Last updateSat, 22 Jul 2017 6am

You are here: Home | Agenda | HIV/AIDS: Big questions | Transmission: Is it just about sex and drugs?

Transmission: Is it just about sex and drugs?

By Mariette Correa

Do we really know what the various forces driving India’s epidemic are? Some studies report that around 23% of medical injections could be using unsterile syringes or needles. WHO estimated that unsterile medical injections accounted for 24% of HIV transmission in India in 2000. But India has focused almost exclusively on the sexual route of transmission. Very little space is left for non-sexual routes of transmission. This has important implications for the prevention programme

The contribution of different routes of transmission to HIV infection in India is rather uncertain. Very early in the epidemic, with a few sex workers infected and drawing on parallels with Africa, the government programme started functioning on the premise that heterosexual sex was the main cause of HIV spread in the country. Accordingly, sentinel surveillance included groups that were perceived to be at risk -- sex workers, people visiting clinics for treatment of sexually transmitted diseases (STDs), injecting drug users (IDUs), and men who have sex with men (MSM). The focus of the prevention programme has been on these vulnerable groups that are at greater risk of HIV infection.

While sentinel surveys do show high prevalence among these populations, calculations of prevalence in each of these populations suggest that the number of infections in all the groups at risk is not large enough to account for anywhere near the majority of estimated cases in the country (1, 2). Unsafe healthcare practices such as unsterile injections and tattooing may be responsible for a significant number of HIV infections, and prevention efforts should address this risk as well.

Determining routes of transmission in India

The National AIDS Control Organisation’s (NACO’s) attribution of cases to different routes of transmission is based entirely on AIDS case surveillance. To get these figures, hospitals and institutions across the country assess and record the route of HIV acquisition for in-patients with AIDS. These institutions report to the relevant State AIDS Control Societies (SACS), which report to NACO. NACO consolidates this data at the national level. To obtain data from the states, NACO develops monthly AIDS case surveillance reporting formats, which the SACS distribute to hospitals reporting AIDS cases. Within hospitals, hospital authorities are responsible for deciding how to collect information for AIDS case reporting.

Reliability of information gathered

As AIDS case reporting is not mandatory, the SACS collect information from only those institutions that cooperate. AIDS cases are both grossly undiagnosed and underreported. Even in the few states in the country where AIDS is notifiable, most doctors do not report AIDS but the underlying disease. Reluctance to report a case as AIDS is partly due to the social and personal implications of identifying HIV-positive individuals, despite the fact that public health surveillance is not meant to identify or track specific cases. In many districts, only one or two government hospitals submit reports. There is minimal reporting, if at all, from the private sector; in fact, very little effort is made to get the private sector to report. Data on AIDS cases are therefore unusable in any meaningful way.

In fact, the conclusion that sex was driving India’s epidemic came in the late-1980s, though up to September 1991, only 96 AIDS cases (including 13 foreigners) had been reported from across the country(3). As of January 1999, 13 years after the first case was discovered, only 6,703 AIDS cases had been reported, of which three-fourths were from Maharashtra and Tamil Nadu (4). Through end-July 2005, a cumulative total of 111,608 AIDS cases had been reported; 70% of these were from Tamil Nadu, Maharashtra and Andhra Pradesh which in no way can give a national estimate of the source of people’s infections.

Given the assumption that sex was almost the sole cause of the spread of HIV, surveillance systems have been geared to perpetuate this fallacy. This has resulted in national estimates that leave very little space for non-sexual routes of HIV transmission. NACO’s figures for cumulative AIDS cases through July 2005 report that 86% of HIV infections in AIDS cases have been from sexual exposure, 2.4% from IDU, 2.0% from transfusion of contaminated blood and blood products, 3.6% from mother to child transmission, and 6.0% from other or not specified risks (5). It gives no estimates for medical injections, or other blood exposures. (Sexual transmission as a category in the formats often, though not always, includes sex between men [5].)

Categories at risk

NACO designs forms for hospitals to use for monthly AIDS case reports. The current form asks hospitals to assign adults with AIDS into one of five ‘risk/transmission’ categories: ‘sexual route’, ‘through blood and blood products’, ‘through infected syringes and needles’, ‘others’, and ‘not specified’. Categories for children up to 14 years are ‘perinatal’, ‘through blood and blood products’, ‘others’, and ‘not specified’. The form does not mention other invasive procedures in healthcare and cosmetic services (such as dental care and tattooing). Moreover, there is no space on the forms to facilitate probing the category ‘others’. Therefore, while doctors sometimes suspect that someone has acquired an HIV infection through another specific exposure, like dental treatment, they have no place to report this information. The category ‘others’ may include an important proportion of cases; at the national level, the joint category for ‘others’ and ‘not specified’ is 6%, larger than any other category except sex. 

The guessing game of risk assignment

In asking hospitals to assign AIDS cases to risk/transmission (which involves the doctor trying to identify what must have been the actual risk of the person’s HIV infection) rather than to exposure categories (based on transmission efficiencies and probabilities of becoming infected through each exposure in India), NACO in effect asks doctors and counsellors to report their best guess about the source of someone’s HIV infection rather than report objectively what they learn about the risks. NACO provides no guidelines on how to classify cases with several risks.

AIDS case surveillance in India does not distinguish between high and low risks. The sexual category includes not only people with high-risk sexual exposures (for example, men who have sex with men and people with known HIV-positive partners), but also cases with lower risks, such as one to several extramarital sexual encounters -- even with condoms -- with a neighbour. Hence, cases with high-risk exposures are ‘lost’ in a large and non-specific category. Similarly, the category ‘through infected syringes and needles’, refers ambiguously to IDU and to medical injections. NACO’s forms do not offer this as a risk category for children, implying that the category for adults is specific to IDUs.

Compare this with surveillance systems in countries like the USA and Canada where they report AIDS cases according to ‘exposure categories’, letting readers consider whether the number of cases assigned to a category reflects its contribution to HIV transmission.

(In many countries, surveillance systems report cases according to a hierarchy of risks reflecting rates of HIV prevalence within groups defined by each risk. For example, in the US (6), Canada (7), Australia (8), and Europe (composite reporting for 52 countries by the European Centre for Epidemiological Monitoring of AIDS [EuroHIV]) (9), the first three categories for adults are: [1] MSM, [2] men and women who report IDU, and [3] men who report both MSM and IDU risks. Thereafter, receipt of blood or blood products (or coagulation disorder) and heterosexual contact are ranked fourth and fifth. In fact, in the US, AIDS cases are assigned to the category ‘heterosexual contact’ only if they “report specific heterosexual contact with a person with, or at increased risk of HIV infection” and have no other risk that is higher in the hierarchy, such as IDU (6).)

Little or no effort has gone into training those responsible for AIDS case surveillance on how to collect information on risks, as the formats are seen as self-explanatory. Since the categories are mentioned on the forms, SACS officials do not feel the need to train doctors and counsellors on reporting routes of transmission. Further, NACO training focuses on detection and clinical diagnosis of AIDS cases.

How hospitals report

Within reporting hospitals, hospital authorities are responsible for deciding how to collect information. There are indications that hospitals have interests that compete or conflict with accurate reporting. Private hospitals may be unwilling to question paying customers about risks. Nosocomial (caused by a medical procedure) and unexplained cases present hospital administrators with a potential conflict of interest. HIV cases that cannot be explained may not be investigated as the hospital may be implicated.

There are also pressures to conform, with hospitals/doctors believing that the reports are just something that they have to routinely send to the SACS. Characteristically, the sexual route is the default category for adults. Despite these reports, counsellors and doctors in virtually every hospital acknowledge that they have come across cases for which they could not explain the route of transmission, or for which they suspected nosocomial transmission.
In the hospital setting, collecting information from patients about risks for HIV is subject to some practical difficulties. Crowded wards limit privacy, and illness weakens memories of exposure many years in the past. Doctors and counsellors understand that they are to identify and report the mode of transmission rather than risk exposures, little appreciating the speculative leap involved in going from information on risks to saying a specific risk was the source of someone’s HIV infection.

Further, they believe that since the sexual route is already firmly established as responsible for the spread of HIV in the country, they do not need to question the clients much. Even when reported sexual behaviour provides no chance for sexual exposure to HIV, doctors may nevertheless attribute cases to the sexual route on the basis of suspected but unreported behaviours. Moral judgements compound the bias towards attribution of cases to the sexual route.

Finally, they underestimate the ability of HIV to survive outside the body and to transmit through parenteral (intravenous or intramuscular injection) exposures, which undermines their attention to parenteral risks. Therefore, while doctors and counsellors often ask about blood transfusions, they ask less often about major or minor surgery or other blood exposures, particularly if a sexual risk has been reported or suspected. NACO provides no clear guidelines about when to report blood exposures. Blood exposures (other than IDU and blood transfusion) are neither included in the formats nor the training given to doctors. Even AIDS cases with blood transfusions and lower-risk heterosexual exposures may be allocated to the category ‘sexual route’.

Reporting from surveillance data

Current practices to compile and report data from AIDS case surveillance are confusing. Specifically, NACO’s national summaries do not include a category for ‘syringes and needles’, as in hospital reporting forms, but instead have a category for ‘injection drug user’. The cases reported in that category at the national level apparently come from the category ‘syringes and needles’ in hospital reporting forms, which includes cases attributed to medical injections.

Why should non-sexual routes be considered?

First, none of the leading AIDS agencies have been able to demonstrate differences in sexual risks that could explain how heterosexual HIV transmission could be so much faster in countries with generalised epidemics versus countries with concentrated epidemics. In the US and Western Europe, despite high-risk heterosexual behaviour on the part of many IDUs and bisexual males, heterosexual transmission has not been sufficient in itself to sustain stable numbers of infections, much less drive epidemic growth.

Second, in India, the distribution of HIV infections across states and communities does not fit the assumption that sexual transmission explains most infections. How do we explain why people in many rural districts in the south are 10 times more likely to be infected with HIV than urban or rural residents in many northern states? For example, a 2003 random sample survey in Bagalkot district, Karnataka, found 6.2% of agricultural labourers -- equal percentages for both men and women -- HIV-positive, compared to 2.9% for all adults in the district (10). Notably, a 2004 survey of sex workers in Chennai found only 4.0% to be HIV-positive, less than among agricultural labourers in Karnataka (11) .

Third, we cannot assume that all HIV infections in high-risk groups are from personal risk behaviours. Why do sex workers in different parts of the country have such a variance in HIV prevalence? Why are sex workers many times more likely to be HIV-positive if they work in Mumbai or Goa than in Chennai or Kolkata? On the other hand, there is a lot of evidence to show that sex workers in developing countries are exposed to unsterile blood exposures and receive injections to prevent and treat STDs, which may account for many of their HIV infections. Hence, the fact that many sex workers are infected in some -- but not all -- communities says nothing about the relative importance of sexual and blood exposures in India’s HIV epidemic.

How significant could the ‘blood’ routes be?

There is much evidence for HIV transmission through blood exposures in India. An incidence study done in Pune from 1993 to 2000 found that STD clinic attendees who received medical injections or tattoos were more likely to become HIV infected (12). A nation-wide study on injection practices conducted by the All India Institute of Medical Sciences in 2003 reported that 23.5% of medical injections reused unsterile or unreliably sterile syringes or needles. Injections were therefore an important risk in transmitting HIV (13). The WHO, in fact, estimates that unsterile medical injections accounted for 24% of HIV transmissions in India in 2000 (14).

Hundreds of HIV infections in India have been reported in children with HIV-negative mothers, and in men and women with no reported sexual exposure to HIV (15) including outbreaks in Pune in 1989 (16) and Mumbai in 1996 (17).

With insufficient attention paid to the nosocomial routes of HIV transmission, the contribution of these routes to HIV prevalence in the country becomes difficult to ascertain. While the evidence points to this contribution being significant, it is surprising that surveillance systems record not a single case in the country through unsafe medical care (except blood transfusions).

Underlying vulnerabilities to the spread of HIV

Acknowledging the contribution of unsafe blood exposures to India’s epidemic should in no way detract from the efforts being made to stop HIV transmission through unsafe sex and through drug injecting behaviour. It only means we need to be aware of all the routes of HIV transmission so that we can work to block all of them.

We are far from understanding the relative strengths of the various forces driving India’s epidemic. There are, in fact, various factors of vulnerability which we need to consider. While poverty has been assumed to be a factor creating vulnerability to HIV, the recent NFHS-3 data specifically states that HIV cannot be equated with poverty (18). It is the fourth quintile of the wealth index that has the highest HIV prevalence -- the category that will approach healthcare, but of the kind most likely to have unsafe injections. We therefore need to avoid direct correlations between poverty and the spread of HIV. While there seem to be links between poverty and HIV transmission among specific marginalised groups, this may not be the case for the general population. That is probably why the more developed states in the country have greater HIV prevalence. Also, HIV in some African countries seems related more to wealth than to poverty.

Migration is another factor that is seen as creating vulnerability to HIV. This could be because of the status of migrants, separation from spouses, and poor access to safe healthcare. Again, NFHS-3 data refutes this assumption.

Human trafficking, powerlessness of marginalised groups, drug use and trafficking, gender disparities and discrimination, illiteracy, and lack of political will could all create underlying vulnerabilities to the spread of HIV.

There are many groups that are considered vulnerable to HIV because of their sexual behaviour -- women in sex work, men who have sex with men, truckers, etc. These are the very groups that have additional exposure to unsafe blood. For example, malaria lancets are reused on sex workers, truck drivers are routinely injected with penicillin as a prophylaxis against STD, the same needles are reused for tattooing sex workers while they wait in long queues at festivals, and due to their stigmatised status, sex workers cannot negotiate for safe healthcare even when they are aware that the doctor may be reusing equipment.

Why underplay non-sexual routes of transmission?

There are various reasons why the contribution of unsafe healthcare to the spread of HIV is ignored. One, spouted by senior AIDS experts, is that people are confused with more than one message, and giving information about safe healthcare might detract from safe sex messages. This ignores the fact that people are bombarded with all sorts of messages by the media and are definitely able to absorb more than one message on HIV. A parallel concern is that a broader focus of HIV messages (read: a complete message) might take resources away from prevention messages on safe sex. Another concern is that people may stay away from healthcare, especially immunisation programmes, if they are scared of getting infected with HIV. This argument leaves people with a choice between unsafe healthcare and no healthcare. Denying people the information that unsafe healthcare can lead to HIV infection violates their right to safe healthcare -- by denying them accurate information on which they can make informed choices.

The insufficient attention given to ensuring safe healthcare is excused by arguments that there is no evidence linking healthcare to HIV infection. This is strange, considering that everyone associated with healthcare in India acknowledges that unsterile practices are common in both the public and private sectors. There is adequate evidence linking unsafe healthcare with HIV infections. India’s increasing reliance on an unregulated private healthcare sector is likely to strengthen this link.

One big hurdle in ensuring the absolute safety of healthcare services is the low perception of risks associated with unsafe blood exposures. This is due to misinformation perpetuated among medical professionals and the general public alike on the survival of HIV outside the human body and transmission efficiency through injections and invasive procedures.

Difficulty in getting attention to the potential of blood transmission is also due to convenience. The sexual and drug injecting routes place the full responsibility on the HIV-positive person, or someone close to him or her, making it to a large extent a moral issue. Addressing the blood exposure route increases the accountability of the government and international agencies. Recognising the contribution of unsafe healthcare to HIV epidemics would mean acknowledging State responsibility for inadequacies in the systems, and revamping them accordingly.  

It is not that the risks of unsafe healthcare to HIV transmission have gone unrecognised.Soon after AIDS was recognised as resulting from a bloodborne virus in the early-1980s, healthcare managers in developed countries, responding to public pressure, cleaned up healthcare systems to protect patients and staff. Even now, UN agencies advise their employees to carry their own syringes and needles with them when they are sent out to work in developing countries.

While in the developed world iatrogenesis as a cause of HIV transmission is completely unacceptable, the same standards are not followed for poorer countries. The inadequate efforts made by international agencies and donors as well as national governments of developing countries themselves to ensure safe healthcare as a fundamental right suggest a “tacit, widespread acceptance of a two-tiered health system: healthcare must observe the highest standards in wealthy countries, but not necessarily in poorer countries” (18). 

Suggestions for improving data on routes of transmission

Despite its limitations, experts and the general public will continue to look first to AIDS case surveillance for information about the contribution of various risks to India’s HIV epidemic. NACO needs to urgently review the AIDS case surveillance system to identify weaknesses and take the necessary corrective measures.

AIDS case surveillance should collect and report objective information on risks rather than subjective opinions on suspected routes of transmission. NACO forms should ask reporting hospitals to assign AIDS cases to ‘risk exposure’ categories rather than to the current ‘risk/transmission’ categories. Forms should present a hierarchy of categories, from higher to lower risk. NACO should provide questionnaires and training to guide doctors and counsellors to ask the necessary questions and to know what evidence is required to classify each case in one category or another.

The current sexual category should be divided into categories for people with specific risks, one high-risk category for MSM and another for high-risk heterosexual exposures (those who have paid or received money for sex, and those who have an identified HIV-positive heterosexual partner). Lower-risk heterosexual behaviour (any non-commercial sexual partners with unknown HIV status) should be recognised in a separate exposure category.

To improve the sensitivity of AIDS case surveillance, formats should be revised to include all the relevant risks. Specifically, the current category for syringes and needles should be renamed ‘skin-piercing exposures’ so as to explicitly recognise all potentially risky invasive procedures in healthcare and cosmetic services (except for receipt of blood and blood products, which remains a separate category). To identify these risks, doctors or counsellors should ask if people have received injections, tattoos, dental care, or other skin-piercing procedures with equipment that may have been reused without sterilisation.

NACO should provide doctors and counsellors with accurate information about the survival of HIV outside the body and about HIV transmission efficiency through skin-piercing procedures. Because invasive procedures are common, HIV prevalence in people with these exposures would not be much greater than in the general population. Hence, this exposure category is low in the hierarchy. AIDS cases with both skin-piercing exposures and another risk ranked higher in the hierarchy would be reported according to the higher-ranked risk.


Currently, AIDS case surveillance collects information from inpatients, which has its difficulties. To improve the reliability of information on risks, surveillance may be extended to outpatients receiving antiretroviral treatment, because health staff have an opportunity to build a rapport with them before asking about risks. It is believed that NACO is currently trying this out and has commissioned a pilot study to gauge the effectiveness of ARV centres as sources of information on routes of HIV transmission.

The enforcement of standard precautions in western countries means that HIV transmission through blood exposures such as medical injections and tattooing is so rare that it can be ignored. Because standard precautions are not standard in India, there cannot be a similar logic for ignoring invasive medical and cosmetic procedures.

When people with HIV infections have had multiple possible exposures to HIV, several research designs are available to estimate percentages of HIV infections from each route. In the ultimate analysis, the need for efficient, objective surveillance systems is imperative to understand the contribution of various factors driving HIV epidemics and the risk exposures that people have in order to plan interventions that reduce people’s risks and contain the epidemic.

(Mariette Correa is an independent consultant who has been involved in HIV/AIDS programming for NGOs in Goa and South Asia. This article is based mainly on a study conducted in 2005 to assess the reliability of AIDS case reporting in determining routes of HIV transmission: ‘Routes of HIV transmission in India: Assessing the reliability of information from AIDS case surveillance’ byCorrea Mariette and Gisselquist David.International Journal of STD & AIDS2006; 17: 731-735)


  1. Gisselquist D and Correa M. ‘How much does heterosexual commercial sex contribute to India’s HIV epidemic?’ International Journal of STD & AIDS 2006; 17: 736-742
  2. NACO. Facts and figures: HIV estimates -- 2003. Available at
  3. World Bank. Project Performance Assessment Report: India: National AIDS Control Project (Credit No 2350). Report No 26224. Washington DC: World Bank, 2003. p 2
  4. Specialist’s Training and Reference Module -- NACO, Ministry of Health and Family Welfare, not dated
  5. NACO: Facts and figures: Monthly updates on AIDS (July 31, 2005)
  6. Centers for Disease Control and Prevention (CDC). HIV/AIDS Surveillance Report, 2004. Vol 16. Atlanta: CDC, 2005.  Available at (accessed January 18, 2006). Pp 32, 43
  7. Health Canada. HIV and AIDS in Canada: surveillance report to June 30, 2005. Ottawa: Health Canada, 2003. Available at (accessed January 18, 2006). p 32
  8. National Centre in HIV Epidemiology and Clinical Research. HIV/AIDS, Viral Hepatitis and Sexually Transmissible Infections in Australia: Annual Surveillance Report 2003. Sydney: National Centre in HIV Epidemiology and Clinical Research, 2003 Available at (accessed January 18, 2006). p 33
  9. European Centre for the Epidemiological Monitoring of AIDS. HIV/AIDS Surveillance in Europe. End-year report 2004, No 71. Available at (accessed January 16, 2006). Pp 21, 32-33
  10. India-Canada Collaborative HIV/AIDS Project (ICHAP). Community-based HIV prevalence study in ICHAP demonstration project area, key findings. Bangalore: ICHAP, 2004. This study includes annexures, including annexure 1: detailed tables
  11. Tamil Nadu State AIDS Control Society. Activities of TANSACS. Chennai: TANSACS, 2005
  12. Mehendale S M, Rodrigues J J, Brookmeyer R S, et al. Incidence and predictors of human immunodeficiency virus type 1 seroconversion in patients attending sexually transmitted disease clinics in India. Journal of Infectious Diseases 1995; 172: 1486-1491
  13. Arora N K,  Mathew T, Devi S R, et al. Assessment of Injection Practices in India (2002-03): An InCLEN Program Evaluation Network Study. New Delhi: All India Institute of Medical Sciences, 2006. Available at
  14. Hauri A J, Armstrong G L, Hutin Y J F. ’The global burden of disease attributable to contaminated injections given in healthcare settings’. International Journal of STD & AIDS 2004; 15: 7-16
  15. Singhal T. ‘Burden of HIV in India due to unsafe injections and blood transfusions’. MSc thesis submitted to University of London, 2002
  16. Banerjee K, Rodrigues J, Israel Z, Kulkarni S, Thakar M. ‘Outbreak of HIV seropositivity among commercial plasma donors in Pune’, India. Lancet 1989; ii: 166
  17. Christiansen C B, Nielsen C, Machucca R. ‘Cluster of HIV-1 infection among children in Indian Hospital in Bombay’. Informal report to WHO, September 1998. Department of Virology, Statens Serum Institut, Copenhagen, Denmark
  18. National Family Health Survey, India. NFHS-3. National Reports.

InfoChange News & Features, January 2008