I wrote this paper to clarify what I said in numerous speeches at the time. As is often the case when I open my mouth and present a topic to an audience, be it large or small, I get asked to put my opinions or views in writing. Sometimes I forget what I said in the speech and thus the written word comes across as different, Perhaps I should write what I am going to say first but that would take the element of surprise (for both the audience and myself) out of the equation making it all boring.
1990 was not any more notable a year than any other but it was the year I entered the giant company environment where I was to last 4 years and 5 months trying to conform. I was far from a conformist (read politically savvy aspirant) and despite numerous attempts by many senior executives (some who quit while they told me it was good if I stayed!) and psychologists I never did adapt. I did learn though about privacy and the need for it in a company,
Since 1990 the insurance world has built far more stringent privacy guidelines after taking the initiative or being legislated and regulated into change. Information continues to be protected and it is not taken for granted. After 14 years since the article I cannot recall any lawsuits over a breach of privacy in regards to the HIV test result. If there have been one or more actions I am unaware and can only assume that lack of public ridicule and press means we did and are doing a good job.
The HIV tests and the masks around it have shrunk in number but the testing amount remains at about $200,000 and the sentinel effect is credited with being the main reason for keeping the routine test. There is a chart showing the costs of AIDS related deaths by type of insurance from the late 1980’s and it would be interesting to see a current chart. Only then could we see what impact all the effort produced.
Ross A. Morton
Written during period with
Manufactures Life Insurance Company
March 28, 1990
In the early 1970’s when I was a mere novice to the world of life reinsurance, I was in awe at the purported depth of expertise and business acumen of my superiors. That quickly became a shattered dream when a humungous claim presented itself to several of the more significant reinsurance companies. One day prior to the anniversary of the policy the insured was found murdered in his home in Oklahoma. The spectre of a premature death claim for $15,000,000 (U.S.) was too much of a shock for the executives of the reinsurers.
A hastily called meeting at the Chicago airport resulted in a hastily concocted statement based on incomplete data about the claim. The message implied to the lay reader that the likely cause of death was suicide and thus under the contract no payment was forthcoming. End of story.
The problem was that someone forgot to await all the information and a subsequent note indicated there were two bullets through the man’s head. This case was eventually settled for one half the amount and ahs become a classic case of overreacting in too urgent a fashion. It also showed that reinsurance executives are human and thus prone to at least one error per decade. I hope I do not commit the ‘90’s only error.
Today, as a human being, I want to convey the message that over the past five years the insurance industry has struggled to find the most acceptable method of risk classification, especially for HIV positives. Beyond the actuarial numbers and the increased payment of claims, (Chart 1), I have witnessed the frustrations of applicant, agent, insurer and reinsurers in trying to cope with the issue of testing and confidentiality. The solutions have emerged as the wisdom of the industry discovered fair and equitable alternatives to just declining to insure certain “profile” individuals. “Wisdom is the ability to discover alternatives; there are many ways to reach solutions.” (From a fortune cookie by Far Eastern Cookie Co. Ltd.)
The past decades have been full of many instances of reducing the price of life insurances for the majority of purchasers. Pure life insurance in the form of term coverage has fallen to nearly 30% of the 1970 price, especially for those in their thirties who are non-smokers. The AIDS issue is almost the only example of a medical impairment that reverses the current trend. In the early days of applicants applying for insurance with a history of auto immune disorders the risks were quite often accepted due to the oversight and ignorance of the lay and medical underwriters. As evidence of rather early and thus unexpected claims emerged the industry found itself in a corner surrounded by the realization of mounting claims that surpassed the expected. How to correctly perform the needed risk classification process became everyone’s quest in all parts of the world that had a life insurance industry.
Insurers are able to provide protection because they can predict, with reasonable accuracy, how many individuals in large, fairly homogeneous group die from year to year. On the basis of this predictability, they assign to a given group those people with similar characteristics….mainly age, sex and health. The information provided by the applicant enables the insurance company to decide the level of risk that person represents and thus the group to which he or she belongs.
Like severe coronary artery disease or recurrent metastatic cancers, AIDS and HIV positive individuals presented the industry with a risk that is could not adequately classify and price for at this time. How to fairly perform the risk classification process, with what tools and at what cost became evolving challenge.
In Britain the industry adopted a series of new questions to be answered by the applicant as well as the use of HIV test. An excerpt from the typical application for insurance is as follows:
“1a) Do you belong to any of the following groups:
i) Homosexual men
ii) Bisexual men
iii) Intravenous drug users
iv) Hemophiliacs (although in the UK it was spelt with an “ae”)”
The representative questions were targeted at those groups of individuals who were deemed to be most at risk. A positive answer to any of the questions was considered as enough reason to refuse to provide a contract of insurance. Failure to answer the question ended any change of insurance coverage and failure to disclose the truth invalidated the policy. The questionnaire was required for all policies where the applicant was a “single/divorced/separated male” but only mid size policies on “married/widowed males” and for extremely large amounts on “females”. Coincident with the questions, HIV testing was mandatory in the aforementioned categories for similar varying amounts.
The following is the form of questioning adopted by the Australian Life insurance industry again in tandem with HIV testing similar to the British example. Should a proposed insured fail to answer yes to all of these type questions the insurance company proceeds to question deeper into the individuals lifestyle or just refuses to proceed with the file. As you can read, the question is extensive and is in addition to questions one would judge are similar in context to those included in the North American style of medical history questions. However, this question goes well beyond what is deemed reasonable in North America.
“6 Since 1980, I have not:
i) Worked as, or engaged in sexual activity with a prostitute
ii) Engaged in male-to-male anal sexual intercourse.
Iii) Injected myself, nor been injected, with a drug which was prescribed for myself by a registered medical practitioner.”
All those involved in underwriting in our market have helped construct a fair method of evaluating the risk presented to the industry. Our standard questions have been tested and refined but are all very similar in context and intend. The “AIDS” questions are included in all applications for all ages, sexes, marital status, etc. The questions reflect a universal need to screen all applicants as we would for other medical impairments, hazardous avocation and occupations. In addition the majority of Canadian life insurance companies now use the HIV test in all applicants for an amount of $100,000 or more.
On reflection, it is only five years ago or less that we did explore a more selective method of testing. Trying to categorize people into classes as the British and Australians do was found to be unacceptable to the industry, the agents and the applicants. As the accuracy of the insurance laboratories improved, their prices fell and the methods of collection expanded, thus the universal testing approach gained acceptance.
The chart you would see from any laboratory represents the flow of an HIV antibody test as the industry, with the assistance of the numerous independent laboratories, strive to ensure only the truly positive are faced with a rejection of life insurance due to the test. As an aside, in some U.S. States the serum antibody testing was not allowed and the so-called “surrogate aids testing” became the test of last resort. The poor predictive accuracy of these test (mainly the T Cell test) of around 10% meant that some U.S insurance companies found themselves declining to issue 9 out of 10 positives unjustifiably. To discriminate and try to pick out the true test on the basis of age, sex, marital status, etc. Might have created a field day for lawyers. Repetitive testing had a predictive accuracy of only 33%, so the industry was still faced with inappropriately declining 66% wrongly.
You all know the final chapter. In 1989 states like California allowed antibody testing for life insurance!
In 1989 one could estimate that 300,000 + blood profiles were completed on Canadians applying for life insurance. The blood profiles all included the HIV antibody test as well as the test for up to eighteen additional blood chemistries. The industry was witnessing about 0.05% positive as you could see from any labs monthly reports. In a representative mid size company, they had four positive HIV antibody tests discovered in the latest calendar year. A quick synopsis of each follows:
Male, married, age 48
Earned income $200,000+ per year
Excellent medical history
Applying for $1,000,000 of permanent insurance
Male, single, age 38
Earned income $100,000+ per year
Excellent medical history
Applying for $1,000,000 of Business insurance with 4 partners
Male, single, age 33
Earned income $25,000+ per year
Admitted homosexual with history of numerous S.T.D.
Admitted to having HIV antibody test several times
Applying for $50,000
Female, married, age 30
New arrival to Canada
Spousal rider $25,000
The last example also points out a practice of a growing number of companies. Individuals of all ages, sex and martial status who have not been a resident of Canada for one year or more are being tested.
The chain of custody for the blood sample, the pre-notification and authorization of the individual and the confidentiality of the record has improved ten fold in the past five years. The applicant now knows they are being tested. The sealed sample follows prescribed handling that strives to avoid loss, replacement and error. The test results, if positive, is communicated directly and under the strictest of control to either the chief underwriting officer or medical director of the life insurance company. That individual then takes on the responsibility of communicating the information that the case has been rejected by the company. This communiqué is forwarded to the applicant in the form of a letter that states that the application for insurance will not be proceeded with do to abnormal findings on one of the tests performed on the applicant. The company offers to send the complete results of all tests to the attending physician for review and discussion with the applicant. If permission is granted the information is sent to the doctor of choice with a recommendation to the doctor that they themselves have the tests corroborated.
In actual practice I have heard that in only 50% of the cases do the insurers receive permission to release the file to a doctor.
Complete new procedures have been developed within the paper storage or file administration areas of life insurance companies. A concentrated effort has been put to the task of ensuring even stricter enforcement of the confidential natures of the declaration regarding health and laboratory results. In some instance all medical evidence from any source has been purged from the financial contract synopsis. This set of confidential information is then restricted to the viewing by medical director or senior underwriting personnel. The professional insurance doctor and underwriter are cognizant of the need for confidentiality. Even in the instance of internal studies and statistics gathering, these prudent professionals use only a numerical reference as opposed to anything that could publicize the individuals identity.
Currently the chain of custody for blood sample and results is closely monitored and controlled. With the recent advent of urine HIV antibody testing the industry is faced with re-evaluating the handling urine samples. Historically the industry pre AIDS allowed agent and branch employees to collect, label and ship the urine sample. The ease and simplicity of the urine collection and the subsequent testing for sugar and albumin was comfortable. Not so for HIV urine test which has left the industry with three yet to be resolved issues:
(1) Only the Elisa will be done on urine. Thus who will do a follow up Western Blot for samples that are positive? Will the follow up request for a blood sample expose the potential of a positive test to the agent and thus fall outside of our controlled confidentiality?
(2) Tampering with urine samples has been evident for some time especially when drugs of abuse are being sought. By involving the agent in the collection of urine samples will we become counterproductive when more than drugs of abuse are being sought?
(3) The urine test will produce more false positives than the blood test. Is the industry prepared to acknowledge that almost all positive tests in an insurance population will be false positives?
The foregoing will be resolved because the industry will accept only what is reasonable as it searches for cost effective testing that is equitable to the insured and does not jeopardize confidentiality.
Stepping away from the HIV testing and returning to the eighteen other tests plus the test done now on urine, the life underwriter is faced with trying to discern the meaning of all the other tests. Unlike HIV testing there currently is no built in confirmatory test in the typical blood profile. In fact the underwriter sees fewer than 50% of the profiles falling within normal.
The following chart, taken from the Journal of Insurance Medicine gives one company’s opinion of the economics of certain actions prompted by a positive test. In the example we are faced with an abnormal GGT. In summary the chart shows a loss of $1,000 + if the company accepts the risk at standard; $50 if the company declines; and $119 if the company charges an extra premium. With those kinds of odds, no wonder the easy way out is to go on vacation and let someone else make the decision.
When the laboratories offered to throw in a test for cocaine on the standard urine test the industry was skeptical. People who are sold life insurance are not likely to be cocaine users. When the laboratories offered a three-month money back guarantee if not satisfied, the life insurance industry said all right, since we have nothing to lose. For only 80 cents per urine test we now are finding a surprising number of individuals who show evidence of cocaine residuals in their urine test positive for the substance cocaine. In fact most statistics reveal a rate of positive that is five times that of HIV infection. A positive cocaine test brings about the same insurance consequence as HIV positive…decline!
The confidential nature of this test like all others is protected with the same vigor. One difference in approach is that some companies will often write a registered letter to the proposed insured stating that a positive test for cocaine was why they were declined.
Billy Martin once said as skipper of the New York Yankees (a USA baseball team) once declared “I feel strongly both ways” when asked to take a position on something that had more than one side. Extending the immortal words of Billy Martin, I feel strongly all ways. The industry is not perfect as no industry is perfect. The public trust that rests in the confidential nature of the files within the insurance industry is well founded today in spite of ever more sensitive data being stored. The industry must protect its financial integrity as mandated by the share holder or par policy holder. On the other hand, while fulfilling that objective, we must continue to protect the individuals privacy and employ only the fairest of testing available. Assessing the risk is paramount to the private voluntary life insurance underwriting process. Risk classification allows the life insurance company to provide individual policies at the fairest prices to the greatest number of people. Fairness by definition means equal treatment for equal risks.
As a professional underwriter I can only hope that the various impairments that reduce our life expectancy or encumber our ability to enjoy our lives become as extinct as the following from three centuries ago:
1. Apoplex and Meagrom 17
2. Bloody flux, scowring and flux 348
3. Consumption 1797
4. Convulsions 241
5. Dropsie and swelling 267
6. Flocks and small pox 531
7. Rising of the lights 98
8. Teeth 470
Taken from Natural and Political Observations by John Graunt a citizen of London in his book on actuarial statistics 1632. Total buried that year was 9535.