Sinha, Gita MD, MPH*; Dyalchand, Ashok MD, MPH†; Khale, Manisha MSc†; Kulkarni, Gopal MSc†; Vasudevan, Shubha BSc†; Bollinger, Robert C MD, MPH*
From the *Johns Hopkins University School of Medicine, Baltimore, MD; and the †Institute of Health Management, Pachod (IHMP), Maharashtra, India.
Received for publication February 7, 2007; accepted October 18, 2007.
Supported by the Department for International Development, United Kingdom. Dr. Sinha's participation was supported by the Johns Hopkins University School of Medicine Clinician Scientist Award.
Correspondence to: Gita Sinha, MD, Division of Infectious Diseases, Johns Hopkins University School of Medicine, 600 North Wolfe Street, Phipps 521, Baltimore, MD 21287 (e-mail: email@example.com).
Introduction: Sixty percent of India's HIV cases occur in rural residents. Despite government policy to expand antenatal HIV screening and prevention of maternal-to-child transmission (PMTCT), little is known about HIV testing among rural women during pregnancy.
Methods: Between January and March 2006, a cross-sectional sample of 400 recently pregnant women from rural Maharashtra was administered a questionnaire regarding HIV awareness, risk, and history of antenatal HIV testing.
Results: Thirteen women (3.3%) reported receiving antenatal HIV testing. Neither antenatal care utilization nor history of sexually transmitted infection (STI) symptoms influenced odds of receiving HIV testing. Women who did not receive HIV testing, compared with women who did, were 95% less likely to have received antenatal HIV counseling (odds ratio = 0.05, 95% confidence interval: 0.02 to 0.17) and 80% less aware of an existing HIV testing facility (odds ratio = 0.19, 95% confidence interval: 0.04 to 0.75).
Conclusions: Despite measurable HIV prevalence, high antenatal care utilization, and STI symptom history, recently pregnant rural Indian women report low HIV testing. Barriers to HIV testing during pregnancy include lack of discussion by antenatal care providers and lack of awareness of existing testing services. Provider-initiated HIV counseling and testing during pregnancy would optimize HIV prevention for women throughout rural India.
With approximately 2.5 million cases, India continues to have one of the highest numbers of HIV cases in the world.1 Approximately 60% of India's HIV cases occur in rural residents.2 Married rural women of childbearing age are considered a highly vulnerable population for HIV acquisition, primarily through their husbands' high-risk pre- and extramarital sexual behavior.2,3
Pregnancy-related health concerns are a major reason for seeking clinical care and, consequently, an opportunity for potentially diagnosing HIV among rural women. Sentinel surveillance and cross-sectional community studies report measurable HIV prevalence among rural pregnant women, ranging from 0.5% to >2% in some regions.2,4 India's National AIDS Control Organization (NACO) policy has thus emphasized expanding voluntary counseling and testing (VCT) and access to prevention of maternal-to-child transmission (PMTCT) therapy for all pregnant women, including in rural settings.2
In this policy context, little is known about HIV risk profiles, access to, or utilization of HIV testing among rural Indian women. The objective of our study was to assess HIV risk profiles and characteristics of HIV testing utilization among rural Indian women of childbearing age, a major target population for rural HIV prevention. Understanding the current features of and potential barriers to HIV testing during pregnancy would help to optimize HIV testing services for rural Indian women.
The study setting was a group of 52 villages comprising 2 rural village blocks and a population of approximately 60,000 individuals in Aurangabad District, Maharashtra, India, where at least 1% of pregnant women test HIV-positive in antenatal clinic sentinel surveillance sites.2 The nearest government-sanctioned VCT center is located approximately 50 km away at the district hospital in Aurangabad City.
The study sample comprised recently pregnant women, defined as having given birth in the previous 12 months. These women were chosen because their recent pregnancy was a marker for sexual activity and because they represent India's HIV prevention target population of women of child-bearing age. Currently pregnant women were excluded from the study to assess women's HIV testing utilization over an entire pregnancy period and to avoid introducing bias into future HIV testing utilization measurements among pregnant women in the community.
From November 2005 through January 2006, 400 recently pregnant women aged 18 years and older were identified through simple random sampling of census-updated demographically labeled households. After obtaining written informed consent, a trained interviewer administered a questionnaire assessing sociodemographic features, utilization of antenatal care, and awareness of and utilization of HIV testing services. History of sexually transmitted infection (STI) or tuberculosis (TB) symptoms in the previous 12 months, based on India's national clinical STI and TB guidelines, was assessed as a marker for clinical HIV-related risk.
Data were entered into a Microsoft (Redmond, WA) database, and all analyses were performed using STATA 9.0 (Stata Corporation, College Station, TX). First, selected sociodemographic and HIV-related characteristics were compared between study respondents and the sample of rural Maharashtrian women responding to the 2001 NACO behavioral surveillance survey (BSS) and the 2001 census of India.5 The BSS sample was chosen as a historical control to assess for significant general or HIV risk-related changes among rural women in Maharashtra. Second, we conducted a comparative descriptive analysis of characteristics of women who did not receive HIV testing versus women who did. We computed odds ratios for the odds of not receiving HIV testing based on the presence or absence of the various sociodemographic and HIV-related factors. Finally, we conducted a descriptive subgroup analysis of women who reported STI symptoms but did not receive HIV testing, because women reporting an STI symptom history were considered to have the highest HIV risk profile among study respondents. χ2 tests for trends in discrete variable proportions, z tests for comparison of independent group proportions, and t tests for continuous variable comparisons were used.
Study Sample Comparison With 2001 BSS Sample
The study sample was compared with the rural Maharashtrian female sample in the 2001 BSS serving as a historical control (Table 1). A significantly higher proportion of study respondents reported general HIV/AIDS awareness, condom awareness, and knowledge of specific HIV transmission and prevention facts. Study respondents' reports of recent STI symptoms or TB symptoms did not differ from 2001 BSS responses. Only 6% of the respondents correctly named an HIV testing facility. Thirty respondents (7.5%) reported receiving antenatal HIV counseling, and 13 (3.3%) reported utilizing HIV testing.
Characteristics of Utilizers and Nonutilizers of HIV Testing During Pregnancy
In a descriptive analysis of test utilizers, all 13 test utilizers reported receiving 1 HIV test during pregnancy (data not shown). Twelve women received testing in a private sector clinical facility. The primary reported reason for testing was receiving a doctor's recommendation for an HIV test. Two women reported mandatory antenatal HIV testing. None of the test utilizers was aware of the existence of VCT.
Sociodemographic and HIV-related factors were compared between women who did not receive HIV testing versus those who did (Table 2). Virtually all women had at least 1 antenatal care visit during pregnancy, regardless of whether they received an HIV test or not. Nonutilizers were 95% less likely than utilizers to report receiving antenatal HIV counseling and 80% less likely to name an existing HIV testing facility correctly.
Only 1 of the 35 women reporting an STI or TB symptom history received HIV testing. In the descriptive subgroup analysis, of the 27 women who reported STI symptoms, 26 (96.2%) reported not receiving HIV testing during pregnancy (data not shown). Compared with the 13 women who did receive HIV testing, these 26 women with an STI symptom history were slightly younger in age (P < 0.01) and had less weekly access to media sources (P < 0.01). Virtually all these women (96.2%) reported seeking antenatal care, and almost half reported seeking medical treatment for their STI symptoms. Only 4 women (15.4%) correctly named an existing HIV testing facility, none were aware of the concept of VCT, and only 1 received HIV counseling in pregnancy.
Although India has recently estimated lower numbers of HIV cases compared with prior reports, HIV seroprevalence among antenatal care-seeking pregnant women remains a major criterion for classifying rural-based districts as having high (>1%) versus low (<1%) HIV prevalence.1,2 Strategies to provide accessible HIV testing for rural Indian women are important for primary and secondary HIV prevention, including PMTCT. To our knowledge, this is the first community-based study reporting extremely low HIV testing among rural Indian women during pregnancy, a major target population for India's HIV prevention efforts.
Compared with the 2001 BSS findings, rural women in our study demonstrated higher HIV awareness and knowledge yet unchanged levels of clinical HIV risk, as indicated by reported STI and TB symptoms. Over the past several years, community and government efforts have increased rural women's HIV awareness.6 Nevertheless, 2 significant barriers to HIV testing seem to exist in this rural setting, namely, lack of discussion of HIV by antenatal care providers and lack of awareness of HIV testing services, including VCT.
Lack of Discussion of HIV by Antenatal Care Providers
Most rural women, including women with a history of STI symptoms, utilized antenatal care but reported no discussion of HIV with their care providers. Conversely, the primary reason for testing cited by the antenatal HIV test utilizers was their care provider's recommendation. In light of India's recently revised HIV seroprevalence estimates, a recent population-based study suggests that HIV sentinel surveillance among urban and rural pregnant women may be biased by differences in HIV testing in public sector versus private sector antenatal care settings.7 Similarly, our study data indicate that even in high HIV-prevalent settings, rural women seeking antenatal care may not be uniformly assessed or counseled for HIV testing during pregnancy. Our study data further highlight the critical role of antenatal care providers to initiate discussion of HIV risk factors and HIV testing with all pregnant women. Given rural women's typically limited access to general health care, pregnancy-related care may be one of the few settings in which HIV risk assessment (during and outside of pregnancy), HIV counseling, and/or testing is accessible and feasible. In support of antenatal provider-led HIV counseling and testing, at least 2 rural antenatal clinic-based studies have demonstrated the acceptability of antenatal clinic-based HIV VCT among rural Indian women.8,9
Lack of Awareness of HIV Testing Services, Including VCT
Few women correctly named an existing HIV testing facility or reported awareness of VCT. Furthermore, only half of HIV test utilizers reported receiving associated counseling. As reported in urban settings, these data suggest that current rural HIV testing services may not be providing the essential elements of VCT.10 Limited socioeconomic status, access to media, and mobility may all contribute to low testing services awareness.11,12 Community- and clinic-based efforts are necessary to increase rural women's awareness of voluntary HIV testing and locations of acceptable testing services.
Beyond the services-related barriers, we have identified social and cultural barriers, including limited autonomy, mobility, or fear of family or community members' reactions, that could limit rural women's HIV testing utilization.12 Lack of available HIV testing services may be another barrier; in this study, respondents named only 5 rural HIV testing facilities in a coverage area of more than 50,000 people. Formal mapping of existing rural HIV testing services would be necessary to validate services availability as a potential barrier to HIV testing, however.
One study limitation is that study respondents were sampled differently from the BSS sample used as a historical control; differences in HIV awareness and knowledge may reflect population factors rather than HIV-related interventions. Additional limitations include reliance on respondent self-report, and the subsequent potential for recall bias, particularly among women who were interviewed up to 12 months since having given birth. Our data indicate that the duration of time since delivery did not influence the odds of receiving HIV testing, which may help to minimize but does not dismiss the concern for recall bias. Finally, the timing of self-reported STI or TB symptoms in relation to the pregnancy was not assessed, and laboratory confirmation was not possible in this study. Symptom onset after delivery rather than during pregnancy could alter considerations for antenatal HIV testing. Given the prevalence of asymptomatic STIs and variability in self-reported symptoms, however, antenatal HIV testing should not be guided solely by self-reported STI symptoms, particularly in communities in which a significant proportion of pregnant women test HIV-positive.13
India's current policy for universal access to PMTCT requires strategies to promote HIV testing among pregnant rural Indian women. Antenatal clinic-based programs to increase rural women's awareness of voluntary HIV testing services, including provider-initiated assessment of HIV risk and discussion of HIV testing, should be developed and prospectively tested for their impact on utilization of HIV testing and subsequent prevention, diagnosis, and PMTCT for women throughout rural India.
The authors additionally thank the following individuals for guidance in study design and manuscript preparation: Manoj Chaudhuri, MS, IHMP; D. N. Chaudhuri, IHMP; Shruti Mehta, PhD, Johns Hopkins Bloomberg School of Public Health; and all rural community residents and leaders who participated in this study.
3. Chandrasekaran P, Dallabetta G, Loo V, et al. Containing HIV/AIDS in India: the unfinished agenda. Lancet Infect Dis. 2006;6:508-521.
4. Arole S, Premkumar R, Arole R, et al. Prevalence of HIV in pregnant women in remote rural areas of Maharashtra State, India. Trop Doct. 2005;35:111-112.
5. National AIDS Control Organization (NACO), Government of India. National baseline general population behavioral surveillance survey (BSS). 2001. Available at: www.nacoonline.org/publication/31.pdf
. Accessed October 7, 2007.
6. International Institute for Population Sciences (IIPS). National family health survey 2007. Available at: http://www.nfhsindia.org/
. Accessed October 7, 2007.
7. Dandona L, Lakshmi V, Sudha T, et al. A population-based study of human immunodeficiency virus in south India reveals major differences from sentinel surveillance-based estimates. BMC Med. 2006;4:31-49.
8. Samuel NM, Srijayanth P, Dharmarajan S, et al. Acceptance of HIV-1 education and voluntary counseling/testing by, and seroprevalence of HIV-1 among, pregnant women in rural south India. Indian J Med Res. 2007;125:49-64.
9. Gupta D, Lhewa D, Viswanath R, et al. Effectiveness of antenatal group HIV voluntary counseling and testing services in rural India. AIDS Educ Prev. 2007;19:187-197.
10. Datye V, Kielmann K, Sheikh K, et al. Private practitioners' communications with patients around HIV testing in Pune, India. Health Policy Plan. 2006;21:343-352.
11. Pallikadavath S, Sreedharan C, Stones RW. Sources of AIDS awareness among women in India. AIDS Care. 2006;18:44-48.
12. Rogers A, Meundi A, Amma A, et al. HIV-related knowledge, attitudes, perceived benefits, and risks of HIV testing among pregnant women in rural Southern India. AIDS Patient Care STDS. 2006;20:803-811.
13. Hawkes S, Mouson L, Foster S, et al. Reproductive tract infections in women in low income, low prevalence situations: assessment of syndromic management in Matlab, Bangladesh. Lancet. 1999;354:1776-1781.
© 2008 Lippincott Williams & Wilkins, Inc.