Under estimating HIV burden in India

[Guest guest]

[The author of this posting wishes to remain anonymous. Moderator]

Dear Forum subscribers,

From a closer analysis itt is quite evident that the figures quoted

as the HIV estimate for the country do not reflect the ground

reality. The entire process of the estimation of the HIV burden of

the country needs to be drastically overhauled and be more evidence

based.

A. Let us look at the ways by which the HIV burden of a country can

be estimated.

1. Encouraging people who may have high-risk behavior to get tested.

This is what Western countries do. Since those who get tested receive

antiretroviral treatment, there is an incentive to get tested and so

depending on the quality of the health care system, nearly complete

enumeration of the cases can be made. This is not suitable in India

since we have nothing to offer anyone testing positive except

discrimination.

2. Population surveys.

What is done is that a sample of a few thousand adults are chosen

and tested and the HIV prevalence results of that sample are used to

estimate the number of infected people.

3. Methods based on the surveillance of pregnant women.

Many African countries do this because it is easy to do. It is cheap

and convenient. But the problem with this method is that since only

the HIV prevalence of pregnant women are available, the HIV

prevalence of the rest of the population has to be derived by making

some assumptions which may not be quite valid. Since men are not

tested, there is no way to know their HIV prevalence. So generally,

the sentinel surveillance survey among pregnant women is supplemented

by small-scale population surveys of the general population to ensure

representation of men and the rest of the population. This will give

a reasonable idea of the HIV burden

B. Problems with estimating HIV burden

1) Generally surveys done on the STD clinic attenders, though useful

in observing the trend of the epidemic are not taken into account in

calculating the HIV burden. This is for several reasons, one of

which, is that the ones who attend hospitals with STDs are extreme

cases and do not represent the people with STDs in the community.

If only the sentinel surveillance data are used for estimating the

HIV burden of the country, many assumptions have to be made.

2) The drawback is that depending on the nature of the assumptions

made, the estimated HIV burden can vary from a low to a high value

and may not have any direct relationship with the true value.

3)The method of calculating the HIV estimate for the country is given

as annexure-1 in http://naco.nic.in/vsnaco/indianscene/esthiv.htm.

4) A careful reading of the process of estimation of the HIV burden

reveals that the method is too dependent on numerous assumptions

which are not verified by direct observation. Hence little confidence

can be placed on any result arrived at by this method.

5) To understand this, let us consider assumption number 3, the most

critical one of the assumptions used in calculating India's HIV

burden. It states " For the purpose of HIV estimation in high risk

population, the rural urban differential will be 3:1 in all the

states. Similarly, for HIV prevalence in low risk population, the

urban –rural differential will be 8:1 in all the states. " Consider

just the second sentence.

6) What this means in simple terms is that it was decided several

years ago when this method of calculation was first introduced, that

for estimating the HIV burden of a state, the HIV prevalence in the

rural population was fixed to be a definite and unvarying fraction of

the urban HIV prevalence.

7) The fraction used is 1/8. The true HIV prevalence in the rural

population and its variation over the years are not taken into

account. In the case of Maharashtra, this would mean that, for

calculating the HIV burden, while one adult in 57 in Mumbai or Pune

is taken to be HIV infected, (1.75% HIV prevalence, 2001), in rural

Maharashtra, only one in 456 adults is HIV infected! The sentinel

surveillance data collected over the years from rural areas in

Maharashtra, which show a far higher figure is not at all taken into

account in the calculation of the HIV burden.

8) So, the burden of infections among the rural population, which

accounts for 75% of the total population of India, is calculated as a

fixed function of the urban HIV on an arbitrary basis.

9) The available data, which shows a rising HIV prevalence in the

rural population, is not taken into account. Many of the other

assumptions made for the HIV estimates are also quite arbitrary,

unverified and most importantly, do not reflect the changing nature

of the epidemic over time.

10) Low risk men are not tested and hence their HIV prevalence

unknown.

11) So, it is quite evident that the figures quoted as the HIV

estimate for the country do not reflect the ground reality. The

entire process of the estimation of the HIV burden of the country

needs to be drastically overhauled and is more evidence based.

______________________