Applying Behavioral Economics to Health Care: a Conversation with Dr. Kevin Volpp


The Pulse spoke with Dr. Kevin Volpp, founding Director of the Center for Health Incentives and Behavioral Economics at the University of Pennsylvania’s Perelman School of Medicine.

Dr. Volpp will be joining one of our conference panels, “The Other 50% of Health: Bending the Health Care Cost Curve via Wellness & Behavioral Economics.” Dr. Volpp discussed some of the key applications of behavioral economics to health care, such as encouraging medication adherence, healthy eating, and smoking cessation.

Pulse: Although basic lifestyle choices surrounding diet, exercise, sleep, and stress play a large role in overall well-being, many of us make poor choices because we have busy lives and careers that disincentivize healthy choices in the short term. What changes to the healthcare system might encourage better lifestyle choices?

KV:  The financing of our health system has historically been all about treatment of disease, but that’s gradually starting to shift. Maybe 3% of our health spending goes into prevention, and the United States ranks first in spending but 26th in life expectancy. Our system is much better at treating preterm infants than preventing preterm deliveries in the first place. The question is, how we might modify benefit design and provider payments?

We spend too much relative to what we’re getting. Our health indices aren’t particularly good, which is related to more challenging social determinants of health in the U.S. than in other countries and challenges in terms of health behavior. There is a steeper income gradient in U.S., and there is a strong link between income and health behaviors. The difference in life expectancy between those in the 1st percentile of income and those in the 99th percentile is about 15 years for men and 10 years for women. There are lots of examples of health behavior challenges, such as gun violence in the U.S. impacting life expectancy statistics.


Pulse: Based on your research, how can we use behavioral economic principles to increase medication adherence? 

KV:  Improving medication adherence is very challenging. There are a series of evidence-based approaches but no magic bullet. For example, we explored changes to the onramp of CVS’ automatic refill program. We used an enhanced active choice model, where we made the decision more salient by highlighting the convenience of automatic refills. Patients would hear “Press 1 if you’d prefer the convenience of automatic refill,” embedded within the refill process.

We also tested synchronized refill programs with Humana. This was another way to make it more convenient by making all refills come on the same date. You as the consumer would get all the medicines you want at home, so the organizational task becomes simpler.

There have been many other studies in this area.  One showed that low-tech reminders don’t work very well in terms of increasing medication adherence. Another demonstrated that lowering copays for high-value medicines can increase adherence by 4-6%. Other studies have explored tying the use of wireless pill bottles that monitor adherence to monetary rewards, but this approach wouldn’t help for chronic conditions, and the results have been mixed.


Pulse: How can we use behavioral economic principles to encourage healthy eating? 

KV:  One example was a study among 8,000 kids at 40 different elementary schools in Utah, where we tried to encourage fruit and vegetable consumption during a time-limited program. Each child who ate a serving of fruits or vegetables during lunch received a 25-cent token to be used at the school store. One group of children was offered the incentive for three weeks, and another group was offered it for five weeks. Relative to an average baseline rate of 39%, providing small incentives doubled the fraction of children eating at least one serving of fruits or vegetables. When we checked in two months after the intervention ended, the consumption rate remained at 21% above baseline for the three-week treatment group and 44% above baseline for the five-week treatment group.


Pulse: Which other areas of wellness might lend themselves to behavioral economic applications?

KV:  Smoking is a big one. Some of our biggest successes have been in smoking cessation, by raising taxes on cigarettes. Every 10% increase in the price of cigarettes leads to 4% in reduction in consumption, and that effect is 2-3x larger in kids, who have less disposable income.

There are implications for benefit design for large employers. Incentives related to smoking cession of about $750 can triple long-term savings in employer contexts. These incentives can take different forms: an employer could charge higher premiums to smokers or lower premiums to non-smokers, or the employer could offer a monetary reward for quitting smoking. All of these measures reduce the amount of cross-subsidies between employees who do and don’t smoke, effectively internalizing the externality.


Pulse: How have you seen social factors, such as socioeconomic factors, education status, etc. affect wellness? What are the implications for facilitating wellness in populations with challenging socioeconomic situations?

KV:  Providers don’t have off-the-shelf solutions, but they’re starting to develop solutions as they move into value-based care. Social determinants affect readmission rates, and at Penn Medicine, they developed a program that’s all around having lay people serve as navigators of system. They basically help people navigate whatever obstacles they’re facing to getting the care they need (transportation, making appointment, etc.). This is in the health system’s own economic interest, because the Medicaid population is at high risk of having bad outcomes, and readmission isn’t lucrative to Penn. Based on some randomized trials, it seems like the program is effective at both improving outcomes and saving money.

Health plans are increasingly thinking about this as well. If you have a fixed reimbursement model, how do you adjust for differences in patient socioeconomic status (as you would be for difference in disease status)? Do you try to adjust provider payment, so that providers are incentivized to care for challenging patients?


Pulse: What other areas of health offer potential opportunities for behavioral economic applications?   

KV:  The use of wearable technology is another area that will be increasingly important in the future but is still in the research realm. Wearable technology could be used to lower future health risk and/or manage chronic diseases, in a compelling combination of technology and behavioral economics.

Rebecca SisselmanComment