According to recent estimates of India's National AIDS Control Organization, 0.36% of the country's population, or between 2.1 and 3 million people, are currently infected with HIV . Prevalence varies greatly across states, with four: Andhra Pradesh, Karnataka, Maharashtra, and Tamil Nadu accounting for the majority (63%) of cases . In Andhra Pradesh, prevalence among antenatal clinic attendees between 2003 and 2006, although declining, is higher than among the other states. It has the sixth highest prevalence of HIV (7.3%) among female sex workers (FSW) , but within its East Godavari district, the prevalence among FSW is over 25% .
With the exception of the northeastern states, most HIV in India is attributed to heterosexual transmission , and considerable prevention effort has focused on FSW. Some models have suggested that sex worker interventions in India would end the epidemic . To reduce HIV risk in any population effectively, however, it is important to understand its sources and implement interventions to address them . HIV prevention interventions for FSW in India have frequently focused on health education, screening and treatment for sexually transmitted infections (STI), and condom promotion and distribution [6–8]. There is, however, increased acknowledgement that they must go beyond such measures to address the multiple contextual, historical, and structural factors that perpetuate the vulnerability of FSW [9–15]. Structural interventions represent a group of interventions particularly suited to this purpose [16,17]. Focusing on the structural context that shapes risk, they may take many different forms, including those that address the unequal distribution of power that can limit the accessibility of health-related resources and increase vulnerability for marginalized populations . Community mobilization represents a particularly important accessibility intervention strategy for addressing HIV risk among FSW, to the extent that it seeks to alter power relations between marginalized and dominant groups and thereby address their vulnerability .
In spite of the growing recognition of the extent to which power impacts on HIV risk among FSW, and the potential for community mobilization interventions (CMI) to reduce risk, there are relatively few empirical analyses of either. The most well known CMI for FSW, in Kolkata's red light Sonagachi district, reported reductions in HIV prevalence among brothel-based FSW [17–20]; however, its replicability remains unclear [19,21,22].
In part, the success of CMI may depend on contextual factors that facilitate or challenge mobilization efforts [23,24]. Also, the concept of community mobilization has been used interchangeably with community participation, empowerment, collectivization, and community development, all of which are meant to convey some degree of engagement with marginalized communities, but each of which may also signify distinct kinds of activities  with varying degrees of effectiveness in challenging power relations. Such activities range from building individual negotiation skills [26–28], to group formation to build or strengthen collective identity or collective efficacy [14,21,29], to advocacy and collective action by FSW on behalf of their peers [21,29,30].
In this paper, we draw from a structural interventions framework that focuses on the structural context of HIV risk to analyse the association between power and condom use practices among FSW in Andhra Pradesh, India, and to analyse the extent to which exposure to a local CMI affects these associations. In particular, we focus on three distinct types of power: collective power; control over sex work; and economic power. We further define collective power according to three dimensions: collective identity (identification with other FSW); collective efficacy (belief that FSW can work together for change); and collective agency (action with other FSW on behalf of FSW). The CMI we study emphasizes the mobilization of FSW through their collective organizing and directs this newly formed collective power to challenge structural factors that place FSW at risk of HIV. It has been implemented as part of the Bill and Melinda Gates Foundation's Avahan initiative .
The data analysed for this paper were collected as part of a larger study, Project Parivartan, of the implementation and impact of a CMI. They come from a cross-sectional survey of FSW conducted from April to June 2006 in the Rajahmundry area of the East Godavari district in Andhra Pradesh. A sample of 812 FSW was recruited using a modified chain referral technique known as respondent-driven sampling (RDS), which was designed to recruit hidden populations [32,33]. Initially, five seeds representing different groups of FSW completed the survey and distributed coupons to up to three members of their FSW network who met the study's eligibility criteria (recruits). Recruits were then given the opportunity to distribute coupons to up to three other FSW in their network. To ensure that the final sample is independent of the initial seeds, it is recommended that at least six recruitment waves are completed and a sample size of at least 500 is generated [34,35]. We completed 22 recruitment waves to generate our sample of 812. All respondents received modest compensation for participation in the survey and for successfully recruiting other FSW into the study.
To be eligible for participation in the face-to-face interview, participants had to report being 18 years of age or older and exchanging sex for money at least once in the year before the interview. Interviews were conducted in the local language (Telugu) by trained interviewers, after ensuring participants' free and informed consent. The survey lasted 90–120 minutes and covered such topics as demographics, socioeconomic situation, sex work history, views on sex worker solidarity and efficacy, condom use with different partner types, participation in FSW and other organizations and in events and activities on behalf of FSW, and exposure to a local CMI. The research protocol was approved by the Yale University Human Investigation Committee, the Duke University School of Medicine Institutional Review Board and the VHS-YRG Care Medical Centre Institutional Review Board in Chennai.
Intervention and study site
The CMI is being implemented in Rajahmundry by an non-governmental organization (NGO) based in India with headquarters in the United States. Rajahmundry is a semi-urban town of approximately 400 000 in the largely rural, economically prosperous district of East Godavari, in Andhra Pradesh. HIV prevalence in the district is among the highest in the state , which is known as a site for sex work . According to the NGO, there are approximately 1500 FSW in its coverage area, which includes eight of the district's 59 revenue ‘mandals’. Like FSW in the state , those in East Godavari are from lower caste groups, economically vulnerable, non-literate, single, separated or divorced , and work in a range of venues including brothels, the street and home [38,40].
The NGO began its work in October 2004, identifying social change agents (SCA) among local FSW to serve as both peer educators and community organizers. SCA, of which there are approximately 70, engage in traditional intervention activities, distributing condoms and promoting condom use among FSW, bringing FSW to the NGO-sponsored STI clinics and drop in centres for services, and providing general health education. They also advocate on behalf of FSW and promote the idea that sex work is work, and that, like other workers, FSW will gain power by joining together. They further encourage the organizing of community-based organizations composed of and led by FSW, of which there were 12 at the time of this writing.
Project Parivartan's research staff worked independently from the intervention. The survey was one of several research activities conducted to assess intervention impact (other activities included field observation, and formal and informal key informant interviews). At monthly NGO meetings, Parivartan staff provided feedback on data collection activities with implications for the intervention, and periodically, they gave presentations on major findings at meetings with both NGO staff and FSW.
We sought to measure three distinct dimensions of collective power. High collective identity was defined as agreeing a little or a lot with the statement ‘you feel a strong sense of unity with sex workers you do not know.’ Those who disagreed a little or a lot were considered to have low collective identity. Those who indicated in response to the statement ‘if there were a problem that affected all or most of the sex work community’ that all or most sex workers would work together to deal with it, were considered to have high collective efficacy compared with those who responded that no one, or some, would work together. Finally, collective agency was defined as attending, in the past 6 months, a public event such as a rally or gathering where the respondent could be identified as a FSW, or going to the police in the past 6 months to speak on behalf of the rights of FSW.
Control over work
We assessed two different aspects of control over work: deciding the type of sex with a client and deciding the amount of money to charge a client. Control was indicated by deciding sometimes, usually or always, as compared with never or rarely.
We assessed economic power in two ways: through a measure with values ranging from 0 to 3 based on whether a respondent indicated that she currently lived in a home with electricity, running water, or bathrooms/toilet; and through a measure of economic independence – whether the respondent was the sole provider for herself and her family, or depended on others for economic support.
In all of our analyses we controlled for age, marital status (currently married versus not), type of sex work (street/lodge-based or other), and literacy (the ability to read and/or write).
For the bivariate analyses, we created three categories of programme exposure: no programme exposure if respondents indicated they had never heard of the CMI; receptive exposure if they had heard of the intervention, received materials relating to it, or spoken to a peer outreach worker or SCA, or another intervention representative; and active exposure if they worked with the intervention, were a member of a community-based organization started by the intervention, or had used one of the services of the intervention (e.g. drop-in centre; STI clinic). For multivariate analyses we created a dichotomous measure of programme exposure (none versus any) in order to test interaction effects of programme exposure with collective power.
Consistent condom use
Consistent condom use with both regular and occasional clients is the primary outcome measuring HIV-related risk behaviour. The respondent was considered to use condoms consistently if, in the past 7 days, she used condoms the last time she had sex with a client (regular and/or occasional) and she always used condoms with clients (regular and/or occasional).
Using a chi-square test, we compared point estimates that had been weighted to take into account recruiters' network size with unweighted estimates. Because there was very little difference between the two, and because in multivariate analysis, regression is fairly immune to weights and unweighted data are preferred , we used unweighted data for all analyses. We conducted preliminary analysis of the data to assess variable distributions, means and standard deviations, and bivariate analysis to determine the relationship between indicators of power and both programme exposure and consistent condom use. To identify possible covariates associated with consistent condom use, we conducted a series of bivariate and multivariate analyses in which all possible covariates that had an association at the P < 0.10 level were considered for inclusion in the final analysis. Then we assessed these for multicollinearity. If two covariates were highly related, we chose the covariate with the largest relationship with the outcome. Through this process we eliminated: caste, religion, number of children, years since first conducting sex work, sole reliance on sex work for income, number of people supported by respondent's income, experience of violence, and current debt (all P > 0.10).
With logistic regression, we assessed the relationship between types of power and programme exposure and consistent condom use. The multivariate analyses used consistent condom use as the outcome and included four cumulative models, all of which controlled for age, marital status, type of sex work, and literacy. The first two models assessed the relationship between different types of FSW power and consistent condom use. Model 1 included measures of collective power and model 2 added measures of control over work and economic power. Model 3 added programme exposure, and the final model (not shown) added interactions between programme exposure and the three measures of collective power. The data were analysed using SPSS 13.0 for Mac OS X.
A total of 803 (98.9%) respondents indicated that they had exchanged sex with an occasional or regular client in the past 7 days; 700 (86%) with an occasional and 636 (78.3%) with a regular client. Table 1 describes the sample.
Table 2 presents results from bivariate analyses of the relationship between different levels of programme exposure and collective power. There is a clear, positive association between the dimensions of collective power and programme exposure, and the association is stronger as the level of exposure increases. Table 3 shows a clear positive association between collective power and consistent condom use, with the largest effect sizes for collective agency.
Table 4 presents results of the multivariate logistic regression models for consistent condom use. Model 1 shows the association for different dimensions of collective power. Collective agency and collective identity are significant predictors of consistent condom use, with collective agency having the strongest effects. Although collective efficacy is not significantly associated with consistent condom use, there is a moderate effect size in the predicted direction.
Although our collective power measures address one aspect of power's association with HIV risk in FSW, power is also signified by the degree of FSW control over sex work and economic resources. Model 2 demonstrates that FSW who have greater control over sex work report much higher consistent condom use, with women who have control over the type of sex 1.75 times more likely and women who decide how much to charge 1.67 times more likely than women with no control to use condoms consistently. Similarly, FSW who are economically better off report higher consistent condom use, and women who are economically dependent are half as likely to engage in consistent condom use. Interestingly, when all types of power are considered together, the association between collective power and consistent condom use is not significant, although effect sizes for collective identity and collective agency remain sizeable (P < 0.10).
Next, to assess the association between exposure to the intervention and consistent condom use, we added the dichotomous measure of programme exposure to the model. Programme exposure is highly associated with consistent condom use, with an adjusted odds ratio for FSW actively exposed to the CMI over twice that of non-exposed FSW. Although the effect sizes for the collective power measures are substantially reduced, the association between consistent condom use and three of the other four measures of control over work and economic power remains strong.
Finally, to explore further the associations among collective power and programme exposure, given that a priority of the CMI is to promote collective power, we accounted for possible interaction effects. We added interaction terms of the three collective power measures with programme exposure one by one to the final model. The interaction terms were not significant for programme exposure with collective identity (chi-square 1.22, P = 0.27) and collective efficacy (chi-square 0.43, P = 0.51), but were significant for programme exposure and collective agency (chi-square 6.62, P = 0.01). Among respondents who reported both programme exposure and high levels of collective agency, the odds ratio of engaging in consistent condom use was 2.5 times that of other FSW.
We have applied a structural interventions framework to understand condom use among a sample of FSW in Andhra Pradesh, India. Accordingly, power inequalities that marginalize FSW and limit their control over their lives are one of the contextual sources of their risk of HIV, suggesting a need for interventions that alter power relations. Collective organizing and action have been tools through which marginalized groups, whether factory workers, gay men, or welfare recipients, have sought and achieved considerable power [42,43]. Increasingly, a similar strategy is being used to prevent HIV among FSW, and the degree of collective power may be an important factor associated with HIV-related risk among FSW. Collective power has, however, been defined in many ways and there is little empirical work distinguishing the differential association of these dimensions with HIV prevention behaviours. Also, collective power is only one of a number of indicators of power among FSW, and even less attention has been paid to issues of control over work and economic power. Furthermore, it is not clear how CMI are associated with these types of power.
Our analyses have shed light on these issues. They demonstrate that two types of power, control over work and economic independence, are significantly associated with consistent condom use in this sample of FSW. FSW with little control over their work and greater economic dependence use condoms less consistently than those with more control over work and those with economic independence, even when they report high collective power and programme exposure. It is likely that FSW who do not feel they can control the type of sex they have with a client will be less able to insist on condom use in the exchange. Those FSW who depend on others for their economic support may feel pressure to earn as much as they can. Under such circumstances, an offer of more money for unprotected sex may be difficult to turn down (65.5% of respondents reported being offered more money for unprotected sex within the last 6 months).
Although FSW undoubtedly benefit in many ways from collective power, our multivariate analyses suggest that these benefits extend to condom use behaviour most, among those who have participated in collective action. The interaction effect, however, suggests that collective agency translates into consistent condom use in the presence of a CMI. At the same time, collective power alone does not necessarily provide greater control over work or reduce economic dependence.
More must be done to address the ways that the conditions of sex work impact on HIV risk. Changes in the conditions of and control over work might take place in a variety of ways, including by placing trash bins near the workplace for the easy discard of condoms, reducing the number of or eliminating middlemen or brokers involved in the sex exchange, creating more protected locations for conducting sex work, or implementing 100% condom use policies. Similarly, activities that promote economic independence and enhance access to economic resources could be useful tools for HIV prevention and may range from the promotion of ration and smart cards, to different community banking and microinsurance schemes, to increases in and the standardization of the price for sex.
Our study is limited by its cross-sectional design. It is difficult to know the nature of the association between collective power and programme exposure. It may be that FSW with high levels of collective agency are drawn to the intervention rather than that the intervention builds collective agency among FSW. Even if this is the case, however, given the sizeable interaction between collective agency and consistent condom use, it seems reasonable to suggest that the intervention helps channel this collective power in ways that promote consistent condom use and that might not happen without the CMI.
Our analysis is also limited by an outcome measure, self-reported condom use, which is subject to social desirability bias. We have sought to minimize this potential bias, however, by using a relatively strict definition of consistent condom use that combines responses to four questions. We compared responses to each of these questions separately with our combined variable and found that each resulted in rates of condom use that were as much as double the rates we found using our measure (analysis not shown).
Similarly, our analysis is limited by our measures of collective power, as we have sought to assess a community-level concept, through data collected at the individual level. Furthermore, we have relied on single variables to measure these complex concepts. Their relatively strong association with our outcome measures suggests their predictive validity, but additional research is necessary to assess the stability of these measures and their predictive power in different samples and contexts.
Finally, it is possible that our collective power variables are surrogate measures of programme exposure. We have sought to minimize this by identifying indicators that do not rely on specific language used by the CMI. When we compared the responses of SCA (with greatest programme exposure) and others who reported active programme exposure, we found no significant differences in their degree of collective efficacy or collective identity (analysis not show), suggesting that our measures of collective power and programme exposure are distinct from one another.
Our analysis supports a structural interventions framework and suggests that power is an important factor influencing HIV-related risk behaviour. Among FSW, different forms of power are associated with consistent condom use with clients, and a CMI that seeks to alter the balance of power appears to be associated with lower risk behaviour. In particular, it is important to promote greater control over sex work and increase FSW access to and control over economic resources. Additional longitudinal and qualitative research is critical for further revealing the relationships among power, community mobilization, and HIV risk among FSW.
The authors wish to thank all of the women who participated in the cross-sectional survey as well as the Parivartan team members who helped design and implement the survey.
Sponsorship: Support for this research was provided by the Bill and Melinda Gates Foundation (K. Blankenship, principal investigator).
The views expressed herein are those of the authors and do not necessarily reflect the official policy or position of the Bill and Melinda Gates Foundation.
Conflicts of interest: None.
1. National Institute of Health and Family Welfare and National AIDS Control Organization. Annual HIV sentinel surveillance country report 2006. Delhi: National Institute of Health and Family Welfare; 2007.
2. Indian Council of Medical Research and Family Health International. National interim summary report – India, integrated behavioral and biological assessment (IBBA), round 1. New Delhi: Indian Council of Medical Research and Family Health International; October 2007.
3. Venkataramana CBS, Sarada PV. Extent and speed of spread of HIV infection in India through the commercial sex networks: a perspective. Trop Med Int Health 2001; 6:1040–1061.
4. Nagelkerke NJD, Jha P, de Vlas SJ, Korenromp EL, Moses S, Blanchard JF, et al. Modelling HIV/AIDS epidemics in Botswana and India: impact of interventions to prevent transmission. Bull WHO 2002; 80:89–96.
5. Blankenship KM, Friedman SR, Dworkin S, Mantell JE. Structural interventions: concepts, challenges, and opportunities for research. J Urban Health 2006; 83:59–72.
6. Jha P, Nagelkerke NJD, Ngugi EN, Prasada Rao JVR, Willbond B, Moses S, et al. Reducing HIV transmission in developing countries. Science 2001; 292:224–225.
7. Bhave G, Lindan CP, Hudes ES, Desai S, Wagle U, Tripathi SP, et al. Impact of an intervention on HIV, sexually transmitted diseases, and condom use among sex workers in Bombay, India. AIDS 1995; 9(Suppl. 1):S21–S30.
8. Singh YN, Malaviya AN. Experience of HIV prevention interventions among female sex workers in Delhi, India. Int J STD AIDS 1994; 5:56–57.
9. Chattopadhyay A, McKaig RG. Social development of commercial sex workers in India: An essential step in HIV/AIDS prevention. AIDS Patient Care STD 2004; 18:159–168.
10. Blanchard JF, O'Neil J, Remesh BM, Parinita Bhattacharjee OT, Moses S. Understanding the social and cultural contexts of female sex workers in Karnataka, India: implications for prevention of HIV infection. J Infect Dis 2005; 191(Suppl. 1):S139–S146.
11. O'Neil JO, Orchard T, Swarankar RC, Blanchard JF, Gurav K, Moses S. Dhandha, dharma and disease: traditional sex work and HIV/AIDS in rural India. Soc Sci Med 2004; 59:851–860.
12. Go VF, Sethulakshmi CJ, Bentley ME, Sivaram S, Srikrishnan AK, Solomon S, et al. When HIV-prevention messages and gender norms clash: the impact of domestic violence on women's HIV risk in slums of Chennai, India. AIDS Behav 2003; 7:263–272.
13. Dandona R, Dandona L, Gutierrez JP, Kumar GA, McPherson S, Samuels F, et al
. High risk of HIV in non-brothel based female sex workers in India. BMC Public Health
2005; Available at: www.biomedcentral.com/1471-2458/5/87
. Accessed: September 2008.
14. Halli SS, Ramesh BM, O'Neil J, Moses S, Blanchard JF. The role of collectives in STI and HIV/AIDS prevention among female sex workers in Karnataka, India. AIDS Care 2006; 18:739–749.
15. Rekart ML. Sex-work harm reduction. Lancet 2005; 366:2123–2134.
16. Blankenship KM, Bray SJ, Merson MH. Structural interventions in public health. AIDS 2000; 14(Suppl. 1):S11–S21.
17. Parker RG, Easton D, Klein CH. Structural barriers and facilitators in HIV prevention: a review of international research. AIDS 2000; 14(Suppl. 1):S22–S32.
18. Jana S, Chakraborty AK, Das A, Khodakevich L, Chakraborty MS, Pal NK. Community based survey of STD/HIV infection among commercial sex-workers in Calcutta (India). Part II. Sexual behaviour, knowledge and attitude towards STD. J Commun Dis 1994; 26:168–171.
19. Jana S, Basu I, Rotheram-Borus MJ, Newman PA. The Sonagachi Project: a sustainable community intervention program. AIDS Educ Prev 2004; 16:405–414.
20. Cohen J. Sonagachi sex workers stymie HIV. Science 2004:304–506.
21. Basu I, Jana S, Rotheram-Borus MJ, Swendeman D, Lee S, Newman PA, et al. HIV prevention among sex workers in India. J Acquir Immune Defic Syndr 2004; 36:845–852.
22. Asthana S, Oostvogels R. Community participation in HIV prevention: problems and prospects for community-based strategies among female sex workers in Madras. Soc Sci Med 1996; 43:133–148.
23. Evans C, Lambert H. Implementing community interventions for HIV prevention: insights from project ethnography. Soc Sci Med 2008; 66:467–478.
24. Cornish F, Ghosh R. The necessary contradictions of ‘community-led’ health promotion: a case study of HIV prevention in an Indian red light district. Soc Sci Med 2007; 64:496–507.
25. Cornish F. Empowerment to participate: a case study of participation by Indian sex workers in HIV prevention. J Commun Appl Soc Psychol 2006; 16:301–315.
26. Ford N, Koetsawang S. A pragmatic intervention to promote condom use by female sex workers in Thailand. Bull WHO 1999; 77:888–894.
27. Busza J, Baker S. Protection and participation: an interactive programme introducing the female condom to migrant sex workers in Cambodia. AIDS Care 2004; 16:507–518.
28. Patterson TL, Semple SJ, Fraga M, Bucardo J, Davila-Fraga W, Strathdee SA. An HIV-Prevention intervention for sex workers in Tijuana, Mexico: a pilot study. Hisp J Behav Sci 2005; 27:82–100.
29. Kerrigan D, Moreno L, Rosario S, Gomez B, Jerez H, Barrington C, et al. Environmental–structural interventions to reduce HIV/STI risk among female sex workers in the Dominican Republic. Am J Public Health 2006; 96:120–125.
30. Jayasree AK. Searching for justice for body and self in a coercive environment: sex work in Kerala, India. Reprod Health Matters 2004; 12:58–67.
31. Chandrasekaran P, Dallabetta G, Loo V, Moore J, Mills S, Saidel T, et al. Evaluation design for large scale HIV prevention programs: the case of Avahan, the India AIDS initiative, AIDS 2008; (Suppl. 5): S1–S10.
32. Magnani R, Sabin K, Saidel T, Heckathorn DD. Review of sampling hard-to-reach and hidden populations for HIV surveillance. AIDS 2005; 19(Suppl. 2):S67–S72.
33. Semaan S, Lauby J, Liebman J. Street and network sampling in evaluation studies of HIV risk-reduction interventions. AIDS Rev 2002; 4:213–223.
34. Magnani R, Sabin K, Saidel T, Heckathorn D. Review of sampling hard-to-reach and hidden populations for HIV surveillance. AIDS 2005; 19(Suppl. 2):S67–S72.
35. Salganik MJ, Heckathorn DD. Sampling and estimation in hidden populations using respondent-driven sampling. Sociol Methodol 2004; 34:193–239.
36. Population Foundation of India. Facts, figures and response to HIV/AIDS in Andhra Pradesh. New Delhi: Population Foundation of India; 2005.
37. Nair PM. A report on trafficking of women and children in India, 2002–2003. New Delhi: Institute of Social Sciences; 2004.
38. Dandona R, Dandona L, Kumar GA, Gutierrez JP, McPherson S, Samuels F, et al
. Demography and sex work characteristics of female sex workers in Andhra Pradesh, India. BMC Int Health Human Rights
2006; Available at: http://www.biomedcentral.com/1472-698X/6/5
. Accessed: September 2008.
39. Family Health International/Development Fund for International Development. Presence of sexually transmitted infections and HIV among female sex workers of Kakinada and Peddapuram, Andhra Pradesh, India. Baseline report series. New Delhi: Family Health International/Development Fund for International Development; 2000.
40. Chandrasekaran P, Dallabetta G, Loo V, Rao S, Gayle H, Alexander A. Containing HIV/AIDS in India: the unfinished agenda. Lancet 2006; 6:508–521.
41. Winship C, Radbill L. Sampling weights and regression analysis. Sociol Methods Res 1994; 23:230–257.
42. Quimby E, Friedman SR. Dynamics of black mobilization against AIDS in New York City. Social Problems 1989; 36:403–415.
43. Piven FF. Challenging authority: how ordinary people change America. Lanham, MD: Rowman and Littlefield; 2006.
© 2008 Lippincott Williams & Wilkins, Inc.