Report on the Global HIV/AIDS epidemic

AIDS in a new millennium: a grim picture with glimmers of hope
In parts of north-east India, widespread injecting drug use provided an easy early
entry-point for HIV. In Manipur, the prevalence of HIV infection among injecting drug
users shot up from virtually nothing in 1988 to over 70% just four years later, and it
has remained at these high levels ever since. Predictably, since almost all injecting
drug users in the state are men, HIV then spread to their wives and girlfriends
through unprotected sex. Around 2.2% of pregnant women in Manipur tested positive for HIV in 1999. In other states, even higher levels of HIV have been recorded
among pregnant women. Most cases of infection appear to have been acquired from
husbands who had been infected in turn by sex workers, themselves part of a longer
chain of transmission. By the mid-1990s, a quarter or more of sex workers in cities
such as New Delhi, Hyderabad, Madurai, Pune, Tirupati and Vellore tested positive
for HIV. In Mumbai, the prevalence of HIV infection among sex workers reached 71%
in 1997.
The Indian Government and individual state governments have launched prevention
programmes to reduce high-risk sex, and there is evidence that in some states these
programmes are resulting in safer behaviour (see Box 2 below). If current prevention
efforts can be scaled up and sustained, India may be able to bring down the rates
of HIV infection in particularly exposed groups and avert a widespread heterosexual
epidemic.
Box 2. A measure of success: changing behaviour in Tamil Nadu
It is important to collect information about the behaviour that spreads HIV as well as
about the progress of the virus itself. It is very difficult to track new infections
because HIV typically causes no symptoms for many years. The proportion of the
population infected - the HIV prevalence rate- is not a reliable indicator of recent
trends in risky behaviour; it encompasses all people with HIV, including those who
became infected many years previously. Information about behaviour can be a useful early warning in countries where the virus is rare or confined to particular groups
and can help explain changes in HIV prevalence in populations where the virus is
already well established. Behavioural information also helps to monitor progress in
promoting safe sexual behaviour and hence shows the impact of national prevention
programmes.
In countries where HIV is widespread, data on sexual behaviour are collected from
men and women in households chosen randomly as representative of the general
population. In countries where HIV infection is less prevalent, it is more common to
conduct surveys among occupational groups - such as soldiers, factory workers or
sex workers - that typify different levels of risk and that can be contacted regularly
over time.
When surveillance systems in the Indian state of Tamil Nadu - home to some 60 million people - showed that HIV infection rates among pregnant women were rising,

13