The decades-long march toward speed and simplicity in life insurance underwriting, powered by advances in technology and data analytics and, more recently, amplified by the challenges of COVID-19, is expanding the footprint of accelerated underwriting.
The primary goal of accelerated underwriting (AU) is to reduce the need for the in-person medical exams and lab tests used in traditional underwriting, thus accelerating the process of policy issuance. By deploying risk-scoring tools and other models made possible by growing access to customer information, insurers can offer alternatives to traditional testing and provide a more client-centric experience. Looking ahead, AU’s sophisticated risk assessment tools offer exciting possibilities to enable faster underwriting within established risk tolerance levels and greater product personalization. To predict where AU may be headed, it helps to understand where it came from.
The Pursuit of Underwriting Efficiency
The goal of speeding the underwriting process is nothing new. Automated underwriting, first developed in the 1990s, brought underwriting into the digital age, shifting much of the life insurance purchasing process from a paper to an electronic format.
Simplified issue (SI) products followed, offering shorter application forms with few medical questions and no medical exam requirement. Its target market comprised primarily younger applicants and those seeking smaller face amounts. SI products carried higher anti-selection risk due to abbreviated underwriting, however, which can result in higher premiums.
AU’s emergence in the mid-2010s signaled a new level of sophistication: the use of third-party data sources to allow some applicants to forego the medical exams and lab tests required in traditional full underwriting (FUW). Underwriters recognized that access to ever-increasing amounts of data could provide a wider window into an applicant’s mortality risk profile. With the development of risk-scoring tools and other models, early AU programs became possible. These programs offered a path for a subset of applicants to be approved through a faster, more seamless process (although not to the same degree as SI). Applicants not demonstrating sufficiently favorable risk characteristics could be routed to traditional full underwriting (FUW).
Throughout these developments, a common theme persisted: the need for increasingly insightful data and appropriate models to provide the best possible mortality protection and risk segmentation.
COVID-19 and its many challenges fast-tracked AU’s ongoing evolution. Early in the pandemic, the number of consumers seeking life insurance ramped up substantially. Medical exams, however, became difficult to secure due to lockdowns and social distancing guidelines. To compensate, insurers turned to alter- native sources of information, such as attending physician statements and digital health records, to assess applicants. The additional evidence enabled policy issuance in cases when the alternative data sources provided a satisfactory level of protective value.
The higher demand for alternatives to traditional evidence in the wake of COVID-19 spurred rapid improvements in the quality and availability of existing and emerging data sources. These developments, in turn, fueled and continue to drive the evolution of AU.
Before AU, faster policy issuance usually meant significant additional mortality risk, resulting in higher policyholder premiums. With more data providing higher protective value now available, carriers are focusing on how more cases can be underwritten in less time at mortality levels close to FUW.
As data sources become richer and hit rates improve, more cases are likely to qualify for AU. Some carriers are incorporating new data sources into existing AU programs to assess each applicant thoroughly already. To the extent the new information can fill a previous void, it might become possible to shift prior rules to produce higher acceleration rates. Further, as the protective value of these newer alternative data sources becomes demonstrably clear, some carriers are expanding the ages and face amounts eligible for their AU programs to reach more of the applicant pool.
Alternative data sources also fuel faster underwriting for cases initially deemed ineligible for AU. In the past, most carriers with AU programs had a binary underwriting process: select applicants eligible for AU and send disqualified cases for FUW with medical exams and life insurance lab panels.
Now, however, some carriers have introduced a middle lane of underwriting between AU and FUW. Applicants not eligible for AU can get a second chance at approval without fluids via an additional review. This process might consist of underwriters re-evaluating the information already collected during the AU process for conditions too nuanced for automatic processing, or ordering additional evidence to fill in data gaps. This allows applicants with more complex medical histories to be considered without needing to provide fluids. For instance, an underwriter may be able to quickly assess and approve a diabetic applicant using the following evidence: a recent normal A1c test result taken in a clinical setting along with a billing code for Type 2 diabetes that does not mention any complication. Whereas, that same diabetic with a billing code for neuropathy may warrant further review.
Carriers are also finding uses outside of AU for alternative underwriting data in place of traditional evidence for defined ages and face amounts. Even if not part of an automated process, alternative data sources can typically be obtained more quickly than scheduling an insurance exam and awaiting results. Mortality impact is highly dependent on the protective value of alternative data being collected. By developing guidelines that ensure sufficient information is available to identify and assess relevant impairments, use of alternative data could potentially produce mortality results close to traditional fully underwritten business.
Maximizing Alternative Data’s Potential
Understanding how information from alternative data sources compares to information typically obtained for traditional FUW can reveal the impact new approaches may have on mortality and other key metrics.
Digital health data in its various forms offers rich sources of alternative underwriting evidence. Prescription drug histories are now regularly used in underwriting, and clinical lab and medical claims data are seeing increased use. While not new to underwriting, attending physician statements (APS) can also be considered an alternative data source when used instead of an insurance exam and labs. Carriers are also actively exploring electronic health records to determine their potential value as an underwriting data source.
A basic goal in using digital health data is often to obtain similar information to that collected through traditional insurance exams and labs. If the same information can be acquired with adequate recency, it can provide similar protective value with corresponding mortality expectations. Nevertheless, while the completeness of these data sources continues to improve, obtaining the exact same information with adequate recency remains a challenge.
For example, information from regular check-ups and wellness visits may lack test results common in traditional FUW. Life insurers will routinely request tests for HIV, cotinine, and liver and kidney function. Unless an applicant has doctor-recognized health issues, this information is unlikely to be available in existing clinical lab results. Surprisingly, even basic height and weight data may not be commonly found in structured alternative data sources. Even if suitable replacement information is available, it is important to consider its recency and completeness.
Although they provide potential cost savings and acquisition efficiencies, alternative data sources may not always speed up the underwriting process. Electronic health records (EHRs), for example, often contain significant amounts of unstructured data, such as handwritten notes or images, which can re- quire substantial time and effort to translate into a usable format. Most insurers using EHRs still rely on human review, and the size of the reports as well as the lack of report uniformity can require painstaking and time-consuming review as underwriters hunt for potentially relevant information.
On the other hand, alternative data sources can contain valuable information beyond traditional evidence that may increase protective value. Physician-directed lab data, for example, could contain results specific to an applicant’s known health issues that might not be captured by insurance labs. Obtaining such lab results over time can also be of value, as it can indicate an applicant’s commitment to improving their health through prescription adherence or lifestyle changes. The potential for finding new information improves even more when adding the data sources for ages and amounts where an APS was not previously required. An applicant’s digital records may also include information on personal or family history undiscoverable through exam or lab results.
Finally, incorporating new evidence sources can support other goals, such as improving customer experience and reducing underwriting expenses. A holistic understanding of the many alternative data sources available can position insurers - and the entire industry - to capitalize on new opportunities while adhering to established risk assessment standards.