The U.S. National Institutes of Health recently awarded University of Pennsylvania researchers more than $2 million to explore the effectiveness of three behavioral economics interventions in persuading Arequipa residents to allow pesticide treatment in their homes. 

Alison Buttenheim, PhD, MBA

"Elimination of Chagas transmission in Peru is an attainable goal" said Alison M. Buttenheim, PhD, MBA, assistant professor in Penn's School of Nursing, "but to reach that goal, we have to boost campaign participation in cities likes Arequipa where transmission is still occurring. Buttenheim and Michael Z. Levy, PhD, assistant professor of epidemiology in Penn's Perelman School of Medicine, are principal investigators on the study, which was funded as of Sept. 1. 

It's important to treat all houses on a block with a pesticide to kill triatomine bugs, which carry the parasites that cause Chagas, said Buttenheim. While an acute form of Chagas can develop shortly after infection, the disease often remains latent for decades, "which makes it a very challenging disease to get people excited about preventing," she said.

Chagas, which affects more than 8 million people, invades smooth muscle tissue in areas like the heart and gastrointestinal tract. There is no effective treatment; so the best prevention is to eliminate triatomine infestation in the home to reduce the risk of bites, said Buttenheim. In Arequipa, however, the refusal rate for household pesticide treatment has been rising, she said.

Reluctance to treat in Arequipa comes more from wealthier communities, who cite objections to allowing spraying in their homes, including lack of time and concerns over the insecticide, according to Buttenheim, who called the treatment safe for people. Typically, residents receive a one-day notice that government workers will be in the neighborhood to spray, she said.

Buttenheim and Levy plan to apply novel tools from behavioral economics that have been tested in small trials and apply them for the first time on a large scale in a randomized controlled trial of three interventions in Arequipa beginning in 2015. 

 
One intervention uses group lotteries, in which groups of six contiguous houses are eligible to win a prize such as a food basket if their lottery number is drawn and they participated in the spray campaign. If every house in the winning group has been treated, however, households will receive a bigger prize, such as a gift card for a home repair store, said Buttenheim. 

The idea behind the group lottery is that people not only won't want to miss an opportunity for a prize, they also may not want to be the one who prevents their neighbors from winning the big prize.

The second intervention involves recruiting a block's opinion leaders, such as the woman who runs the daycare center or the corner store, to recruit neighbors to help the community by allowing spraying.

In the third intervention, health workers will show up two to three weeks before spraying, rather than the day before, to seek a household's commitment to schedule a treatment in advance, and then encourage the household to make a plan for the spray day. People tend to think they'll be less busy in two or three weeks than they are now, Buttenheim explained. 

"This is straight out of the behavioral economics playbook," which shows people are more likely to take action if they make a plan, she said. This intervention, if successful, would be very easy for campaigns to incorporate, she noted.

Researchers will compare the outcomes of the three interventions, part of a five-year study covering 1,450 households in one district.
The study may have lessons that researchers can apply in the United States, where Levy works on the bed bug problem and Buttenheim studies childhood vaccination refusal. 

- Dinah Wisenberg Brin

The researchers also found that many people are willing to become donors even if they previously declined, suggesting that authorities should ask repeatedly.

Despite the wide use of active-choice, yes-or-no framing, its effectiveness on donor registration hadn't been empirically tested previously, according to an August 2014 National Bureau of Economic Research working paper by Judd B. Kessler, assistant professor of business economics and public policy at The Wharton School at Penn, and Stanford Prof. Alvin E. Roth, who won the 2012 Nobel Memorial Prize in Economics.

The researchers examined the data and found that California's recent switch from an opt-in to an active-choice organ-donor question likely decreased registration rates. They also ran separate experiments that similarly found no increase in registrations under an active-choice framing, and that people appear more hesitant to support donating the organs of a deceased relative who said no in an active-choice, yes-or-no scenario.

 "We provide empirical evidence on organ donor registration rates using a 2011 policy change in California that switched the organ donor registration question at the (Department of Motor Vehicles) from an opt-in frame to an active choice frame. We find that the switch did not increase registration rates and likely decreased them given the positive secular trend observed in other states," Kessler and Roth wrote in Don't Take 'No' For An Answer: An Experiment With Actual Organ Donor Registrations.

In addition, they reported on their "field in the lab" experiment on whether changing the way the choice is presented affects real organ donor registration decisions. The researchers gave people whom the Massachusetts Department of Transportation previously had asked to register as organ donors the chance to change their donor statuses. Again the researchers found that an active-choice presentation doesn't increase, and may decrease, organ donor registration rates.

"One of the reasons why failure to observe an increase in organ donor registration under the active choice frame is particularly concerning for deceased organ donations is that the frame of the organ donation question might affect the decisions of next-of-kin," Kessler and Roth wrote.

A second experiment involved a survey that indicates family members would be more willing to donate the organs of a deceased relative who had not opted in to a registry than they would be in the case of a relative who had declined to register when given a yes-or-no choice.

The results suggest that active choice "may decrease the transplantation rate through suggesting to next of kin that unregistered donors had actively chosen not to donate," the paper says.

Ask and Ask Again

On the positive side for those hoping to boost donor registration, the researchers found people who previously had been asked were 22 times more likely to add themselves to the registry than to remove themselves - bringing another dimension to the don't-take-no-for-an-answer concept. "This suggests the effectiveness of making a repeated appeal for organ donor registration. In addition, we find that providing more information about organ donation increases registration rates," they wrote. Policymakers shouldn't assume that a "no" is a final answer, they said.

More than 10,000 U.S. patients a year die while awaiting organs, and more than 120,000 sit on waiting lists for life-saving transplants. A large majority of organs transplanted in the United States come from recently deceased donors. Only 48 percent of U.S. adults are registered as organ donors, and the donation rate among eligible donors is less than 100% even with transplants approved by next of kin, the authors said.

In an interview, Kessler said the researchers were surprised the yes-no question didn't increase registration.

"We think states should potentially reconsider this framing that they have of the yes or no," Kessler said. More research is needed on the best way to ask and on the information provided to people when they are asked, he said.

Currently, motorists are asked every few years when they renew their licenses. Perhaps they can be asked more frequently and in different contexts, such as during physical exams, when paying taxes, or on phone bills, Kessler said.

Justin E. Bekelman, M.D. is Assistant Professor in the Departments of Radiation Oncology and Medical Ethics & Health Policy at the Perelman School of Medicine. He is also a CHIBE Affiliate Faculty member and a Fellow at the Leonard David Institute for Health Economics. Dr. Bekelman leads research programs in cancer health services, comparative effectiveness, and treatment delivery and payment reform. He applies methods from epidemiology, clinical trials, and health economics, in combination with approaches from public policy, to assess the effectiveness of cancer treatments and promote high value cancer care.

How did you decide to be a radiation oncologist and a health services researcher?

I took a non-traditional path in my medical career. After I graduated from college, I worked in the Defense Department as an aid to the Undersecretary of Defense for Policy, following which I worked in the private sector for several years as a management consultant. I went to medical school because I knew that I wanted to help people - both by caring for individual patients and by figuring out ways to improve health care on a more macro level. During my rotations I found caring for cancer patients to be particularly gratifying, and I enjoyed the challenge and effectiveness of using radiation therapy to treat tumors.  My research, which focuses on how government policy and private sector incentives impact cancer care and outcomes, combines my interests and experiences in patient care, government policy, and the business world.

What is the relationship between your research and your clinical work?

The questions I face as a practicing physician inform my research. Many of my projects are inspired by conversations with my patients, who typically have prostate, bladder, or testicular cancer. For example, what is the best treatment for older patients with locally advanced prostate cancer? We know from randomized trials that radiation combined with hormone therapy leads to longer survival than hormone therapy alone. But, there is little evidence about the treatment of subgroups of patients not included in the trials, such as elderly patients or those with aggressive cancers discovered by PSA screening. We formed a cross-disciplinary team of clinicians, health economists, and biostatisticians, several of whom are affiliated with CHIBE, to determine whether the findings in the randomized clinical trial literature apply to prevalent patient populations who were not included in the studies. The results are provocative. We hope the research will help practicing physicians consider carefully the treatment options they discuss with their patients.

Can you elaborate on the challenges of implementing clinical research?

Good evidence is only a starting point for improving how cancer is treated. For example, four prospective randomized controlled trials in patients with early-stage breast cancer have shown that a three-week course of radiation is the same in terms of side effects and cancer control as the commonly used, more expensive, and longer duration five-to-seven week treatment schedule. Additionally, the three-week regimen allows patients with breast cancer to return to work or home sooner. Yet, the shorter course of radiation is only used in 30% of eligible patients in the United States. In comparison, it is used in over 70% of patients in Canada and for the majority of patients in the United Kingdom. Why might physicians continue the status quo in the face of reliable evidence that it's not the most patient-centric course of treatment? Some may harbor residual concerns about the toxicity of the shorter treatment; it's also human nature to be averse to change. Being a practicing oncologist, I understand how hard it is to change physician behavior, but also how important it is that physicians treat patients in the most patient-centric fashion.

What does motivate changes in physician behavior?

I think a combination of evidence, training, local norms of care, and incentive structures all affect how physicians treat cancer. Studies dating back over 40 years have shown that physicians respond to financial incentives. The best incentive structure perfectly aligns the interests of physicians and patients around the quantity and quality of medical care delivered. Yet, it's very challenging to design payment policy to achieve that ideal. For example, oncologists and hospitals generate revenue when they prescribe chemotherapy drugs. The current Medicare payment system for chemotherapies favors expensive drugs even when therapeutically equivalent, cheaper alternatives are available. This gives oncologists financial incentives to prescribe higher-priced chemotherapies. In fact, medical spending on cancer treatment was $125 billion in 2010, second only to heart disease. The renewed interest in health care payment reform prompted by the Affordable Care Act gives us an opportunity to examine how payment structures have influenced treatment in oncology and also to design experiments that test the effects of new payment models going forward.

How does payment reform figure into your research program?    

Starting in 2007, a large national health insurer changed their payment structure for chemotherapies by increasing profit margins on therapeutically-equivalent generic drugs in order to compete with the higher margins on branded drugs. If generic chemotherapies were prescribed more, an oncologist's or hospital's overall revenue may have declined, but their profit margin would have increased. The insurer introduced the new fee schedule gradually, so some practices were under the old fee schedule and some the new fee schedule. This set up a natural experiment that we are exploiting to examine the effect of chemotherapy reimbursement changes on cancer treatment patterns, quality, and costs - including the financial burden of out-of-pocket expenses for patients.  

Do you have any advice for those just beginning their research careers?

Be persistent.  The nature of medical research is that it will reveal unexpected roadblocks or findings. For every successful grant application you will have plenty of failures; I certainly have. My patients are a source of great perspective and renewed inspiration in my research. They are the reason I do my work, and every time I fail or hit a roadblock, I remind myself of that.  Be passionate. Research requires long hours and it is unsustainable without genuine, deep interest. Finally, have fun. I work on research projects with a diverse and fun group of collaborators, and a big part of why I love my work is that I get to collaborate with truly great people.

-Christine Weeks

Principal Investigator: Justin Bekelman, MD

 Over the past several years, a large, national commercial insurer implemented reimbursement changes to incentivize the use of generic chemotherapies, presenting a natural experiment which will be leveraged to examine the effects of chemotherapy reimbursement policy changes on treatment patterns, quality and costs. 

Funded by: The American Cancer Society

Incentives, Information, and Impulse: A Field Experiment on Food Choice 

Principal Investigator: Saurabh Bhargava, PhD

This project aims to investigate the economic and psychological determinants of food choice using a large and novel panel dataset of actual food purchases and a series of randomized experiments on real-world consumers through collaboration with the world's largest online food delivery firm. Specifically, the project seeks to understand the role of financial stress on food choice through analysis of archival data, and the role of (i) Incentives, (ii) Information (and the context within which such information is provided), and (iii) Impulsivity on food choice through a series of field experiments.  

Funded by: NIA Penn Roybal Center on Behavioral Economics and Health

The Fresh Start Effect: Temporal Landmarks Motivate Healthy Behaviors 

Principal Investigators: Katherine Milkman, PhD; Jason Riis, PhD; Hengchen Dai

This project will determine what types of life transition and calendar events or temporal landmarks are most likely to motivate healthy behaviors and how to leverage these temporal landmarks to enhance peoples' engagement in healthy activities.  

Funded by: NIA Penn Roybal Center on Behavioral Economics and Health

Choice Architecture and Organ Donation 

Prinicpal Investigators: Judd Kessler, PhD; Alvin Roth, PhD

This project will investigate how choice architecture affects decisions to register as organ donors among a representative sample of residents from the state of Massachusetts. 

Funded by: Medical Ethics and Health Policy

A Randomized Trial of Expanded Choice Sets to Increase Completion of Advance Directives 

Principal Investigator: Scott Halpern, MD, PhD

This study will examine whether expanding choice sets can increase the completion of and choices within advance directives among patients with end-stage renal disease (ESRD) presenting at DaVita Dialysis Centers in Philadelphia. 

Funded by: Medical Ethics and Health Policy, CHIBE

The Impact of Nonlinear Pricing on Portion Size of Unhealthy Food Purchases 

Principal Investigators: Julie Downs, PhD; George Loewenstein, PhD

This study will test the differential impact of nonlinear pricing on (over) consumption of healthy versus non-healthy food items. 

Funded by: NIA Penn Roybal Center on Behavioral Economics and Health