A “nudge unit” at the University of Pennsylvania is helping bridge the gap between the study of human behavior and the practice of medicine, devising ways to improve quality of care and patient safety.

The Penn Medicine Nudge Unit is the first behavioral design team embedded within a health system, so the concept is not widely known in healthcare circles.

The “nudge” refers to subtle changes to the way information is presented that can significantly influence how decisions are made or people behave, explains Mitesh Patel, MD, MBA, MS, director of the Penn Medicine Nudge Unit. A nudge unit is a behavioral design team comprised of people with skills in behavioral economics, psychology, and related fields.

That team takes a systematic approach to determining when to use a nudge (a change in a process or structure) to create the desired result, along with exactly how to implement it and roll it out, Patel says. Nudge units were first developed for use in government. Today, governments around the world employ nudge units to create various changes and improvements, Patel says.

The Penn Medicine Nudge Unit launched in 2016 as the first such program in healthcare. Since then it has launched more than 50 projects in multiple departments at Penn Medicine, and it also offers an annual symposium for other healthcare organizations interested in creating nudge units. The Penn Medicine Nudge Unit works closely with the Penn Medicine Center for Health Care Innovation, whose goal is to reimagine how Penn Medicine provides medical care, and the University of Pennsylvania Center for Health Incentives and Behavioral Economics (CHIBE), which is a research group that tests how economics can influence health policy.

The unit is conceptually and physically a part of the Penn health system, with a steering committee that includes leadership from IT, clinical care, and behavioral economics. The nudge unit regularly presents findings to the health system CEO, chief innovation officer, and other top executives, Patel says.

“We’re constantly working with different departments. We have a lot of active projects and a team of about 20 people that includes project managers, research coordinators, and data analysts,” Patel says. “But we also have partners in every clinical specialty. For example, we have an attending in cardiology who leads all our cardiology projects, and we have two oncology fellows who are leading a couple projects in cancer.”

The Penn Medicine Nudge Unit was launched in part to address a particular problem at Penn Medicine, Patel notes. The health system was ranked last in the region for generic prescribing and wanted to encourage clinicians to prescribe the low-cost alternative more often (when appropriate).

“When people type into the electronic health record, they often think in terms of brand names, so they type in Lipitor. Or, if a patient comes in and asks about a drug they need and you search for it, the first one that came up was the brand name, with the generics listed at the bottom,” Patel says. “It nudged you to select the brand name because everyone is in a hurry and typing in the drug orders quickly.”

Patel’s team changed the electronic health record so that generics became the default choice. The clinician was required to opt out of that choice to prescribe the brand name medication. “Whether you wrote the order as Lipitor or the generic atorvastatin, it would go to the pharmacy as the generic, unless you wrote it for Lipitor and also clicked the box that says ‘dispense as written,’” Patel explains. “We first tested this in a small setting, then rolled it out to all the ambulatory clinics across all specialties in all of Penn Medicine. The generic prescribing rate went from 75% to 99% almost overnight.”

That prescribing rate has been maintained for three years, which brought a savings of $32 million for just the top 75 most prescribed medications. The beauty of it all was that the change only took about an hour to implement, Patel says.

“It required very little effort for such a big impact. All it took was getting the stakeholders aligned and realizing that the system was set up to make it harder to do the right thing,” he explains. “If we just changed the default, that would have a significant impact on physician behavior.”

Many Successful Projects

Not all nudges are that simple or quick, but the experience showed Penn Medicine the value of such an approach. The nudge unit has seen many other successes since then, such as changing the referral rate to cardiac rehab from 15% to 85% by creating an opt out pathway. The unit also addressed unnecessary imaging for end-of-life cancer patients, reducing that by half simply by changing the default order sets in the electronic health record.

Other improvements include increasing flu vaccination by 10% and cancer screening by 20%. One project started with a survey of internal medicine and general surgery residents, asking how often they ordered inappropriate tests and why. The nudge unit survey also asked for ways to reduce inappropriate ordering. More than 80% of those surveyed said they ordered unnecessary tests, with almost 50% saying they did so every day.

The reasons they cited included a lack of transparency about the cost of tests and a healthcare culture that did not encourage or reward restraint when ordering tests. Those reasons were used to begin addressing the way faculty support the more prudent ordering of tests when mentoring residents.

Collaboration Avoids Roadblocks

Of course, simply having an idea for a better way to accomplish a task is not as easy as actually implementing it. Lots of quality improvement professionals offer suggestions for these kinds of changes, but adoption is not so easy.

What makes a nudge unit effective? According to Patel, it takes a collaborative approach based on data and research, rather than one person going to another department and telling them to change something.

“The steering committee with leaders from all fields in the health system work with experts in economics and behavioral science. All our ideas are vetted through them first,” Patel says. “Most of our ideas come from frontline clinicians. We get ideas from executive leadership about projects and goals they want us to work on. We also come up with our own ideas.”

In vetting the ideas, the steering committee determines if the issue is addressed within the health system already and whether the change aligns with the goals and needs of the system, Patel says.

If the project is deemed appropriate for the nudge unit, analysts are assigned, and team members work with the right clinical leaders to implement the change.

“It’s really a systematic, structured group approach, rather than thinking about how to do this ad hoc. That’s what most health systems do and what we were doing before our nudge unit,” Patel says. “The steering committee helps prioritize projects and directs us to the right people to work with. We typically will set up a six-month timeline to get all the stakeholders aligned and decided if it’s a good project to continue.”

Stakeholders Must Be on Board

A fundamental principle is that all stakeholders must be aligned, Patel says. For instance, a project in cardiology will not move forward without support from the division chair. But the nudge unit also involves frontline clinicians in developing the intervention, Patel notes.

“We’ll either show them mock-ups at the beginning or we’ll have a prototype they can play around with and give us feedback,” he says. “The most effective nudges are the ones embedded in their workflow. Often, in order to make that happen, we have to see them using it in everyday practice. We get top-down and bottom-up support before we roll anything out.”

Aligning stakeholders can be the biggest challenge and what requires the most time, Patel reports. The nudge unit addresses that challenge by focusing mostly on changes that can be supported with national guidelines and accepted best practices.

“Another challenge is with IT. Sometimes, the electronic health record is not set up to do things the way we want. We may find that it is not possible to set up a default or include a note where we want to,” Patel says.

Full Unit Not Required

Obviously, the Penn State Nudge Unit achieves great improvements, but it is a substantial formal unit within a major health system. Can any of the same tactics be applied on a smaller scale?

“Many people are already tinkering in the electronic health record and the way their communications are designed. Most institutions get started by finding an area of work that is of high importance to the health system and there’s a good opportunity to implement a nudge,” Patel notes. “Some institutions will need help with behavioral economics, and others will need help with implementing the change with IT. Others might face more of a challenge with evaluating the results.”

Penn has developed the Nudge Unit Collaborative to help hospitals learn from the experience of the Penn State Nudge Unit as well as the work of other hospitals. (More details about the collaborative are available online at: http://bit.ly/2Mk3xZX.)

“It’s a great way to see what has worked at other places and what hasn’t worked,” Patel offers. “When I’m thinking about starting a new program or implementing a nudge at my hospital, I can increase my chances of success by seeing what others have done. You also can get feedback from others about the nudge you’re trying to implement.”

Patel expects to see more nudge units in healthcare soon. “I think what has gotten people’s attention is how simple, low-cost interventions can have a huge impact on patient care and really help to align goals,” Patel says. “These nudges often make clinicians’ jobs easier because they’re doing workarounds to get the result they know is right. Nudges are all about making the evidence-based choice the easy choice. I think nudge units are something that will really take off for hospitals and health systems.”