Notes in the Margin - 4/2/24
Expertise, capacity assessments, AI, assisted suicide, and technology
Capable of More: Some Underemphasized Aspects of Capacity Assessment
Some clinicians approach the assessment of patients’ decision-making capacity woodenly. As Nicholas Kontos and colleagues argue, though, there’s so much more to this that can be used to the patient’s benefit. Clinicians shouldn’t just seek to assess capacity, but to access capacity. That is, they should give patients every opportunity to demonstrate their optimal decision-making. There are non cognitive factors, like emotions, that factor into decision-making and thus the capacity assessment. And finally, the capacity assessment should also answer the question, “If this person is incapacitated, what has incapacitated them?” That is, an outcome of incapacity should also carry with it a diagnosis explaining it.
discusses the importance, even necessity, of treating large language models (LLM) like humans to receive better responses. “Anthropomorphism is the future, in ways good and bad. An increasing number of humans already feel they have deep bonds with AI, with unpredictable results for our own interactions - helping some people while damaging the human relationships of others. Given the trend, treating AIs like people seems like an inevitability, so figuring out how to do it in safe, productive ways may be better than the alternatives.” But what formative effects will this have on the real people themselves?We can’t afford a taboo on assisted dying
Matthew Parris states bluntly where the logic of assisted suicide and euthanasia (ASE) leads. If ASE is a medical intervention, and we use medical interventions to remediate social ills, why not use it to dispense with a “burdensome” population of people? Don’t worry - this won’t be imposed on them unwillingly. It will be important, argues Parris, that people come to believe these things for themselves, to choose them on their own. That’s the kind of culture Parris and others want to create. This is John Hardwig’s argument about a duty to die all over again, but in a new context: the world is becoming more accepting of ASE now. The ideas are descending the ivory tower and stepping out into hospitals, clinics, and homes.
“Is life still giving us more pleasure than pain? How much is all this costing relatives and the health service? How much of a burden are we placing on those who love us? How much of a burden are we placing on ourselves? … But it’s not as if these questions are new: they already haunt and have always haunted many afflicted by intolerable misery, indignity or suffering. That’s simply how people think: it’s natural. It’s right. … It will be a healthy development. In this country the future holds an almost cosmic struggle between, on the one hand, the old world with our aging populations and inflexible economies, and, on the other, the raw and unbridled energies of an emerging, young, nimbler and very different world…”
Technology and History: “Kranzberg’s Laws”
Melvin Kranzberg was the founder of the journal Technology and Culture and a professor who studied the history of technology. In this essay, he provides six things he had learned after working in the field for thirty years. They appear to remain true today. I’d like to incorporate them into an essay on technology and medicine, but for now, I commend it to you unvarnished by that mediation. He cites many historical examples to back up his claims.
A quick comment on the duty to die argument.
One of the ideas here is that healthcare for the sick elderly is expensive, and so expensive that we cannot afford to treat them.
Here's how the argument goes: The Canadian healthcare system is under substantial stress. (IT IS.) So doesn't Medical Assistance in Dying (MAID) relieve this stress in part by obviating the need for expensive end-of-life care? IT DOES.
But how much does the Canadian health care system save via MAID?
A Back-of-the-Envelope (BoE) calculation of those savings. Suppose that about 10,000 Canadian cancer patients die by MAID each year (roughly correct). Suppose that each would live one more year if they hadn't been euthanized (very uncertain). Suppose that year of care would have cost $33,000 (possibly low, but that's based on a recent study). Suppose that MAID costs nothing (untrue, but close enough for government work).
Then MAID saves the Canadian healthcare system
10,000 patients/year * $33,000/patient = $330 million/year.
That's a lot of money, no question. However, the Canadian Institute for Health Information reports that Canada spent $340 billion on healthcare in 2023. So the BoE estimated savings resulting from MAID was < 0.1% of 2023 total healthcare spending. If my BoE estimate was too small by a factor of 5, then the savings were < half a percent.
TLDR summary: MAID saves Canada money, but not so much that we have no choice other than to euthanize the sick and elderly.