Improving Health Outcomes With Personalized Apps

Source : Battelle

Advances in sensors, smartphones, communications, and analytics give developers new tools to build creative and powerful applications that promise to improve health outcomes and reduce healthcare costs. According to the CDC, more than 75 percent of healthcare costs are due to chronic conditions so applications that help manage blood-pressure, diabetes or weight loss can make a significant impact. Technology has an important role to play, but to capture the most value for the patient and providers, the applications need to address three key elements: scalability, relevance, and stickiness.

Scalability:

Applications need to scale across tens of thousands of patients but a singular approach won’t serve the needs of the individual. Segmentation can help and choosing the right parameters are important, such as people’s priorities, attitudes and values. Effective segmentation informs us about attitudes towards healthcare risks, health status, and behavior, among others, and will help us predict how people will behave when making a decision that could affect compliance (or some other health behavior). Segmentation helps us “meet the patient where they are” in terms of attitudes, communication, priorities, healthcare consumption, needs, etc., and gives us insight into human behavior and context.

Relevance:

“Every patient is the only patient” is the goal, tailoring the app-patient interaction specific to the needs of the patient (communication style, health literacy, cognitive ability, behaviors, context aware, physical ability, external factors, etc.). Challenging to do, yes, and an approach using analytics to build user models that are based on machine learning models can improve relevance over time based on repeat successes. For instance, suggesting to an 80-year-old woman with arthritic knees that she walk 10,000 steps will likely be ignored, but suggesting aquatic therapy may be more appropriate (and the app will tell her where the nearest pool is). Over time, the system will build a library of “success triggers” that are matched to individual signals to enhance relevance.

Stickiness:

Long-term engagement and completion is difficult because behavior change is hard. However, with the right data collection and insights extracted by analytics, the app can respond with personalized messaging that “nudge” patients in the right direction. Every day, we make hundreds of “micro decisions” that add up to certain health related behaviors (should I take the stairs or the elevator? should I eat a hamburger or a salad?) An app that delivers recommendations that are relevant and personalized will help patients master their behaviors over time