
651.601%0.987%
661.714%1.072%
671.836%1.166%
681.969%1.271%
692.117%1.389%
702.289%1.529%
712.487%1.688%
722.710%1.861%
732.959%2.047%
743.239%2.252%
753.567%2.493%
763.940%2.773%
774.345%3.086%
784.783%3.432%
795.265%3.821%
805.821%4.277%
816.458%4.799%
827.166%5.368%
837.947%5.981%
848.814%6.658%
859.785%7.426%
8610.875%8.305%
8712.092%9.312%
8813.443%10.454%
8914.927%11.731%
9016.545%13.139%
9118.294%14.675%
9220.168%16.333%
9322.164%18.106%
9424.275%19.989%
9526.367%21.891%
9628.401%23.782%
9730.336%25.627%
9832.127%27.389%
9933.733%29.033%
10035.420%30.775%
10137.191%32.621%
10239.050%34.579%
10341.003%36.653%
10443.053%38.852%
10545.206%41.184%
10647.466%43.655%
10749.839%46.274%
10852.331%49.050%
10954.948%51.993%
11057.695%55.113%
11160.580%58.420%
11263.609%61.925%
11366.789%65.640%
11470.129%69.579%
11573.635%73.635%
11677.317%77.317%
11781.183%81.183%
11885.242%85.242%
11989.504%89.504%

Author: Joseph Moore

Enough with the smarty-pants Dante quote. Just some opinionated blogger dude.

3 thoughts on “5 Levels of Risk Analysis: We’re All Going to Die”

  1. Call me a rabbit; I’ll continue to wear it proudly. Your presentation of risk analysis spread over broad populations with Level V addressing only two potential factors: physical health and risky behaviors seems to me useless on a practical level (which is where I, anyway, live.) Take, for example, driving- a fairly risky activity – you seem to argue that the risk of driving is only expressed broadly across the population vs. statistics on death from other causes. In the practical world, the risk of driving varies based on all sorts of external factors. The risk of driving to the grocery store alone on a day with temperate weather is utterly different that the risk of driving 60 miles to the nearest metropolis, which is different than driving to the metropolis after 8 pm on St. Patrick’s Day, driving 200 miles while sleep deprived, with screaming children, etc. etc. et c.

    1. The example you give for driving is a good one for the personal behaviors point. It boils down to assessing which personal behaviors affect outcomes. As in COVID, there are generally two subpopulations to consider for driving risk: there are those who drive carefully, keep their cars up, and stay out of dangerous driving situations – and those who don’t. (Of course, there are people in-between, but, as a former underwriter, I can tell you it’s surprising how small that group is.) The first group of careful drivers has a low background risk of accidents and death, a large part of which is getting clobbered by members of the second group. The second group – young males, drunk drivers, road ragers, cell-phone users – account for a very disproportionate percentage of ‘accidents’.

      With COVID, however, the two main groups do not break down along lines of behavior. About 2/3rds of attributed deaths are among the 0,5% of the population in nursing homes. Not in a nursing home? Like a good driver, your risk is much smaller. But it has nothing to do with your behavior.

      But good drivers wear seat belts! And smart laymen wear masks! Is that the proper analogy? Seatbelts do have a comparatively minor but real effect on reducing death and injury in car crashes, validated by years of accident data. Up until 15 months ago, no one anywhere had any data supporting the idea that masks *as used by the general public* reduces overall risk of airborne viral infection. No data demonstrating that masks work *as used by the general public* have yet to be produced. Yes, I looked at the “70 studies” – at least, I read each of the first five until I reached the point of incredulity: as a scientifically literate person, I know 1) studies in themselves are just claims; 2) those claims rest on evidence and logic; 3) most studies are ‘wrong’, meaning the evidence and logic do not sufficiently or at all support the conclusions. That’s why science demands criticism and replication of study results before claims can be taken seriously.

      What am I, or any scientifically literate person, looking for, in a mask study? Let’s list it:

      1.) Testing the actual claim. Saying masks work in clinical setting says nothing about if they work when put on by fumble-fingered non-professionals who have not scrubbed up and gloved up, and don’t change their masks (putting the old on into hazardous waste) every 20-30 minutes.
      2) Responses to obvious criticisms. Study authors are often meticulous about the challenges they think they can answer, but silent on those they can’t. So, when a study ignores such things as logic of mask disposal – if they work, used masks are now loaded with deadly virions, and are biohazards themselves! Now what? – and focus on one, and only one, aspect of risk, we’re talking magic, not science.
      3. Valid methodology. This one gets a little more technical, but, for example, meta-studies – studies of studies – need to account for a boatload of differences in each of the base studies they’re analyzing AND show that they are not cherry-picking studies they like. For starters.
      4. Significance: studies all involve uncertainty – of method, of data, even in the basic math. Those uncertainties don’t go away just because I really, really want them to. To be meaningful, a study’s results must be significantly larger than the cumulative amount of uncertainty. One study of the legendary 70 showed something like a 2.5% reduction in spread rate across several very different population, where the counting had been done by different, or at least unexamined methodologies. That’s a very small difference claimed against a very large amount of uncertainty. And reeks of cherry-picking.

      Conclusion: looking at the data coming out state by state, there is NO, as in NO, difference in outcome that follows mask use, lockdowns, compliance assumptions, or anything else at all. All that rigmarole simply didn’t matter. To see this, step one: stop listening to professional panic mongers, and look at the data yourself. If you don’t feel qualified to interpret the data yourself, I suggest William Briggs. http://wmbriggs.com/ Or me.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: