People often make irrational decisions, and behavioral economics accounts for that aspect of human nature and finds ways to motivate them to act constructively. The applications for health care are significant, says David Asch, MD, one of the field’s leading experts.
A challenge in designing a rational healthcare system arises from the fact that the most important components in the system—people—often act irrationally.
That’s the view of purveyors of behavioral economics, a field that applies psychological insights to understand and guide human behavior.
The potential of behavioral economics to positively affect health care is proving to be significant, said David Asch, MD, executive director of the Penn Medicine Center for Health Care Innovation.
“Human behavior is on the final common pathway of every single advance in health care,” Asch said. “Even an incredibly effective drug or an incredibly effective vaccine requires a physician to prescribe it and a patient to take it, and those things aren’t quite as simple as they may seem.
“Something that helps us motivate the clinicians and the patients would tremendously advance health care and let us take the scientific advances and deliver them to patients more effectively. What behavioral economics does is it effectively supercharges those approaches.”
Traditionally, Asch noted during a presentation at HFMA’s 2018 Thought Leadership Retreat in Washington, D.C., public health campaigns and healthcare provider initiatives have relied on the notion that given the right information, people will make the right decisions. Examples include nutrition and calorie labeling and informed consent.
Asch takes issue with that concept, noting that information alone provides a shaky foundation. Most people already know that smoking can be harmful, for example, so educational campaigns such as the one being promoted by the Centers for Disease Control and Prevention may not make an impact.
A more advanced step, which also may fall short of the mark, is to use standard economic approaches. The idea is that offering financial incentives will prod people to act the right way. But human nature does not frequently translate to actuary-style decision making.
Behavioral economics goes beyond standard economics, contributing “an understanding of how people actually make decisions, which is often not so rational,” Asch said. “They make their decisions based on emotion and framing and things like that, and that’s not what’s taught in medical school.”
Providing a Nudge
The Penn Center for Healthcare Innovation differs from many provider-based innovation centers, Asch said, in that it goes beyond tech-transfer or IT initiatives to examine how care delivery processes can be improved.
Behavioral economics is central to this effort. The Center established what Asch said is the nation’s first “nudge unit” at an academic health system. The unit’s purpose “is to find gentle but effective ways to change the behavior of either our clinicians or our patients toward more optimal clinical outcomes,” he said.
On the clinician side, a nudge could be used to increase referrals to cardiac rehabilitation for patients who have had a heart attack, which has been established as a best practice but typically requires various steps on the part of the referring physician. Conversely, a nudge might be deployed to dissuade a physician from utilizing unnecessary diagnostic testing.
For patients, nudges can promote adherence to exercise programs or medication regimens, or visits with a physician.
“There are so many elements of health care that are dependent on the patient’s behavior, and of course patients are highly motivated to improve their behavior, but each of those things is hard,” Asch said. “So rather than just tell people, ‘Go do that,’ we should recognize that it’s hard. How do we make it a little bit easier to remember? How do we make it a little bit easier to do? How do we keep people from falling into the traps that humans naturally fall into?”
Designing the Right Incentives
At Penn Medicine, one situation that needed addressing, Asch said, was the tendency of clinicians in the health system to prescribe brand-name drugs when generic versions—vastly cheaper yet similarly effective—were available. “All the clinicians knew that they should be prescribing the generic drugs, and we’d had meetings with clinicians, we did everything possible, yet the level of prescribing of generic drugs was flat for each of these agents,” Asch said.
The information systems team then implemented a small but crucial change in the physician order entry system, making the generic version of a drug the default choice. Physicians had to opt out to access the brand-name variety. The difference in prescribing patterns was immediate and substantial, and indicative of the potential of targeted nudges to effect change.
“It didn’t restrict anybody’s liberty—the physicians could use the brand if they wanted to,” Asch said. “But remember, these were physicians who wanted to use generics, they were just in the habit of using brand-name drugs. So we got them out of their habit, we greased the skids for them to do what they fundamentally wanted to do, which we thought was also better patient care.
“If you just thought that education or financial incentives were the way to go, you would never have thought of that particular approach.”
Providing default options is one example of a nudge that is based in behavioral economics because it recognizes the status quo bias that is hardwired into human beings. Others include:
Lottery-based rewards: State lotteries capitalize on the innate regret aversion of the human mind, Asch noted. Advertising campaigns target people who play the lottery regularly, imploring them not to let this be the week that they don’t play their number—because this just might be the week that their number wins. Such campaigns are effective despite the impossibly low odds of winning the lottery.
Providers can use the same trick on patients in tandem with pill-bottle technology that allows medication adherence to be monitored remotely. A lottery can be held each day or each week, with, say, a $100 prize going to the patient whose number is drawn at random. But patients are ineligible to win if they did not take their medication. The fear of getting a message that their number was drawn and they would have won if they taken their medication is a powerful motivator.
“They don’t have to get the regret message, they just anticipate it,” Asch said.
Straightforward incentive structures. Asch thinks the concept of value-based insurance design (VBID) is flawed because of its complexity. He cited a study in which almost 6,000 patients who’d had a heart attack were randomized into one of two groups. The first owed standard copayments for statins and other preventive medications, while the second had their copayments waived.
Although medication adherence was four to six percentage points higher in the second group, there was no difference in time until first major vascular event or revascularization.
The results indicate what Asch described as the “asymmetric” nature of VBID. Copayment increases can be effective at discouraging people from taking a medication or seeking a form of treatment that is considered low-value. But copayment decreases, as seen through the lens of behavioral economics, may be futile as a value-oriented strategy.
“A copayment decrease is meant to encourage people who aren’t taking a drug to take it,” Asch said. “But if I’m not taking the drug, I never see a copayment decrease. I don’t even notice that. I don’t go the drugstore. The whole problem is that I’m not going there, so the copayment decrease can be invisible to me.
“That helps explain why value-based insurance design is asymmetric. The effects of increasing costs are much more potent than the effects of decreasing cost. We should just pay attention to that as we think about these programs.”
Finding the Sweet Spot
As to whether behavioral economics is better directed towards patients or toward clinicians, the evidence suggests that the answer is: both.
Asch led a study that consisted of four groups, each of which was seeking to achieve LDL cholesterol reduction in patients: Physician incentives ($1,024 per patient who achieved his or her cholesterol target after 12 months), patient incentives ($1,024 for achieving the target), shared incentives ($512 each to the physician and patient for achieving the target), and a control group.
The results showed that patient incentives or physician incentives alone did not produce significantly greater cholesterol reduction than the control group. But shared incentives did boost results.
Asch noted that the effectiveness of pay-for-performance models among physicians had never been examined before that 2015 study. “It’s just always been assumed to work,” he said.
The results should not have been surprising. “It was like we accidentally discovered the doctor-patient relationship,” he said. “It takes two to tango. It takes a physician to prescribe and a patient to take it, and in a lot of strategies to improve health care, we target one or the other. We don’t think about using them together. We’ve all been taught about the clinician-patient relationship and how important it is. Why aren’t we using this effectively?”
A Boost for Value
The push to make health care more value-focused generally hinges on the idea that by altering financial incentives for patients and providers, policymakers can change behavior.
That approach, Asch said, is too reliant on standard economics when behavioral economics offers more-targeted methods.
“From experience and frankly from common sense, just dangling a financial incentive in front of a patient of in front of a hospital or health system creates some motivation for change. But it doesn’t tell them how to do it.
“It’s actually very hard to change your behavior if you’re a patient. ‘I know I shouldn’t eat the potato chips. I didn’t need more motivation. It’s just, how do I avoid them?’ And a hospital that’s used to operating a certain way, it’s like turning a battleship—it’s not so easy to turn on a dime.
“These financial incentives that come from approaches to value-based care are important because you want your incentives aligned, but I don’t think they’re ever going to take us the full way. So we need to think about what else we need to apply to be in concert with this. It has to be a campaign.”
Nick Hut is managing editor of Leadership.
Interviewed for this article: David Asch, MD, MBA, executive director, Penn Medicine for Health Care Innovation.
This article is based in part on a presentation at HFMA’s 2018 Thought Leadership Retreat.