Subscribe to Decision Science News by Email (one email per week, easy unsubscribe)
GET $7,500,000 TO INVESTIGATE NUDGES IN MEDICINE
People who have suffered heart attacks can improve their chances by taking aspirin and other medicines. There is a great deal of research on this, yet, we still don’t see 100% of hospitals prescribing these drugs, and the rate of prescription varies from region to region. In general, the US government has noticed “surprisingly modest behavioral response of health care providers and health care systems to information concerning treatments or procedures judged to be superior”.
How can we get the health practitioners to make decisions in line with evidence?
The goods new for you, decision-science-researching reader, is that the National Institutes of Health (NIH) and Agency for Healthcare Research and Quality (AHRQ) are looking to give away $15 million dollars to fund two projects on the effectiveness of nudges on health care practices. The executive summary and parts of the background section make for some interesting reading. Sorry about all the abbrevs.
This NIH Funding Opportunity Announcement (FOA), supported by funds provided to the NIH and AHRQ under the American Recovery & Reinvestment Act of 2009 (“Recovery Act” or “ARRA”), Public Law 111-5, invites applications proposing clinical trials using the principles of behavioral economics to enhance the uptake of the results of comparative effectiveness research (CER) among health care providers in their practice. For this FOA, applicants must propose controlled trials that randomize units (whether individuals or clusters such as practices, hospitals, or larger units) to conditions, resulting in a randomized clinical trial (RCT) or cluster randomized trial (CRT). Research to foster the uptake of CER is seen to be necessary given the surprisingly modest behavioral response of health care providers and health care systems to information concerning treatments or procedures judged to be superior in CER trials. An additional possible benefit is that some behavioral economic interventions to promote the uptake of CER (e.g., those that rely on manipulating a provider’s default options) could be more cost effective than conventional approaches including some pay for performance schemes (P4P). For the purposes of this FOA, the definition of comparative effectiveness research will adhere to that adopted by the Federal Coordinating Council given at http://www.hhs.gov/recovery/programs/cer/cerannualrpt.pdf. Behavioral economics refers to the interdisciplinary efforts involving cognitive and social psychologists, decision scientists, and other social scientists together with economists to model economic decision-making and consequent actions. The approach is inclusive, since at its heart it tries to take into account what is known about how people actually make decisions rather than relying on the assumption that economic agents are fundamentally rational in the sense of expected utility theory (see, e.g., Kahneman and Tversky’s (1979) work on Prospect Theory and Kahneman’s (2003) Nobel lecture). It is hoped that this line of research will lead to significantly greater consideration of CER by health care providers and therefore enhance the quality of the nation’s health.
From the Background section:
Comparative effectiveness research (CER) holds significant promise to improve health care quality and potentially lower costs. It appears that knowledge of which procedures and treatments are comparatively effective may not be sufficient to change critical provider practices and crucial patient behaviors. For example, although the prescription of aspirin, beta blockers, and ACE inhibitors/ARBs after acute myocardial infarction (AMI) has been shown to be extremely effective in clinical trials, strongly endorsed by professional societies such as the American College of Cardiology, and used as a quality indicator by government organizations including the Centers for Medicare and Medicaid Services (CMS), rates of prescription for these drugs in hospitals following AMI show substantial regional and institutional variation and are still below 100% according to the 2008 AHRQ National Healthcare Quality Report (NHRQ). Even when comparatively effective treatments are prescribed, adherence to treatment can be disappointingly low. For example, approximately 50 percent of all AMI patients stop taking prescribed statins within two years of their event as late as the beginning of this decade (Jackevicius et al., 2002). Among asthmatics, only 32% took their preventive asthma medicine daily. Similar adherence problems exist among diabetics, resulting in poor health outcomes. Fewer than 60% of all adults age 40 and over with diagnosed diabetes have their blood sugar, cholesterol, or blood pressure under optimal control. Only 40.1% receive all three recommended services for diabetes, including an HbA1c measurement, a dilated eye examination, and a foot examination. (2008 AHRQ National Healthcare Quality Report)
It is generally presumed that both providers and patients respond to incentives and disincentives to change their behaviors, but to date, efforts to incentivize the uptake of CER have had only modest success. This funding opportunity seeks applications that will investigate whether the principles of behavioral economics could enhance the uptake of the results CER among health care providers and thus improve the health of patient populations. …
In the context of this FOA, behavioral economics refers to the interdisciplinary efforts involving cognitive and social psychologists, decision scientists, and other social scientists together with economists to model economic decision-making and consequent actions. The approach is inclusive, since at its heart it tries to take into account what is known about how people actually make decisions rather than relying on the assumption that economic agents are fundamentally rational in the sense of expected utility. As a field, behavioral economics seeks to understand how human social, cognitive, and emotional factors affect economic decisions. It considers the values assigned to all aspects of a choice, including, but not limited to monetary factors. In addition, behavioral economics acknowledges the important role that a specific context (or frame) may have on decisions, and takes into account people’s apparently irrational preferences (e.g., losses count more than gains, an object that is owned is more valuable than the same object that is not owned). For a recent review of behavioral economics from an economic perspective, Dellavigna (2009) is useful; from a psychological standpoint, Kahneman and Tversky (2000) and Kahneman (2003) provide useful data and historical context. There is growing evidence that such approaches may hold more promise than approaches based on either conventional theories of behavior change or neoclassical economics. The application of approaches from behavioral economics to the healthcare field could be valuable in the development of incentives or disincentives to motivate sustainable changes in provider and patient behavior.
It should be noted that the use of conventional economic incentives to affect provider behavior, including the uptake of CER, has been the subject of considerable research. Perhaps most germane to the topic of this FOA is the literature on “pay for performance”, also known as P4P. The logic of P4P is clear: rather than paying physicians or other health care providers (just) for the specific, billable, services they provide (which naturally incentivizes the ordering of more tests and procedures), they should be paid based on patient outcomes or on their achievement of other objective milestones that should be directly related to patient outcomes. In a classical economic context, it would be a puzzle if physicians and other treatment providers did not align their practice with the procedures or guidelines for practice that are incentivized. Strikingly, however, the evidence for the effectiveness of P4P schemes are often quite modest (reviewed by, e.g., Petersen et al., 2006). Although there are explanations in part for some of these incentive failures (e.g., the principal-agent problem), it seems clear that the incentive system could be improved. Further, many behavioral economists would argue that key improvements could be made not only in the design and delivery of incentives but also in the construction of the decision environment for the health care provider.
To date, implementation of behavioral economic approaches to change decision-making and behavior has focused primarily on economic topics such as behavioral finance (but see, e.g., Volpp et al., 2009 for a recent trial involving smoking cessation). The underlying ideas would seem to have much broader applicability. Behavioral economic interventions are generally of two basic types: one can restructure the choice environment, or manipulate the individual’s perceived incentives. One notable example of the former was Choi, Laibson, and Madrian’s (2004) intervention to increase retirement savings participation. By changing the default action to “contribute” they relied on behavioral inertia to maintain that level of participation. This is an example of altering behavior by manipulating the “choice architecture” that confronts individuals in daily life (see also Thaler and Sunstein, 2008). By structuring choice architectures to subvert individuals’ entrenched biases to stick with the status quo and discount future benefits, a well-developed system of so-called asymmetric paternalism (Loewenstein, Brennan, and Volpp, 2007) could provide interesting opportunities to induce change in provider behavior with respect to selecting a comparatively effective treatment while preserving a clinician’s freedom to choose an alternative treatment when the CER-recommended treatment is counter-indicated. One relevant example of the use of a default option approach that has been successfully implemented (albeit not in the context of CER per se) can be seen in places where statute or policy allows generic equivalents to be substituted for brand name drugs by pharmacists unless a physician specifically notes (or checks off a box denoting that) the prescription is to be “dispensed as written” (DAW). Here, the “transaction cost” of over-riding the default is almost zero, but the effect on generic dispensing rates can be quite large. In particular, generic drug utilization rates varied from 37 percent to 83 percent among Medicare Part D plans (Levinson, 2007), and it would appear that some of this variation is attributable to systemic factors that could be manipulated.
In addition to these more passive, environmental manipulations, behavioral economists have explored the manipulation of incentives to alter behavior. There has been particular interest in the use of deposit contracts, lotteries, and other monetary contingencies to effect health-related behavior change (e.g., Volpp et al. 2008). (Also note that some self-imposed commitment devices can be at least modestly effective at nearly zero external cost, e.g., Ariely and Wertenbroch, 2002). The effect of these devices is generally to allow individuals to overcome their own behavioral inertia, or to make continued compliance with recommended courses of action more attractive. Some of these techniques are similar in spirit to P4P, but the design of the incentives can be very different, and reflects what is known by psychologists about people’s preferences, and how those preferences can be manipulated. Like P4P, however, there can be concerns about the efficiency of providing incentives to reward behavior that would occur in any event, and questions concerning the overall cost effectiveness of monetary incentives. Trials supported by this funding opportunity will, of course, have the option to directly compare rates of uptake in incentivized and non-incentivized conditions, potentially leading to estimates of the marginal cost of the incentive.
More information at:
And please, if you get $7,500,000, please consider yourself nudged to donating 10% to Decision Science News.
Photo credit: http://www.flickr.com/photos/nirbhao/2573135240/