Medical and Behavioral Data Together Help Life Insurers Uncover Hidden Risks

The quest for enriching life underwriting risk assessment with innovative types of evidence is part of the insurance industry DNA. Underwriters are accustomed to the introduction of new types of information and figuring out how to optimize their application in the underwriting process. We’ve got prescription history, which can help not only identify potential medical issues but also give a sense of how the applicant is managing chronic conditions. Or Motor Vehicle Records (MVR) that show us an applicant’s driving history—and no surprise that those with a poorer driving record have an increased mortality risk from an automotive fatality.

We just devised a new data combination to explore and conquer. Forward-looking carriers are already using them and achieving great results.

Riskier behavior in one aspect of a person’s life can be considered as a proxy for unsafe behavior in other aspects of it. Interactions between medical and behavioral data can be highly revealing in terms of risk. Take smoking as another example. The choice to smoke (or use other tobacco products) is a behavioral choice that increases mortality and susceptibility to a multitude of impairments, such as lung cancer, emphysema, coronary heart disease and others. This example is still a generally intuitive conclusion about the relationship between the behavior and the resulting medical risks. But we should all ask ourselves: Are there interactions that are not as intuitive? 

The answer is a resounding “YES.”

Hidden risks may be already in your book

At LexisNexis® Risk Solutions, our hypothesis has been that looking at medical and behavioral data separately doesn’t empower carriers to understand the full picture of risk. To prove our hypothesis, we embarked on an internal study linking medical diagnoses, credit, public records and driving history, and assessed it against actual mortality results. The findings of the study are compelling and reveal interactions across the data types in ways that were not visible before:

• Among life insurance applicants with asthma, 60% have average or better-than-average mortality risk¹
• Among life insurance applicants with sleep apnea, 50% have average or better-than-average mortality risk¹

So, think about this: There are hidden risks that you are taking on today. On the other side, there are risks that actually perform better than you expect. Seeing those extra risks can be transformative for the industry—and you don’t have to take my word for it.

In the evolution of the art and science of life underwriting, the industry has a long track record of using new strategies to assess risks. Not that long ago, it was all paper-based and heavily reliant on physicians’ statements. The industry has added data types over time, like paramedical exams, lab testing, MVRs, prescription history and more. Insurers have increased their segmentation by introducing concepts: smoking classifications, build tables, preferred underwriting—all trying to further refine and drive higher precision in risk classification. Most recently, we’ve seen the introduction of predictive models, decision engines and, in some instances, fully automated offers. Through that lens, our industry has immensely evolved in complexity.

Start it up

Now it’s time—and the opportunity—to introduce the interactions of medical and non-medical information to your risk assessment go further than before. I recently conducted a webinar with Matt Stull, senior director and head of data science. The webinar is now available on-demand, and goes deeper into the work we’ve been doing to integrate digital health data into the LexisNexis Risk Solutions ecosystem. We show our study files and details of the results. We also share some ideas on how to get started. I hope you check it out, and we look forward to talking to you soon about how we can support you on your journey to use this new combined view of data.

¹ Source: LexisNexis® Risk Solutions, Internal Insurance Mortality Risk Management Study, 2024