Secondary Logo

Journal Logo

EPIDEMIOLOGY AND SOCIAL

Prevalence and determinants of HIV infection in South India: a heterogeneous, rural epidemic

Becker, Marissa La; Ramesh, BMb,c; Washington, Reynold Gb,c,d; Halli, Shivab,c; Blanchard, James Fa,b,c; Moses, Stephena,b,c,e

Author Information
doi: 10.1097/QAD.0b013e328012b885

Abstract

Introduction

There are an estimated 5.7 million persons living with HIV/AIDS in India [1], and since the first case of AIDS was reported in India in 1986, the epidemic has grown steadily [2]. There is considerable heterogeneity in the epidemic pattern within the country; most states report HIV prevalence of less than 0.5% among antenatal clinic attenders, while four states have prevalence among antenatal clinic women of 1% or more [2]. Heterosexual sex is felt to be the predominant mode of transmission in the southern states and accounts for 85% of the infections in the country as a whole, although intravenous drug use plays a large role in driving the epidemic in the northeastern states [1]. While the HIV epidemic in India was initially thought to be concentrated in large urban areas such as Mumbai and Chennai, there is increasing evidence that the epidemic is now becoming established within rural populations [2]. Karnataka, a southern state, is estimated to have an HIV prevalence in the adult population of 1.6% [3], suggesting that there are about 500 000 persons living with HIV/AIDS in the state. Within Karnataka, HIV prevalence varies substantially across districts, ranging from just over zero in some districts to upwards of 3% in others [3].

Bagalkot, a district in the northern part of Karnataka, is a largely agricultural district with a population of 1.6 million people, 69% of whom live in rural areas. Antenatal clinic surveillance data have suggested that Bagalkot is a high prevalence district, with an estimated HIV prevalence of 2.6% in 2004 [4]. Some potential explanations for why Bagalkot is particularly vulnerable to HIV infection include a large volume of female sex work within the district and the fact that sex workers tend to be younger, have higher rates of illiteracy and receive more clients than their southern counterparts [5]. This sex work pattern is probably linked with the presence in this region of northern Karnataka of the traditional Devadasi system of sex work. The Devadasi tradition dates back several centuries and involves the dedication of young girls through marriage to different gods. Girls were traditionally required to perform various temple duties, including the provision of sexual services to priests. Over time the practice has become more commercialized, but sex work associated with the tradition is socially accepted and culturally sanctioned [6,7].

As part of the India-Canada Collaborative HIV/AIDS Project (ICHAP), which was funded by the Canadian International Development Agency in partnership with the Government of India and the National AIDS Control Organisation, Bagalkot district was selected for the implementation of a demonstration project for HIV prevention in rural areas. As part of the Bagalkot demonstration project we conducted a baseline socio-behavioral and HIV survey to better understand HIV transmission dynamics and program needs. In this article we describe the prevalence of HIV in the general population in Bagalkot, assess the demographic, socio-economic, behavioral and medical factors associated with HIV infection, and discuss the implications for HIV prevention programs.

Methods

We conducted a community-based study of individuals aged 15–49 years in three of the six sub-districts that comprise Bagalkot District, denoted herein as sub-districts A, B and C. The study was undertaken in these sub-districts to provide baseline information with which to plan a comprehensive HIV prevention program, which began in the three sub-districts, and was later expanded to cover the remaining sub-districts. Using a probability proportional to estimated population size (PPES) method of sampling, 10 villages in Bagalkot were randomly selected. Twenty urban blocks from six towns were also selected, using a systematic random sampling procedure. A complete census of households and individuals in selected villages and urban blocks was carried out. Through this census, a list of all persons aged 15–49 years was prepared for both the rural and urban areas. Certain background characteristics such as age, sex, occupation and literacy were also ascertained in the census. The sample was stratified to ensure approximately equal numbers of men and women, and equal numbers of urban and rural dwellers. Using sex, age and marital status as stratification variables, random samples were selected from each of the rural and urban lists.

Following informed written consent, each respondent was interviewed using a standardized, pre-tested questionnaire. The interview was conducted at the individual's home by a trained interviewer in the local language (Kannada). Information was collected on socio-demographic characteristics and sexual behavior of the participants. The interview was followed with blood collection by a trained nurse. Specimens were transported to the nearby Karnataka Institute of Medical Sciences (KIMS) in the city of Hubli for testing. Serum samples were tested for HIV-specific antibodies using an enzyme-linked immunosorbent assay (ELISA), the Detect HIV 1/2 system (BioChem ImmunoSystems, Montreal, Canada). Reactive specimens were confirmed using a second ELISA test, Genedia HIV 1/2 ELISA 3.0 (Green Cross Life Science Corporation, Yongin-shi, South Korea). Samples were considered positive when both ELISA tests were positive. Samples were also tested for syphilis using the rapid plasma reagin (RPR) method and if positive, confirmed with the Treponema pallidum hemagluttination assay (TPHA) test. Herpes simplex virus type 2 (HSV-2) testing was conducted on a stratified sample of 25% of all specimens, using the ELISA-based Kalon Biological Kit (Kalon Biological Ltd, Aldershot, UK). Participants were enrolled from April 2003 until September 2003. Institutional Review Boards from the University of Manitoba, Winnipeg, Canada and St. John's Medical College, Bangalore, India approved the study.

Of the 6703 individuals selected for the study, 4949 agreed to participate (73.8%). Reasons for non-participation included permanent (1.5%) and temporary (6.9%) migration from the area during the time between the census and the survey. A further 15.8% of individuals refused to participate at all. There were 2% of selected individuals who were excluded because they were mentally incapable of responding, deaf, mute or deceased. A total of 4008 of those interviewed also agreed to provide a blood sample (81.0% of participants, 1864 rural and 2144 urban). Reasons given for refusal to give blood included fear of giving blood in general and fear of knowing one's own HIV status due to the stigma related to HIV/AIDS.

Stata 7 (Stata Corp., College Station, Texas, USA) software was used for both univariate and multivariate analysis to evaluate potential risk factors for HIV infection. Multiple logistic regression analysis was performed to calculate adjusted odds ratios and 95% confidence intervals (CI).

Results

Socio-demographic characteristics

The socio-demographic characteristics of the respondents according to their participation in the study are shown in Table 1. There were similar numbers of men and women and of urban and rural dwellers in the selected sample, but female and urban dwellers were more likely to participate than male and rural dwellers. The majority of the participants were less than 40 years of age and over 60% of respondents were married. Illiteracy rates were very high among the respondents. Over 80% of the participants were Hindu and the majority classified themselves as high or middle-caste Hindus. Of the participants in the survey who gave blood samples (4008), females and urban respondents were more likely to give blood than male and rural dwellers: 61% of females compared with 59% of males [odds ratio (OR), 1.1; 95% CI, 1.0–1.2) and 65% of urban participants compared with 55% of rural participants (OR, 1.5; 95% CI, 1.4–1.7). In addition, a larger percentage of respondents from sub-district B gave blood than from the other two sub-districts (OR, 1.7; 95% CI, 1.6–2.0). There were significantly fewer illiterate participants who gave blood compared with those who were literate and fewer agricultural labourers who gave blood compared with other occupations. In addition, significantly fewer low-caste compared with higher-caste Hindus gave blood (OR, 0.76; 95% CI, 0.65–0.89).

Table 1
Table 1:
Characteristics of respondents according to participation in the survey (%),N = 6703.

Sexual behavior

Tables 2 and 3 summarize the sexual behavior characteristics of the respondents. The median age at first sexual intercourse was 22.5 years for men and 15.6 years for women. Rural respondents were more likely to report their first sexual experience at a younger age than their urban counterparts (median of 21.8 years for rural men versus 23.3 years for urban men, and median of 14.9 years for rural women versus 16.4 years for urban women). Table 2 shows reported specific sexual behaviors. Less than 3% of men and less than 1% of women reported having had more than one sex partner during the previous 12 months. Significantly more men than women reported ever having had multiple sex partners, 12.2 and 1.7%, respectively. This proportion was also significantly higher in rural areas in comparison with urban areas. Almost 70% of respondents overall had never seen a condom, as shown in Table 3. Significantly more rural and female respondents reported never having seen a condom. Only 9% of respondents overall reported ever having used a condom and less than 2% reported using a condom during last sex.

Table 2
Table 2:
Characteristics of total respondents who have reported specific sexual behaviours (%),N = 4949.
Table 3
Table 3:
Characteristics of respondents regarding condom awareness and use (%),N = 4949.

Prevalence and determinants of HIV infection

The overall prevalence of HIV infection among study respondents was 2.9%, as shown in Table 4. This compares closely to the 2003 sentinel surveillance data among antenatal clinic attenders in which HIV prevalence was 2% in the Bagalkot District Headquarters Hospital and 4.3% in a sub-district hospital [8]. Several factors were associated with HIV infection in univariate analysis. The HIV prevalence of 3.6% in rural areas was significantly higher than in urban areas at 2.4%. A significant difference in prevalence was also seen among the different sub-districts within the district. Among both men and women, sub-district C had a significantly higher HIV prevalence than the other sub-districts. This was particularly true among men, where HIV prevalence was 6.4% in comparison with 1.1 and 3.0% in sub-districts A and B, respectively. Heterogeneity in HIV prevalence was even seen at village level: in three villages sampled, HIV prevalence ranged from 0 to 1.5%; in four other villages, it ranged from 2.8 to 3.4%; and in the three remaining villages, it ranged from 6.3 to 8.2%.

Table 4
Table 4:
HIV prevalence and determinants,N = 4008.

HIV prevalence among men was higher than among women, although this difference was not statistically significant (3.3 versus 2.5%; P = 0.10). Men aged 25–39 years had higher HIV prevalence than men at other ages. No significant association between HIV infection and age was seen for women. Those respondents who were widowed, divorced, separated or deserted were significantly more likely to test HIV positive (9.4% in comparison with 3.0% if currently married or 1.1% if never married). A lower HIV prevalence were found among individuals who had received education to a higher than secondary level than among those individuals who had lower levels of education. Business persons (OR, 0.48; 95% CI, 0.27–0.85) and salaried employees (OR, 0.44; 95% CI, 0.21–0.92) had lower prevalence than those in other occupations, specifically labourers (data not shown). Travel due to work was not found to be a significant risk factor for HIV (OR, 1.2; 95% CI, 0.77–1.9). HIV prevalence also varied significantly by religion. Muslims had the lowest prevalence at 0.9%, followed by Hindus at 3.3% and then other religions at 5.0%. In addition, low-caste Hindus had higher HIV prevalence (5.7%) than either higher-caste Hindus (2.7–3.0%) or non-Hindus (1.3%). This difference was particularly apparent among female low-caste Hindus, with an HIV prevalence of 7.2%. Several behaviors were also found to be significantly associated with HIV infection. In particular, HIV prevalence was 7.2% among respondents who reported ever having more than one sex partner compared with 2.6% among those who did not, and women who reported having received money for sex were significantly more likely to be infected with HIV. Men reporting ever having symptoms of sexually transmitted infections (STI) were almost three times more likely to be HIV positive, but there was no such association observed among women. In addition, men who reported more than 10 medical injections in the past year and women who reported more than five medical injections had higher HIV prevalence levels than those individuals who received fewer injections. The association observed with reporting more than 10 medical injections in the past year was also significant in the overall group.

A multiple logistic regression model was constructed, including key demographic variables and other variables that were significant predictors of HIV infection in univariate analysis: age, rural/urban residence, occupation, marital status, education level, caste, self-reported sexual risk behavior and number of injections received (Table 5). Significant predictors of HIV infection included gender, sub-district, age, marital status, education and caste. Men aged 30–39 years, persons residing in sub-district B or C, persons whose marriage had dissolved, women who reported receiving more than five injections in the past year, those reporting 10 or more injections in the past year in the total group, and Hindus (particularly lower-caste Hindus), were at higher risk of HIV infection. Unmarried persons and those with higher levels of education were at lower risk for infection.

Table 5
Table 5:
Determinants of HIV infection by multiple logistic regression,N = 4008.

In order to understand the effect of differential participation in the study on HIV prevalence by socio-demographic characteristics, prevalence rates weighted for differential participation were calculated using a simple standardization technique. Weights were calculated separately for: (1) those who gave both interviews and blood samples; (2) those who gave only an interview and not a blood sample; and (3) those who neither gave an interview nor a blood sample. In general, the weighted prevalence rates were slightly lower than the unweighted rates (data not shown), suggesting that non-participants were not at higher risk of HIV infection, but the differences for most sub-groups of interest were not very great.

Other sexually transmitted infections

Of the sample of 901 persons tested for HSV-2, 18.9% were antibody positive. HSV-2 prevalence was significantly higher among rural dwellers than among urban dwellers (23.5 versus 15.4%; P = 0.002). There was no statistically significant difference in seroprevalence between men and women. Many of the same characteristics that were associated with HIV infection were also associated with HSV-2 infection. These included, sub-district, age, education, martial status, religion and caste. For example, HSV-2 prevalence followed a similar pattern to HIV infection with the following differences seen between the sub-districts: sub-district C 23.9%, sub-district B 17.7%, and sub-district A 16.3%. There was a strong association between HSV-2 and HIV infection: 8.0% of HSV-2 positive persons were HIV infected, versus only 1.6% of HSV-2 negative persons (OR, 5.2; 95% CI, 2.3–11.5). This association was particularly significant among HIV positive men: 11.5% of HSV-2 positive men were HIV infected, versus only 1.6% of HSV-2 negative men (OR, 7.8; 95% CI, 2.7–22.7). The syphilis positivity rate among the participants was low at 0.4% overall, with no significant difference between men and women. This compares with antenatal surveillance data from 2003 for the state of Karnataka that found an overall median syphilis prevalence of 0.74%. There was no significant association seen between syphilis and HIV infection.

Discussion

The HIV epidemic is highly heterogeneous, so understanding the epidemic patterns at local levels is important for developing effective program responses. Based on the HIV prevalence of 2.9% identified in this study, Bagalkot is a high prevalence district in the Indian context. Within the district, the prevalence in rural areas was significantly higher than in urban areas. The rural nature of the epidemic has been observed elsewhere in India [1,2,9], and has important implications both for prevention programs and for provision of care to HIV-infected persons. Large variations in HIV prevalence across sub-districts, and even between villages, were also identified in Bagalkot. This suggests that vulnerability and risk for HIV transmission may be greater in some villages than in others, and may be due to individual or community-level factors. Further research is required to arrive at a better understanding of why some communities seem to be at higher risk than others, and tools to rapidly identify higher risk areas need to be developed. This would allow for rapid identification of priority areas on which to focus appropriate preventive interventions.

Other than place of residence, significant risk factors identified for HIV infection included gender, age, marital status, education and caste. These factors have been shown to be associated with a high risk of HIV infection in other studies [10–12]. Persons who were widowed, divorced or separated were more likely to be HIV positive, as has been reported in studies from Africa [10,13] and India [12]. The association seen with marital status may result from infection through a spouse who was initially infected and subsequently died of his/her illness, or because once identified as being HIV positive, his/her spouse abandoned the family. In addition, women who were widowed were more vulnerable to sexual exploitation and this may lead to increased HIV risk. A study from Tanzania reporting a higher HIV prevalence among such women suggests that it was attributable to a greater number of sex partners [10]. We also identified low-caste Hindu women as having a higher risk of HIV infection. In Bagalkot district, this group of women were more likely to be Devadasi[5], and this very vulnerable group is important in terms of directing prevention efforts. In a study in rural Zimbabwe assessing patterns of movement and risk of HIV infection, mobility was associated with risk of HIV infection for those who migrated between rural locations [14]. In our study, mobility was not strongly associated with HIV infection. This suggests that HIV transmission may be sustained through local sexual networks in this area, and may not be dependent on continuous introduction from distal sources. Much more work is required to elucidate the interaction between migration and HIV transmission dynamics in rural areas of India. The association found between HIV infection and history of medical injections has been reported previously [15], and in a cross-sectional study it is not possible to ascertain whether it is a true risk factor for acquisition, or is a marker for increased likelihood of receiving medical treatment for an HIV-related illness, or for another condition, such as a sexually transmitted infection that is associated with HIV infection [16]. Prospective studies are required to make causal inferences in this regard [17]. A strong association between HSV-2 seropositivity and HIV infection was identified in this study, as has been reported elsewhere [18–23]. This finding has implications for STI services in India, including diagnostics and treatment. There are ongoing studies evaluating intermittent and suppressive antiviral treatment of HSV-2 as a strategy for prevention of HIV [18], and the outcome of those studies will have bearing on policy in India.

We found that both access to condoms and condom use in the general population was low. Almost 70% of respondents had never seen a condom and this percentage was significantly higher among rural and female respondents. Reported condom use was also low at less than 10%. In a study from New Delhi, almost 50% of those persons who had pre or extra-marital sex reported never using a condom [24]. A study assessing non-use of condoms among female sex workers in the southern Indian state of Andhra Pradesh showed that about half of the female sex workers did not use condoms consistently with their clients [25]. Furthermore, non-brothel-based female sex workers were much less likely to use condoms consistently [25]. Some of the significant predictors of condom non-use included lack of knowledge that HIV could be prevented and no access to free condoms [25]. These misconceptions and lack of understanding persist despite an epidemic nearing almost 20 years duration, and despite many education and intervention programs. This suggests that current programs may be limited in coverage or that there needs to be a change in the way that programs are delivered. In particular, it is important to ensure that education programs are delivered to rural, largely non-literate populations.

There are a number of limitations to this study. Questionnaires were administered face-to-face in households, and it is likely that this may have led to under-reporting of particularly sensitive items, like sexual behavior. When data on sexual behavior were collected in the same areas of Bagalkot District using a polling booth survey methodology [26], which is a way of collecting sensitive behavioral information in an anonymous fashion, reported levels of risk behavior were much higher. In addition, female, urban, and middle and high-caste Hindus were more likely to participate in the study. However, the analysis of unweighted HIV prevalence versus HIV prevalence weighted to take into account non-participation rates, suggests that non-participation did not have a major influence on HIV prevalence, and that non-participants were not at higher risk of HIV than participants.

This study highlights the challenges in dealing with an epidemic that is rural, dispersed and highly heterogeneous. Further studies need to be conducted in order to better understand the nature of the epidemic and of HIV transmission dynamics at the micro-level in rural communities. It is also important to understand how high-risk villages or communities can be rapidly identified in order to guide program resource allocation. It is important to understand the role of migration in sustaining a rural epidemic, and the impact that migrating female sex workers and male clients may have on rural HIV epidemics. An HIV epidemic as heterogeneous and diverse as India's will require a response that is both comprehensive in terms of coverage of vulnerable populations, and is of sufficient scale to address the epidemic in the vast rural areas where the majority of the population lives, and which as yet are largely neglected by prevention programs.

Acknowledgements

We would like to thank the many communities and individuals in Bagalkot District that supported and participated in this project, and especially Ms H. S. Usharani, Mr P. Nagaraj and Mr C. R. Soragavi for their enthusiasm and dedication.

Sponsorship: This work was supported by a grant from the Canadian International Development Agency. M.B. is the recipient of a Canadian Institutes of Health Research (CIHR) fellowship award. S.M. is the recipient of a CIHR investigator award. J.B. is supported by a Canada Research Chair in Epidemiology and Global Public Health Research.

References

1. Joint United Nations Programme on HIV/AIDS (UNAIDS). 2006 Report on the global AIDS epidemic. Geneva: UNAIDS; 2006.
2. Solomon S, Chakraborty A, D'Souza Yepthomi R. A review of the HIV epidemic in India. AIDS Educ Prev 2004; 16(Suppl A):155–169.
3. Karnataka State AIDS Prevention Society (KSAPS), Government of Karnataka, and India-Canada Collaborative HIV/AIDS Project (ICHAP). Karnataka HIV status reports: report on HIV sentinel surveillance 2004. Bangalore: KSAPS & ICHAP; 2004.
4. Population Foundation of India (PFI), Karnataka State AIDS Prevention Society (KSAPS), India-Canada Collaborative HIV/AIDS Project (ICHAP) and Population Reference Bureau (PRB). HIV/AIDS in Karnataka, situation and response. December 2004. Bangalore: PFI; 2004.
5. Population Research Centre (PRC), J.S.S. Institute of Economic Research. Female commercial sex workers in Karnataka: a baseline survey 2002. Dharwad (Karnataka, India): PRC; 2003. Report no. 120.
6. Blanchard J, O'Neil J, Ramesh BM, Bhattacharjee P, Orchard T, 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.
7. O'Neil J, 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.
8. Karnataka State AIDS Prevention Society/India-Canada Collaborative HIV/AIDS Project. Report on HIV sentinel surveillance 2003. Bangalore: KSAPS & ICHAP; 2003.
9. Solomon S, Kumarasamy N, Ganesh AK, Amalraj RE. Prevalence and risk factors of HIV-1 and HIV-2 infection in urban and rural areas in Tamil Nadu, India. Int J STD AIDS 1998; 9:98–103.
10. Barongo LR, Borgdorff MW, Mosha FF, Nicoll A, Grosskurth H, Senkoro KP, et al. The epidemiology of HIV-1 infection in urban areas, roadwide settlements and rural villages in Mwanza Region, Tanzania. AIDS 1992; 6:1521–1528.
11. Mukhopadhyay C, Nath G, Gulati AK, Mohapatra SC. Prevalence of HIV among low and high risk population of eastern part of northern India. J Commun Dis 2001; 33:136–142.
12. Mehta SH, Gupta A, Sahay S, Godbole SV, Joshi SN, Reynolds SJ, et al. High HIV prevalence among a high-risk subgroup of women attending sexually transmitted infection clinics in Pune, India. J Acquir Immune Defic Syndr 2006; 41:75–80.
13. Ryder RW, Ndilu M, Hassig SE, Kamenga M, Sequeira D, Kashamuka M, et al. Heterosexual transmission of HIV-1 among employees and their spouses at two large businesses in Zaire. AIDS 1990; 4:725–732.
14. Coffee MP, Garnett GP, Mlilo M, Voeten HACM, Chandiwana S, Gregson S. Patterns of movement and risk of HIV infection in rural Zimbabwe. J Infect Dis 2005; 191(Suppl 1):S159–S167.
15. Becker ML, Reza Paul S, Ramesh BM, Washington R, Moses S, Blanchard JF. Association between medical injections and HIV infection in a community based study in India [Letter]. AIDS 2005; 19:1334–1336.
16. Auvert B, Sobngwi-Tambekou J, Taljaard D, Lagarde E. Authors' reply. PLoS Med 2006; 3:e67. DOI: 10.1371/journal.pmed.0030067
17. Schmid GP, Buve A, Mugyenyi P, Garnett GP, Hayes RJ, Williams BG, et al. Transmission of HIV-1 infection in sub-Saharan Africa and effect of elimination of unsafe injections. Lancet 2004; 363:482–488.
18. Celum CL, Robinson NJ, Cohen MS. Potential effect of HIV Type 1 antiretroviral and Herpes Simplex Virus Type 2 antiviral therapy on transmission and acquisition of HIV Type 1 infection. J Infect Dis 2005; 191(Suppl 1):S107–S114.
19. Hook EW 3rd, Cannon RO, Nahmias AJ, Lee FF, Campbell CH Jr, Glasser D, Quinn TC. Herpes simplex virus infection as a risk factor for human immunodeficiency virus infection in heterosexuals. J Infect Dis 1992; 165:251–255.
20. Reynolds SJ, Risbud AR, Shepherd ME, Zenilman JM, Brookmeyer RS, Paranjape RS, et al. Recent herpes simplex virus type 2 infection and the risk of human immunodeficiency virus type 1 acquisition in India. J Infect Dis 2003; 187:1513–1521.
21. Wald A, Corey L. How does herpes simplex virus type 2 influence human immunodeficiency virus infection and pathogenesis? J Infect Dis 2003; 187:1509–1512.
22. Kar HK, Jain RK, Sharma PK, Gautam RK, Gupta AK, Sharma SK, et al. Increasing HIV prevalence in STD clinic attendees in Delhi, India: 6 year (1995–2000) hospital based study results. Sex Trans Infect 2001; 77:393.
23. Kaul R, Kimani J, Nagelkerke NJ, Fonck EN, Keli F, MacDonald KS, et al. Monthly antibiotic chemoprophylaxis and incidence of sexually transmitted infections and HIV-1 infection in Kenyan Sex Workers: a randomized controlled trial. JAMA 2004; 291:2555–2562.
24. Kumar A, Mehra M, Badham SK, Gulati N. Heterosexual behaviour and condom usage in an urban population of Delhi. India AIDS Care 1997; 9:311–318.
25. 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; 5:87–97.
26. India-Canada Collaborative HIV/AIDS Project (ICHAP). The hidden epidemic: HIV/AIDS in Rural Karnataka – a situation assessment in Bagalkot District. Bangalore: ICHAP; 2004.
Keywords:

HIV; seroprevalence; epidemiology; risk factors

© 2007 Lippincott Williams & Wilkins, Inc.