Population Health Analytics: It’s the intervention that matters!

June 9, 2015

The real value of population health management lies not in the data it generates but in the interventions it drives, according to Jay Rajda, MD., medical director of Aetna's Innovation Lab.

Dr. RajdaPopulation health management is an oft-cited priority for health plans and providers in the post-reform era. But its real value lies not in the data it generates but in the interventions it drives, according to Jay Rajda, MD., medical director of Aetna Innovation Labs.

Rajda, in a June 5 presentation at the America’s Health Insurance Plans' Institute 2015 Conference in Nashville, Tennessee, noted that financial predictive models identifying future cost/ utilization are not useful to clinicians. Instead, to make intervention more effective, the industry needs a holistic data view and actionable insights.

“Value is generated from the interventions, not from the data-analytics itself,” Rajda told Managed Healthcare Executive.  “It is important, therefore, to focus on actionable insights.”

The application of advanced analytics to large data sets is becoming more prevalent in healthcare as the volume and variety of data produced expands and the cost of sophisticated analytic solutions drops, noted Rajda.  “Personalized risk prediction, focus on opportunities for intervention, and personalized communications to maximize effectiveness are all required to meet today’s needs for population health management,” he said.

Rajda’s presentation focused on three areas of population health management: 

  • Identifying risk at the population level to guide the development of population health management programs;

  • Identifying risk at the individual level to guide placement/ engagement in programs;

  • Identifying the most effective means of engagement at the individual level.

Read: How healthcare executives can turn data into action

According to Rajda, while most conventional risk stratification models used for care management outreach prioritization focus on future costs and future utilization, “we have found that the prediction of overall risk is essential but not sufficient to target individuals with the greatest opportunity.”

Instead, Aetna has developed a model that focuses on the modifiable component of risk-and the opportunity for intervention-at the individual level.  “This allows us to focus our care management resources on the members that have the greatest opportunity for clinical improvement, while also providing our care managers with the most actionable insights for intervention, to be able to mitigate the modifiable component of risk,” said Rajda.

NEXT: Predicting metabolic syndrome

 

Rajda highlighted Aetna’s work on the prediction of metabolic syndrome as an example of focusing on actionable insights at the population level. “Using the model we developed, we are able to predict the emergence of not just metabolic syndrome, but also the individual clinical components of the diagnosis of metabolic syndrome, both at the population, as well as the individual level,” he noted. “This enabled us to develop targeted intervention programs that are likely to be most successful at reducing the incidence of metabolic syndrome at the population level.”

The most effective interventions are those in which the communication is tailored to the individual, noted Rajda. Aetna has developed a system for using data to segment its population into groups with like behaviors, attitudes and lifestyles, so it can optimize communication channels and message effectiveness.