Are you aware of any interventions that have been done using the self-determination theory (or other theories) as the basis for deciding which intervention or action is taken to help that patient?

I am not aware of anyone who’s used SDT as the basis to triage people to a specific intervention or approach, although I love the idea and think it has promise. I’m reminded of the Consumer Reports study Seligman did on psychotherapy in the 90s where he essentially concluded that all methods were roughly equally good, because people self-select into the right approach for them (eventually). It would be great if we were able to quickly, efficiently, and accurately direct people to the right interventions based on which basic psychological needs they most want supported.

Two theories I have seen used for triage are the transtheoretical model (sometimes known as stages of change) by Prochaska (https://www.prochange.com/transtheoretical-model-of-behavior-change) and the Patient Activation Measure (PAM) by Insignia Health (https://www.insigniahealth.com/products/pam-survey).

The good news is, there is a large body of research, much of which you can find on selfdeterminationtheory.org, that examines SDT-based interventions versus non, and generally finds support for an SDT-based approach to behavior change. Many of those interventions were developed for the research and are not commercial, but they provide an evidence base for bringing the approach into commercial products.

One commercially available program that is SDT-based and worth looking at is Pivot by Carrot (https://pivot.co/), which focuses on smoking cessation. We also relied on SDT to inform our approach when I worked for Johnson & Johnson Health and Wellness Solutions group; two examples of interventions we built incorporating SDT include the My Health Skills HealthMedia digital coaching programs, and the J&J Official Seven Minute Workout app.

Do you prefer segmentation or personalization? If personalization, how do we convince pharma to pursue this path?

If you picture a continuum that goes from generic to personalized, I see segmentation as somewhere in the middle. Segmentation allows us to learn a little bit about a person and then make inferences about which group they belong to, and offer communications and tools that are generally right for people in that group. Personalization requires us to learn a lot more about a person so we can offer a truly individual set of resources.

Personalization is more effective for changing behavior. There’s a robust body of research that shows that when health coaching is personalized, people are more receptive to it, more likely to remember what they learn, and more likely to follow through. However, segmentation is a lot more practical because we often don’t have the access to people to get enough data to do personalization. So I think you need to think about the business context and balance what it will cost to get personalization-quality data for what potential benefit. A lot of times segmentation is adequate for what pharma would like to do (for example, recruit people to enroll in a program).

As for how to convince pharma to consider personalization more often despite the work it takes to do it right, I think it’s all about building a convincing evidence base. The more researchers and companies are willing to invest to see how personalized digital interventions affect health outcomes and costs over both short and long time periods, the more compelling it will become to consider personalization as an approach.

Can you mention some Digital Health apps which has been tremendously successful in changing patient’s behaviors long term? And is there a common explanation for their success?

One digital health intervention that is performing very well over time is Omada Health. They’ve been able to demonstrate meaningful results across a variety of biometrics. The Care for Your Health program for chronic condition management from J&J has strong cost savings results over an 8+ year period—I shared that information in the webinar as one of the results from behavior change based programs. And for slightly shorter term results, the ImagineCare intervention piloted with Dartmouth Hitchcock and now live in Sweden showed excellent results for people with a variety of chronic conditions. Finally, Scale Down was recently acquired in part due to its strong results in helping people lose weight and keep it off.

I think there are two common explanations for the success of these interventions. First, they are all somewhat personalized, whether it’s in the actual intervention delivered, or in the feedback people receive as they follow the intervention steps. We know that people respond well to personalized approaches, and technology allows us to deliver that in a scalable fashion. Second, these interventions integrate well with the live healthcare professional. In both ImagineCare and Omada, a live nurse or coach is actually an intended part of the intervention, while Care for Your Health and Scale Down both facilitate communication with an HCP. Giving people flexibility about how to receive support can help them stick with a behavior change for a longer period of time.

Of course, each of these interventions does several other things well, but those are the two commonalities I see.

Which patient types are most suitable for digital health like OneTouch Reveal?
I believe that pretty much anyone can be matched with a digital health tool if they see it as providing value for them. That means you need to consider where the patient has gaps in their experience (which is one function of the journey map process—to highlight differences between the ideal and actual patient experience). Some of those gaps will be addressed by OneTouch Reveal, such as having difficulty sorting through the high volume of data collected during daily SMBG and figuring out patterns. Other times, it may require helping the patient to consider the experience gap differently. “You are testing your sugar regularly, and working hard to make healthy changes, but the results aren’t there. Maybe this tool will help us figure out why.” Or, “You’re hitting your targets and doing great, but spending a ton of time on charting and logging. You could maintain your results while reducing your work with this tool.”

That said, I always consider access when deciding what type of intervention is right for someone. In particular, can people afford the technology they need to access the tool? And do they have the infrastructure they need for it to work, for example reliable high-speed internet access? Sometimes the first step is solving for access issues, then providing the intervention.