Measuring vulnerability and identifying determinants of vulnerability are key to designing interventions for marginalized groups like sex workers. The current study introduces a new approach of measuring vulnerability among female sex workers (FSWs) by adopting a multidimensional poverty measurement framework. A multidimensional vulnerability index was created from four dimensions and 16 indicators using a dual cut-off approach. The study found that 55% of FSWs were multidimensionally vulnerable with 48% of intensity in vulnerability. The overall value of multidimensional vulnerability index was 0.265. FSWs in Maharashtra were most vulnerable (82%). Lack of financial security contributed mostly to FSWs’ vulnerability. Further, compared to less vulnerable FSWs, multidimensionally vulnerable ones were more to engage in behaviors that put them at risk such as inconsistent use of condoms with clients, alcohol consumption, engaging in anal sex with clients and experiencing sexually transmitted infections. Findings suggest that structural, social and financial vulnerabilities of FSWs need to be addressed concurrently.
Citation: Mahapatra B, Bhattacharya R, Atmavilas Y, Saggurti N (2018) Measuring vulnerability among female sex workers in India using a multidimensional framework. PLoS ONE 13(9): e0204055. https://doi.org/10.1371/journal.pone.0204055
Editor: Geng-Feng Fu, Jiangsu Provincial Center for Disease Control and Prevention, CHINA
Received: May 11, 2018; Accepted: August 31, 2018; Published: September 25, 2018
Copyright: © 2018 Mahapatra et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data are available at the following Dataverse URL: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/SBN5PI.
Funding: This paper was written as part of an evaluation grant funded to the Population Council by the Bill & Melinda Gates Foundation (grant # OPP1110266). The donor had no role in study design, data collection, and analysis, decision to publish, or preparation of the manuscript. YA who works with the Bill & Melinda Gates Foundation contributed to this paper on her individual capacity and opinion expressed in the paper does not necessarily reflect the official policy and position of the Bill & Melinda Gates Foundation.
Competing interests: The authors have declared that no competing interests exist.
Vulnerability among female sex workers (FSWs) is a widely researched area [1–5]; however, the definition and use of the term “vulnerability” vary considerably with a clear agreement that HIV vulnerability is different from HIV risk [6, 7]. While HIV risk refers to behaviors or situations that directly increase the probability of HIV transmission, HIV vulnerability refers to the context governing individual’s ability to prevent oneself from getting HIV infection . These vulnerabilities can be in various forms ranging from personal attributes to financial, environmental and cultural norms including stigma and discrimination at various levels. Empirical research has shown that vulnerabilities increase FSWs’ HIV risk-taking behaviors including experience of violence, poor negotiation skills to use condom, inability to refuse clients for sex and lack of access to HIV prevention services [8–11]. Given the HIV epidemic in India is a concentrated one and FSWs remain as one of core source of transmission, the current HIV prevention efforts directed towards FSWs need to continue. However, to make the on-going prevention efforts more successful and sustainable, there is a need to address the different vulnerabilities faced by FSWs and hence, enhancing their ability to adopt safe free behavior without any compulsion or fear. The first step in this direction is identifying vulnerable FSWs and measuring the degree of vulnerability faced by them.
Vulnerability measurement revolves around two frameworks: unidimensional and multidimensional. A review of existing literature suggests that most researchers have adopted a unidimensional framework to identify vulnerability [7, 12–14]. Only two Indian studies have considered the multi-dimensional nature of vulnerability by creating a composite index of vulnerability. One of these studies considered three factors (solicitation in street-based setting, experience of violence and financial debt) to create a composite measure of vulnerability . The other study considered six factors (literacy, duration in sex work, mobility for sex work, an additional source of income other than sex work, currently being in debt, and having children) to create an index of vulnerability. However, both these studies were limited by the number of indicators considered for creating the index, a fact which has been acknowledged by the authors themselves.
While multidimensional approach is more comprehensive than unidimensional approach, the former one needs to be aggregated appropriately to estimate vulnerability accurately. In multidimensional approach, vulnerability can be defined in two ways: union approach (vulnerability in any indicator) and interaction approach (vulnerability in all indicators). While union approach over-estimates vulnerability, interaction approach under-estimates. Therefore, this paper addresses this issue of aggregation by employing a “dual cut-off” approach in a multidimensional framework. Moreover, we estimate the intensity of vulnerability in addition to incidence of vulnerability and estimate contribution of each indicator to the overall vulnerability enabling policymakers to identify points of intervention. This study further examines the effect of vulnerability on FSWs’ HIV-related sexual risk behavior.
Materials and methods
The study was conducted in four southern (Andhra Pradesh, Telangana, Karnataka, Tamil Nadu) and one western (Maharashtra) Indian states. The states of Andhra Pradesh and Telangana were considered as one sampling domain in this study. These states are historically known for high HIV prevalence, and large number of FSWs. The HIV prevalence among FSWs in Andhra Pradesh/Telangana and Karnataka is around 6% each, 7% in Maharashtra, 1% in Tamil Nadu . The living condition of FSWs in these areas was one of the poorest with a large proportion of FSWs being exposed to harsh socio-environmental stress such as lack of proper health care and security. Nearly half the FSWs in Maharashtra, two-thirds in Andhra Pradesh and Telangana, one-fifth in Karnataka and two-fifth in Tamil Nadu depended solely on earnings from sex work to make a living . Moreover, a considerable proportion of FSWs in these states reported experience of general stigma (ranges from 21–42%, national level: 27%) and discrimination at health facility (ranges from 15%-32%, national level: 21%) . Also, a considerable higher proportion of FSWs from these states experienced physical and sexual violence as compared to the national average.
Data and study design
Data were drawn from a cross-sectional survey conducted in 2015 among 4098 FSWs across five Indian states . FSWs who were 18 years or older and had sex in exchange for cash or kind in the month preceding the survey were included in the survey. The study was conducted among 4098 FSWs who were selected using a three-stage sampling procedure to ensure appropriate geographical representation. The detailed sampling process is published elsewhere . All interviews were conducted by trained female investigators with verbal and written skills in the local language of each state. A structured survey schedule was used for collecting data using face-to-face interview techniques.
The institutional review boards (IRBs) of Population Council and Sahara, center for residential care and rehabilitation, reviewed and approved the study procedure and tools. In accordance with the protocol, written consent was obtained from all respondents prior to their participation in the survey. The interviewers read the complete script of the consent form to the respondent and explained if there was any doubt about any aspects of the survey. Participants who could not read and write, the consent process was explained in the presence of a witness (either program staff or fellow sex worker) and verbal consent was taken for such participants. Both the IRBs approved the oral consent in the presence of a witness for respondents who could not read and write. All the interviews were held in a private location specifically hired for the survey, or in a location convenient to the study participants.
Multidimensional vulnerability index (MVI).
- H is the vulnerability headcount ratio and refers to the proportion of FSWs who are multidimensionally vulnerable;
- A is the intensity of vulnerability and refers to the average weighted vulnerability experienced by the multidimensionally vulnerable and computed as a division of censored vulnerability score upon number of people who are multidimensionally vulnerable.
- and MVI denotes the total vulnerability possible which is experienced by the vulnerable FSWs. The value of MVI ranges from 0 to 1; 0 indicating nobody is multidimensional vulnerable and 1 indicating everyone vulnerable in the study population.
Key steps for computing MVI involves selection of dimensions and indicators of vulnerability, defining indicator level vulnerability cutoffs to identify when an FSW is vulnerable in an indicator, choice of weights to ensure different indicators are accounted for their relative importance, and identifying a second-stage cutoff to determine when an FSW to be considered as multidimensionally vulnerable.
Selection of dimensions and indicators and identifying vulnerability cut-offs for indicators.
We used four dimensions and 16 indicators to capture different aspects of vulnerability. Table 1 provides the details of each dimension and indicators as well as the rationale behind considering them as indicators of vulnerability. The four dimensions of vulnerability considered are personal attributes, financial security, social protection and social support and network. Personal attributes aim to capture disempowering attributes or factors that put lives of FSWs at a greater risk than their counterparts and include four indicators. The dimension of financial security contains five indicators reflecting economic situation of FSWs. Similarly, dimension on social protection includes four indicators reflecting FSWs’ access to various social schemes, food security, and awareness of basic rights. The dimension of social support and network has three indicators, each reflecting depth of FSWs’ social network. Lack of social network can exclude FSWs from social gatherings and jeopardize their ability to demand any need. Given that the validity and stability of the index rest upon the intuition behind these indicators and their actual relevance for the study population, all the indicators were thoroughly reviewed, discussed and assessed for their programmatic and scientific relevance. Each of these indicators was defined in a way to generate dichotomous variables with “1” indicating vulnerable condition; else “0” indicating non-vulnerability.
Weighting of indicators.
All dimensions and indicators within each dimension were assigned equal weights. This means each dimension received a weight of 1/4. Weight for an indicator within a dimension was calculated as reciprocal of number of indicators in the dimension multiplied by number of dimensions. For example, if there were five indicators in one dimension, then weight for indicator would be 1/4*1/5 = 0.05. The weights were multiplied with each indicator (coded as 0 and 1) to generated weighted indicator values. The weighted values of all indicators were summed to generate a weighted vulnerability score.
Identifying multidimensionally vulnerable FSWs.
FSWs whose weighted scores were above a certain threshold level k were considered multidimensionally vulnerable. The selection of the threshold value is a normative decision considering the indicators included and how important they are in defining vulnerability [19, 21]. This study has four dimensions, thus having a threshold value of k = 0.25 would qualify as multidimensional vulnerable. However, after a series of robustness test, we choose to have a threshold of 0.33 which means an FSW is multidimensionally vulnerable if she is vulnerable in more than one-third of weighted vulnerability score. We checked the robustness of the level by comparing the state-specific proportion of multidimensionally vulnerable FSWs (Fig 1) and MVI values (Table 2) at different threshold values. If they hold a consistent pattern, they were considered as robust. The consistency was checked using Kendal Tau-b coefficient of rank correlation which compared the ranking of states for MVI estimates at different threshold values . More detailed steps of calculation of multidimensional indexing and its properties can be found elsewhere [19, 20].
HIV-related sexual risk factors
HIV-related sexual risk factors were assessed in terms of inconsistent condom use with clients, alcohol consumption, anal sex in last 12 months, ability to negotiate for condom use with clients and experience (self-reported) of sexually transmitted infections (STIs) in past six months. FSWs were asked separately about the frequency of condom use with occasional and regular clients in past one month with response options of ‘always’, ‘most of the time’, ‘sometimes’ or ‘never’. FSWs who used a condom “always” with both clients were considered consistent condom user; otherwise considered inconsistent user (S1 Appendix). Ability to negotiate for condom use was derived from a question on any instance in past six months when the FSW wanted to use a condom but could not do it with response categories “no” and “yes”. Those responded “yes” were considered to have poor ability to negotiate; else considered to have the ability to negotiate for condom use. To determine the experience of STI symptoms, participants were asked whether they had experienced any of the following symptoms in the six months prior to the survey: ulcers/sores in the genital area, swelling in the groin area, and frequent painful urination. Those responding “yes” to any of these symptoms were classified as having experienced STI symptoms (Yes = 1, No = 0). FSWs’ treatment seeking behavior in last 12 months, engaging in anal sex in last 12 months, alcohol consumption, and current use of any family planning method was assessed using single-item questions. Information on socio-demographic variables like marital status (currently married, never married and formerly married), education (no formal education, formal education), and duration in sex work (continuous) was assessed using single item questions.
Univariate analysis was used to present the percent of FSWs who are vulnerable in different indicators. Bivariate analysis was used to present the levels of different HIV-related sexual risk factors by the multidimensional vulnerability. A series of multiple logistic regression models were fitted to examine the effect of multidimensional vulnerability on different HIV-related sexual risk behaviors. In each of these logistic regressions, multidimensional vulnerability was the key independent variable whereas socio-demographic factors were used as covariates and indicators on HIV-related sexual risk behaviors were dependent variables. Results were presented in the form of percentages, adjusted odds ratios (AOR) and their corresponding 95% confidence interval (CI). All the analyses were performed using STATA 13.1 (StataCorp., TX, USA).
Surveyed FSWs were, on an average, 34 years and practicing sex work for about seven years. A little above two-fifths (42%) of FSWs were having no education, and about three-fifths (62%) were currently married (Data not shown).
Intensity and headcount of multidimensional vulnerability.
More than half (55%) of FSWs were multidimensionally vulnerable (Table 3). The proportion of FSWs multidimensionally vulnerable was highest in Maharashtra (82%) and least in Karnataka (35%). The intensity of vulnerability was 48%, which suggests that multidimensionally FSWs were vulnerable in nearly half of the weighted vulnerable indicators. The intensity of vulnerability was highest in Maharashtra (53%). In the other three states, intensity of vulnerability among vulnerable FSWs was almost similar. The overall MVI was 0.265 with the highest level in Maharashtra (0.436) and least in Karnataka (0.152).
Contribution of different factors to multidimensional vulnerability
The dimension of financial security contributed most to FSWs’ vulnerability (31%), followed by social support and network dimension (26%) (Fig 2). Personal attributes contributed 23% to overall vulnerability, whereas contribution of social protection was only 20%. In terms of individual indicators, the highest contribution to overall vulnerability was from lack of community coherence (12%), followed by lack of any insurance (10%), no exposure to legal awareness training (10%) and having two or more dependents (10%) (Table 4). Lack of savings or investment contributed about 8% to overall vulnerability.
The major contributors to vulnerability was also prevalent in the study population as indicated by the proportion of vulnerable FSWs for different indicators among multidimensionally vulnerable sub-sample (Table 4). For example, A little more than half (53%) did not possess any insurance, about two-fifths had no financial savings or investments (43%) and had two or more dependents (41%) and had no exposure to legal awareness trainings (42%) and lacked community coherence (39%). The proportion of FSWs in these indicators also remained high in the overall study population as can be seen in Table 1. For example, FSWs who had no insurance was 87% in the overall sample and reduced by 30% when the sample is restricted to multidimensionally vulnerable FSWs.
Association between multidimensional vulnerability and HIV-related sexual risk factors
Multidimensionally vulnerable FSWs were more likely to report negative health behaviors compared to those multidimensionally non-vulnerable ones (Table 5). For example, inconsistent use of condoms with clients in past one month was almost twice higher among multidimensionally vulnerable FSWs than non-vulnerable ones (AOR: 1.75, 95% CI: 1.35–2.26). Similarly, multidimensional vulnerability was positively associated with inability to use condom even when FSWs wanted to use it (AOR: 1.58, 95% CI: 1.22–2.04), practice of anal sex (AOR: 1.42, 95% CI: 1.12–1.79) and alcohol consumption (AOR: 1.51, 95% CI: 1.29–1.77). FSWs who were multidimensionally vulnerable were also less likely to go for HIV testing in past two years and STI check-up and counseling services in past 12 months.
Measurement of vulnerability and identifying the context leading to vulnerability is important for strategic planning for any intervention. FSWs in India face multiple vulnerabilities by the nature of their profession; hence, identification and measurement of vulnerability in a robust way is necessary for reaching the vulnerable population. As the extent and nature of vulnerabilities vary considerably depending on the definition, measurement approaches and interpretation , it can be perplexing at-times to stakeholders and implementers. In this context, this study suggests a more comprehensive way of measuring vulnerability among FSWs by adopting an approach used mostly to measure poverty. The approach has not only helped in integrating the multiple dimensions and indicators of vulnerability but also provided the intensity of vulnerability alongside the prevalence of vulnerability . Moreover, this approach is statistically more superior compared to other approaches of computing index as has been demonstrated in the computation of multidimensional poverty index . The study found that a large share (55%) of FSWs in the study area were multidimensionally vulnerable. FSWs in Maharashtra were most vulnerable (82%) and least in Karnataka (35%). Lack of financial security contributed mostly to FSWs’ vulnerability. Specifically, lack of possession of any insurance scheme, and lack of savings or investment in gold, property or land contributed mostly to FSWs’ vulnerability.
Multidimensionally vulnerable FSWs were also more likely to be at increased HIV risk and less likely to go for STI check-up and HIV testing than those multidimensionally non-vulnerable. This reflects the robustness of our approach in identifying vulnerable FSWs. Moreover, it indicates that reducing vulnerabilities among FSWs would eventually contribute to a reduction in their risk-taking behavior. This is in-line with the past research in India which suggests that addressing vulnerabilities among sex workers could lead to increase in their safe sex behaviors [4, 6, 7, 10, 24, 25]. While sex workers enter into sex work due to increasing financial burden, trafficking and coercion, their vulnerabilities increase after getting into sex work due to the non-conducive policies and institutional arrangements. In a patriarchal society, their power of negotiation with clients (for money, type of sex and use of condoms) gets weakened to the social and financial vulnerabilities they are facing simultaneously. Therefore, addressing these vulnerabilities will empower FSWs and hence, the ability to negotiate for sex will increase [5, 9, 24, 26, 27].
The study also found a negative relationship between vulnerability and STI check-up and HIV testing which is contradicting the earlier study in Mumbai, India which suggested that high vulnerability of FSWs (though measured differently than this study) linked to higher service uptake. The difference can be due to the fact the earlier study was conducted among individuals who were part of the targeted HIV prevention intervention and the intervention prioritized providing services to the highly vulnerable ones . Therefore, the earlier study showed a positive relationship between vulnerability and service uptake.
The vulnerability among FSWs was primarily owing to lack of access to financial security. It is well-known fact that FSWs in India face challenges of a different nature compared to women in general population to access the financial services or take benefits of social schemes . The clandestine nature of sex work and stigma associated with sex work practice has limited FSWs’ access to different government services including health and social services . Moreover, the fact that majority of sex workers enter the industry due to financial burden and hence, addressing financial and social vulnerabilities of sex workers is utmost important to improve the ability to negotiate with clients, brothel owners and pimps [27, 29]. Also, evidence suggests that FSWs with high financial vulnerability and lack of access to social entitlements are more likely to engage in risky sexual behavior [2, 6, 10, 26–28, 30]. Therefore, addressing the financial needs and improving financial inclusion and capability among FSWs will not only address vulnerabilities but also improve the likelihood of safe sex practice among them.
The contribution of individual factors within each dimension varies considerably across dimensions. Within the domain of financial security, lack of insurance contributed most (32%), followed by contribution from lack of savings or investment in gold, land or property (26%), lack of alternative income other than sex work (20%) and not having saving accounts in bank or post office (16%). Loan from an informal source contributed only 6%. While in previous vulnerability research loan has been considered as a key factor [7, 15], our study showed that other economic indicators are more relevant to assess FSWs’ vulnerability than only financial debt. Within the social protection dimension, the contribution to vulnerability was mostly due to lack of awareness on legal rights (49%), followed by lack of possession of ration card (21%) and citizen identity card (17%). Food insufficiency contributed the least (13%) to social protection dimension which again in contrast to previous research which argues for food insufficiency as the key vulnerability. Rather, it is lack of owning a ration card is a more important contributor of vulnerability among FSWs in India.
Most of the existing research has treated indicators of personal attributes as covariates while undertaking analysis involving vulnerabilities of FSWs. In this study, instead of considering them as covariates, we considered them as key factors that contribute to FSWs’ vulnerability. The findings suggested that 23% of the vulnerabilities among FSWs were due to their personal attributes. The indicators within personal attributes that contributed most is degree of dependency (42%) and being mobile/migrant for sex work (34%). The role of degree of dependency on FSWs’ vulnerability has been highlighted in prior research also 25. While previous research in India suggests that place of solicitation is the key factor that decides FSW’s vulnerability and exposure to risk, we found that contribution from street-based solicitation is considerably low (4% to overall vulnerability). This is primarily due to consideration of multi-dimensional framework as against a unidimensional approach considered by previous studies. Similar to the dimension of personal attributes, the absence of social support and network contributed about one-quarter to overall vulnerability level. The indicator on social coherence (measured as community members did not come together to help FSW at the time of crisis) contributed most to the social support and network dimension. Given the increasing importance of community mobilization interventions among FSWs, this finding reiterates the need for such intervention.
The study’s findings should be interpreted in the light of following limitations. First, there may be social desirability bias in some of the behavioral measures such as condom use due to their prior exposure to HIV prevention programs. Second, the relationship between vulnerability and HIV-related sexual risk factors are associative in nature given that the analysis was conducted on a cross-sectional data. Like the multidimensional poverty measurement, the multidimensional vulnerability measure is based on certain normative decisions. One can always argue the rationale behind the inclusion or removal of certain indicators or categorization of an indicator within a specific dimension. We completely acknowledge the possibility of refinement in the measurement process and see this is one of the first steps in adopting a multidimensional framework to measure vulnerability.
The study findings have important policy implications for programs working with sex workers. First, vulnerabilities of FSWs should not be seen from the context of sex work alone, rather, the structural, social and financial vulnerabilities need to be addressed to provide individuals a safe and enabling environment. Second, intervention programs need to adopt a multi-pronged strategy to address multidimensional vulnerabilities given that the vulnerabilities vary in nature and scale. The intensity of vulnerability and their contribution to overall vulnerability level can be a first step in developing intervention strategies to reduce vulnerability. The interventions should specifically design programs to address financial vulnerabilities by facilitating access to formal financial services and helping them to increase their saving habits and make investments in different movable and immovable assets. Some of the notable interventions in India are Songachhi project in Kolkata, and the Pragati intervention in Bangalore, where addressing the financial vulnerabilities have been at the forefront which has led to improved HIV risk behaviors [26, 31]. Lack of awareness of legal rights is another area which needs to address as with proper knowledge on different rights, they may be more resilient to exploitation . In addition, given the increasing emphasis of providing insurance to all individuals by Government of India, and lack of insurance a key contributor to vulnerability, interventions working FSWs should devise strategies that help sex workers’ getting access to insurance. Finally, the effort on building community collectivization should continue as it can reduce the vulnerabilities among FSWs. In conclusion, this study demonstrated how a multidimensional approach to measuring vulnerabilities among FSWs may help in identifying a range of factors that can be prioritized by interventions. Based on this study the importance of measuring vulnerabilities from an aggregate framework is emphasized because a vulnerability which may come out as important in a single-dimensional study may prove to be less important when putting in a multidimensional frame. Further, the study showed positive associations between vulnerability and HIV-related sexual risk factors suggesting that addressing vulnerabilities of FSWs would eventually lead to better HIV risk behavior among them.
We are thankful to Prof. Sanjay Kumar Mohanty, International Institute for Population Sciences, Mumbai, India and Dr. Anrudh Jain, Population Council, New York for their guidance and comment on an initial draft of the paper. We also acknowledge the useful feedback provided by Dr. Madhusudan Battala and Dr. Sangram Kishor Patel to an earlier draft.
- 1. Ad Guzman. Reducing social vulnerability to HIV/AIDS: Models of care and their impact in resource-poor settings. AIDS Care. 2001;13(5):663–75. pmid:11571013
- 2. Evans C, Jana S, Lambert H. What makes a structural intervention? Reducing vulnerability to HIV in community settings, with particular reference to sex work. Global Public Health. 2010;5(5):449–61. pmid:19507079
- 3. Chakravarthy JBR, Joseph SV, Pelto P, Kovvali D. Community mobilisation programme for female sex workers in coastal Andhra Pradesh, India: processes and their effects. Journal of Epidemiology and Community Health. 2012;66(Suppl 2):ii78-ii86. pmid:22945909
- 4. Parimi P, Mishra RM, Tucker S, Saggurti N. Mobilising community collectivisation among female sex workers to promote STI service utilisation from the government healthcare system in Andhra Pradesh, India. Journal of Epidemiology and Community Health. 2012;66(Suppl 2):ii62-ii8. pmid:22493478
- 5. Blankenship KM, West BS, Kershaw TS, Biradavolu MR. Power, community mobilization, and condom use practices among female sex workers in Andhra Pradesh, India. AIDS. 2008;22:S109–S16. 00002030-200812005-00010. pmid:19098471
- 6. Gupta GR, Parkhurst JO, Ogden JA, Aggleton P, Mahal A. Structural approaches to HIV prevention. The Lancet. 2008;372(9640):764–75. (08)60887-9.
- 7. Ranebennur V, Gaikwad S, Ramesh S, Bhende A. Addressing vulnerabilities of female sex workers in an HIV prevention intervention in Mumbai and Thane: experiences from the Aastha project. HIV/AIDS (Auckland, NZ). 2014;6:9–18. PMC3933664. pmid:24600249
- 8. Ramesh S, Ganju D, Mahapatra B, Mishra RM, Saggurti N. Relationship between mobility, violence and HIV/STI among female sex workers in Andhra Pradesh, India. BMC Public Health. 2012;12(1):764. pmid:22967276
- 9. Bharat S, Mahapatra B, Roy S, Saggurti N. Are Female Sex Workers Able to Negotiate Condom Use with Male Clients? The Case of Mobile FSWs in Four High HIV Prevalence States of India. PLOS ONE. 2013;8(6):e68043. pmid:23840806
- 10. Pillai P, Bhattacharjee P, Ramesh B, Isac S. Impact of two vulnerability reduction strategies—collectivisation and participation in savings activities—on HIV risk reduction among female sex workers,. Banglore: Karnataka Health Promotion Trust (KHPT), 2012.
- 11. Greig FE, Koopman C. Multilevel Analysis of Women's Empowerment and HIV Prevention: Quantitative Survey Results from a Preliminary Study in Botswana. AIDS and Behavior. 2003;7(2):195–208. :1023954526639. pmid:14586204
- 12. Gurnani V, Beattie TS, Bhattacharjee P, Mohan H, Maddur S, Washington R, et al. An integrated structural intervention to reduce vulnerability to HIV and sexually transmitted infections among female sex workers in Karnataka state, south India. BMC Public Health. 2011;11(1):755. pmid:21962115
- 13. Saggurti N, Jain AK, Sebastian MP, Singh R, Modugu HR, Halli SS, et al. Indicators of Mobility, Socio-Economic Vulnerabilities and HIV Risk Behaviours Among Mobile Female Sex Workers in India. AIDS and Behavior. 2012;16(4):952–9. PMC3618410. pmid:22186960
- 14. Greener R, Sarkar S. Risk and vulnerability: do socioeconomic factors influence the risk of acquiring HIV in Asia? AIDS. 2010;24:S3–S11. 00002030-201009003-00002. pmid:20926925
- 15. Travasso SM, Mahapatra B, Saggurti N, Krishnan S. Non-paying partnerships and its association with HIV risk behavior, program exposure and service utilization among female sex workers in India. BMC Public Health. 2014;14:248–. PMC3995596. pmid:24621082
- 16. National AIDS Control Organization (NACO). National Integrated Biological and Behavioural Surveillance (IBBS), India 2014–15. New Delhi, India: NACO, Ministry of Health and Family Welfare, Government of India, 2015.
- 17. Avahan-3 Baseline Survey [Internet]. Harvard Dataverse. 2018. Available from: https://doi.org/10.7910/DVN/SBN5PI.
- 18. Patel SK, Battala M, Adhikary R. Food Insufficiency, Violence and HIV Risk Behaviors among Female Sex Workers in India. Journal of Food Security. 2016;4(5):104–11.
- 19. Alkire S, Foster J. Counting and multidimensional poverty measurement. Journal of Public Economics. 2011;95(7):476–87. https://doi.org/10.1016/j.jpubeco.2010.11.006.
- 20. Alkire S, Foster J. Counting and multidimensional poverty measures. Working Paper No 7 [Internet]. 2007.
- 21. Tsui K-y. Multidimensional poverty indices. Social Choice and Welfare. 2002;19(1):69–93. https://doi.org/10.1007/s355-002-8326-3
- 22. Alkire S, Santos ME. Measuring Acute Poverty in the Developing World: Robustness and Scope of the Multidimensional Poverty Index. World Development. 2014;59(Supplement C):251–74. https://doi.org/10.1016/j.worlddev.2014.01.026.
- 23. Alkire S, Foster J. Counting and multidimensional poverty measures. Working Paper No 32 [Internet]. 2009.
- 24. Gaikwad SS, Bhende A, Nidhi G, Saggurti N, Ranebennur V. How effective is community mobilisation in HIV prevention among highly diverse sex workers in urban settings? The Aastha intervention experience in Mumbai and Thane districts, India. Journal of Epidemiology and Community Health. 2012;66(Suppl 2):ii69–ii77. pmid:22760223
- 25. Shahmanesh M, Patel V, Mabey D, Cowan F. Effectiveness of interventions for the prevention of HIV and other sexually transmitted infections in female sex workers in resource poor setting: a systematic review. Tropical Medicine & International Health. 2008;13(5):659–79. pmid:18266784
- 26. Patel SK, Prabhakar P, Jain AK, Saggurti N, Adhikary R. Relationship between Community Collectivization and Financial Vulnerability of Female Sex Workers in Southern India. PLOS ONE. 2016;11(5):e0156060. pmid:27227998
- 27. Saggurti N, Verma RK, Halli SS, Swain SN, Singh R, Modugu HR, et al. Motivations for entry into sex work and hiv risk among mobile female sex workers in India. Journal of Biosocial Science. 2011;43(5):535–54. Epub 2011/06/10. pmid:21729360
- 28. Moret W. Economic Strengthening for Female Sex Workers: A Review of the Literature. FHI360, 2014.
- 29. Scorgie F, Chersich MF, Ntaganira I, Gerbase A, Lule F, Lo Y-R. Socio-Demographic Characteristics and Behavioral Risk Factors of Female Sex Workers in Sub-Saharan Africa: A Systematic Review. AIDS and Behavior. 2012;16(4):920–33. pmid:21750918
- 30. Reed E, Gupta J, Biradavolu M, Devireddy V, Blankenship KM. The Context of Economic Insecurity and its Relation to Violence and Risk Factors for HIV among Female Sex Workers in Andhra Pradesh, India. Public Health Reports. 2010;125(4_suppl):81–9. pmid:20629253
- 31. Jana S, Singh S. Beyond medical model of STD intervention—lessons from Sonagachi. Indian Journal of Public Health. 1995;39(3):125–31. pmid:8690494
- 32. Ganju D, Patel SK, Prabhakar P, Adhikary R. Knowledge and exercise of human rights, and barriers and facilitators to claiming rights: a cross-sectional study of female sex workers and high-risk men who have sex with men in Andhra Pradesh, India. BMC International Health and Human Rights. 2016;16:29. PMC5112884. pmid:27855692
- 33. Ramesh BM, Moses S, Washington R, Isac S, Mahapatra B, Mahagaonkar SB, et al. Determinants of HIV prevalence among female sex workers in four south Indian states: analysis of cross-sectional surveys in twenty-three districts. AIDS. 2008;22:S35–S44. 00002030-200812005-00003. pmid:19098478
- 34. Shahmanesh M, Cowan F, Wayal S, Copas A, Patel V, Mabey D. The burden and determinants of HIV and sexually transmitted infections in a population-based sample of female sex workers in Goa, India. Sexually Transmitted Infections. 2009;85(1):50. pmid:18684856
- 35. Sarkar K, Bal B, Mukherjee R, Saha MK, Chakraborty S, Niyogi SK, et al. Young age is a risk factor for HIV among female sex workers—An experience from India. Journal of Infection. 2006;53(4):255–9. pmid:16386307
- 36. Dandona R, Dandona L, Gutierrez JP, Kumar AG, McPherson S, Samuels F, et al. High risk of HIV in non-brothel based female sex workers in India. BMC Public Health. 2005;5:87–. PMC1208909. pmid:16111497
- 37. Ntumbanzondo M, Dubrow R, Niccolai LM, Mwandagalirwa K, Merson MH. Unprotected intercourse for extra money among commercial sex workers in Kinshasa, Democratic Republic of Congo. AIDS Care. 2006;18(7):777–85. pmid:16971288
- 38. Shannon K, Goldenberg SM, Deering KN, Strathdee SA. HIV Infection Among Female Sex Workers in Concentrated and High Prevalence Epidemics: Why A Structural Determinants Framework Is Needed? Current opinion in HIV and AIDS. 2014;9(2):174–82. PMC4286412. pmid:24464089
- 39. Blankenship KM, Friedman SR, Dworkin S, Mantell JE. Structural interventions: concepts, challenges and opportunities for research. J Urban Health. 2006;83. pmid:16736355
- 40. Punyam S, Pullikalu RS, Mishra RM, Sandri P, Mutupuru BP, Kokku SB, et al. Community advocacy groups as a means to address the social environment of female sex workers: a case study in Andhra Pradesh, India. Journal of Epidemiology and Community Health. 2012.
- 41. Cornish F, Priego-Hernandez J, Campbell C, Mburu G, McLean S. The impact of Community Mobilisation on HIV Prevention in Middle and Low Income Countries: A Systematic Review and Critique. AIDS and Behavior. 2014;18(11):2110–34. PMC4196137. pmid:24659360