[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries
.

HIV positives

Three years ago I went to a regional meeting of the Royal Statistical Society in Liverpool. The subject was the prevalence of HIV infection and AIDS in the UK and, speaking of past projections, the speaker said, ‘The Institute of Actuaries has made a mess of this. Their predictions were way off the mark’. The Institute gave a range of projections of course, but now lower projections are accepted in valuation bases. When the speaker personally criticised my actuarial hero, Professor David Wilkie, I had to speak in his defence. I said, ‘If the Institute had been clairvoyant and been able to predict accurately the HIV and AIDS figures that we see today, do you think that anyone would have changed their behaviour?’

Raising awareness
I am sure that the alarmist projections from the Institute of Actuaries and elsewhere did a good job. That, and some high-profile deaths, put fear into the people who were most at risk and made them consider safe sex. Hence, the much lower figures that we see today. Without those projections, we might have had something akin to the Black Death but, of course, the figures, for various reasons, are still alarmingly high and at epidemic levels in some countries.
There is still no cure for AIDS, but the trend towards safe sex, the fact that the average period between becoming HIV-positive and developing AIDS has been extended, and the fact that AIDS victims are living much longer thanks to medical advances, have all meant that the UK figures are constantly improving.
How does this affect life and health insurers? After an uncertain start in the 1980s, it has become standard practice to ask applicants to have an HIV test when the sum assured is over a certain level or when the applicant is considered to be at risk. Each office is free to determine its own limits, but the levels (which have increased in recent years) tend to be around £250,000 sum assured for single males and £1m sum assured for married males and females. Hence, relatively few married males and females are asked to have HIV tests as a matter of course. A lifestyle questionnaire asks about sexual persuasion and anyone who says he is homosexual may be sent for an HIV test.

Risk assessment
The industry has become more lenient when it comes to assessing AIDS risks. Ten years ago a homosexual applicant with a negative HIV test would have been charged an additional £5 per annum per each £1,000 sum assured someone wanting a £100,000 mortgage contract might have to pay an additional £40 a month. Now many offices quote standard terms and as this is a competitive even sheep-like industry, all offices may be doing the same in a few years’ time.
My question is simple: are the life offices going too far? Because there has not been the expected run of AIDS claims, have the offices been right to relax their requirements? I am adding this note of caution because some figures from Diagnostic Technologies Corporation (DTC) give considerable cause for concern. DTC supplies HIV-testing kits to many UK offices and arranges for the results to be tested. With DTC’s kind permission, we are able to reproduce their statistics for the past five years in table 1.
It does not need advanced statistical testing to demonstrate that the results for 1999 are out of line with the other years and are surprisingly high. A binomial test indicates that there was no notable change in the statistics between 1995 and 1998, but the 1999 result does indicate a significant increase. Similarly, if we take the normal distribution for the first four years, we find that the 1999 data point is not within the 99% confidence interval.
There was a drop in the number of tests received by DTC in 1998, but this returned to its previous level in 1999. There is no significance in this. DTC has contracts with various life offices and they can change from year to year, thereby affecting the number of tests received.

Distortion and discrepancies?
My first question was to ask whether the laboratory assessment had changed in some way. Has some new method of analysis caused a larger number of samples to be categorised as HIV-positive? I am informed that nothing has changed at all.
Second, I asked about duplicates. If one HIV-positive individual had sought cover with ten offices, then the figures would be distorted. However, there are no duplicates among the HIV positives during 1999 or, indeed, in any of the other years.
Third, I wondered whether the offices had changed their limits and hence their approach to requesting HIV tests during 1999. DTC has no information as to why an applicant has been asked to have an HIV test, but there has been no major relaxation in the automatic limits during 1999.
The indication, therefore, is that something is happening and we need to determine what it is. Up to mid-February, DTC had a further two HIV positives, but this does not enable us to make any assumptions about Y2K. DTC does not have all the ages and sexes of the applicants who are HIV-positive over the five-year period. Prior to 1999, 15 were known to be male and six female. Only three out of 21 lives were over 40, the eldest being 43. The breakdown for 1999 is shown in table 2. This information is, quite simply, terrifying.
Some of these people would not have been sexually active at the time of the first advertising campaigns and may have been lulled into a false sense of security with the lower figures of recent years. Where are the alarmist projections now that we need them? Has the safe sex message not got home to younger people?

Where now?
I don’t think it has. There was recently a new report on Radio 4 saying precisely this and, apparently, STD clinics are busier than ever. The message of safe sex has been swamped by stronger messages, also aimed at young people, that promote rampant promiscuity, which is now called ‘lifestyle’. London is inevitably the centre of such things, but the club and university cultures mean that the lifestyle message is widespread throughout the UK and is most unappealing.
It is important to keep monitoring HIV test statistics and perhaps obtain further figures from other laboratories serving the industry. Ideally, we would like to know the difference between the lives thought to be at risk and the lives whose tests were requested automatically. Also, because the automatic HIV-testing limit is so high for females, why did the various underwriters consider the females above to be at risk?
Quite possibly, the life offices are targeting the wrong lives for automatic HIV-testing. Maybe the limit should be £250,000 for all males and females under age 45, and £750,000 over that age. It would be intriguing to have the views and experiences of fellow actuaries on this issue.
And there’s another reason why offices need to be cautious about the assessment of HIV risks. If you log on to gaymoney.com, you will read the following advice: ‘Get your insurance in place before you test for HIV. Test anonymously so that you have all your facts before you take action; do not do that first test through your doctor.’ Good advice, to be sure, but if that isn’t an invitation to non-disclosure, then I don’t know what is.

00_05_09.pdf