First, do no harm (medical research edition)

A plea for integrating the reduction of medical costs in major medical research programmes.

The National Institutes of Health explain the fundamental driver of the trend increase in health costs much better than I could:

In the past 40+ years, NIH funded research has successfully reduced the mortality and morbidity of once acute and lethal diseases and conditions by finding ways to improve treatment — even in later stages. These advances have moved what had been to acute to chronic diseases, to diseases that are chronic and manageable. These chronic diseases now form the largest component of health burden.

National Institutes of Health FY 2010 President’s Budget, page 3

A touch of hubris here – the NIH is much the biggest fish in the pond, but not the only one – but pardonable. In short: we die of things less, and live longer as permanent patients.

The great medical achievements of the NIH and its fellows worldwide are leading us (see my previous post) to an economic and social catastrophe. On current trends in medical costs, either all rich countries go broke sooner (USA) or later (rest of OECD), or we have extensive rationing of cutting-edge medicine by the market or the state, or we just leave the discoveries unused and settle for an iron rice bowl. (To be complete, or something turns up.) In homage to Jonathan Swift, who 300 years ago imagined the horrors of immortality without a stop to aging, let’s call this unsustainable situation a Struldbrug box. What’s the NIH plan for getting us out out of it?

The next sentence gives their reply:

Biomedical research is the key to transform medicine from the curative health care paradigm of the past where we intervened late in the natural history of a disease, to a preemptive model in which the onset of disease is significantly delayed or even never allowed to develop. 

A bit lame, isn’t it? As commenter npm pointed out on my previous post, prevention is no guarantee of cheapness. Instead of expensive lifelong treatment, we may end up with expensive lifelong prevention. The model for optimism is the polio vaccine; the reality may be more like insulin therapy for diabetes. Scarred by over-hyped and premature announcements of silver bullets for cancer etc just round the corner, the NIH no longer dares highlight the powerful word cure. But can someone explain to me why the sort of deep knowledge that would give us a one-shot preventive vaccine against AIDS might not just as likely provide a six-pill cure? Either way, the cost problem is solved.

In the long run, possibly. But before that we are all broke. As far as medical costs are concerned, the current research strategy amounts to “that’s not my department“. I can’t find evidence that the NIH, which spent $30.6 bn last year, takes medical costs at all seriously. It’s a vast and complex organisation, and presumably there are people inside it who worry about the problem, but it’s not evident in the public face. Cost control is not mentioned in the mission statement, nor does it seem a be a specific funding area. Googling for “cost reduction” in the grants area drew a blank, “low-cost” returned only 22 hits.

I did find one programme out of hundreds that really fits the bill: significantly in the outlying, engineer-driven technology area:

RFA-HG-10-012: Revolutionary Genome Sequencing Technologies

The $1000 Genome. The timidity shows in the headline number. DARPA would have put $100: and why not? Moore’s law continues to operate; electronic widgets will become not only cheaper, but much cheaper over time. Even a mere SF reader like me can see that you make the sequencer an automated solid-state microfluidics array, shove in a fast wifi connection so you can updump the massive parallel processing to the cloud…) The Pentagon spends research pin money on robot infantry and armed dolphins. The NIH is stll thinking inside its (large, well-furnished, highly esteemed) box.

The NIH isn’t alone. The UK Medical Research Council is a high-class outfit that funded Watson, Crick, Wilkins and Franklin to crack the DNA code (guess the gender of the one who didn’t get the Nobel prize. Update: Commenter Warren Terra points out that Franklin died before the prize was awarded for the DNA code, so it’s not the Nobel committee that should be blamed for the lack of proper credit.) Annual grants run at £758 million. The website shows the same pattern. There’s no strategic objective mentioning cost control. Searching the website, “low-cost” – which you would use to publicise results – returned 27 hits, “cost reduction ” – words you would put in an objective – nil.

My third guinea pig was the Wellcome Trust, SFIK the world’s largest medical research charity. It was founded and richly endowed by one of the distant British progenitors of the GSK pharma behemoth, and awarded £628m in grants in 2010. Again, nothing about cost reduction in the strategic objectives. The search hits predictably gave only 2 for “cost reduction”, neither in the grants area, but 148 for “low-cost”. When you look at these, many of them come from projects to tackle Third World health problems, such as TB diagnostic kits.

This fits. Researchers on African diseases think about costs all the time. The Gates Foundation funds field programmes in poor countries not research, so it’s only an analogy, but the website returned 222 hits for “low-cost”. There’s a new initiative led by Norway to reduce the cost of the latest malaria treatments from $6-10 a head by an order of magnitude to <$1. This would be pointless in rich countries, but for an African peasant living on $1 a day, a $10 treatment for his child may be out of reach. A treated mosquito net in Africa, cost £4.

The malaria people are the model for rich countries. We are all African peasants in that we live under binding resource constraints, and best-practice medicine is becoming unaffordable everywhere under these constraints. The result is avoidable damage to health. Hippocrates’ first principle – which I used as the title for this post – requires that physicians take the problem seriously. So here is my recommendation to the NIH, MRC, Wellcome, and anybody else in the trade who might be listening:

Medical costs are a health problem.

This looks a tautology but isn’t. Medical costs are also correctly seen as an economic problem since medical care has opportunity costs. The patient, her fellow-insured and the taxpayer all have other things they could do with the money to pursue their individual and collective happiness. Formally speaking, this would still be true if we all only paid $1 a year, but the effect would be trivial. Today, in Ougadougou and Brighton and Santa Monica, the crowding out is frighteningly high and getting worse all the time. The economic problem has health effects, since nobody does or can give lexical priority to all health expenditure. Those who pay for medical services do not in practice pay for state-of-the-art care for all (though I guess the Nordics still come close). Doctors will regret these choices, but they are part of the world we live in.

Unafforded care has health consequences. These are large and serious. A 2009 Harvard study put the annual mortality associated with lack of health insurance in the USA at 45,000. In countries with universal care the mortality from lack of access is lower, but not negligible. In the UK there are surely deaths from rationing: for instance, there are too few oncologists. You can in principle measure these deaths and morbidities, even if it isn’t done enough.
By this route, the economic constraints and choices behind unafforded medical costs penetrate the membrane guarding the medical world and become a set of health problems: and meeting health problems is what that world does for a living.
The response generally preferred by the profession, like others, is to grumble about the choices their environment makes, to blame the deciders for the results, and to try to lobby away the constraints. Understandable, but inadequate. First, it’s not going to work, entirely. Second, the current pattern of treatment and its cost burden also reflects choices made in the past within the medical world over organisation (mega-hospitals vs. primary care) and in research (under-investment in malaria and STDs). Doctors are much more likely to have low-cost treatments and diagnostics if we look for them systematically and purposefully than by looking under the streetlights of current programmes and priorities. As Pasteur and later Fleming said, “fortune favours the prepared mind”.

So my recommendation 2 follows from 1 and the Hippocratic oath:

Making medical innovation affordable should be a major priority for medical research.

These people know their trade, and I’m sure they can think up better ways of carrying out my recommendation than I can. I’ll still kick a few cans down the road in my third and final post in this series.
Postscript
To anticipate two objections.
1. It’s unethical to internalise the necessary and external evil of rationing in medical research, which should single-mindedly devote itself to expanding human capacities for intervention to promote better and better health.
Reply: (a) Bollocks to the ivory tower. Tell that to the African peasants. Medical research is part of medicine, and must address the needs of patients in the real world.
(b) Rationing is an evil insofar as it creates ethical dilemmas and unreasonable burdens for carers and patients. To the extent these harms can be avoided, they should be (Hippocrates). The choice in research between a cost-reduction project (for at least equal health outcomes) and a treatment-improvement project (involving equal or higher costs) is not a true dilemma; as they can both be assessed against the metric of ultimate improved health outcomes.
2. It would be terrible to subject all medical innovation to a cost metric. Wouldn’t that mean that you would never spend money on say better treatments for spina bifida?
Reply: Quite right and I am not proposing any such thing. The idea is to bend the overall cost curve for applied medical innovation, not in every area. Success in reducing costs in any area will reduce the cost pressure on others. By the same token, it would be absurd to impose a top-down cost ceiling for innovation in advance. No research plan survives contact with the enemy, and serendipity will rule the actual payoffs.

Previous post
[Update] Final post in this series

Author: James Wimberley

James Wimberley (b. 1946, an Englishman raised in the Channel Islands. three adult children) is a former career international bureaucrat with the Council of Europe in Strasbourg. His main achievements there were the Lisbon Convention on recognition of qualifications and the Kosovo law on school education. He retired in 2006 to a little white house in Andalucia, His first wife Patricia Morris died in 2009 after a long illness. He remarried in 2011. to the former Brazilian TV actress Lu Mendonça. The cat overlords are now three. I suppose I've been invited to join real scholars on the list because my skills, acquired in a decade of technical assistance work in eastern Europe, include being able to ask faux-naïf questions like the exotic Persians and Chinese of eighteenth-century philosophical fiction. So I'm quite comfortable in the role of country-cousin blogger with a European perspective. The other specialised skill I learnt was making toasts with a moral in the course of drunken Caucasian banquets. I'm open to expenses-paid offers to retell Noah the great Armenian and Columbus, the orange, and university reform in Georgia. James Wimberley's occasional publications on the web

19 thoughts on “First, do no harm (medical research edition)”

  1. > In the long run, possibly. But before that
    > we are all broke.

    I guess I am unsure what we are supposed to be doing with all this wealth we would save by _not_ spending it on extending life and providing medical care? [And I’ll note that by “extending life” I don’t mean endless months in an ICU at the very end; I have signed an advance directive and made similar arrangements with my family). Setting aside arguendo the (very real) possibilities of peak oil and environmental catastrophe, we in the first world have incredibly wealthy and abundant economies. Should we continue to spend enormous amounts of our wealth on utterly useless military toys and similarly useless Wall Street/City shennanigans? What exactly is wrong with a collective social decision to spend more of it on medical care?

    Cranky

  2. Exactly Cranky. What are people for? If a human-made economy does not serve humans, it needs to be discarded.

  3. Well, look at current US politics. The battle over Medicare is precisely over the opposition of the Republican Party – reflecting the regrettable but we have to assume genuine opinion of many Americans – that since the combination of rising medical costs and social insurance = higher taxes, and higher taxes are unacceptable as a lexical priority, then you have to desocialise medicalcare of the elderly. If they fail, then the taxes rise. If they succeed, you don’t get the higher taxes but you get inexorably rising private costs, from a higher base so the doomsday is brought forward. Also, the opposition to the higher taxes required by public heath programmes is already preventing necessary investments in transport infrastructure and green energy. I don’t think you have to wait 10 years and health costs reach 25% of GDP to see real and nasty crowding-out effects in the USA.

    Mathematically, the problem is Malthusian: a part that grows faster than the whole indefinitely ends up consuming it all. Something will have to give, probably messily.

    Ethically, the problem is that medical consumption is not a good, merely a necessary evil. Hospital stays and operations and tests and pills and shots are not fun. By themselves they are minor bads. The things they will crowd out include the genuine goals of consumption: love, improving the mind, conviviality with family and friends, travel, sport, enjoyment of nature, and the arts (make your own list). A world in which these are eaten up by permanent medicine is the dystopia of the Struldbrugs.

  4. Mathematically, the problem is Malthusian: a part that grows faster than the whole indefinitely ends up consuming it all. Something will have to give, probably messily.

    If you look at it in a certain way, these policies – never mind their human-origin morals or ethics – are profoundly green and earth-friendly.

    That’s right: their policies may ease the human burden on Mother Earth by lowering population. Whoda thunk it?

  5. > Ethically, the problem is that medical consumption is not a
    > good, merely a necessary evil. Hospital stays and operations
    > and tests and pills and shots are not fun. By themselves they
    > are minor bads. The things they will crowd out include the
    > genuine goals of consumption: love, improving the mind,
    > conviviality with family and friends, travel, sport, enjoyment
    > of nature, and the arts (make your own list). A world in
    > which these are eaten up by permanent medicine is the dystopia
    > of the Struldbrugs.

    Ah, you have a tremendous number of assumptions and inferences packed into that short paragraph, and to me at least your conclusion (“dystopia of the Struldbrugs”) needs far more justification than you have provided. Even on the surface two problems are apparent: First, while I understand what you mean when you say medical services are not a “good”, they are nonetheless employment for those who provide them, who provide the support services, who build & maintain the facilities, manufacture the supplies, etc. with our economy in the process of being destroyed by outsourcing and offshoring a large industry which by its nature cannot be offshored is not a bad thing.

    Second, you seem to assume that long-term management of conditions will automatically “eat up” the “genuine goals of consumption”. But you leave out the possibility that such care will make possible “genuine goals” that weren’t possible before the treatments were available. E.g. my grandmother’s lens replacement surgery: using technology and surgical techniques unavailable just 20 years ago she can now enjoy vastly better sight than her mother did for the last 30 years of her life (and in fact reported better vision than she had experienced since teenagerhood). It did cost her two days of surgery, two weeks of recovery, and two months of putting in eyedrops (thanks Medicare!), but that is hardly a dystopia. Yes, it is possible to get to the point where the cost/benefit of medical treatment gets too high, but that’s not an automatic argument against all treatment up to that point.

    Cranky

  6. There is a basic problem here: these major funders of research are not doing what you want them to be doing. But you seem to think it’s their problem, and I’m not so sure. It is true that the NIH doesn’t devote most of its resources to making health care cheaper, or maximizing the availability of existing cheap health care – at least in part because that is not their mission. The NIH and its peers fund basic research and discovery; their main mission is the function of the human body, not the function of the health care system. You want money to be spent on and attention to be paid to a variety of worthwhile goals, but have chosen to express that desire in the form of criticizing the fact that the role you envision is not filled by the NIH and its peers, and that this isn’t where the money they spend is going. In short: you’ve called for the already endangered bulwark of research excellence to be further ravaged so that its resources may be shifted away from its current efforts and onto new and different areas. Not an especially good idea, in my opinion.

    Your substantive and specific complaints (about high-throughput sequencing, for example) are merely an unfortunate display of ignorance that is not really central to your main point, and are therefore best ignored.

    PS Still, to nitpick on one of your parentheticals: your snide aside about Rosalind Franklin was incorrect. The question of whether Franklin should have shared in the Nobel Prize awarded to the discoverers of the structure of DNA is a complete non-issue, as she died more than four years before that Prize was awarded and the Prize is never awarded posthumously. Franklin was treated abysmally, Watson continues to be a massive prick on this issue as he is on most issues, and I recommend that anyone interested in the discovery of the structure and function of DNA and in the role of Rosalind Franklin read the masterwork of the recently sadly deceased Horace Freeland Judson, The Eighth Day Of Creation – or they might read Brenda Maddox’s book specifically on Franklin, The Dark Lady Of DNA, which has been highly praised but which I haven’t read. But your little bit of snark was misplaced.

  7. This is a large subject — but here is my one idea: as the recipient of some fairly invasive cancer screening services, a first step would be to increase efforts to diagnose and screen cheaply (empahsis on cheaply). The equivalent of pap smears (instead of CAT scans and MRIs and colonoscopies) to either find disease or pre-screen people who need further intervention would be a big step towards reducing use of imaging technology and identifying cancers when they are most amenable to surgical intervention rather than increasingly expensive “chronic” medicines. Surgery works if the mass is contained — it’s the diagnostic tools that are lacking to find it an early enough stage.

    Secondarily, it just seems like most of our big “medical” problems are lifestyle problems. Medicine just doesn’t have a lot to teach us about curing or avoiding those, and if we were more honest and special interests were less entrenched, we would no longer see them as diseases in need of cure so much as lifestyle problems in need of better public planning.

  8. In homage to Jonathan Swift, who 300 years ago imagined the horrors of immortality without a stop to aging, let’s call this unsustainable situation a Struldbrug box.

    The Greeks beat him to this by more than 2 millenia.

    Maybe our healthcare system will become a cicada. No healthcare for anyone at all, except every 17 (or 19) years, when those who have survived the intervening period will take all that they have saved up during that time and put it toward a physical.

  9. Warren: Thanks for the correction on Franklin, post updated.

    You also write: “The NIH and its peers fund basic research and discovery; their main mission is the function of the human body, not the function of the health care system.”

    If that’s so, why is it called the “National Institutes of Health?
    The actual mission statement reads (my emphasis):

    NIH’s mission is to seek fundamental knowledge about the nature and behavior of living systems and the application of that knowledge to enhance health, lengthen life, and reduce the burdens of illness and disability.

    It’s a historical accident that so much basic biology now gets funded under the medical label, not only in the US. I don’t suggest changing this anomaly because it works. But it’s absurd to say that knowhow about the delivery of innovation in actual health care is inherently outside the missions of such bodies. In practice the topic is neglected as an operational choice that may be reasonable, but which they should be expected to justify.

  10. I’m not sure why you think the Republican Party’s position reflects “the genuine opinion of many Americans,” but as far as I can tell most Democratic, Independent and even Republican voters are strongly supportive of Medicare. This is likely why the Republicans accused the Dems of threatening Medicare as their central theme for the 2010 elections – which would have been an odd strategy if these Medicare was unpopular. The present effort to attack Medicare under the Ryan plan is premised as saving Medicare, and they are shrieking in horror at the suggestion that what they are doing is anything but that. For some reason, replacing Medicare with vouchers and calling it Medicare is being treated by journalists as something other than a trick. This need to dissemble about what they are up to only makes sense given that Medicare is wildly popular.

    I think it is fair to say that this is an elite driven phenomena, as conservative Republicans opposed the creation of Medicare (as they did Social Security) and eventually were forced to stop admitting as much because the program was too popular, and now they are looking for ways to undermine it that won’t cause a backlash because they hope no one will notice.

  11. Re James Wimberley’s reply to other posters, above: “reflecting the regrettable but we have to assume genuine opinion of many Americans…” I am very much not convinced of this. I think when your average Republican voter figures out that the choice is between health care for aging parents, and keeping one’s house, they will get back with the program of traditional Medicare very very quickly. As they should. It’s probably already happening. Ryan’s plan just sounded good in the two-paragraph summary most people read in the first day or two after it came out.

    As for the rest of your post, you raise important questions. But I think you should raise them much more briefly, and oftener! What else might we be doing? Barbara has some good ideas. We haven’t even *started* trying to prevent ill health here. Not even close.

    Also, the specific attack on the NIH is misplaced. What you’re talking about is simply not what they do. How is a researcher trying to solve a puzzle supposed to factor in the cost of the solution she hasn’t discovered yet? I think this would lead to self-censorship. Why not let her do her job and we’ll worry about how to fund it, or not, later?

    But again, very important subject.

  12. New Scientist: “I, algorithm: A new dawn for artificial intelligence”
    http://www.newscientist.com/article/mg20927971.200-i-algorithm-a-new-dawn-for-artificial-intelligence.html

    This is why Domingos places such faith in automated medical diagnosis. One of the best known is the Quick Medical Reference, Decision Theoretic (QMR-DT), a Bayesian network which models 600 significant diseases and 4000 related symptoms. Its goal is to infer a probability distribution for diseases given some symptoms. Researchers have fine-tuned the inference algorithms of QMR-DT for specific diseases, and taught it using patients’ records. “People have done comparisons of these systems with human doctors and the [systems] tend to win,” says Domingos. “Humans are very inconsistent in their judgements, including diagnosis. The only reason these systems aren’t more widely used is that doctors don’t want to let go of the interesting parts of their jobs.”

    Automate.

    You got some white spots on your back…
    You type in your symptoms. The algorithm says with 99.9% certainly it is the fungus Tinea versicolor…
    A online prescription shows prices for the topical tube of medicine from sellers around the world.
    You place your order. UPS delivers. A week later the spots are gone.

    That’s how you bend the cost curve.
    The problem of course is no one wants to empower you and I to such an extent…

  13. Re: artificial intelligence — I agree that it should be used to inform medical diagnosis, although I expect that its true potential is to enable health care personnel lower in the medical hierarchy to make at least tentative diagnoses, for a number of reasons.

    Also, further re diagnostic opportunities: my father died of pancreatic cancer, and nearly everyone who has it dies because it is almost never discovered in time to do surgical treatment to remove the entire mass, followed by chemotherapy. Those who do get it treated in time (Justice Ginsberg) survive for many years.

    When a colleague of mine died of it I went back to look at what is going on the world of pancreatic cancer research, and I found the following — researchers believe that, upon diagnosis, the average pancreatic cancer is at least 15 years old — that means there are at least 7-10 years in which it could have been diagnosed before it metastasized. Other common GI cancers like colon and esophageal (fastest growing cancer among Caucasian men in the U.S.) are probably similar. Imagine going to a doctor for an annual check up and once every two or three years beginning at age 35 or 45 getting a diagnostic panel that includes a blood or stool test for the robust markers of common GI cancers — the test doesn’t even have to be perfect. If you get it every other year, it just has to be good enough to get it mostly right because chances are even if it was missed this time, it will get it next time.

    Even mammography is very invasive as a SCREENING tool — for one thing, it delivers a dose of radiation, requires a separate appointment etc. Colonscopy is way to invasive to be the kind of screening tool it is sold as.

    In other words, right now, we are focused on imaging as screening and massive investments in management of diseases at the back end stage where they have become incurable. The aim should be for less invasive screening that allows for surgical intervention at an early stage when it has a good chance of being curable. This will also cost much less money.

  14. One problem here is that “low-cost” has two different meanings, depending on which end of the process you’re looking at. There’s low production cost (which is crucial in, say, the malaria-control example) and there’s low delivery cost, which is true in the curve-bending exercises but is more about market forces and willingness to pay than about, say, the marginal cost per treatment-unit of stuff being produced. Even if the NIH discovered (and turned over to a firm for commercial development) a cure for most cancers/strokes/alzheimers that cost a buck a unit in chemicals and marginal production equipment to produce, in the modern environment it would cost $5K a dose until the patent ran out. (And that’s not necessarily profiteering, just the way big new drugs get to market and distributed to the public.)

    Many (not quite all) of the local incentives for everybody involved in providing health care are in the direction of increasing costs, regardless of the “actual” cost of the individual treatments involved. Until that changes, changes in the cost of individual interventions are going to have marginal effects at best.

  15. paul, you are so right about that. That’s why surgical intervention usually ends up costing less than medical (i.e., pharmaceutical) intervention. All the incentives are stacked in the direction of coming up with useful but costly interventions. If you saw the front page of today’s NYT, you will see that new drugs for mealanoma are considered to be a breakthrough for extending people’s lives by “several months.” I didn’t read through for cost information, but think about that. A few months is hailed as if it were the equivalent of a cure, not to say immortality! It’s striking when you are outside the system — the bubble these researchers live in is thick indeed.

  16. Barbara: the piece doesn’t mention cost, as it turns out, but it’s also somewhat deceptive in another way: that “several months” is apparently a mean value, with many patients entirely unaffected (except for massive lost quality of life for the months they had left) and others apparently getting 6-24 months. A big part of effectiveness review, if it’s properly done, will be in figuring out the subsets of patients for whom certain treatments work and those for whom they don’t. There are a whole lot of very expensive placebos being given out there. (And back when I was following this stuff closely, patients typically wanted about 4 months of increased survival to compensate for the hassle and side-effects of then-current chemotherapy. That was independent of cost issues, just based on quality of life.)

  17. “That’s why surgical intervention usually ends up costing less than medical (i.e., pharmaceutical) intervention.”

    Unless something gets deliberately left behind in the body, in which case the same sort of regulations as are applied to drugs kick in, and suddenly everything gets incredibly expensive again. (Which is why Lasik is so much cheaper than corneal ring implants, despite the latter requiring much cheaper equipment, and being completely reversible.)

    The chief difference, I think, isn’t just the patents, but the fact that you don’t have to do a massive study to get the FDA to approve a surgical approach, most of which lead to rejections.

  18. Brett, you make a fair point in your last paragraph. I have become to wonder whether many of our legal protections for IP are now just as likely to detract from as to contribute to the common good.

  19. Barbara, if you were to relax protection of ip in the form of patents, while keeping the regulatory process the same, you’d make things worse, as the patents are the only reason anybody can sometimes afford to leap through all the flaming hoops. The problem is the flaming hoops. Countries with far fewer of them than the US suffer no more problems.

Comments are closed.