By Michael D. Frakes, Jonathan Gruber, and Anupam Jena
National Bureau of Economic Research, July 2019
Stemming from the belief that the key barrier to achieving high-quality and low-cost health care is the deficiency of information and medical knowledge among patients, an enormous number of health policies are focused on patient education. In this paper, we attempt to place an upper bound on the improvements to health care quality that may emanate from such information campaigns. To do so, we compare the care received by a group of patients that should have the best possible information on health care service efficacy — i.e., physicians as patients — with a comparable group of non-physician patients, taking various steps to account for unobservable differences between the two groups. Our results suggest that physicians do only slightly better in adhering to both low- and high-value care guidelines than non-physicians – but not by much and not always.
There is widespread belief in health care that the deficiency in information and medical knowledge among consumers is the key barrier to achieving high-quality health outcomes and lower costs of care. An enormous number of health policies follow from this assumption, including efforts to educate patients about health, both at a population level through large-scale public health campaigns (e.g., campaigns to encourage exercise and healthier diets) and at an individual level through provider-led patient education (e.g., education of patients about the inappropriateness of antibiotics for the common cold, the importance of vaccination, etc.), decision-making support, and so on. In addition to information-based policies intended to improve health, other important demand-side approaches intended to improve the efficiency of health care (e.g., high-deductible health plans) focus on increasing direct consumer engagement in the purchase of health care, relying heavily on the notion that informing patients about the costs and benefits of health care services can steer patients towards higher-quality, more efficient care.
The widely prevailing view that information scarcity is a key obstacle to achieving high quality health outcomes raises the basic question: is it simply inadequate medical knowledge and information among patients that has limited the utility of various demand-side approaches to health improvement? If so, it remains possible that improvements could be made along both cost and quality lines should policymakers and providers encourage even greater levels of information disclosure to patients, a point not only stressed by patient-centered health care enthusiasts promoting shared decision-making models, but also intrinsic to the many public health efforts that aim to use patient education as a primary way to improve health. In this paper, we attempt to shed light on whether eliminating the scarcity of information in medical knowledge – an extreme form of information disclosure – could plausibly lead to the delivery and receipt of higher value patient care.
A natural way to explore the importance of scarcity of medical knowledge and information is to look at the care received by a group of patients that should have the best possible information on health care service efficacy: physicians. The decisions of physicians about which type of health care to receive would likely place an upper bound on how well non-physicians could do in selecting their health care treatments if fully informed about the costs and benefits of different types of health care interventions. With the exception of Johnson and Rehavi (2016), who focused on a single health care service – cesarean sections – there is no work which has been able to study the role of physicians as patients.
We are able to address this shortfall by using a source of data previously unexploited in economics research: data on physicians in the Military Health System (MHS). The MHS provides health insurance for all active duty military, their dependents, and retirees. Care is provided both directly on military bases and purchased from an off-base network of contracted providers. The MHS is one of the largest sources of health care spending in the U.S., with spending of over $50 billion per year. We have gathered data on the complete claims records for all MHS enrollees over a ten-year period. Importantly, this includes the claims data for MHS physicians when they are treated as patients themselves (drawing from records of over nearly 35,000 military physicians). The ability to observe physicians as patients provides us with a unique and powerful opportunity to answer the question: do especially well informed patients elect to receive higher value medical services?
For these purposes, we evaluate the quality of health care services received by physicians and non-physicians, focusing on a set of services for which objective, evidence-based standards exist. In particular, we assess whether physicians receive more services deemed “high value” by the relevant medical literature and fewer services deemed “low-value,” in each case relative to the less-informed comparison group of similar non-physician patients. An example of low value care is a chest x-ray before eye surgery, a very low risk procedure for which pre-operative diagnostic testing is not recommended. An example of high value care would be statin therapy for patients with cardiovascular disease. If deficiency of medical knowledge and information is an important reason why demand-side interventions to improve patient health have met with limited success, we should expect that physicians as patients would exhibit markedly higher levels of adherence to high-value health care services and lower levels of low-value care compared to otherwise similar non-physician patients.
Our results suggest that physicians do only slightly better than non-physicians – but not by much and not always. Across most of our low-value settings, physicians receive less low-value care than do non-physicians, but the differences are modest, and generally amount to less than one-fifth of the gap between what is received by non-physicians and recommended guidelines. The results are slightly more mixed in the case of the high-value care analysis, with some evidence suggesting that physicians appear to receive high-value care at roughly the same rate received by non-physicians and some evidence suggesting that physicians do slightly better than non-physicians. These results provide a rough boundary on the extent to which additional information disclosure (beyond prevailing levels) can be expected to improve the delivery of health care in the U.S. Relatedly, these findings suggest that, despite the threat to the optimality of the health care system posed by information asymmetries between physicians and patients (Arrow 1963), most of the explanation behind the over- and under-utilization of low- and high-value services likely arises from factors other than informational deficiencies of patients.
Moreover, in the case of low-value care—where we do find modest physician effects — we attempt to shed light on the mechanisms underlying any information channel. We do not find consistent evidence to suggest that the observed average physician effects arise from the patient’s choice of a low-intensity physician or from a reduced scope for physician induced demand. Given these findings, it is likely that some of the modest physician effects we estimate arise simply from informed patients making higher quality choices at the moment of the care decision itself.
Our paper proceeds as follows. Part II reviews the existing literature on the impacts of health care information initiatives and provides a conceptual overview behind our approach of observing physicians as patients as a means of exploring the bounds of the quality gains that may arise from greater information disclosure. Part III provides a background on the Military Health System and discusses our data and empirical methodology. Part IV presents the results of our analysis and Part V concludes.
Discussion and Conclusion
We attempted to explore the upper bounds of the gains that may ensue from consumer-driven health care approaches that draw upon greater information-disclosure to patients by exploring the care choices of physicians when acting as patients themselves relative to the clinical choices of non-physician patients. In an attempt to isolate the information-component to the differences between these two groups of patients, we took various approaches to ensure comparability between physician and non-physician groups. Our results consistently suggest that, at most, superior patient knowledge is associated with only modest improvements in the quality of care selected during medical encounters. Perhaps one interpretation of these findings is that patients remain generally deferential to the care recommendations of their treating physicians, even in the case of near fully-informed patients. This interpretation would be consistent with other recent findings in the health economics literature, including recent research by Chernew et al. (2018) showing that referring physicians are dramatically stronger determinants of where patients receive MRIs relative to patient cost-sharing factors. Should concerns over the quality of care received by patients remain—as is suggested by the large gaps demonstrated between prevailing rates of adherence to high- and low-value care guidelines and recommended rates of adherence — our results are informative on the limitations that may come with information-focused, demand-side solutions.
Despite our efforts to ensure comparability across our treatment and control groups, concerns may remain over the generalizability of our findings. In particular, among other external validity issues, some may be concerned that we tested for the impacts of information in an environment characterized by very low levels of cost-sharing. The influence of information, however, may interact with financial considerations of this nature. Consider, for instance, preoperative testing and assume, as is the case with the Military Health System beneficiaries, that copayments for these diagnostic procedures are negligible. Even though such tests may render insufficient benefits to justify their full social costs and even if patients are aware of this, informed patients may nonetheless undertake the tests anyway given the negligible financial implications. In contrast, informed patients may decide to forego these tests when cost-sharing is more substantial.
This consideration does not necessarily threaten the external validity of our low-value care findings—at least not overly so—in light of the fact that the costs to undergoing low-value care extend beyond just direct financial outlays. For instance, in the case of pre-operative testing by the patients in our sample, we demonstrate in the Online Appendix that much or most of these instances of pre-operative testing necessitate an entirely separate visit to a medical provider on a separate day, an outcome that likely imposes costs and inconveniences of another nature on the affected patient. Accordingly, if we still estimate modest physician effects in the face of low-value care decisions that carry potentially significant non-cost-sharing consequences, such modest findings may indeed generalize to populations facing higher levels of cost-sharing.
Further reinforcing this point, the costs associated with receiving low-value care may extend beyond cost-sharing dollars in light of the possibility of physical harms and discomfort associated with the care itself. This is very likely true with the case of cesarean delivery, especially considering the recovery period involved. These physical consequences are, of course, likely to be far less significant in the preoperative testing setting, especially in the case of blood draws.
Finally, it is important to note that this particular generalizability concern does not extend to the high-value care analysis. If anything, the fact that cost-sharing is lower in the MHS should only elevate the role of information in encouraging greater guideline adherence. In this light, our finding of little to no improvements in high-value-care adherence in fully-informed patients is likely to generalize to settings with higher levels of cost-sharing.
The entire 49-page paper can be downloaded for free at the following link:
Doctors aren’t much better at picking the best medical treatments than laypersons
By Michael Hiltzik
Los Angeles Times, July 8, 2019
In recent years, the idea has spread that forcing consumers to pay more for healthcare — giving them “skin in the game” is the usual mantra — will prompt them to become more discerning medical shoppers.
The goal is to improve the efficiency of the healthcare system by saddling consumers with higher costs if they opt for less useful or more overpriced services. Experts have identified numerous flaws with this concept, as we’ll get to in a bit.
But a new paper by Michael D. Frakes of Duke, Jonathan Gruber of MIT and Anupam Jena of Harvard raises an important and oft-overlooked point: Even if consumers have better information about the efficacy of particular treatments, will they be any good at using that information?
The answer is: probably not.
The authors tested the theory behind “consumer-driven healthcare” by examining the behavior of the best-informed consumers of all: doctors. (Their sample of physicians as patients came from the massive Military Health Service.) What they found was that doctors are better at picking high-value treatments and avoiding low-value care — but “not by much and not always.”
The message is that “purely relying on consumer cost-sharing and high deductibles won’t get us to the best outcome,” Gruber says.
The authors’ examples of low-value care included caesarean sections, which are judged by experts to be vastly overused; and pre-operative diagnostic tests for low-risk surgeries, such as chest X-rays prescribed for patients undergoing eye operations. The high-value examples included comprehensive diabetes care, statin use for patients with cardiac risk, and child vaccinations.
“There really should be 100% consumption of high-value and 0% of low-value care,” Gruber told me. “On those benchmarks, doctors don’t do a whole lot better than non-doctors.”
The finding is important because it suggests there’s a limit on how much more efficient the healthcare system can become by making consumers better informed. If doctors, the best-informed medical consumers, can’t consistently be steered to the most cost-effective treatments, then obviously average consumers will do much worse.
“If our approach is just to get more information to patients,” Frakes says, “that may not give us a great deal of return.”
This study isn’t the first to cast doubts on the theory that high deductibles and co-pays alone will lead consumers to make better and more cost-effective choices on healthcare. In 2015, a team from UC Berkeley and Harvard examined what happened when a large company shifted from an all-expenses-paid health plan to a high-deductible plan in a single year.
They found that total medical spending fell by as much as 13.8%, but that consumers cut spending on “potentially valuable care” such as preventive services as well as potentially wasteful care, such as X-rays and scans. They found “no evidence of consumers learning to price shop after two years in high-deductible coverage.” That confirmed a 1993 study by the Rand Corp. that also found that consumers reduced spending on good treatments and bad alike.
The latest study found that physician families did reduce their rate of C-sections, but only modestly and not enough to bring the rate into line with what medical experts say is proper. (Medical professionals say the rate, which is about 30%, should be only 10% to 15%.) Physician patients received unnecessary pre-surgical diagnostic tests nearly 20% of the time, less than non-physicians but still much too often.
That led the authors to conclude that “even the best informed patients do not make any less use of low-value health services.”
As for high-value care, physician patients failed to receive comprehensive diabetes care 30% of the time. There was virtually no difference between doctors and laypersons in receiving post-cardiac care, or in vaccinations for diphtheria/tetanus/pertussis and hepatitis. There were “small improvements” in compliance for child vaccinations for chicken pox, polio and measles/mumps/rubella.
The findings don’t show that there’s no place for deductibles and co-pays in the U.S. healthcare system. “There’s no step forward in healthcare that doesn’t have pros and cons,” Gruber says. “Absent any consumer cost-sharing, healthcare will be overused; the potential con is that people will forego both more valuable and less valuable care.”
That points to a more nuanced system of deductibles and charges, he says. Often this is termed “value-based healthcare,” in which flexible cost-sharing steers consumers to more effective care. That means lower co-pays for generic drugs, higher charges for unnecessary tests or higher-cost treatments not shown to have better results than lower-cost alternatives.
“The right answer for our system is not to tell patients they can’t have service X,” Gruber says, “but to say service X isn’t cost-effective — if you want it you have to pay for it. That’s controversial and hard, but the lesson from this paper is that we should bring expert analysis into cost sharing, rather than relying on consumers to figure it out.”
By Don McCanne, M.D.
Please excuse the fact that today’s Quote of the Day selections are relatively long, but they are because it is imperative that we clearly understand the fundamental health policy implications that can be extrapolated from this NBER report.
For a few decades much attention has been directed to the concerns about the very high health care costs in the United States in a system that is delivering mediocrity. The debate has centered around whether public policies – the government – or consumer-directed health care – the market – can deliver greater value in health care – higher quality at lower costs.
This study is important because it places the provider of health care – the physician – in the role of the health care consumer – the patient. Nobel laureate Kenneth Arrow showed us long ago that information asymmetry – the fact that the health care provider has information that the health care consumer does not – is a major flaw in the theory that free markets are more effective in improving quality and reducing costs when it comes to health care. The fact is that free markets do not work effectively in health care.
Government regulation and public administration have been shown to be effective in improving value in health care, much to the chagrin of conservative/libertarian ideologues. To counter the perceived “government takeover” of health care there have been efforts to reinforce the effectiveness of markets, especially though consumer-directed health care policies that theoretically place the patient in charge.
One major effort has been to place more information in the hands of the patient to enable informed decisions to be made by the patient in the health care marketplace. This landmark NBER paper is important because it essentially eliminates information asymmetry as a significant factor by observing the behavior of the most informed of patients – physicians themselves – in the health care “marketplace.” They show that these highly informed health care shoppers – physicians – do not do much better than comparable uninformed laypersons. The minimal benefit, if any, is so small that we cannot rely on expanding available information for the patient-consumer to use the market more effectively to improve quality and reduce costs. This study shows that giving every single patient a medical school education would not have much of an impact.
Another effort to increase the effectiveness of health care shopping in the marketplace has been to increase price sensitivity of the patients. Supposedly, when all things are otherwise equal, the patient will go for lower prices. Here we run into real problems.
Health care has become too expensive. Patients need some form of health care coverage to prevent financial hardship. Insurance products have been designed to expose patients to costs through deductibles and other cost sharing. This really doesn’t have much impact on health care prices since the ultimate costs are not determined by the patient-consumer but rather by nontransparent business arrangements between the health care providers and the third party payers such as the private insurers. The patients may not even be aware of the actual prices unless they eventually see them in the insurers’ explanation of benefits.
So since the deductibles and other cost sharing arrangements do not have much influence actual prices, then what impact do they have? Deductibles do cause patients to forgo some recommended care. It is often glibly stated that, since both low-value care and high-value care are forgone because of cost sharing, it is a trade-off, as if that is perfectly acceptable. Well, it isn’t. Forgoing high-value care is always wrong and indicates that the health care financing system is not functioning properly. So what about low-value care? By definition, it is not care that is totally without value, so it is care that should be seriously considered in the management of the patient. It is only no-value care that should never be rendered. More about this in a moment.
The consumer-directed advocates have recommended health savings accounts to increase the shopping power of the patient-consumer. But they are a gimmick that unfairly provides regressive tax benefits to higher-income individuals. More important, families with inadequate disposable income cannot benefit from an empty health savings account. They are not a satisfactory solution to the financial barriers created by high deductibles.
There is considerable talk today about value-based health care. It is often suggested that the patient should pay more for low-value care, but less or nothing at all for high-value care. Obviously this is an attempt to counter the negative impact of high deductibles. So let’s look at this.
A high performance health care system should be designed to enable people to get the care they should have. Thus high-value care should have no financial barriers, thus no deductibles or other coast airing. Care that is of no value should never be offered and thus no payments should ever be made for it. Regarding the more difficult problem of low-value care, which may be entirely appropriate, some would suggest that patients should pay larger amounts out-of-pocket for lower-value care. But rather than the patients making those decisions based on the amount of out-of-pocket spending, the providers are in a better position to decide the value based on cost effectiveness. (Conservatives often say that cost effectiveness should never enter into the decision, but they are wrong. You cannot give the health care delivery system a blank check, where anything goes.) Regardless of high-value, low-value, or no-value, if pricing decisions were left to patients, they would price high-value and low-value care too low and no-value care too high. If pricing decisions were left to providers, everything would be (and already is) priced too high.
The beauty of a well designed single payer financing system is that it uses very effective tools to get pricing right (global budgets, negotiated rates, bulk purchasing, etc.). Today’s lesson is that if physicians are not much better health care shoppers than anyone else then it is time to dump the consumer-driven market ideology and move forward with a proven method of financing health care equitably so that we all receive the care we should have: single payer improved Medicare for all.
Stay informed! Visit www.pnhp.org/qotd to sign up for daily email updates.