• *Catastrophic Care*: Chapter 10

    Patients increasingly have access to websites designed for providers and other health care professionals. The NNT (which stands for “number needed to treat”) is a particularly valuable corrective to much medical information, reversing the perspective on treatment efficacy that patients usually get in health care advertisements. For example, the site will translate a pharmaceutical company’s technically correct claim that its drug reduces the risk of a specific illness by 40 percent into a more meaningful fact: a hundred patients will have to be treated with the drug before a single patient can expect to benefit.

    Quoting this passage from chapter 10 of Catastrophic Care gives me the opportunity to do three things.

    1. Here’s the link to The NNT. And, Goldhill is right to hype it. It’s very good.
    2. One of the creators and champions of The NNT is David Newman. I’ve mentioned him before. He also wrote a book about evidence-based medicine for a lay audience: Hippocrates’ Shadow. I recommend it.
    3. In my first post on Catastrophic Care, I made two stylized points about the health system: (i) It’s not responsive to consumers. (ii) It’s not responsive to evidence. I can understand how patient control of payment — Goldhill’s thrust — could address the first of these. I’m far less convinced it can do very much about the second. After all, patient satisfaction and clinical quality are two different things. It does not surprise me at all that patients can be subjectively happy with objectively bad care. A good bedside manner and attention to amenities and convenience go a long way. Patients deserve that. But they deserve good clinical quality too. If they can’t judge it, I’m not convinced they’ll buy it. In fact, I’m more convinced they’ll buy something that makes them feel good in the moment even if it they know it’s not objectively better. Americans (probably all humans) buy a lot of stuff that’s of low quality or even bad for them, but makes them feel good right now. That doesn’t matter a great deal if the product is a movie or a dinner or a clock radio (though, to those with modest resources, yes, it does matter). But it does matter a great deal if the product is a diagnosis or a cardiac procedure. This is not an argument for the status quo, merely an argument to be cautious about the benefits and realistic about the limitations of patient control of payment.

    Other posts about Catastrophic Care are book here.


    • Frakt deserves a lot of credit for commenting in depth on Goldhill’s book after Frakt first admitted he is tired of health reform and the health care debate, a sentiment most experts in health care probably share after all these years. We’ve done health care reform, so let’s put our efforts in implementation of the reform we got not a hypothetical reform we didn’t. Senator Baucus created a kerfuffle last week when he said implementation was a train wreck. That was the context of the ACA question at Obama’s news conference today. Obama’s long response says all one needs to know: his response was all about universal insurance coverage and its implementation. If ACA and implementation are judged solely on universal insurance coverage, then Obama’s is unquestionably correct. The problem is that many folks had expected more, a whole lot more, from health care reform. To them I would say you weren’t listening, for Obama was clear in the 2008 campaign that universal insurance coverage was his goal, signaled in the campaign in his opposition to the mandate and since his election to the compromises he made during Congressional debates and during implementation. Obama believes health care reform is about universal insurance coverage, while Goldhill believes health care reform is about eliminating insurance. It’s a Mars-Venus thing.

    • On the “not responsive to evidence point,” I can’t help but wonder if you think that a patient’s preference for care that is satisfying in some dimension but has no objective health benefit is (A) a problem that warrants state intervention in all cases, or (B) only in those cases where the state is paying all or part of the bill?

      If it’s (A) then what is the appropriate public policy with regards to the litany of alternative modalities such as homeopathy, reiki, chelation therapy, chiropractic interventions for non musculoskeletal problems, etc, etc, that have zero proven clinical benefit but patients spend many billions on every year? If it’s (B) does that include all services paid for with pre-tax funds?

      • Translation, dissemination, and implementation of evidence-based medicine is an exceedingly hard problem. There are all manner of perverse incentives that make it so, as well as human nature and culture. I would not presume to know how to solve such a complex problem. I would hope that those that devote hundreds of pages to the subject of tearing down our existing system (which may be warranted) and proposing a totally new one (again, which may be better than what we have), would also devote a few pages to this issue.

        I could say more, but I know you well enough to know you will never be satisfied. I’ll spare myself the grief. You can glean what you wish from the posts on this blog.

        Meanwhile, what are your thoughts?

        • “Translation, dissemination, and implementation of evidence-based medicine is an exceedingly hard problem. There are all manner of perverse incentives that make it so, as well as human nature and culture.”

          The solution for the rentier class will soon be to get the human (physician) out of the equation. This is why the further we accept left-brained business control of medicine, the further we lose the humanity of healthcare in the first place.

          More profit can be extracted with expert systems coupled with cheap employees carrying out the expert guidelines. The seduction is that evidence is without bias, that it is free from human nature and culture. This is dangerous.

          From Iain McGilchrist:

          “Guidelines are based on what has been gleaned about generalities. But I have never seen a single patient who is a generality. What I know is that a different patient – every single patient of mine – requires a different package of therapy, or medication, or both. And the differences are things that could never be demonstrated by the blunt instrument of a clinical trial, which can only identify the grossest differences in general terms, and often fails to identify what is grossly obvious to any experienced clinician. Such evidence is, in any case, notoriously open to manipulation in a variety of ways, so that it should neither be dismissed, nor blindly accepted as the best path to sound understanding.”


        • -It may just be a peculiar fixation of mine, but given that the clinical efficacy of virtually every therapeutic intervention offered by alternative practitioners is zero, and the American public spends staggering sums of money obtaining care from practitioners who are making claims about the clinical benefits of the care that they are delivering that are demonstrably false – I’m puzzled by the lack of attention and scrutiny that advocates of health system reform pay to this issue.

          I’m an intermittent reader of this blog so I don’t know if this is a particular concern of yours or not, or it’s something you’ve addressed or not, but it’s a puzzling feature of the field in general.

          If I were to set about crafting a utopia, I would simply extend the rules regarding the standards of evidence and professional liability that physicians and pharmaceutical companies have to comply with to anyone practicing any brand of medicine. I’d also make anyone who accepts public funds or funds subsidized with tax exemptions for care delivered on the basis of claims that are demonstrably false liable for defrauding the public. People could still waste their own money on woo to their heart’s content, but they’d have to do it exclusively on their own dime. Homtheopaths could still sell sub-yoctomolar dilutions of any herb they wished, but they’d have to prove any claims about the health benefits of their products as extensively as Pfizer does, etc, etc, etc.

          -When it comes to your concerns about patients choosing care that is emotionally satisfying in some dimension but which has no clinical benefit, there’s ample reason to believe that coupling lifestyle risks to premiums and provider payment to demonstrable clinical efficacy would provide both the means and the motive for both patients and providers to consider the costs, benefits, and efficacy of the care that they are getting or delivering much more carefully than they are under the third party payor regimen that prevails now.

          It wouldn’t be a panacea, but I’d argue that the reason why providers and patients don’t make better use of the information that’s already out there concerning cost and efficacy is simply that they don’t have to. Patient control of payment isn’t a panacea, and anyone looking for evidence need look no further than the local alternative medicine clinic, but patients do care quite a bit and and do pay attention to what things cost, and providers care just as much about what they get paid for and what they don’t, so any incremental move towards a system where the price-based incentives are stronger for both providers and patients would move us further towards a system that’s responsive to both consumers and evidence.

          Value based plan design along whatever harmonic convergence exists along the Frakt-Goodman axis would certainly help move things in that direction. I think that the current political/institutional momentum is moving us in precisely the opposite direction, but perhaps that will change in time.

    • Clinical quality is best judged statistically, rather than by patients. When you have a good or a bad outcome, how can the patient know if it was a result of good or bad doctoring, as opposed to luck? A professional will only know if they follow the case closely.

      I remember a case from the UK of a pediatric cardiologist who was slower than average at open-heart surgery. It was hard to say that any one case was the result of any error, not the complication of the case, but statistically, he was slower (meaning babies were on the bypass machine longer) and more babies died in his OR than elsewhere.

      • This is an example of gathering outcomes data.

        Atul Gawande talks about surgeon self-regulation, and the disincentives for both surgeon and colleague to act when assessing performance.

        This is a moral discussion about the threshold at which a surgeon or practitioner should stop practicing surgery and the ramifications of this, both to patient and the surgeon’s livelihood. Should surgeon’s livelihood even matter? In a throwaway society where value is only in replaceable skill I’d say most would not care. That is sad and a waste of myriad human potential.

        Should the cutoff be strict and applied categorically or should it be consequentialist, and by applying strict cutoffs in a rural area where more people would die for lack of surgeons.

        Is the surgeon picking more difficult patients or assigned more difficult patients to harm their statistics, or choosing easier patients to boost their statistics?

        Surgical intervention ( whether invasive or minimally invasive) is in some ways easier to track than medical intervention. The issue is how to interpret the data and balance access and results.

        The harder question: Is surgical intervention actually required in the first place? and are there incentives to intervene other than the underlying condition.