17th annual eyeforpharma Philadelphia (Commercial, Digital and Patients)

Apr 16, 2019 - Apr 17, 2019, Philadelphia

800+ pharma leaders join together to discuss how to revolutionize the patient experience – and, accordingly, their commercial performance.

The Science Behind Solidarity

Patient communities benefit the entire healthcare ecosystem. Behavioral science creates the conditions necessary for them to thrive



Humans are social creatures. We also share common feelings, emotions and responses when put into similar situations. Such psychology can be used to bring people together, and, moreover, help them create positive change. In the healthcare space this is seen in the power of patient communities, such as patient support groups, where people gain huge benefits from having a network who understand their health concerns and can provide advice or just a friendly ear.

From in-person support groups through to online chatrooms, these communities offer a safe space for people to better manage their diseases and conditions, as well as benefiting the entire healthcare eco-system – often acting as a first stop for teams putting together new research projects and, increasingly, in a patient-centric environment, a way for pharma to find people with lived experience to enhance their business practice. But when designing such spaces, how can leaders ensure communities reach their full potential?

By harnessing the new science of behavioral change, says Sean Young. As a scientist, entrepreneur and medical school professor, he sits at the exciting juncture between medicine, psychology and technology. Teaming behavioral science, prediction technology and big data, he explores the science behind human behaviour, and, moreover, how it can be changed, looking to groups like Alcoholics Anonymous and Weight Watchers as testament to the “power of community.”

Young juggles Executive Directorships of the University of California Institute for Prediction Technology and the UCLA Center for Digital Behavior with working as a medical school professor with the UCLA Department of Family Medicine. It’s this work which focuses on changing and predicting health behavior, leaning on mobile technology and big data.

“I study digital behavior and prediction technology, or how and why people use social media, mobile apps, and wearable devices,” says Young. “I help people and businesses apply this knowledge to predict what people will do in the future ­– in areas like health, medicine, politics, and business – and to change how they will do it.

“There are three different types of behavior. What I call automatic, burning and common behavior, A, B and C behavior. It's important to know these three different types of behaviour, because there are different set of tools that we can use for changing each one. That’s a breakthrough within behavior change research, that not all behavior is the same and there are different ways we can change different types of behavior.” Fusing classic behavior change, which traditionally focused on teaching people methods and how to ingrain habits, with these more modern findings, offers new ways to enhance group work.

A large part of Young’s own work has been building something he calls the HOPE Intervention – Harnessing Online Peer Education – which is made up of an online community behavior change community. His research was published in his best-selling book Stick With It, which drills into the science behind human behavior and making lasting change.  

Ingredients for success 
Successful community building can be distilled down into six ‘ingredients’, he says, which each address people’s fundamental psychological needs; the need to trust, to fit in, for self-worth, for a social magnet, to be rewarded and to be empowered. By focusing community development around these buckets, people feel supported, good about themselves and part of something bigger than themselves – all of which feeds into a healthy space in which to make meaningful change.

Peer-to-peer support is central, with the aim of creating a domino effect of cascading change. “We found that people in our community are about two to three times more likely to change their behavior,” says Young. “We leverage peer communities; we teach diffused behavior change to peer role models, on how to change other's behavior, then they'll talk with other participants and, over a 12-week period, we find the people they're talking to end up becoming peer role models themselves and help to change other patients’ behavior, so we get these waves of network effects.”

To get this off the ground, they start with 15 per cent of the community as peer role models for the three months, a time in which a community will normally be up and running. Identifying the right peer role models sets the tone – they need to share the same demographics and health as the rest of the community, be interested in helping others and have the capacity to be respected by their peers.

The trained peer leaders are charged with building trust with the members who have been added to the group, publically posting provided content, as well as reaching out in private messages to encourage participation among those who aren’t being active. Slowly, over the weeks, this technique has been shown to create sustainability, where community members no longer need the peer mentors to guide them, taking up the mantle themselves.

Imbibing patient communities with these ideas should change the trajectory of the community being built, by setting it on the desired path and then, hopefully, it should become self-sufficient. It’s a method that seems to get results – after three months, more than 90 per cent of the people in HOPE communities remain engaged, and after 18 months, it’s typically more than 80 per cent.  

“With our HOPE online behavior change communities, we've really created a science of firsts,” says Young. “We have to understand patient needs, so we interview them, figure out what they need and what type of behavior change we want to accomplish, and then tailor the community based on that information. If we were just to open up a patient community and assume that will change patient behavior, it absolutely won't. It's really important to use the correct science.”

The more specific you can make a platform, the more effective you can be in engaging in changing behavior. “There’s always this trade-off in trying to make something broad enough and scalable enough,” says Young. “Both can be achieved, it's just that it's usually a conversation with the product team to figure out the best balance of the two. One of the first things we do is figure out what type of behavior we’re trying to change and then we tailor the community based on that.”

Data mining
Such control is possible thanks to the plethora of information available from digital tools and devices, while, at the same time, a rapidly increasing number of statistical and computer science approaches for being able to understand said data are being developed.

“With social media data, every day there are approximately 500 million tweets on Twitter, 70 million posts on Instagram and, in the United States, about 700 million snaps sent on Snapchat,” says Young. “That's a lot of data, much more than we've ever had. So how do you make use of that information?”

This wealth of personal information now being posted online, in particular on social media, can be mined to great effect. A study run at UCLA among students, which tracked their sleep and activity, found that Twitter data could be used to predict their stress and their sleep, Young says. There have been similar pilots run around social media discussions about public health, HIV and STDs, crime and cyber bullying.

“We have advanced statistical tools, machine learning, and deep learning using artificial intelligence, so we have the tools to be able to look through billions of social media posts, as well as wearable device data and medical records data,” says Young. “Our group is looking at this specifically from the lens of behavior and asking how we can use data to better understand and predict people's behavior. Because if we can understand and predict something, then we'll know how to intervene and change it.”

The next step, he says, is to take a deep dive into the quality of the data being posted publically online. How much of what people post can you take to be the truth? Secondly, ethically, are people comfortable with scientists and psychologists mining their information to predict their behavior? And thirdly, how accurate are the predictions which are being made based on this data – both now, and looking forward into the future? As researchers continue to grapple with these questions, they can be sure of one thing – there is only going to be more data and more potential avenues to explore in this arena, with exciting prospects for patients.



17th annual eyeforpharma Philadelphia (Commercial, Digital and Patients)

Apr 16, 2019 - Apr 17, 2019, Philadelphia

800+ pharma leaders join together to discuss how to revolutionize the patient experience – and, accordingly, their commercial performance.