Late in 1996, when scientists developed the first reliable genetic test for mutations in BRCA1 and BRCA2, the two genes associated with breast and ovarian cancer, few observers could have predicted the controversies this medical breakthrough would trigger. One particularly heated debate centers around the issue of genetic testing and an individual’s right to privacy. Already, concerns about insurance and job discrimination against consumers whose test results could become public have led 34 states to prohibit life and health insurers from using genetic information in their underwriting decisions. On February 8, 2000, U.S. President Clinton signed an executive order prohibiting federal agencies from collecting genetic information from their employees or using such information to make hiring, promotion, or placement decisions. These laws in turn have prompted insurers to warn about the dangers of ‘adverse selection’ — the process by which consumers make insurance decisions based on risk characteristics (such as a hereditary predilection for cancer, heart disease, or other major illness) that are known to them but not revealed to the insurer. Access to this knowledge allows these consumers to obtain insurance at cheaper rates than they would otherwise qualify for. For example, individuals who learn through genetic testing that they are at much greater risk for cancer may buy more than the typical amount of life insurance, which to them looks inexpensive since it is priced at rates set for average risks. Individuals whose testing shows they are not at risk may be more likely to purchase less life insurance. These two forces, the argument goes, combine to increase the aggregate mortality of the insurance purchasers. But insurers who do not have access to these test results are unable to identify which individuals are at higher risk. They then have to increase premiums for everyone, driving those at lower risk out of the pool. Eventually such behavior leads to a spiral of increasing prices, decreasing numbers of policies and potential insolvency for insurers. The debate over access to genetic screening information will only become more intense in the next few years as progress in the Human Genome Project continues and as the ability to genetically test for major illnesses becomes more prevalent. While detailed genetic test data are currently available only for breast and ovarian cancers, advances in pinpointing hereditary bases for many other diseases, such as colon cancer and Alzheimer’s Disease, have recently made headlines. In the midst of this debate, Wharton professor
The model developed by the authors starts with the available information on breast and ovarian cancer. It is known, for example, that about one in nine women in the U.S. will develop breast cancer over her lifetime, and that one in forty will die from the disease. About 1.79% of women will develop ovarian cancer and more than 60% of them will die from it. While the vast majority of these cancers are the result of diet, lifestyle, environmental exposures and other factors, some cancers are inherited. Studies show that a twofold to threefold increase in the risk of breast cancer development has been associated with a family history of breast cancer in a mother or sister. For ovarian cancer, the risk is multiplied by 5.4.
A small percentage of women (estimated in a range of one woman out of 833 to 2.3 percent in some ethnic groups) have a mutation in one or two dominant breast cancer susceptibility genes, called BRCA1 and BRCA2, the researchers note. Women with a BRCA gene mutation are at extreme risk of developing breast or ovarian cancer.
The 1996 development of a genetic test for detecting these mutations raised a red flag for the insurance industry. Insurers want a requirement that "all test results provided to individuals also be made available to insurers," says Lemaire. "This request for a ‘level playing field’ contrasts with opposite efforts by consumer groups to increase the privacy protection of genetic information."
As noted above, a number of states already have introduced laws prohibiting insurers from using genetic information. Even more restrictive — and potentially devastating to the insurance industry, Lemaire says — is a bill introduced in March 1999 by Maine senator Olympia Snowe. The legislation, backed by Clinton and included in the GOP Patients’ Bill of Rights, would block insurers from denying coverage or setting premiums based on genetic information or family history. "Should that bill pass, my prediction is that you are talking about major disruptions in the insurance market," Lemaire says.
Lemaire and his colleagues – Temple University professor and Wharton PhD candidate Krupa Subramanian, Wharton professors John C. Hershey and Mark V. Pauly, and Katrina Armstrong and David A. Asch from Penn’s School of Medicine – used data from studies done at the Hospital of the University of Pennsylvania, one of the few medical centers in the U.S. where women can get tested for the gene mutation. Their goal was to quantify the impact of family history and BRCA1 and BRCA2 mutations on mortality and on the costs of term insurance. These analyses included data from previous large epidemiologic studies that allowed Lemaire and his colleagues to estimate, for instance, the probability that a woman whose mother had breast cancer at age 56 will develop the disease by age 40. The analyses also allow the researchers to recommend the insurance surcharge that would make prices actuarially fair for women at higher risk because of their family history.
Overall, the conclusions that Lemaire and his colleagues reach should ease the concerns of insurers. The researchers find that the average adverse selection cost in an insurance portfolio is expected to be lower than 10%. Furthermore, this cost is "likely to be compensated by the overall long-term trend of decrease in mortality rates that stands around 0.5 percent per year nowadays. Therefore we believe that adverse selection is a problem that is controllable by insurers."
These conclusions, however, are dependent on insurers applying "very tight underwriting standards," the researchers note. For insurers to control this problem, consumers would need to provide a "detailed family history of all their first-degree relatives, with ages at onset of any cancer. Applicants’ statements need to be carefully checked by underwriters. If companies fail to correctly identify the family history of the applicant, then adverse selection costs could become unbearable." Similar consequences could result if companies are prevented by law from requesting the applicants’ family history.
"We are certainly pushing insurers to investigate much more seriously an applicant’s background," notes Lemaire. "If an insurer cannot use genetic test results then it should be much more serious about investigating the next best thing, which is family history. Especially important for applicants with a history of cancer in their family is the age of onset. If your mother had breast cancer in her 60s, the good news is there is very little for you or the insurance company to worry about. But if your mother had cancer at age 25, then that is a cause for concern, for you as well as the underwriter."
A similar set of issues was raised a decade ago in the area of life insurance coverage for people with AIDS. "AIDS did not bankrupt the insurance industry but there were fights in every single state" over what information could be required from consumers, Lemaire says. "Eventually insurance companies won the right to ask individuals if they have ‘ever tested positive for antibodies to HIV.’ Meanwhile, the situation continues to evolve, especially since survival rates for people with HIV are constantly improving."
Lemaire and his colleagues make several other recommendations in their study. They suggest, for example, that in cases where individuals are buying large term life insurance policies (those exceeding several hundred thousand dollars), insurers "be allowed to use genetic test results for the underwriting of such policies in return for a ban on the use of test results for policies with a reasonable amount."
Lemaire points out that the recommendations of the research are valid only for term insurance, the most widely sold policy in the U.S. At the end of 1998, more than 134 million employer-provided term life insurance policies were in force, with a combined face amount just short of $5 trillion. Approximately 25 million individually-bought term life policies were in force, with a combined face amount nearing $3 trillion. Similar actuarial models need to be developed to analyze the possible adverse selection effects of genetic testing on health care insurance and nursing home policies.
The main results of the research have already been published in the December 1999 issue of the Journal of Risk and Insurance. A technical actuarial note will appear in the April 2000 issue of the North American Actuarial Journal.
