January is a time for two things: resolutions and predictions. We’ll leave the former to you and your will power. But here, we make three bold predictions for the life insurance industry in 2024:
If you haven’t been following the development of the Trusted Exchange Framework and Common Agreement (TEFCA) network, you should. The TEFCA network will enable the sharing of patient medical data, facilitated by Qualified Health Information Network (QHIN) connectivity brokers. While this exchange framework was primarily designed to support patient care, it will enable the life insurance industry to access more electronic health records (EHR) data at lower prices.
While the TEFCA coverage level is already impressive, it could reach nearly 100% in the coming year, which will go a long way in overcoming one of the biggest obstacles to automated, accelerated and fluidless underwriting: access to the data.
We may be overly optimistic about the pace of adoption, but the TEFCA network has the potential to change the power dynamics of data acquisition in life insurance.
When we started, the industry was skeptical of predictive analytics. Advanced machine learning–based predictive models appeared to be a “black box,” and underwriters were reluctant to subordinate their expertise for a score that they didn’t fully understand.
Despite this skepticism, life insurers have expanded their use of predictive analytics. Algorithms are involved in virtually every aspect of the customer journey, including identifying leads, assessing risk, processing policies, acting as customer service representatives and handling claims.
As data becomes more readily available (see prediction #1 above), data scientists are applying their craft in new ways. Early efforts focused on making efficient “all-cause” mortality assessments on relatively simple cases. Now, they’re finding opportunities to optimize routing decisions—or even to replace inefficient rule sets with more sophisticated machine learning–based algorithms. For example, they examine medical conditions that have traditionally required an underwriter’s assessment, then use machine learning techniques to mirror that human judgment, enabling more cases to stay in accelerated underwriting without sacrificing mortality protection.
Impairment-specific predictive models use data to quickly assess the presence and severity of issues such as diabetes, heart trouble or psychiatric conditions. We’ve seen the results, and they are astounding.
As underwriting becomes more automated, it will also be more flexible, adaptable and responsive to changes. Ordering evidence will be more tailored to the individual. The process may be exponentially more complex, but it will also be unique and streamlined for each applicant and still improve efficiency and speed in the underwriting process.
The downside of this approach is that the process will be less predictable—and producers and customers like to know what to expect. Fortunately, based on our first prediction about the TEFCA network, life underwriting will rely much more heavily on EHRs, because they’ll be much more accessible. As a result, most applicants will have a similar underwriting experience, where delays due to a paramedical exam or an attending physician statement will become increasingly rare.