Traditional mortality risk measures - such as cholesterol, blood pressure, BMI, tobacco use and family history - often misclassify applicants because they only provide part of an individual’s health profile and fail to spot important individualised measures such as resting heart rate, heart recovery rate, sleep and activity versus inactivity levels.
Built on 12 years of health data spanning over six million life years, Klarity has developed a new risk scoring tool that produces individual-level mortality scores to predict and classify risks. The model incorporates third-party data obtained from smartwatches and other wearable devices that track physical activity, heart rate and sleep patterns.
WTW’s collaboration with Klarity included testing the efficacy of the model’s mortality score predictions on data from the U.S., leveraging data from the National Health and Nutrition Examination Survey (NHANES). WTW’s analysis found that Klarity’s model can more clearly identify individual mortality risk profiles, enabling improved risk segmentation and insurers to align pricing more accurately than traditional underwriting metrics alone.
For example, some residual non-smokers actually had risk profiles similar to risks classified as preferred and, in some cases, best preferred based on traditional criteria alone. This means these individuals could qualify for better rates. Use of the model also helps to flag more extreme outliers within each class. For example, the model flagged a smaller percentage of applicants who may be considered preferred risks under traditional underwriting metrics but showed hidden risk factors. This suggests they may not actually exhibit mortality aligned with preferred risk and, in some cases, even residual standard risk pricing.
Will Cooper, Founder and CEO of Klarity, said: “By integrating AI-driven insights with diverse health and behavioral data, we’ve built a model that not only enhances underwriting accuracy but also strengthens customer engagement and loyalty. Our collaboration with WTW not only validates the model’s performance in North America - it also reveals how many applicants are under- or over-classified by traditional methods. This opens the door to more accurate, inclusive and dynamic underwriting.”
Mary Bahna-Nolan, Senior Director, Insurance Consulting & Technology, WTW, said: “The life insurance industry has a unique opportunity to harness the power of data to deliver more personalised outcomes that reflect real-world health habits. Klarity’s model is a prime example of how predictive analytics - coupled with data representative of an individual’s health indicators and habits such as movement or activity, heart rate and sleep - can redefine risk assessment and improve mortality prediction. Eventually, this will open the door to more personalised pricing and rethinking customer experience and engagement.”
With 62 million U.S. consumers using a fitness tracker in 2024, projected to increase to over 92 million by 20291, smartwatches and other wearable technologies have entered the mainstream. Analysing data from wearables provides real-time insights that can help predict an individual’s lifestyle and health risks, improving the precision of risk assessments and pricing of life insurance policies.
|
