Timothy Taylor finds a report from the OECD. Taylor writes,
The report divides the evidence into three main categories: wasteful clinical care (care that either provides very low value or can even be counterproductive to health); operational waste (like paying excessively high prices or overusing expensive inputs like brand-name drugs); and governance-related waste (like ineffective or unnecessary administrative expenses)
Ten years ago, in Crisis of Abundance, I concluded that the main issue was the first: medical procedures with high costs and low benefits. Taylor lists these examples from the OECD report:
Imaging for low back pain.
Imaging for headaches.
Antibiotics for upper respiratory tract infection.
Dual energy X-ray absorptiometry (used to measure bone mineral density).
Preoperative testing in low-risk patients (electrocardiography, stress electrocardiography, chest radiography).
Antipsychotics in older patients.
Artificial nutrition in patients with advanced dementia or advanced cancer.
Proton pump inhibitors in gastro-oesophageal reflux disease.
Urinary catheter placement.
Cardiac imaging in low-risk patients.
Induction of labour.
Cancer screening (cervical smear test, CA-125 antigen for ovarian cancer, prostate-specific antigen screening, mammography).
Caesarean section.
Note that cancer screening is on the list. Cancer screening is something of a sacred cow in the U.S. In fact, as I point out in my book, even something as widely advocated as colonoscopy to screen for colon cancer is likely to have a very high cost per life saved.
Taylor concludes:
In many cases, decisions about what medical care to receive and how to deliver that care fall into a gray area. It’s often not 100% clear whether a certain procedure was needed, or not needed; not 100% clear that an error was made, or whether a reasonable judgment call was made; or whether a certain administrative act is wasteful, or whether it is reasonable oversight that reduces the risk of poor care and holds down costs. But the report makes a persuasive case that a substantial share of health care spending, not just in the US but in all advanced economies, is not doing much to improve health.
This recalls what I wrote in Crisis of Abundance in chapter 3.
It is not true that health care is a black-and-white proposition in which services are either utterly necessary or else utterly unwarranted. . .Services that fall in the gray area are services that offer some benefits but which are not absolutely necessary.
To reduce the use of high-cost, low-benefit procedures, the parties paying for health care have to engage in rationing. If the patient is paying, then the patient will self-ration. If the government is paying, then the government will ration.
Proton pump inhibitors have vastly improved my life.
That raises a good question. How are we measuring the benefits, anyway? Figuring out consumer surplus is hard enough, especially without normal prices and markets. We just can’t observe the “quality adjustments” directly, so maybe we are underestimating benefits.
Maybe we are drunks looking for keys under street lamps because that’s where the light is by focusing so much on “lives saved”.
I stopped eating black beans on a regular basis and the heartburn almost went away completely. Had I sought medical help they might have operated on my sphincter or something.
That would have been horrible. You might have lost your ability to speak.
You have no idea how much I hold back now that I could just let go.
My heartburn is not really related to diet, so maybe I am the exception.
Was worth a try. I accidentally stopped eating one thing and it went away 90%. I wouldn’t have believed it had it not happened.
Would you personally pay $800/year for a prescription?
(My recollection is that statins were $2-4/day in the first half of their life cycle.)
If so, then you would self-select to take them. If not you’d self-ration.
Adding to the list of examples:
Costly physical therapy as the first step in the care “progression” for just about any orthopedic issue, and often prescribed at a PT center owned by the orthopedist. On three separate occasions, I’ve gone through a time-consuming series of PT visits that proved pointless based on later diagnoses.
Sending you to PT can be a way of wearing you down so you don’t seek more expensive treatment.
Approaching the problem from this point of view is going down a rabbit hole. The market should figure all this out for us. Experts shouldn’t be looking at this stuff and asserting top-down reforms.
Unfortunately there are no market mechanisms. There is no one who says no thank you. Patient data is really doctor data. With different intellectual property laws and different market mechanisms, imaging that cost $1200 might cost $99. Or maybe even $19.99. Often, as you proceed across multiple doctors, they re-image just because it is easier than dealing with getting data from another doctor. There just is no pressure on anyone to do better.
Personal anecdote. My wife talked to an MD about a procedure. His initial guidance was to obtain three items (scan / path / I forget). When informed we were self-pay and not insurance, his response was “go ahead and get A. I can use the B you already have. C we can wait on and decide later.
There are difficult issues in the provisioning of healthcare (trauma for out-of-area residents, high-cost/early-onset genetic disease).
But there is no question market mechanisms are grossly underutilized in healthcare.
So, reimbursements distort incentives. Doctoring is also applied research. There needs to be a disaggregation of the incentives and the doctoring expertise, and then the outcomes need to be fed back to the insurance companies. We also need more doctors. So, what if we had insurance company protocols that could be applied by nurse practitioners. Going outside the protocols requires a doctor who then must provide the feedback data to the insurance company for judging outcomes based on presentation and demographics. On the other hand, local knowledge trumps central knowledge, so while you might be willing to bet on a low-value procedure and should have that option, if it doesn’t work you won’t get full reimbursement for it. Shouldn’t insurance companies be interested in things that improve outcomes? Why do they seem unconcerned with net outcomes? Or is this part of the problem, that a lot of their payouts are to people who are no longer paying in?
Insurance companies want to reduce claims and increase revenues.
They would pay zero claims if they could still get a premium, but nobody would pay the premium.
So they cover what they need to in order to get decision makers who manage health plans to fork over premium dollars.
If “what makes decision makers fork over premium dollars” and “what improves outcomes” aren’t the same then you will find a difference in coverage vs outcomes.
In my experience decision makers who fork over premium dollars have little more then a vague set of desires and complaints from the various stakeholders they answer to. Their primary wish is that they could just get back to doing whatever it is they do for a living rather then try to understand a health plan.
Also, the average time a policyholder stays on a given plan is something like 3 years. So (a) it’s hard for insurers to gather clean long-term data and (b) there’s not much incentive to invest for all but the shortest time horizon. Also, I believe the claims data they see doesn’t have much clinical data (e.g., results of tests) so that makes it harder for insurers to develop targeted policies.
So I just went through a meeting that went like this:
GM: The group client want Rosuvastin on a lower tier with a $0 copay. It’s cheap everywhere else, we aren’t competitive.
Me: It’s cheap at wholesale, but for complicated reasons XYZ we contract drugs based on this particular metric and Rosuvastatin happens to be a weird drug where there is a big difference between real world cost and what we have to pay the pharmacies for it. As such the difference between Rosuvastatin and Simvastatin, while maybe only a few dollars in real life, is like $30/script for us. We really don’t want to cover it on Tier 1 $0 copay because it will cost your plan $X.XX PMPM and they probably aren’t going to like that.
GM: We need to cover it on Tier 1 $0 copay because that is what everyone else is doing. The members that take this drug are making a big deal about it.
Me: Well, if we raise premium $X.XX PMPM then every member is going to complain. Why can’t you try to switch these people over to any of the other Statins? They all do the same thing. They’ve always been a bunch of scam me-to drugs.
GM: It doesn’t matter what they do. There is a perception that they are better.
Me: There is a perception that they do more because these are rich group plans where the people taking these drugs haven’t had to pay the full cost. If they did, I bet their opinions would change.
GM: Well explain that to client X. All he knows is that a lot of his members are complaining. I have to make this sale.
Me: Well I don’t know what would make the sale or not. I just know this is a very expensive subsidy to a small group of people who could easily get what they want by changing behavior.
Me: Also, why do we need $0 copay, is $2 really that much. And why do you want a new formulary. We can only align our pricing with one formulary for this product. If you break the groups into several formularies our pricing and benefits will be out of alignment, which means big premium increases for them. We spent all last year fixing that.
GM: Well client X wants ABC, and client Y wants DEF, and client Z wants…
Me: Yeah I get that but people can’t get everything they want. If they really understood the cost I’m not sure they would.
GM: They care about cost but individuals cause more noise then (a slightly higher then inflation premium increase). I need to make this sale. If I don’t make the sale there will be hell to pay.
Me: Well, you can do whatever you want but I’m adding it to the list of reasons I’m going to give my boss why we didn’t beat cost trend on your block this year. (Just like last year.)
GM: Also, whenever a pharmacy U&Cs a drug on the STARS list, even though the person is adherent, the fact that it adducted cash means that CMS thinks they weren’t adherent on our plan. So that hurts our STARS, even if it has nothing to do with reality. We need to price below U&C so we get STARS (STARS is supposed to be a quality measure).
Me: Well, we price all the other statins well below U&C. Can’t we just switch people over.
GM: We can’t get everyone switched over right away, so those still taking the drug will look bad on STARS. We can’t afford to miss STARS for a year or we will miss plan and not get bonuses.
Me: (Well then I guess we will need to make financially dubious decisions of no clinical value in order to do better on a clinical metric).
Relying on the unknowledgeable to ration is almost certain to be a mistake. It is an exceptionally bad mistake when what is being rationed is access through ability to pay, leading both to avoidance by the poor and excess use by the rich. Nor is excess use the greatest problem. Only in some advanced and specialized areas like imaging is use in America higher. Doctor visits are fewer and shorter, nurses fewer, hospital stays shorter. For this we have much higher prices.
Prices started going up exactly when the government got involved.
I’m running out of patients, hardy har…and can’t find a doctor to overprescribe me pain meds.
The alternative is payer (insurance/gov’t) rationing based on some combination of expert conventional wisdom and non-price factors like ‘are you willing to sit in the ED for 10 hours’. Ability to pay is the worst way except for all the others.
Every poll should include inquiry into the balance of cost v benefit. Everyone wants more of … whatever…, until they’re asked to pay for it, either in cash or quality.
Seems to me that any poll result on any topic is pretty much misleading (dishonest?) unless that balance is reflected somehow.
https://www.cato.org/blog/support-obamacares-most-popular-provisions-plummet-quality-care-threatened