HIV: The numbers controversy revisited

[Guest guest]

HIV: The numbers controversy

By Sandhya Srinivasan

Do we have 2 million or 20 million HIV-positive in India? Or is there a

plateuing of the epidemic? Speculative and alarmist figures about the number

of Indians affected by HIV/AIDS have added to public confusion and affected

the programme's credibility. This is the first in a series of articles on

the issues and controversies surrounding HIV in India

HIV and AIDS have been the centre of controversy since the early-1980s, when

immune-deficiency illnesses were identified in otherwise healthy men in the

US. The controversies have continued over the last two decades to the

present day, even as huge amounts of money are being spent, internationally

and in India, on prevention programmes, vaccine research and drugs believed

to keep infection under control without cure.

This series of articles on HIV and AIDS will examine all these controversies

and place several questions in perspective. The first big controversy has to

do with the numbers. How accurate is the method by which official estimates

and predictions of HIV infection are made? What is their relative

significance when compared to other health problems in India? And does it

make a difference? Other articles in this series will ask whether HIV

prevention and treatment for AIDS should be a vertical programme or

integrated into the health services; whether research into a vaccine will

benefit India or whether Indians will be used as guineapigs; whether

mandatory premarital testing is possible and ethical, and whether patients

have the right to treatment or the right to refuse treatment. There are

other issues too, but first, the numbers controversy.

In November 2002, a US government report predicting that India would have 20

million people with HIV in a couple of years received much press publicity

as well as a lot of flak. This was only the latest in a series of

speculative and alarmist figures that have added to public confusion and

affected the programme's credibility. For some years now, India has been

mentioned in international reports as leading in the number of people with

HIV, sometimes the first in the world, sometimes the first in Asia. In 1996

we were told there were between three and five million people with HIV in

India. In 2000, up to three lakh were said to have died as a result of AIDS

in India. Indian government sources have announced figures disputed by

international organisations, and vice versa, and all of it has been

confusing. For most of us average readers, this would make our eyes glaze

over.

Estimates of HIV infection are based on screening at designated sentinel

surveillance sites in the country. The number of sites has grown over the

years from 55 in 1994, to 180 in 1998 and 232 in 2000. As of August 2001,

there were 320 sentinel testing sites under the supervision of the National

AIDS Control Organisation (NACO). These include STD clinics and prenatal

clinics, as well as other sites to look at the prevalence of HIV among

injecting drug users, sex workers and men who have sex with men. The testing

is meant to be anonymous and unlinked - the test results cannot be traced

back to any individual.

Blood samples collected at these centres are tested for HIV. Based on

various calculations - agreed upon by epidemiologists and bio-statisticians

at NACO, WHO and UNAIDS -- estimates are arrived at on the prevalence of HIV

among populations with high risk behaviour as well as the general

population. The prevalence of HIV among the pregnant women surveyed is used

to estimate the prevalence of HIV in the general population in that

particular geographical area. All these figures are used to estimate the

number of HIV infections at the national level.

All that sounds very good, but there are a number of problems. For one, even

NACO admits that the surveillance machinery needs improvement. Does the

testing at these sites present a reasonably accurate picture of HIV

prevalence in the population in various parts of the country? If it is bad

now, it was worse earlier when there were fewer centres, and what do we make

of the figures from those years? Third, who knows about the quality of data

collection in states with abysmal health infrastructures and related

facilities? And finally, the estimates have a low, medium and high estimate

and the official estimate is always the high estimate. We must presume that

the models and calculations by which the estimates are arrived at, are

appropriate.

Estimates for the last four years might also raise further questions.

According to the National AIDS Control Organisation's website, the estimated

number of adults living with HIV was 3.5 million in 1998, 3.7 million in

1999, 3.86 million in 2000 and 3.97 million in 2001.

If those figures are reasonably accurate, one might wonder what they mean.

The number of new infections each year was 0.2 million in 1998-99, 0.16

million in 1999-2000, and 0.11 million in 2000-2001. So is there a gradual

decrease in new infections? One might argue that the decrease is even more

significant since the country's population has gone up over this period.

NACO suggests that " over a period of time, the new infections may reduce to

a negligible number, which is an indicator of the plateauing of the

epidemic. "

If indeed there has been a drop, should we believe that an effective AIDS

prevention programme is responsible? Or should we believe that this is the

result of a natural decline in the epidemic? How accurate are these figures

anyway? It is interesting that the government acknowledges the limitations

of the information from its existing surveillance system.

Others might argue that all this quibbling about numbers is irrelevant, that

it doesn't matter whether there are 2 or 20 million people with HIV --

surely the problem is obvious enough, and we should be concentrating our

energy on preventing more people from getting infected, on providing

treatment to those who are already HIV-positive, and on addressing the many

associated socio-economic, ethical, and legal dilemmas that HIV has created.

It is also argued that HIV should not be compared to other illnesses because

it affects all other diseases. So, for example, more people will die of TB

because of HIV. Also, as a virus transmitted mostly by unprotected sex, it

affects young people at the most productive stage of their lives. This will

have enormous economic consequences, to families as well as the overall

economy. Incidentally, the prevalence of HIV in India is relatively low but

the large population results in a high overall number which could have a big

impact.

There are many reasons why numbers are relevant. For one, inflated figures

can be used to take resources and energy away from other equally important

public health programmes. At the same time, underestimates can result in a

problem being ignored. When inaccurate figures are challenged and proven

wrong the programme can backfire. Finally, people have a right to accurate

estimates of all health problems and their relative burden - on people's

health, the health care system, individuals' finances and the economy. They

also deserve access to health and welfare services that meet their needs. It

is not a good idea to quote high numbers without ensuring accessible and

affordable health services to all who need them.

Source: http://infochangeindia.org/features97.jsp

(InfoChange News & Features, May 2003)