Last month, I was honored to speak on a webinar with Dr. Susan Weaver, the former Chief Medical Officer of Blue Cross and Blue Shield of North Carolina, Jason Cooper, the Vice President and Chief Analytics Officer of Horizon Blue Cross Blue Shield of New Jersey, and Dr. Jan Berger, CEO of Health Intelligence Partners. We had a productive discussion on how clinical leaders and analytics teams at health plans can apply machine learning to solve some of our biggest challenges in healthcare.Susan, Jason, Jan and I highlighted three areas where machine learning is having an impact in our industry. These include places where we can better identify patients at risk and increase intervention effectiveness – both of which help to improve clinical outcomes and lower healthcare costs.
Our discussion focused on new member risk, metabolic syndrome and end of life care as three examples. Below is a summary of the key learnings for these three topics as discussed in our webinar.
- New Member Risk: As new individuals sign on for health benefits, there’s a segment of the population that has not been in our healthcare system for much of their adult lives. Employers, health plans and providers will likely not have comprehensive information associated with these individuals relative to their health status, any chronic conditions they may have and other relevant information. As such, targeting early interventions can be challenging.
- Metabolic Syndrome (MetS): A very important characteristic of MetS is understanding which individuals are at risk and how individuals will progress as it relates to the condition. Machine learning identifies which individuals are at risk in a given population and the population health benefits of improving specific MetS factors. Additionally, machine learning methods predict ROI and efficacy of MetS interventions for specific MetS factors. Finally, machine learning provides continuous feedback on adoption and engagement in interventions and has demonstrated proven savings within the first 12 months for individuals. An important peer-review study on this topic recently appeared in the American Journal of Managed Care.
- End of Life: Machine learning also plays an important role in end of life care – helping individuals and their families. With this information, individuals and caregivers can have a more informed conversation about end of life care. Machine learning also helps to change socio-behavioral aspects in the role of the clinician. It can help create certainty on timing, and thus prevent providers from over-treatment of their patients during their final days. Machine learning gives guidance on the right timing for when clinicians should recommend a transition to end of life care.
There are many more examples of important healthcare challenges at health plans where you’ll find machine learning making an impact today. These include medication adherence, preterm birth, and comparative effectiveness, to name a few. We’ll be discussing these additional topics and associated machine learning insights in future blog posts.
Colin Hill is chairman and CEO of GNS Healthcare. He can be reached at Colin.Hill@gnshealthcare.com.