Last updateThu, 15 Jun 2017 11am

You are here: Home | Agenda | HIV/AIDS: Big questions | Prevention of HIV transmission: Do we know what works and what doesn't?

By Mariette Correa

We know that HIV prevalence has stabilised or dropped in some parts of the country and amongst certain groups of the population. We know, for instance, that prevalence amongst female sex workers in Maharashtra has dropped from over 54% to 23%. And that prevalence in the general population in Tamil Nadu has dropped from 1% to .5%. But do we know why? An analysis of prevention efforts in India, the successes and failures, throws up more questions than answers

It is clear that the prevalence of HIV has stabilised in some parts of the country and populations. In Karnataka and Maharashtra, for instance, prevalence at antenatal clinics (ANCs) has remained 1.25% between 2003 and 2005, and in the same period remained nil in Bihar and Assam. National data on HIV prevalence among female sex workers (FSWs) shows a decrease from 10.3% to 8.44% from 2003 to 2005. There have been significant drops in HIV prevalence among FSWs in Maharashtra, for instance, with prevalence of 54.29% in 2003 dropping to 23.62% in 2005. In Andhra Pradesh, figures for FSWs have dropped from 20.00% in 2003 to 12.97% in 2005. There is also a decrease nationally amongst injecting drug users (IDUs) from 13.3% to 10.16% between 2003 and 2005. But simultaneously, prevalence has increased in some parts. In Nagaland, rates of STDs have gone up from 0.98% to 3.50% between the years 2003 and 2005.

We attribute these successes or failures to assumptions about the forces that contributed to HIV spread in the first place and the factors that contributed to its reduction.

Our baseline data for the general population regarding HIV prevalence is weak. This is true not only for the estimates of those infected but for the sources of transmission too. However, data from ANC clinic attendees, which has been discredited as a source on which national estimates are based, may not be useful for estimate projections but could be used to get some sense of trends in local populations. More importantly, what we do have is data from sentinel surveillance sources which give us information about specific population groups. These annual surveys form the basis on which comparisons can be made regarding HIV prevalence in specific populations.

However, we are still uncertain about the relative contribution of factors driving the HIV epidemic in the country. Conventional wisdom and national statistics on the proportion of factors contributing to the spread of HIV are based on data generated through AIDS case surveillance which has completely faulty methodologies. Given this, when there are noticeable declines or increases in HIV prevalence in various parts of the country, it becomes difficult to attribute these changes to specific factors.

Further, it is impossible to gauge the success of individual programmes, since any drop in HIV prevalence is the result of an interplay of factors. However, one can broadly credit some efforts when there are significant drops in prevalence in certain areas. It is clear through ANC surveillance over several years that HIV prevalence has dropped in Tamil Nadu (from over 1.00% in 2000 to 0.5% in 2005). This success is officially attributed to various factors including the expansion of groups covered under prevention and care and support; establishment of consortiums to bring different stakeholders together; setting up of antiretroviral treatment centres and integrated counselling and testing centres. Further, the intersectoral approach and the strategy of moving well beyond the government machinery have contributed to the decreasing prevalence. Consortiums or working groups which have served to bring all agencies together on the same platform have helped to reduce duplication of tasks and facilitate sharing of information. However, it is not clear which factors played a predominant role or whether there were additional factors responsible for declining prevalence.

Again, in Tamil Nadu, despite the drop in prevalence in the general population there are noticeable increases in HIV prevalence among MSM populations (from 2.40% in 2002 to 6.40% in 2005). The reasons for this increase become unclear in a scenario where there are so many factors helping to reduce HIV prevalence overall.

The credit for successes (drop in HIV prevalence) is assumed by the AIDS sector, ignoring other influences. For instance, organisations working to reduce trafficking of women into prostitution are contributing a great deal to the reduction of HIV among sex workers. However, it is only the condom-distributing organisations, working under the AIDS umbrella, that are credited with success, if any, when HIV prevalence in a sex worker community reduces.

Conversely, failures to reduce prevalence are attributed to various socioeconomic vulnerabilities that are beyond the control of AIDS programmes.

With specific population groups (like FSW, MSM, IDU) we may be more aware of the factors contributing to the spread of the epidemic and can plan our efforts accordingly. Still, prevention efforts seem to be donor-driven and straitjacketed. There is a focus on one route of transmission -- for example, sex among sex workers, forgetting that sex workers also have to access healthcare and are exposed to unsafe invasive procedures. In fact, in the overall prevention scenario the inadequate attention given to safety in healthcare is serious cause for concern.

Sentinel surveys show a drop in HIV prevalence among some sex worker populations, and it is likely that the credit goes to interventions in the area. However, it is unclear why some interventions work and others do not when the package of services may be similar. Why are there significant drops in HIV prevalence among some sex worker sites and not in others? In some places, like the Sonagachi project in Kolkata, maintaining the HIV prevalence below 5% over the years despite a state average of 6.80% in 2005 has been attributed to an effective community development approach to HIV/STI prevention. However, the fluctuating sentinel surveillance data across the country for sex worker sites makes the role of HIV prevention interventions difficult to assess.

Successes are also attributed to what was assumed to be the route of HIV transmission; other factors contributing to HIV spread and consequently to reduction in the rate of increase in HIV prevalence are ignored. To give a more or less universal example in the national context, most success stories are attributed to safer sex programmes. The fact is that much of HIV has been spreading iatrogenically and increasing the safety of healthcare in some areas could have contributed significantly to reduction in HIV spread.

In fact, the risks of unsafe invasive procedures (in healthcare, and in tattooing) to which sex workers are exposed have been documented. When there are improvements in healthcare for sex workers, this is not even considered a factor contributing to decline in HIV rates. Sentinel surveillance shows a drop in prevalence in Maharashtra from 54.29% of sex workers infected in 2003 to 23.62% in 2005 at the sentinel sites. It becomes impossible to say how much of these reductions have been due to safer healthcare practices. For instance, a private practitioner in Sangli reported that he was forced to shift from glass to disposable syringes and needles as the sex workers (a large part of his clientele) insisted that he make the shift. They were concerned about their own safety as well as that of other patients.

Attributing the drop in HIV prevalence among specific groups or states solely to the AIDS programme ensures of course that resources keep pouring in. In fact, one of the reasons for the outcry (voiced by many) over the new NFHS-3 estimates of infections in the country is concern that funds for HIV will be reduced. When the inflated figures were in existence, they were hardly ever questioned despite the fact that the basis for these estimates was inherently biased. The national figures were based largely on prevalence of HIV among ANC attendees, and these figures were extended to the overall population. UNAIDS figures have been discredited in many countries, especially following the more rigorous and objective demographic and health surveys (see Global Reports of AIDS epidemics 2005 and 2006).

Preventing HIV spread: Analysing strategies

Prevention of HIV infection depends on a number of factors -- knowledge of how the virus spreads, power equations between sexual partners, the status of women and disadvantaged groups in society, the actions of healthcare professionals and institutions, and so on.

Where are prevention efforts in India focused? What are the hurdles in implementing them? Can we gauge their success?

Preventing heterosexual transmission

Two main strategies are adopted to prevent heterosexual transmission: the ABC approach (Abstain or Be faithful to one partner; if not, use a Condom), and treating sexually transmitted infections.

But the ABC approach ignores the realities of women’s position in patriarchal societies, the limited negotiation spaces that women have to insist that their partners use condoms.

Further, efforts to prevent heterosexual transmission have been sporadic and only a part of larger HIV prevention programmes. Prevention programmes are expected to focus on ‘high-risk’ groups -- that is where the money is. So if you are not a sex worker, injecting drug user, man having sex with men, or closely associated with these groups, you are likely to be left out of prevention programmes. As there are no specific programmes that focus on prevention of HIV transmission in marital or long-term relationships, the success of these efforts becomes difficult to gauge.

The second strategy of prevention amongst heterosexuals, which is treating STDs, does reduce the risk of contracting HIV through unprotected sex (provided the treatment itself is safe). However, the direct association between HIV transmission and STD prevalence becomes difficult to establish. In fact, NFHS-3 findings show that the reported prevalence of STIs is much higher in states with the lowest HIV prevalence: “The reported prevalence of STIs or STI symptoms varies substantially across states, ranging from a low of 2% among women in Goa to a high of 25% among women in Assam, closely followed by Madhya Pradesh (23%). In addition to Assam and Madhya Pradesh, states with a prevalence of 15% or higher among women are Bihar, Tripura, Rajasthan, and Uttar Pradesh. States, in addition to Goa, with prevalence below 5% are Karnataka, Andhra Pradesh, Nagaland, Meghalaya, Tamil Nadu, Himachal Pradesh, and Maharashtra. Prevalence among men is highest in West Bengal (11%), followed by Tripura (10%) and Orissa (9%). States with prevalence among men of less than 2% are Karnataka, Mizoram, Tamil Nadu, Haryana, Andhra Pradesh, and Nagaland.”

States with the highest levels of HIV prevalence have very low reported prevalence of STIs. Also, STD prevalence rates around the world do not correspond to HIV prevalence patterns. In 2002, Bangladesh combined a low HIV prevalence (0.1%) with an STD prevalence of 60% (1) while in the same year a study in Zimbabwe showed a prevalence of 2%, 1% and 2% respectively for gonorrhoea, chlamydia and syphilis, with an HIV prevalence rate of 26% (2).

Further, several studies have shown greater association between STD treatment and HIV than between STDs and HIV (3). One survey showed that HIV prevalence for those with STDs and taking injections was almost double the HIV prevalence for those with STDs and not taking injections (4). It has also been convincingly argued that associations between HIV and sexual variables could be confounded by medical exposures to treat STDs (5) and that expanded STD treatment without attention to injection safety could, ironically, increase rather than decrease HIV incidence (6).

Following recent studies in Africa which reported that male circumcision can reduce HIV infection, there has been much interest in and support for this method as a preventive technique. However, these studies have conflicting findings: they show that while in some age-groups circumcision reduced HIV infections, in other age-groups it was associated with an increased risk of HIV infection. Further, circumcision under unsterile conditions may actually be responsible for HIV infection. Nor are all the studies conclusive. One study that attributes lower HIV prevalence in Muslim men than Hindu men in Kolkata to circumcision shows no significant difference in STD prevalence (7).  Therefore, while HIV prevalence was significantly higher among Hindu men, the prevalence of syphilis and gonorrhoea were not significantly different between the two religious groups. Further, the prevalence of syphilis and gonorrhoea among HIV-positive and HIV-negative men were not significantly different.

Preventing HIV from sex among MSMs

In concentrated epidemics, where there is high HIV prevalence among specific populations, the risks of transmission within those populations are better identified than in generalised epidemics. Prevention efforts are also therefore more focused. The decreases in prevalence among, for example, MSM in the US and Australia could be attributed to the combined efforts of the MSM community, care and support services, and to some extent, national policy.

The situation in India is different. First, while the AIDS programme strives to reduce HIV transmission among MSM, this work is thwarted by the fact that the activity for which this population is targeted is illegal. Despite efforts made by organisations to repeal Section 377 of the IPC, which criminalises “unnatural” private consensual sex between adults, this law is still untouched. Section 377 affects HIV prevention efforts and criminalising predominantly homosexual acts in effect provides moral and legal sanction for continued social discrimination against sexual minorities. It is clear that India’s laws on homosexuality threaten human rights and encourage the spread of HIV.

Further, while the NACO guidelines have been in place to address this population, many State AIDS Control Societies (SACS) are reluctant to follow these guidelines. Also, due to their marginalised status, not many MSM groups have either been formed or come forward for support across the country. With poor mapping in place to estimate the size of the population and the lack of adequate sentinel sites, it is difficult to assess whether there have been actual changes in prevalence over the years.

If one looks at the sentinel data from the years 2003 to 2005, Maharashtra shows a decrease in HIV prevalence rates among MSM (16.80% in 2002, 18.80% in 2003, 11.20% in 2004 and 10.40% in 2005) and Tamil Nadu shows an increase (2.40% in 2002, 4.20% in 2003, 6.80% in 2004 and 6.20% in 2005). But there are wide fluctuations in the sentinel survey data. For example, Delhi shows HIV positivity among MSM of 27.42%, 6.67% and 21.60% in the years 2003, 2004 and 2005 respectively, and Goa’s surveillance data for MSM shows HIV prevalence of 9.09%, 1.68% and 4.90% for the years 2003, 2004 and 2005 respectively. These fluctuations are cause for concern and raise questions about the methodology adopted. Informal reports from NGOs (that are expected to ‘provide’ the MSM population for testing during these surveys) show that some have been pressurised to meet the target requirements for fear of their projects being discontinued, and have therefore included any (read heterosexual) men in their sample.

Preventing mother-to-child HIV transmission

Most mother to child HIV transmission can be prevented. In 2001, NACO and the health departments began testing pregnant women and offering medical interventions to protect children from HIV infection through perinatal transmission. Antiretroviral drugs reduce HIV transmission to 10% (down from 33% otherwise). This rate can be further lowered with better combinations of antiretroviral drugs, which are available through the private sector, along with caesarean delivery and avoidance of breastfeeding.

The PPTCT programme in India includes single-dose nevirapine to the mother and later to the baby.

However, successes in the West on the prevention of perinatal transmission have in fact relied more on combination therapies than on nevirapine. In the United States and Europe, where today fewer than 2% of babies born to HIV-positive pregnant women have the virus, it was mainly due to routine use of Zidovudine during pregnancy.

Even in Botswana, which has successfully reduced the rate of HIV transmission from mother to child to less than 4%, the regimen has been dual drug treatment -- four weeks of AZT, and then a single dose of nevirapine at birth to mother and child. The success in Botswana is also attributed to political support and policy decisions, like opt-out testing (the testing of all pregnant women for HIV unless they refuse); providing HIV test results in 20 minutes to expectant mothers; and giving dual drug treatment for HIV-positive women. The success of Botswana is particularly creditable considering 34% of its pregnant women are HIV-positive.

However, the global scenario on prevention of perinatal transmission remains bleak. According to the Global AIDS Report 2006, a mere 9% of pregnant women in low- and middle-income countries were offered services to prevent transmission to their newborns in 2005, as compared to 7.6% in 2003. In 2005, the percentage of HIV-positive pregnant women who received prophylactic antiretrovirals was 9.2. Since these are global figures and considering that most HIV-positive women in developed countries receive ARV to prevent their babies from getting HIV infection, it is evident that developing countries fall far short of adequate coverage.

Preventing HIV from injecting drug use

Given the limited success of de-addiction programmes, prevention programmes relating to injecting drug use educate injecting drug users on sterile injecting equipment -- avoiding sharing needles,syringes, water, or drug preparation equipment; using syringes obtained from a reliable source; sterilising previously used equipment in boiling water or disinfecting it with bleach before reuse. Injection drug users and their sex partners are also advised to take precautions, such as using condoms consistently and correctly, to reduce the risks of sexual transmission of HIV.

NGOs supported by SACS or international agencies have comprehensive harm reduction programmes which include community outreach, abscess management, safe injecting, needle syringe exchange programmes, and oral substitutes. Buprenorphine and methadone are commonly used opioid substitutes given under medical supervision to injecting drug users. The former is being increasingly used among NGOs working with IDUs in different parts of the country. Opioid substitution therapy is also part of the next phase of the National AIDS Control Programme (NACP III). For those who can and are willing to give up drugs altogether, detoxification, rehabilitation and other therapies are available in many parts of the country.

While some NGOs have been partially successful in reducing harm to IDU, it is again unclear how much the success of individual organisations has accounted for changes in HIV prevalence in various states. Sentinel surveys show an increased HIV prevalence amongst injecting drug users in Delhi (7.23, 14.40, 17.60 and 22.80 for the years 2002, 2003, 2004 and 2005 respectively) over the last few years, whereas there have been drops in Maharashtra (39.42, 24.90, 29.20 and 12.80 for the years 2002, 2003, 2004 and 2005 respectively), Manipur (39.01 in 2002, 21.00 in 2004) and Nagaland (9.56 in 2002, 3.22 in 2004), till 2004, with slight increases again in 2005 (24.10 and 4.51 in 2005 in Manipur and Nagaland respectively). Much of the success can be attributed to various NGO programmes as there have been practically no government interventions (apart from some support by SACS and from the Ministry of Social Justice and Empowerment) for programmes for IDU. This is despite the fact that IDU continues to be a criminalised activity, with constant police harassment impeding the efforts of organisations working to prevent HIV among this population.

Preventing HIV from other blood exposures

Unsafe invasive procedures in healthcare contribute significantly to the spread of HIV. But this is accorded low priority, making it difficult to contain this route of transmission. Much effort has been made over the years (especially through the AIDS programme) to train doctors and nurses on standard biosafety precautions; but healthcare workers and facilities often do not have the necessary equipment.

Some changes have been made over the years to improve injection safety, including shifting from glass to disposable syringes and needles and the promotion of auto-disable (AD) syringes. The government has introduced AD syringes in all centrally-sponsored immunisation programmes.

Because of the risks that healthcare providers face in administering care, the government has a programme where health providers are given medication after a possible risky exposure to HIV (post-exposure prophylaxis) to prevent them from getting infected. This is available to all doctors and nurses working in government health facilities. Unfortunately, not many access this facility even after accidental injuries as the procedures are often cumbersome. The medications are not always available, requisitions need to be made to the authorities, and confidentiality of the healthcare worker is not always protected. In the more remote areas, medical staff are not even aware that these facilities exist, nor are there enough doctors trained to administer the regimen across the country. 

Irrespective of the resources that flow into HIV prevention and how much this may detract from other public health priorities, we will be a long way from fighting the epidemic in India without clear evidence-based research on the proportionate contribution of various factors driving the epidemic.  

(Mariette Correa is an independent consultant who has been involved in HIV/AIDS programming for NGOs in Goa and South Asia)


  1. ‘Low HIV prevalence and high risk behaviours: Adolescents in urban Bangladesh’. Muna L, Cleland J. Int Conf AIDS. July 7-12, 2002; 14: abstract no WePeE6505
  2. Ref: ‘National survey of STDs and HIV prevalence among residents in rural growth point villages in Zimbabwe’. Woelk G B, Kasprzyk D, Montano D E, Mutsindiri R, NIMH Collaborative HIV/STD Prevention Trial NI. Int Conf AIDS. July 7-12, 2002; 14: abstract no WeOrC1270
  3. National Family Health Survey, India. NFHS-3. National Reports.
  4. Gisselquist D, Potterat J J, Brody S, Vachon F. ‘Let it be sexual: how healthcare transmission of AIDS in Africa was ignored’. Int J STD AIDS 2003; 14: 148-161
  5. Gisselquist D, Rothenberg R, Potterat  J, Drucker E. ‘HIV infections in sub-Saharan Africa not explained by sexual or vertical transmission’. Int J STD AIDS 2002; 13: 657-666
  6. Minkin S F. ‘Iatrogenic AIDS: unsafe medical practices and the HIV epidemic’. Soc Sci Med 1991; 33:786-790
  7. Gisselquist D, Potterat J J. ‘Confound it: latent lessons from the Mwanza trial of STD treatment to reduce HIV transmission’. Int J STD AIDS 2003; 14: 179-184
  8. Talukdara A, Khandokara M R, Bandopadhyaya S K, Detels R. ‘Risk of HIV infection but not other sexually transmitted diseases is lower among homeless Muslim men in Kolkata’. AIDS 2007, Vol 21 No 16

InfoChange News & Features, January 2008