Safety in Numbers

Chapter 5

The presentation of official statistics

In Britain, data on positive HIV antibody test results and AIDS diagnoses are sent voluntarily and confidentially to the Public Health Laboratory Service (PHLS) Communicable Disease Surveillance Centre (CDSC) or to the Communicable Disease (Scotland) Unit (CD(S)U) by clinicians and laboratory microbiologists. Since 1987, the Department of Health has then released up-to-date AIDS statistics by press release each month, and more detailed figures for both HIV infection and AIDS each quarter. A more thorough analysis is prepared by the PHLS in the form of AIDS/HIV Quarterly Surveillance Tables, which are circulated primarily to public health officials involved in AIDS.

While the PHLS' role is purely professional, the Department of Health press releases usually contain several pages of editorial comment, sometimes including a direct response to the statistics from an appropriate Government figure. In this way the Department of Health is able to direct attention to the areas which most concern it, and to present a political perspective on the data. These press releases and the media coverage they generate have proved to be a highly effective way of emphasising the small but growing number of cases of HIV transmission through heterosexual sex, at the direct expense of appropriate attention to the very much greater and increasing number of cases among gay men.

The Department of Health's editorialising mostly consists of useful explanations of important considerations in the interpretation of the figures. For example, the press releases highlight the likelihood that AIDS diagnoses are somewhat under-reported, and that individuals who have elected to take the HIV antibody test are probably only a minority of those currently infected with HIV. However, the comments on the current incidence of HIV in particular risk groups betrays something of the relative importance attached to each of these discrete 'mini-epidemics'. For example, between 1988 and 1990 the quarterly press releases commented that the HIV statistics "show [the epidemic's] spread outside the homosexual and bisexual groups" (31), when during this same period 70 - 72% of positive HIV antibody tests were attributed to sexual intercourse between men. (32) Likewise, while no comment is made on the fact that during the quarter March - June 1988 191 gay or bisexual men tested HIV-positive, 28 new infections by heterosexual transmission elicit the concerned observation that:

 

the number of heterosexual infections is growing... the potential for HIV infection spreading into and becoming established in the general low-risk population is therefore steadily increasing. Evidence from other countries underscores the potential importance of heterosexual spread, and the government's campaign at educating the general population, particularly the young and sexually active, about the continuing danager (sic) of HIV infection, forms another essential part of the strategy for the prevention of spread of HIV in this country. (33)

In October 1990 this standard press release form was changed. Now the editorial comment was broken down under just five headings: "AIDS cases"; "AIDS in heterosexuals"; "Region"; "HIV antibody positive persons"; and "HIV in heterosexuals". (34) Bizarrely, all the information about incidence in particular risk groups was now presented under the "AIDS in heterosexuals" or "HIV in heterosexuals" headings, including any comments on incidence among gay men. This was at least an honest way of presenting the Department's obvious view that the only interest in the statistics was the extent to which they reflected heterosexual transmission. Thus the section "AIDS in heterosexuals" contained 21 lines on the 117 new AIDS cases during the previous three months which were attributed to heterosexual HIV transmission, while in the same section the 886 new cases among gay or bisexual men merited only 26 words. And again, under the heading "HIV in heterosexuals", the 407 new positive tests among heterosexuals during the previous year are highlighted as a 57 per cent increase, while newly detected infections among gay or bisexual men are referred to only as a 41 per cent increase, with the actual number, 866, omitted entirely from the press release. (35)

As Simon Watney has pointed out:

 

it is routine in Britain to find wholly misleading analogies being made in the Press, and elsewhere, between percentage increases in HIV or AIDS cases among heterosexuals, and the overall cumulative statistics, with the false implication that HIV is being more widely transmitted amongst heterosexuals than amongst gay men or drug users. It now seems that an earlier ideological tendency to deny any risk to heterosexuals has been replaced by a regular tendency to overstate risk from unprotected vaginal intercourse. (36)

Much of the responsibility for this tendency doubtless lies with the Department of Health and its press releases. The statistics for 1990 offer a typical case study. In January 1991, newspaper reports of the latest British statistics carried headlines such as "Aids virus infection by normal sex rises" (37), "AIDS risk growing for girls" (38), "Aids spread greatest in heterosexual community" (39) and "MUMS-TO-BE IN NEW AIDS STUNNER". (40) Most of these reports quoted the Department of Health's press release saying that "Cases of AIDS among heterosexuals are still increasing faster than in any other group", and only two out of the eight stories even mentioned the latest totals among gay men.

The statistics contained in the press release did indeed reveal continuing HIV transmission among heterosexuals: during 1990 the total number of persons infected through heterosexual intercourse rose by 61% from 766 to 1237, and cases of AIDS in people infected by this route increased by 98% from 135 to 268. The press release also pointed out that "[AIDS] cases in women have increased rapidly during 1990, by 78 per cent from 114 to 203". The editorial notes contained only one sentence on AIDS in gay or bisexual men, which noted that "The number of these cases rose much more slowly during 1990, by 42 per cent from 2326 to 3295". As if in fear that this sentence might distract editors' attention from the heterosexual statistics, the press release quickly added, still under the heading AIDS In Homosexual Or Bisexual Men:

 

Worldwide, the main route of spread is vaginal intercourse. WHO recently forecast that up to 10 million children could be infected by their mothers by the end of the century. The Government's campaign aimed at educating the general population, particularly the young and sexually active, about the continuing danger is an essential component of the strategy for the prevention of spread of HIV in this country. (41)

Given this emphasis, it is hardly surprising that the press reports focussed so disproportionately on heterosexual transmission. A closer look at the figures would have revealed that the total for women with AIDS included those infected through drug use or blood or tissue transfer such as transfusions, and that AIDS cases among women infected through heterosexual sex had in fact increased from 52 to 97 cases, accounting for less than half of the total number of women with AIDS. A more reasonable summary of the breakdown of people with AIDS would have reported that no less than 80% of cases reported by the end of 1990 were among gay or bisexual men. 70% of people known to be HIV-positive were gay or bisexual men, with 1,411 testing positive in 1990. This means that in the space of just one year, a far greater number of gay men tested HIV-positive than the total number of male or female persons infected through heterosexual sex throughout the entire British epidemic. Yet the editorial notes on HIV statistics focussed exclusively on the 471 new reports of heterosexual transmission and made not a mention of continuing transmission among gay men. In this way, the Department of Health' presentation of the statistics has helped to set an agenda in which the evidence of small-scale but increasing heterosexual transmission is allowed totally to obscure the continuing reality of the British epidemic and its disproportionate impact upon gay and bisexual men.

Likewise, in May 1991 the Junior Health Minister Virginia Bottomley announced the first results from the Medical Research Council's programme of anonymised HIV surveys, in which tests were performed on blood taken for other purposes from about 44,000 people who attended 27 selected ante-natal clinics or 6 genito-urinary medicine (STD) clinics. (42) Prevalence among pregnant women ranged from zero to about one in 200 in one inner London clinic; at the STD clinics in London, prevalence among heterosexual women was one in 500 and among heterosexual men one in 100. Among gay or bisexual men at clinics, prevalence was no less than one in five. "These results," said Mrs Bottomley, "...provide valuable information to help us target our response to the epidemic and show the extent to which HIV is a problem in London". (43) Yet in virtually all the ensuing press attention, Mrs Bottomley and her colleagues such as the Chief Medical Officer Sir Donald Acheson focused solely on the heterosexual statistics, saying nothing about the shockingly high levels of infection among gay men. Thus the quote from Sir Donald provided in the official press release claimed that "The results of these surveys... reinforce our earlier messages that HIV and AIDS is increasing in the heterosexual population and is a problem that faces us all." He went on to summarise safer sex advice as "Avoid casual sex and try to stick to one faithful partner" - advice which is likely only to increase the risk of infection for a gay man with perhaps a 20% chance of choosing an infected partner.

The specific profile of HIV infection in Scotland is frequently held up as a example of an epidemic which has evolved to the extent that it now predominantly affects heterosexuals. As part of the UK, Scotland is seen as offering a compelling example of the imminent threat to the nation as a whole. Statistics published in September 1992 were reported in The Independent newspaper under the headline "Heterosexual intercourse 'commonest HIV source'", blurring any distinction between Scotland and the rest of the UK, while the story itself argued that "Although the figures refer only to Scotland, they will cause considerable concern in relation to future UK studies of HIV infection". (44)

These figures showed that in the first six months of 1992, a larger proportion of newly detected cases of HIV infection were attributable to heterosexual transmission than any other route. However, the real picture is somewhat more complicated. First, the actual number of heterosexuals taking the HIV antibody test has increased faster than the number of gay and bisexual men or drug users. (45) This means that even if prevalence within all risk groups had remained absolutely stable, the proportion of cases seen among heterosexuals would nevertheless increase. The most recent figures reveal that compared with 1990, the number of non-drug-using heterosexuals who took the test increased by 44% in 1991, while the proportion who tested HIV-positive actually decreased from 0.7% in 1990, to 0.5% in 1991. (46) Once again, however, these very low prevalence rates are allowed to overshadow those among gay and bisexual men. In 1991, a gay man taking the HIV antibody test in Scotland was ten times more likely to be infected than a non-drug-using heterosexual.

It is generally believed that the Scottish epidemic continues mostly to affect injecting drug users, and indeed, a cumulative total of 57% of cases of HIV infection in Scotland have been among drug users. However, the introduction of needle exchange schemes in the mid-1980s appears to have had a dramatic impact on new infections in this population, such that the proportion of cases among drug users has fallen from 80% in the years up to 1985, to 18% in 1992. By contrast, throughout the 1980s gay and bisexual men constituted a growing proportion of positive test results, rising from 18% by the end of 1985, to 41% in 1990. (47)

Carelessness with the HIV/AIDS statistics can thus lead to entirely mistaken conclusions about the current epidemic. For example, in February 1992 the highly-regarded AIDS Newsletter published by the Bureau of Hygiene and Tropical Diseases in London reported that during 1991 "the greatest increase in [AIDS] cases was recorded in the transmission category sexual intercourse between men and women (up 66.2% from 266 to 442)...". (48) The clear implication is that the 176 new cases of AIDS among heterosexuals in 1991 represented the largest increase in any transmission category. In reality, the same set of statistics revealed that 959 gay or bisexual men had been diagnosed with AIDS during the same twelve months - over five times as many as among heterosexuals. It is through mistakes such as this that inaccurate perceptions of the current shape of the epidemic are formed, with serious consequences for the establishment of prevention priorities.