Via Ari Friedman:
If driven by belief in a positive outcome, the placebo effect would seem to require deceit. After all, if one knows one is taking a sugar pill, why should one think anything good would happen? And yet, at least two studies suggest non-deceiving placebos are possible. I discuss them in my latest AcademyHealth post.
Let me just quote from the email:
The Translation and Communications Interest Group is sponsoring a Twitter chat with Dr. Austin Frakt on Wednesday, February 25 from 2 – 3 p.m. ET. Dr. Frakt is the lead author of the 2013 HSR Article of the Year, Plan–Provider Integration, Premiums, and Quality in the Medicare Advantage Market. This study is the basis for the 2015 Student Competition: Presenting Research in Compelling Ways, which will be held during a special session at the Annual Research Meeting in Minneapolis on June 15, 2015. Dr. Frakt will answer questions from students related to his work and the competition, using the #ARMStudent hashtag.
Whether a bit of economic methodology illuminates some noneconomic domain is an empirically open question. It can’t be held relevant on the general principle that human agency just has this utility-maximizing or decision-theoretic structure, because that’s the psychological view I find goofy. Nor can it be held irrelevant on the grounds that we are not homo economicus all the way down, in one setting after another. We’re not. But we don’t know that we’re never at all that outside markets, either.
Some of the analytic structures of modem economics-rent seeking, clearing markets, prisoners’ dilemmas, coordination games, opportunity cost, principal-agent problems, and so on-ought to be in the toolkit of any self-respecting social or political theorist. But no self-respecting theorist will limit her stock of tools to those being retailed by economists, however aggressive their salesmanship, else she’ll fall down on the job, lamentably, over and over again.
I thank Nicholas Bagley for sharing the paper. I’ll draw from it again in a future post.
Related, you should read Megan McArdle’s take on the externalities Obamacare might address. I will come back to them and respond to some of McArdle’s points another time, though I’ve written about Obamacare and externalities before (here and here, for instance).
But, just to put McArdle and Herzog together for a moment, McArdle wrote,
What if knowing that you and your partner are doing naughty things a few feet from my bed causes me severe mental anguish? That’s surely an externality, so why don’t we take notice of it? The answer is that we don’t simply say “OMG, negative externality! Quick, government, kill it with fire!” Because in that way lies madness. As it’s easy to see, reducing this negative externality itself has a negative externality, which falls on the folks who enjoy doing naughty things a few feet from your head.
Related, Herzog wrote,
Suppose June is horrified at the very thought that the lesbians down the hall are engaging in oral sex. Yes, they’re consenting adults, acting in private. But their doing so, she thinks, is disgusting. Their actions move her to a lower indifference curve. Maybe a much lower indifference curve: maybe the thought makes her physically ill. Maybe this is a visceral reaction, or maybe it flows from her deepest religious commitments. Either way, it’s a social cost they don’t have to internalize. […]
Maybe the lesbians should have to offer June financial compensation in order to go about their business. (It being their business, in this view, can’t be at bottom any claim of individual rights or personal autonomy. It’s just something they want to do, and their wants have no better standing than others’ wants that they not do it. […])
Ultimately, Herzog and McArdle come to similar views on externalities: when and how they arise in economics arguments doesn’t itself hinge on their economic import. Other principles come into play. Again, more later.
The following originally appeared on The Upshot (copyright 2015, The New York Times Company).
If you or a loved one is having a heart attack, your most pressing concerns probably include how quickly you can get to the hospital and the quality of care you’ll receive. You’re probably not thinking about the hospital’s board room, even though quality of care for heart attacks and many other conditions may be determined in large part by decisions made there.
”Most board members are community leaders, serving on the board to support fund-raising goals,” said Ashish Jha, a Harvard physician. “They don’t think it’s their job to hold management accountable for performance. Board members often feel like clinical quality is physicians’ jobs, and they don’t want to step on doctors’ toes.”
The trouble with this perspective is that boards, and other hospital management, can influence care in ways that individual physicians cannot. They can promote protocols that ensure that crucial information is conveyed to the right people at the right time. They can establish systems so that equipment and supplies are available when needed. They can set expectations for a culture of high performance, not just from individuals but from teams of them that must work together. And they can require quality to be monitored against goals with incentives to push it toward those targets.
”I’m a much better doctor in a well-managed hospital where the systems are in place to help me do my best work,” Dr. Jha said. “Even a great chef can’t produce a good omelet with eggs that are stored in the freezer or the stove doesn’t work reliably.”
Each hospital is a bit different, but generally board members are recruited for either their fund-raising prowess (if the hospital is nonprofit) or for expertise in a specific field like finance or regulatory compliance (if the hospital is for-profit), Dr. Jha said. Few board members are medical professionals. Most boards perpetuate themselves — in other words, the board itself, or its chairman, invites new members to join, rather than holding an election among a larger group of stakeholders.
In general, hospital boards do not view themselves as institutional champions of quality. According to work by Dr. Jha and his colleague Arnold Epstein, only 20 percent of nonprofit hospital board chairmen reported that the board was one of the top forces for quality at their hospitals. At hospitals with low-quality scores on standards established by Medicare, only 11 percent did so. Only half of boards view clinical quality as one of their top two concerns. In contrast, financial performance was a top priority for about three-quarters of hospital boards. The analysis examined the association of boards’ priorities with a wide range of evidence-based measures of quality, including those for heart attack care.
Troublingly, most hospitals boards can’t accurately assess their institution’s quality. There’s a Lake Wobegon effect: More than half of hospitals with low quality thought they were actually above average.
Almost all hospital boards have the power to hire and fire the chief executive officer, and the management that the C.E.O. provides reflects the board’s priorities. Jonathan Kalodimos, an economist with the Securities and Exchange Commission, found that when hospital boards exert stronger governance over management, as measured by level of involvement in setting compensation levels, more effective medical treatment is provided, leading to lower heart attack mortality rates. But when the chairmen of only 44 percent of hospital boards choose clinical quality as a top priority for evaluating C.E.O. performance, that may signal that quality is not paramount. Hospitals with higher quality are twice as likely to emphasize it in C.E.O. evaluations by their boards, according to Dr. Jha’s work.
Hospital C.E.O.s are highly paid, in general, earning annual salaries of just under $600,000, on average. But those salaries are not related to clinical quality, according to work by the Harvard physician Karen Joynt and colleagues. Instead, C.E.O. salaries are higher at large teaching hospitals, at institutions with higher patient satisfaction, and at hospitals that use more advanced technology.
In turn, C.E.O. leadership and management decisions affect quality. A health economist at Oregon Health & Science University, K. John McConnell, and his colleagues found that hospital management practices adopted from manufacturing and technology sectors — such as “lean” methodologies developed by Toyota — were associated with better care and lower 30-day mortality from heart attacks. These management practices include eliminating inefficiencies and variations, fostering collaboration, setting targets and tracking progress toward them. They can reduce the time it takes between when a heart attack patient arrives at a hospital and when he’s treated, improving outcomes. Other work found good management is associated with better quality of care in intensive care units.
If board and management leadership are so important for quality, how can we foster more of it? One way is to promote competition. Work by Mr. McConnell and colleagues found that hospitals facing greater competitive pressure have better management practices, though the relationship is of modest strength. This is consistent with other evidence reviewed by the health economists Martin Gaynor and Robert Town. They found that, in general, when hospitals consolidate, reducing competition, quality suffers.
The market should promote quality another way. Hospitals whose heart attack patients survive longer have a greater market share that grows over time, a study by the Harvard economist Amitabh Chandra and colleagues found.
This doesn’t mean we should sit back and let the market as it exists today be the sole arbiter of quality. Policy can play a role as well, working with the market, not against it. One way is to base payment to health care providersmore on quality and less on volume. In theory, this should align what hospital boards tend to focus on — financial performance — with what we might prefer they focus on — quality. Though many attempts at pay-for-performance have yielded disappointing results, one reason may be that the incentives (and penalties) were not large enough.
The American health system is known for its high cost. But what’s just as troubling, if not more so, is the mediocre quality it delivers for that cost. It’s natural to think that direct caregivers — doctors, nurses, and other technicians and assistants — bear the entire burden of providing better care. But research shows that we should consider the environment and culture in which they work, which is shaped in large part by their institutions’ boards and management. If we don’t foster a focus on quality at high levels, even the best health care workers may not provide the best care.
In Health Affairs, James Robinson claims we have:
Insurers are tightening coverage criteria and managing utilization more aggressively. Hospitals are pushing back on the prices charged for supplies and equipment. Physicians increasingly are being paid through methods that discourage prescription of costly treatments. Consumers are being asked to pay more for their care, both through higher premiums and through higher cost sharing at the time of service. […] The new era will challenge the technology industry to improve the value of its innovations, defined as emphasizing performance improvements that justify the prices charged.
The life sciences industry has won every battle but appears to be losing the war against the use of comparative clinical and cost-effectiveness research. The industry successfully lobbied Congress for provisions limiting use of comparative clinical evidence and banning the use of cost-effectiveness analysis for Medicare coverage policy. The Centers for Medicare and Medicaid Services (CMS) is also prevented from using these data in its reimbursement policies. But drug and device firms must submit comparative data when seeking coverage and reimbursement from payers in Europe and other nations. It may be difficult for these firms to avoid providing analogous dossiers to payers in the United States.
In some cases, they already do. The pharmacy and therapeutics committees of major insurers often require drug companies to supply studies and observational data on the clinical performance of their new products; committees increasingly use these data when deciding which products to include or exclude from formularies. Firms must also supply data on the cost implications of their new products, extending beyond unit price to include cost per course of care; cost for the covered enrollee population; and, in some cases, indirect costs through impacts on workforce productivity. If the firm cannot prove that its product is meaningfully superior in safety and effectiveness, insurers may assume that it is equivalent and not superior to products already on the market. They then make coverage decisions on the basis of price.
Data on clinical and cost performance are increasingly demanded by hospitals and integrated delivery systems as they accept capitation and other forms of prospective payment. Leading hospitals maintain technology assessment committees that draw on staff physicians and surgeons to help inform purchasing decisions with respect to implantable devices and other technologies. These committees not only assess the potential clinical and cost implications of a new technology relative to the existing inventory but also serve the cultural function of encouraging physician-leaders to consider trade-offs between price and performance. Comparative performance data may also be used in purchasing decisions by accountable care organizations (ACOs) serving the Medicare fee-for-service population, even if CMS is unable to use them directly for national coverage policy. […]
Some hospitals are beginning to insist that prices for implantable devices not exceed a defined percentage of the procedure revenues they obtain from insurers and that supply prices follow procedure prices downward as insurers drive harder bargains.
All this sounds plausible, but also aspirational. Studies that demonstrate improving cost-effectiveness would help the case, as well as potentially identify which of the various organizational models play key roles.
Unless you’ve been living in a snow cave for a month, you’ve heard about Boston’s very (very, very, very) snowy winter, which is breaking all kinds of records. Last night and into today, we got another foot or so, on top of the six or so dumped over the past three weeks. The snow accumulations just went from amazing to unfathomable.
I was already in awe of the snow before the latest dump. Snow banks towering over cars is one thing, but massively thick ones that reduce two lane roads to one are another. The standard line is that we have nowhere to put more snow. Even I’ve said as much, but it wasn’t really true at my house. I still had ready capacity … before today.
I just came in from over four hours shoveling roofs, walkways and the driveway. (According to MyFitnessPal, I burned over 2,000 calories doing so. I’ll be eating a lot today.) And, that was aided by a Bobcat. I don’t know where it came from or who was driving, but it saved me several hours of work on the driveway. I only needed to scrape down to asphalt and clean the margins.
That was hard enough. To make new capacity, I had to snowscape the big pile in front of my garage. I climbed up it on snowshoes to push snow off the top to make room for more.* To more easily deliver new snow to the top, I built some stairs by pounding in long two-by-fours.
With this much snow, I had to cut some corners. I closed our front steps. Though we could still get down in an emergency, they’re no longer ready for regular traffic or deliveries.
The snow is level with the bottom of some of our first floor windows. I can walk right up to the edge of our garage roof. The driveway is passable, but barely. The avalanches off our solar panels are thunderous. Keeping the hot water heater and furnace vents clear is a daily job. (On the plus side, though they don’t melt the snow directly beneath them, they do so a few feet out, making room to stuff in more snow.)
More snow is in the forecast. I have only one more place to put it: inside the garage itself. I just might.
* The pushed snow dumps into the neighbor’s driveway, but a portion they said they’re not going to use. Their second car buried there won’t move for months. (It’s also where I push garage roof snow.)
For the Washington Post, Peter Whoriskey reports:
The nation’s top nutrition advisory panel has decided to drop its caution about eating cholesterol-laden food, a move that could undo almost 40 years of government warnings about its consumption. […]
The finding follows an evolution of thinking among many nutritionists who now believe that, for healthy adults, eating foods high in cholesterol may not significantly affect the level of cholesterol in the blood or increase the risk of heart disease.
According to my reading of Gary Taubes’ Good Calories, Bad Calories, this is, at long last, consistent with the available evidence. However, the Post article makes no mention of any consideration of the role of carbohydrates, which probably is causing Taubes’ head to explode. (If you don’t know, his read of the evidence is that it is far more consistent with the idea that first world diseases — diabetes and obesity among them — are due to overconsumption of carbs. Though he’s also quick to admit that the evidence about anything in nutrition is not as strong as it could or should be.)
By the way, Good Calories, Bad Calories (GCBC), though a seemingly thorough tour of the history of nutrition and its (abysmal) evidence base, is not an easy read. It’s very long, repetitive, and doesn’t do a good job of helping the lay reader comprehend the physiological mechanisms, to the extent they’re known (or hypothesized). Your mileage may vary.
I see that Taubes has written a more recent book, which my be better in these regards. Or, perhaps, one of his articles or those of others offers a more accessible treatment of the carbohydrate hypothesis, but I don’t know. You can email or tweet at me if you’ve seen something particularly good and relatively brief in this area.
One of the most interesting bits of meta from GCBC is the idea that maximal promotion for public health purposes must gloss over scientific uncertainty. Articulate just a little bit of daylight between the science as known and the science as we’d like to believe and getting people to “do the right thing” becomes substantially harder. (Imagine, in today’s environment, articulating a single shred of uncertainty about vaccination, for instance. It’d be a disaster, as we’ve already seen from fraudulent claims of harm from vaccines.)
This tension comes through in Whoriskey’s reporting: “the problem in nutrition stems from the arrogance that sometimes accompanies dietary advice.” That arrogance exists, in part, because expressing uncertainty can water down impact. When one believes that one is likely on the right side of science, even with the uncertainty (and even when one is wrong about it), if one wants people to change, one cannot be the least bit wishy-washy. “We kinda think perhaps cholesterol is [or carbs are] not so good, maybe” will get you nowhere, as accurate a reflection of the science as that may or may not be.
Science is hard enough. Evidence-based public health with uncertain science is often even harder. If there’s a good way out, short of just keeping mum until the science is “sufficiently certain” (whatever that means), I’m not sure what it is.