Nothing in random in any mystical causative sense. Many things are unpredictable. And, indeed, unpredictable is what people mean when they say random. That, or unknown cause.
I was reminded of this after pointed by Wrath of Gnon to the paper “Deaths by horsekick in the Prussian army – and other ‘Never Events’ in large organisations” by J. J. Pandit in Anaesthesia.
I don’t disagree, in general, with the author’s broad conclusion, but I think his views can be improved by the switches I mentioned.
Background is Moppet and Moppet wrote a paper discussing “Never Events”, which are “serious adverse incidents like wrong site surgery”, events which M&M say “follow” a Poisson distribution.
In attempting to clarify this, Pandit reminds us of the work of Ladislaus Bortkiewicz, who cataloged the number of times each year Prussian cavalrymen were kicked to death by their horses. The numbers are in this picture:
Most years (counting by cavalry units) have no deaths, some had one, fewer had two, and so on as you see. Now this:
Bortkiewicz went on to explain how this is a formal mathematical description of random, rare events. In other words, any real distribution (e.g. Prussian horsekicks) that resembles a Poisson in turn can be readily assumed to have arisen through chance events and not as a result of intent or design (e.g. inherent systems failure in the military). Similarly, if Moppett and Moppett have concluded that NHS Never Events are also Poisson-distributed and so random, then does it follow that they have no ‘cause’ and, in turn, that they cannot be prevented? We discuss below why this does not follow.
As I said, Pandit sees the essential problem, but missed his chance (good pun!) to go a bit farther.
He opens his gambit with this: “A common Oxford entrance interview question asks candidates: ‘What is the opposite of “random”?‘”
You, dear reader, must say “predictable or with a known cause.” Let’s see why.
Now nothing follows a Poisson distribution, but that probability model can mathematically represent the uncertainty you have in some event, like number of wrong surgeries or horse kicks—or anything you assign.
This implies the event is not completely unpredictable, because if it was then the best you could say is “I do not know.” And take in the strictest sense: you know nothing. Since you can or do assign a Poisson—or perhaps even deduce it from known premises about the event, a very rare thing to do—you are saying you do know something, but only enough to quantify the chance of the event.
The event is in common parlance random. Which only means you cannot foresee whether it will happen, or when, or precisely how. You allow only that it can, and with this-and-such uncertainty.
But when the event happens, as Pandit sees, “If we are the patient (or the family), then none of the statistics matter: the event has happened to us and naturally we are always concerned.”
And then there must be a cause of the event. A doctor screwed up, or a nurse did, or somebody put the wrong information into the chart, or all three, or more, or any other form of incompetence: some form of incompetence or error.
Blame can and must be assigned. It is no good to say “Oh, well, the event was a Never Event because it followed a Poisson.”
Something always causes everything to happen, even if we can never know the cause in advance, or even after, as with quantum mechanics.
Here’s his conclusion:
Moppett and Moppett have nicely shown that the distribution of Never Events in the NHS – just like that of deaths by horsekick in the Prussian cavalry over a century ago – follows a quasi-Poisson distribution. This means that the events, when viewed in totality, are rare and random. There is no reason to suspect any fundamental systems failure across the NHS or in any one hospital based on the Never Event data alone.
I disagree. Because, as I said, every wrong surgery (or whatever) represents an error, something happened that should not have, and we should learn why.
A Poisson, or other probability model, is only useful to make predictions of the future, given the assumption nothing known has changed causally.
Buy my new book and learn to argue against the regime: Everything You Believe Is Wrong.
Subscribe or donate to support this site and its wholly independent host using credit card click here. For Zelle, use my email: matt@wmbriggs.com, and please include yours so I know who to thank.
You make an important point – because the conclusion “This means that the events, when viewed in totality, are rare and random. There is no reason to suspect any fundamental systems failure” is *dangerously* wrong
– especially when (as likely) this conclusion is implemented by a managerially-controlled bureaucracy.
For example, wrong site surgery (e.g. amputating the right instead of left big-toe, or leg) is only rare because the staff involved are 1. competent, 2. trained, and 3. adhere to well-designed procedures to prevent this error.
Once ‘managers’ have absorbed the idea that ‘science has shown’ that wrong site surgery is innately rare and random; then the strong bureaucratic incentives to employ cheaper/ worse/ more-diverse staff; reduce their training standards; or/and simplify/ abolish preventive procedures – would certainly increase the frequency of this error.
And very soon wrong site surgery would cease to be ‘rare’, and the cause of its increase would have been perfectly predictable *if* people had remembered that (as you say) even very rare events always have causes, and often these causes can be known.
Even in quantum mechanics, the cause never follows the event. Even in general relativity, the cause never follows the event, nor even appears to do so for any observer.
Oh, how I abhor the state of education on physics. Even among professional physicists. The bullies at the Copenhagen conference are never to be sufficiently damned for their mysticism and deliberate obfuscation.
Reality is, at heart, simple geometry. Note that geometry seldom has a simple mathematical model. (For example, space is Euclidean. Spacetime, however, is hyperbolic. This is because time is imaginary. It must be, to be at 90 right angles to every spacial dimension.)
As our esteemed host repeats often, events have causes. That we may not perfectly know these causes in advance does not mean they do not exist. Those who say otherwise are guilty of magical “thinking”, in other words, non-thought or nonsense.
As used to be said, not a sparrow falls but is known to the mind of God.
I’ve never been kicked by a horse but I did have a neighbors pony deliberately
fall down and roll over on top of me when I was 12yo. This after he
failed to rub me off his back against the fence railing. I’m not positive
but I don’t think that’s one off behavior for a Shetland pony, though I could be
wrong. When he broke into the feed shed and ate himself to death on
oats I can’t say I was heartbroken.
Brings to mind a Paul Newman movie “The Verdict,” about a medical malpractice trial. The crux of the matter was misreading the patient’s medical chart. Excellent drama and performance by Newman.
As a side note, there are elements of the staging (one scene in particular) that wouldn’t make the screen today. In fact, I doubt many films today measure up to many subtleties displayed, as movies have “messages” and go way over the top in pursuit of an agenda–especially fulfilling The Current Thing.
The Prussian cavalry could have amputated the horse’s hind legs and replaced them with wheels. Other solutions may exist.
So, if I read this right, Moppet and Moppet (probably commissioned by the British gov’t or one of its partners,) “wrote a paper” for NHS apologists to use, which looks at seemingly ‘random’, meaning unpredictable, events, within the NHS, such as “wrong site surgery” (oops)…..
….and by using the “Poisson distribution” model, demonstrate (so they think) that these “Never Events”, (which aren’t “never” events, because they do in fact happen, albeit “randomly”), have no cause.
Giving them the benefit of the doubt, (because who IS that stupid?) they want to say there’s no “systematic” cause, in order to get the system, in this case the NHS, off the hook. In other words, because these events follow the Poisson curve, they aren’t something the NHS can predict or prevent in advance.
I believe this is known as the “s**t happens” defense, which is why there’s malpractice and other kinds of insurance.
It’s hard to believe that these Moppet-heads don’t recognize a “cause” for an event (ie malpractice), simply because it was not predictable in advance. What they really want us to believe, is that the NHS is SO “systematically” awesome, that medical malpractice is as rare as a Prussian soldier getting kicked by his horse. They are THAT good. Can we have an NHS here, too, daddy?
This assumes the NHS reports all wrong site surgery……………
After reading both papers quickly, I’d guess that M&M didn’t disagree with Pandit.
A random variable, by definition, is a function that assigns values to outcomes. It is called a random variable because its value cannot be known with certainty in advance. For instance, let X be the number of Never Events next year. The value of X cannot be known with certainty… regardless of whether or what causes exist. Hence, past data are used to model/ approximate the uncertainty.
Interesting, what does it mean to say that events of Never Events are random? Does it mean (1) that we don’t know whether they will occur for certain in advance, (2) that we don’t know what causes the events, or (3) that the events are causeless as implied by Pandit? I say (1). However, I suspect the question has been considered by some philosophers.
“A common Oxford entrance interview question asks candidates: ‘What is the opposite of “random”?‘”
Deterministic.