Conceptually the loss-of-a chance doctrine recently reaffirmed in Rash v. Providence Health & Services appears to make sense. The typical facts in such cases include (1) a usually fatal disease (e.g. certain cancers); (2) that was diagnosed later than was possible with proper care (or that a less effective treatment was used); and where (3) the limited chances of survival decline further with each successive stage of the disease’s progression. Not wanting to “provide a ‘blanket’ release from liability for doctors and hospitals any time there was less than a 50 percent chance of survival, regardless of how flagrant the negligence“, yet unable to come up with a sound reason why a plaintiff ought to be able to recover for an act or omission which probably did not cause the course of her disease to be altered, some courts made the erosion of the chance of survival the harm rather than the subsequent death. With that the causation dilemma seemed to disappear. Meanwhile, a mechanism for disincentivizing  (via the imposition of tort liability) the provision of anything less than optimal care, even to those unlikely to benefit from it, is created. One problem with the approach is that chance, especially in this setting, is not a thing that can be lost. Another comes from encouraging doctors to treat probability distributions instead of people.

Chance is a word imbued with powerful meanings. Often wrapped up in it are ideas about fate, destiny, fairness and even justice. Take the case of a simple coin flip that settles controversies from who kicks-off to who owns a $125,000 car. We may dispute the circumstances of the flip but never the outcome. Somehow, once in the air and spinning, fate, destiny, justice, karma or whatever hands down its unappealable judgment which is promptly revealed for all to plainly see. This idea of chance as a proxy for justice (or perhaps as a ward against injustice) is a particularly old one. Consider Jonah 1:7 :

Then the sailors said to each other, “Come, let us cast lots to find out who is responsible for this calamity.” They cast lots and the lot fell on Jonah.

Of course in the age of “Big Data” chance is supposed to be about the attempt to quantify our uncertainty. When we say “the odds are 50-50″ what we’re really saying is that we don’t have access to any information that would lead us to believe that one side is more likely to come up than the other. In this sense chance may be considered a measure of our ignorance of the mechanisms and/or variables that determine which side comes up.

Now the fact that it’s unappetizingly about uncertainty and ignorance wouldn’t be a good reason not to compensate someone who lost a chance like the one depicted in the coin toss scene from “No Country for Old Men”. There’s one chance you wouldn’t want to lose. Only in such a pure instance of chance can it become a thing you can lose; and that must be the concept of chance imagined by courts like the one that authored Rash. Unfortunately that’s not at all the sort of chance we’re talking about when we talk about the chance of surviving cancer.

Where do estimations of the chance of surviving cancer for five years come from? Obviously from other people and not the newly diagnosed. And did those other people all experience identical survival intervals? No. Even the graph of late stage pancreatic cancer patients has a long tail of the very lucky few. Consequently, any estimation of the central tendency of those other people, usually the median but sometimes the average survival time, homogenizes the experience of all the patients and produces a mathematically “typical” patient with an experience unlike any of the individual patients. Whereas the gas station cashier in the coin toss scene had the opportunity to save his life by choosing “heads”, to seize the opportunity presented by the graph of the survival experience of patients undergoing a new treatment the cancer patient would somehow have to be able to choose to be the “typical” patient; and that would mean being able to choose to have whatever currently unknown genetic and epigenetic makeup is responsible for the slightly improved “typical” survival time – which is impossible. You can’t buy that chance, and neither can you lose it.

The remaining argument for the loss-of-a-chance doctrine is that disincentivizing doctors from providing anything other than the treatment with the longest “typical” survival time at the earliest possible date would save some unidentifiable lives and so produce a benefit to society as a whole. This is where we wade into the widening controversy swirling around the use of statistics, despite (or rather because of) ignorance of underlying mechanisms and variables, to determine treatment. On one side are those who hold the view that “it is obsolete for the doctor to approach each patient strictly as an individual; medical decisions should be made on the basis of what is best for the population as a whole“. The idea here is that if earlier or a newer treatment has shifted the survival curve in the direction of longer survival in a subset of people with the disease then earlier or newer treatment across all people with the disease will surely save lives.

On the other side are those who point out that the medical journals (and law books) are littered with examples of treatments which demonstrated a pattern of better outcomes in a small population but which showed no benefit or worse outcomes once they were widely prescribed. That many researchers, doctors and pharmaceutical companies ”find some pattern in their data and they don’t even want to consider the possibility that it might not hold in the general population” is a well-known phenomenon.

As for our take on the controversy all we can say is that until the underlying mechanisms of cancer are elucidated inferring treatment from statistics is pretty much all we’ve got … but often that ain’t sayin’ much. Hopefully in the not too distant future physicians will look back on our current era and shake their heads at the thought of the primitives who settled upon cancer treatment options essentially by casting lots. That being said, to anchor liability on the claim that the slight positive shift in the probability distribution calculated for a small sample of likely terminal patients (in turn premised on the dubious assumption that patients can be thought of as so many balls in a quincunx machine getting chemotherapy an infinite number of times) will also be seen in a much larger sample of completely different likely terminal patients seems more than just a bit of a stretch.

Consider also the following: if the loss of the (imagined) chance is the harm, why don’t the people who lost the chance at the new treatment, but who responded to the old treatment anyway, have a claim? They lost a chance and that’s a harm after all. And what would the damages be for the harm? They’d be the same as they would be for the person who lost a chance at a treatment that probably wouldn’t have made a difference anyway, right? So why is it that some who are harmed have a claim while others who sufferer the identical harm do not? Because the loss-of-a-chance doctrine is incoherent.

In Rash the appellate court ultimately affirmed the dismissal of plaintiff’s claim because her expert couldn’t quantify the chance she had lost. That’s just another example of a court falling into the trap of believing that assigning numbers to things, even to things that are not things, makes them “scientific” so that, as here, damages may be “accurately” calculated. Yet it’s vital to the assumption that doctors are able to sell and patients are able to buy the “typical ” (mean or median) outcome of a treatment that actually yielded a wide range of outcomes, none of which were precisely “typical”. And the illusion of accuracy created by multiplying the quantified chance of the “typical” patient from small study by the value of someone else’s life to determine her damages is just that. But so it goes with the loss-of-a-chance doctrine.

However far science pushes back the shadows to reveal how the universe really works, chance retains its place as a somehow essential and inescapable aspect of our lives. Perhaps, as ably argued by a colleague recently when we were outlining this post, Garth Brooks nailed it when he sang “I’m glad I didn’t know, the way it all would end, the way it all would go. Our lives, are better left to chance, I could have missed the pain, but I’d have had to miss the dance”. Or maybe chance, once revealed as uncertainty, is actually the driving force behind mankind’s quest for truth. That’s my take. But whatever it is it’s not something you can buy at the doctor’s office.