Healthcare Analytics and Big Data: Essential Ingredients to Improving Population Health
Consider Kyle and George, two hypothetical members in their early 50’s who appear similar clinically – they are both moderately overweight and have indicators of pre-diabetes. If you could only afford to outreach to one of them with a message about getting healthier, who would you choose? Maybe you like the name Kyle better than the name George, but basically it would be a random choice without more information. And what would you say?
There are many definitions of Big Data; to me it is data captured from all angles that can be combined to generate insights, predict future health outcomes, and ultimately improve outcomes through custom interventions.
Capturing data from all angles includes widening, deepening and lengthening the data:
Widening the data by combining clinical data with demographic and consumer data. Is the member in a rural or urban area? Who also lives in the member’s household? Is the member on a fixed income? Is the member more likely to buy a hamburger or a lean piece of fish?
Deepening the data by increasing the granularity of what is captured, such as genome testing or data captured from wearable devices. Does the member record data on a fitness tracker? What types of activities does the member do and how often does the member do them?
Lengthening the data by incorporating trends and trigger events. Did the member just get married, move, have a child, or get divorced? Is the member’s weight moving up or down? Did the member recently visit a doctor, have surgery, or visit the ER?
Weaving this data together provides insight into the member as a person, not just a patient. And it is the person who will take the path of least resistance. The more interventions align with each member’s specific path of least resistance, the greater chance the member will change.
Big data is an essential ingredient to total population health because it enables custom interventions:
Each intervention is personalized for the member with the right message, delivery channel, and timing.
Outreach is focused on those members most likely to benefit from the intervention as well most likely to be receptive to the intervention.
Back to Kyle and George… From collected consumer data, you discover more about each of them:
Kyle is a busy, urban tech-savvy single professional who is competitive and enjoys watching football on the weekends. Now you know that Kyle is motivated by competition and likes technology. From the data, we know that others like him have lost weight and improved their health through the use of an online platform that included incentives, competition, and rewards.
Kyle’s personalized intervention-
Message: “Show your friends that you’ve still got it in just a few clicks a day.”
Delivery channel: Electronic
Timing: Contact in the evening
George is a rural family man who enjoys hanging out in the local bar with his childhood buddies. George is more traditional and would likely be motivated to set an example for his family and show off for his friends. From the data, we know that others like him did best when they felt accountable to a person they did not want to let down.
George’s personalized intervention-
Message: “Show your kids (and friends) that dads don’t require beer bellies”
Delivery channel: Health coach
Timing: Contact in early evening
In this scenario, you fortuitously find by reaching out to Kyle electronically, you are able to afford outreach for both Kyle and George!
In future posts, we will continue to follow Kyle and George—learn from their data and behavior, and discover what interventions will get them on the right path towards a healthier future. We’ll highlight other applications of big data, such as the use of predictive models to supplement when clinical data is not available and other applications of big data in the healthcare industry.
Watch this video to learn how Health Dialog’s proven identification and engagement methods guide individuals on the path to improved health and wellness: