Digital Health Tools Offer New Opportunities for Personalized Care
Tracking individual behaviors, preferences, motivational profiles to create a “behavioral phenotype” can be used to tailor digital health interventions such as exercise programs. How people use their smartphones — for example their app use, music consumption, and day- vs. nighttime use — can reveal personality characteristics that could, with patients’ permission, be used in the future to identify behavioral phenotypes and help target health interventions.
Digital health has rapidly become a core component of the delivery of health care in the United States. Health systems have implemented new technology platforms to allow for more virtual visits. Patients are increasingly using smartphones to receive health information and communicate with care providers. Health insurers and large employers are providing incentives for using wearables and other digital health devices to better track daily health behaviors such as physical activity, weight, and medication adherence.
While these changes offer significant promise, digital health interventions often fail to create sustained behavior change for many patients. A key reason is that most such programs provide a single offering to thousands or even millions of people who vary in many ways. We know that each patient is different — not only in terms of their medical history but also their personality, motivations, and values — and those differences can be amplified when it comes to making decisions about their health. While a “one-size-fits-all” approach may work for some, it surely will not work for all patients.
Other industries that went digital long before health care have developed approaches to offer more personalized experiences to consumers. For example, Netflix uses your viewing behavior to suggest content that is tailored to your interests. Amazon can tell from your past purchases if you are more likely to buy a children’s book or a science fiction novel. Searching Google for the word “weather” produces different results if you are in Philadelphia than if you are in Los Angeles. These companies track individual attributes including behaviors, preferences, and experiences to create a “behavioral phenotype” that is then used to predict how to tailor information so it provides a better experience.
We believe that behavioral phenotyping could make digital health solutions more personalized and effective. To demonstrate its potential, members of our group conducted a series of studies with overweight and obese adults from 40 U.S. states. Participants were recruited from a large consulting firm for a six-month program that tested the use of gamification to increase physical activity levels. Daily step counts were tracked by wearable devices that transmitted data to a remote-monitoring platform. Participants were randomly assigned to one of four groups and selected a daily step goal. In the control group, they received feedback from the wearable device but no other interventions. The other three groups were also entered into a game that used behavioral nudges and social incentives. The game differed for the three groups in that it was designed to enhance either support, collaboration, or competition.
After six months, all three versions of the game increased physical activity levels by more than the control group. But enhanced competition performed the best and was the only version that led to sustained changes in physical activity during the three-month follow-up period after the game ended.
Based on these initial results, one might assume that since the competition game design worked best, it should be deployed to the entire population. But in a follow-up study, we found that taking an individual’s behavioral and psychological characteristics into account led to dramatically different results.
Prior to starting the program, participants completed surveys on their personality traits, risk preferences, and social support. They also used a wearable device to establish baseline physical activity measures. We used a statistical technique called latent class analysis to identify hidden patterns in the data that linked individuals together through a common behavioral phenotype. Three distinct groups emerged.
A little more than half of people were designated as “extroverted and motivated.” These individuals were more likely than others to be more outgoing and gritty, and to exercise self-efficacy (an ability to carry through on goals). The competitive game worked well for this group, but the enhanced support and collaboration games did not. However, during the follow-up period activity levels in the extroverted/motivated group returned to normal, even among those who received the competition game. Digital health programs targeting this phenotype either need to extend the program duration or combine it with other approaches to sustain behavior change.
A second group was designated as “less active and less social” and included people who had lower social support, were more introverted and less open, and had lower baseline physical activity levels (as revealed by wearable data). All three forms of the game worked equally well for this group. Importantly, activity levels remained higher after the game ended, meaning this group formed a longer-lasting habit — the key goal of the program. Since all three versions of the game worked, participants in gamified activity-boosting programs could be asked to choose the experience that they prefer.
The third group was designated as “at-risk and less motivated.” These people had lower grit and self-efficacy, higher levels of neuroticism, higher health and safety risk-taking, and poorer sleep quality. Unfortunately, none of the versions of the game worked for this group. Though disappointing, knowing this information ahead of time could help to redirect efforts towards other approaches that are better suited for people in this group such as engaging them with health coaches or community health workers.
What can we learn from these new findings? Just as other industries tailor their digital offerings, health care providers could improve digital health efforts by delivering experiences that are more personalized and precisely tailored for each patient. How people use their smartphones — for example their app use, music consumption, and day- vs. nighttime use — can reveal personality characteristics that could, with patients’ permission, be used in the future to identify behavioral phenotypes and help target health interventions. Wearables such as the Apple Watch and Fitbit could also be tapped for information about phenotype.
Changing behavior is always going to be hard but using behavioral phenotyping to personalize the design of digital health solutions could make it a little easier.
Digital Health Tools Offer New Opportunities for Personalized Care
Research & References of Digital Health Tools Offer New Opportunities for Personalized Care|A&C Accounting And Tax Services
Source