The Mortality Association of CEA in an Insurance Population
In a recent Journal of Insurance Medicine article¹, Drs. Rigatti and Stout used data from over 300,000 individuals who had their blood tested at CRL during the insurance underwriting process to determine how strongly carcinoembryonic antigen (CEA) was associated with mortality risk.
The study encompassed an average of 6.6 years of average follow up, during which 6,084 subjects died based on review of the Social Security Death Master File. Mortality ratios were evaluated in several different ways (actuarial models and Cox models) and demonstrated significantly elevated mortality ratios. This was especially true in the early durations, where the actuarial VBT-based mortality ratios were 20, 17 and 12 in the first 3 durations for those with CEA levels above 10mg/dl compared to those with levels below 5 mg/dl in the same age group. In Cox models controlling for age, sex, smoking, cholesterol, albumin and BMI, CEA remained a strong and significant predictor of mortality.
Although the mortality ratios were very high, it should also be kept in mind that the group of applicants with CEA levels above 10 mg/dl is very small. Only about 0.2% of non-smokers and 0.9% of smokers have values this high. Values above 5 mg/dl are more common, of course, but still only about 2% of the overall population.
Despite the above evidence which clearly shows how strongly predictive of mortality CEA is in an insurance-buying population, the decision to use CEA is made more difficult by its lack of use in the clinical world. Clinically CEA is regarded as a marker of colon cancer, but is not used for screening, only to monitor colon cancers for recurrence. This is likely because CEA has not shown itself to be valuable in a clinical screening context. The fact that it is shown to be very predictive in an insurance context may be surprising. It could be that the overall low mortality risk in the insured population causes the mortality effect of CEA to be magnified — or it could be a signal of anti-selection at play. Nonetheless, CEA is one example of how clinical and insurance medicine can differ significantly at times. Though this difference can at times, cause difficult conversations between insurers, field agents, customers and clinical physicians, it is not an unfamiliar issue. Insurers commonly treat other laboratory tests more or less stringently than clinical physicians (e.g. GGT, albumin) or may run tests that reveal a previously unsuspected and severe diagnosis (Hepatitis B/C, HIV, etc.). The decision to use CEA in an insurance context should thus be based on an individual carrier’s tolerance for these downstream effects, as well as a thorough analysis of the actuarial implications.
About the Authors
Steven J. Rigatti, MD is a consulting medical director with Clinical Reference Laboratory, with 12 years’ experience in the life insurance industry. He is the current chair of the Mortality Committee of the American Academy of Life Insurance Medicine.
Robert L. Stout, PhD is the cofounder of CRL and serves as Chief Scientific Officer and Laboratory Director for CRL’s General Laboratory. He provides scientific direction and research leadership for all of CRL with particular focus on research for CRL’s Insurer Services business. Dr. Stout, who has been with CRL since 1983, holds more than ten U.S. Patents in science, including one for immunoassay method and apparatus development (U.S. Patent #4,414,324) and one for methods of determining chronic Hepatitis C infection (U.S. Patent #10,051,253). He is the author of numerous research and review articles and is a regular presenter at scientific meetings and conferences. Dr. Stout holds a B.S. in Biochemistry from California State University and a Ph.D. in Biological Chemistry from U.C.L.A. School of Medicine.