Imagine if health care costs could be dramatically reduced, and outcomes improved without any heavy lifting – no bills would need to be passed, no policies approved, and no major restructuring required. What if we could simply will people to make decisions that resulted in better care and a healthier population?
“Decisions are affected by emotions, bias, social context. The solution is design,” David Asch, MD, MBA, executive director of Penn’s Center for Health Care Innovation, recently said at the Healthcare Financial Management Association’s annual conference.
The idea that better decisions can be made simply by guiding people to them is the principle behind the Penn Medicine’s Nudge Unit, which officially launched last year. Under the direction of Mitesh Patel, MD, MBA, the Nudge Unit was the first of its kind to apply lessons learned from behavioral sciences to design and test small interventions – or “nudges” – to steer providers and patients toward better decisions to improve health care delivery and outcomes.
“When the Nudge Unit started, we had only a handful of projects that were looking how very small adjustments might affect change,” Patel said. “In the past year we’ve completed 11 projects, several of which have impacted care in a substantial way, and looking at the year ahead, we have almost 30 projects that include experts from almost every area of the health system.”
In the health care arena, a nudge is a change in the way choices are presented or information is framed that changes a patient’s or provider’s behavior without restricting choice. For example, one of the projects the team completed earlier this year showed that programming electronic health records to alert physicians when a patient was eligible to receive a flu vaccination increased vaccinations by 37 percent over the previous year.
Translated to the national population, the simple prompt asking the provider to “accept” or “decline” the vaccination could prevent millions of flu cases every year. Another project, which used the same technique (known as “active choice”) saw a 35 percent increase in cancer screening tests. Following the success of the pilot projects, these prompts have since been expanded to all Internal Medicine clinics across Penn Medicine.
Looking to the year ahead, Patel says the Nudge Unit is seeing considerable growth and interest from partners across the health system. Projects both on-going and up-coming are using some of the most basic principles of behavioral economics – gamification, financial incentives, and default settings – to tackle costly health interventions and some of the leading causes of death and other health risks, including statin and opioid use, cardiovascular disease, diabetes, and readmissions.
“Cardiovascular disease is the leading cause of death in the United States, partially because more than 50 percent of adults don’t get enough exercise, and partially because those at risk or diagnosed with the disease aren’t getting the right kind of help,” Patel said. “We’re working on projects that address a swath of health conditions, but several that we’ve completed or are continuing to work on apply what we know about behavioral economics techniques to address this particular issue.”
It’s All Fun and Games
Gamification, where game design elements are added in non-game contexts, is increasingly being used in digital health interventions to promote changes in health behaviors such as physical activity. In one study being conducted with help from the Nudge Unit, the team partnered with the Framingham Heart Study to determine whether using collaboration, accountability, and peer support could improve walking.
Participants in the Be Fit study enrolled with one or two family members and, using a wearable device or smart phone, established a baseline step count. Groups then selected their own step goals, which became their daily walking target.
In addition to receiving daily feedback on whether their goal was achieved, participants in the intervention arm worked with their family members to earn points and progress through levels. Teams started with 70 points each week, and one teammate was selected at random each day. If the teammate met their walking target the day prior, the team kept their points. If by the end of the week, the team had at least 50 points, the team would move up a level. If not, the team moved down a level.
Consistent with results of previous studies using social networks to change health behaviors, the results of the new study showed that by introducing these “gaming” elements, families significantly increased physical activity, adding further evidence to suggest that using gamification and social networks may be a realistic approach to improving healthy lifestyles.
Take the Money and Run
Cardiac rehabilitation is a program supervised by a medical professional that encourages patients to take steps to improve heart health through exercise, reducing stress, and adopting healthy living habits such as eating better and managing other risk factors. Though it’s well known that cardiac rehabilitation reduces risk by 20-30 percent in patients diagnosed with the condition, currently only about 33 percent of patients are enrolled in such programs.
Working with behavioral economics experts in the Nudge Unit, one team led by Neel Chokshi, MD, MBA, medical director of Penn’s Sports Cardiology and Fitness Program, is testing the effectiveness of a home-based, remote-monitoring program using wearable devices and financial incentives to increase physical activity.
“An incentive is not an incentive; a dollar is not a dollar. It all depends on how these programs are designed, how they’re delivered,” said Kevin Volpp, MD, PhD, director of the Penn Center for Health Incentives and Behavioral Economics, of a 2015 study examining how different types of financial incentives – penalties vs. rewards – impacted smoking cessation rates among participants. He work over the years has demonstrated, for example, that the fear of having money taken away is more powerful than the possibility of a reward.
The new 16-week study gave participants daily step goals, a pre-loaded virtual account with $14, and feedback each day on whether they’d reached the goal or not. For the first eight weeks, the daily step goal was increased by 15 percent each week to a maximum goal of 10,000 steps. After the initial ramp-up phase, participants were asked to maintain that daily step goal for another eight weeks. During the 16 weeks, each day during the week that the participant did not meet their daily step goal, $2 was removed from their account.
The results of that study are pending, but based on previous research showing that financial incentives can affect change, the group hopes to find that the intervention increased physical activity, and that a combination of financial incentives and daily feedback in conjunction with using wearable devices can be an effective method of improving behaviors among patients diagnosed with cardiovascular disease.
Winning by Default
As that team focuses on using financial incentives to improve CR participation, another “nudge” team is looking at whether merely changing default referral settings in electronic health records for all CR-eligible patients from opt-in to opt-out would improve participation rates. Once implemented, CR-eligible patients will automatically be referred to CR while in the hospital. By automatically assigning these patients to a referral, the group hopes to see an increase in rates of CR referrals and attendance.
Speaking at the HFMA conference, Asch sited other programs that have had significant success using this technique of requiring people to opt-out instead of opting-in. For example, countries in which people opt into becoming an organ donor have drastically lower rates of survival among those needing transplants, while countries that have an opt-in policy have survival rates often upwards of 98 percent.
Over the past year, Patel has examined whether changes to default settings might make a difference in a number of ways. In one study, his team worked to implement a system-wide EHR prescribing change in which a generic-equivalent medication would be prescribed unless a provider specifically selected an opt-out checkbox. Results of the study showed the overall generic prescribing rate increased from 75.3 percent before the change, to 98.4 percent. Patel says the study demonstrates the effectiveness of EHR default options to change provider behavior and could drastically reduce unnecessary spending.
“It's not always easy to tell physicians how to practice medicine, or to tell patients how to live healthier lives, but the work we’ve done in the past year shows unequivocally that a subtle nudge often can be more effective than a shove,” Patel said. “We’re constantly influenced by our surroundings – our family and friends, environments, emotions. Defaults, incentives, and other ‘nudges’ are already affecting the choices we make. All of these studies are just working to make sure we're nudging in a healthier direction.”