We characterized the human immunodeficiency virus (HIV) care continuum among men who have sex with men and persons who inject drugs across 27 sites in India. The primary barrier to engagement was awareness of HIV-positive status.
Keywords: HIV care continuum, India, men who have sex with men, people who inject drugs
Abstract
Background. We characterize the human immunodeficiency virus (HIV) care continuum for men who have sex with men (MSM) and persons who inject drugs (PWID) across India.
Methods. We recruited 12 022 MSM and 14 481 PWID across 26 Indian cities, using respondent-driven sampling (September 2012 to December 2013). Participants were aged ≥18 years and either self-identified as male and reported sex with a man in the prior year (MSM) or reported injection drug use in the prior 2 years (PWID). Correlates of awareness of HIV-positive status were characterized using multilevel logistic regression.
Results. A total of 1146 MSM were HIV infected, of whom a median of 30% were aware of their HIV-positive status, 23% were linked to care, 22% were retained before antiretroviral therapy (ART), 16% had started ART, 16% were currently receiving ART, and 10% had suppressed viral loads. There was site variability (awareness range, 0%–90%; suppressed viral load range, 0%–58%). A total of 2906 PWID were HIV infected, of whom a median of 41% were aware, 36% were linked to care, 31% were retained before ART, 20% had started ART, 18% were currently receiving ART, and 15% had suppressed viral loads. Similar site variability was observed (awareness range: 2%–93%; suppressed viral load range: 0%–47%). Factors significantly associated with awareness were region, older age, being married (MSM) or female (PWID), use of other services (PWID), more lifetime sexual partners (MSM), and needle sharing (PWID). Ongoing injection drug use (PWID) and alcohol use (MSM) were associated with lower awareness.
Conclusions. In this large sample, the major barrier to HIV care engagement was awareness of HIV-positive status. Efforts should focus on linking HIV testing to other essential services.
Clinical Trials Registration. NCT01686750.
The now established benefits of early antiretroviral therapy (ART) for individual health [1, 2] and secondary benefits to partners [3] and the population [4, 5] have prompted the ambitious UNAIDS (Joint United Nations Programme on HIV/AIDS) target of “90-90-90” by 2020 [6]. Achieving this target requires that 90% of persons infected with human immunodeficiency virus (HIV) persons be aware of their status, 90% of those aware of their status receive ART, and 90% of those receiving ART achieve viral suppression. Importantly, 90-90-90 needs to be achieved in all populations, including men who have sex with men (MSM) and persons who inject drugs (PWID) in resource-limited settings.
Data on HIV care continuum outcomes from resource-limited settings are limited and derived primarily from clinic-based populations, where the denominator represents persons who have already accessed HIV care [7–10]. Moreover, these data primarily reflect experiences of heterosexual populations, in which outcomes tend to be superior to those of MSM and PWID in high-income settings [11, 12]. This difference may be more pronounced in resource-limited settings, where higher levels of stigma and discrimination further hinder access.
India is home to the third largest number of HIV-infected persons globally [13]. India's HIV epidemic has been historically driven by heterosexual transmission. Consequently, most programs have targeted heterosexual populations, where prevalence and incidence have subsequently declined [14, 15]. In contrast, HIV prevalence and incidence among PWID and MSM seem to be increasing, or at best stable. These groups have not been a major focus of the National AIDS Control Program until recently [15]. In India, both homosexuality and injection drug use are punishable offences driving individuals to remain hidden. Moreover, in India as in many settings, although HIV testing, ART and harm reduction services are provided by the government, they are delivered in geographically segregated centers, further complicating access [15–17]. Using respondent-driven sampling (RDS), a strategy designed to produce unbiased prevalence estimates in hidden populations [18, 19], we characterized the HIV care continuum among large community-based samples of MSM and PWID from across India.
METHODS
Study Setting
This study was conducted in 27 sites (12 MSM and 15 PWID) across 26 Indian cities (17 states) as the baseline assessment of a cluster-randomized trial (ClinicalTrials.gov identifier NCT01686750). As described elsewhere, cities were selected to represent regions with varying HIV epidemic stages (MSM) and varying drug use epidemic stages (PWID) (Figure 1) [20–22]. In each city, a local partner was identified that maintained a MSM/PWID drop-in center and provided some HIV prevention services (eg, opioid agonist therapy [23, 24]). A single study site was established in each city and targeted either MSM or PWID, except in New Delhi where both an MSM and a PWID site were established (not operational at the same time.)
Figure 1.
Weighted site-level human immunodeficiency virus (HIV) prevalence among 12 022 men who have sex with men (A) and 14 481 persons who inject drugs (B). Weighted HIV prevalences were calculated using respondent driven sampling (RDS) II weights.
Study Population
Eligibility criteria included: (1) age ≥18 years; (2) provision of informed consent; and (3) possession of a valid referral coupon. For MSM, additional criteria were (1) self-identification as male and (2) report or oral or anal sex with a man in the prior 12 months. For PWID, an additional criterion was self-report of drug injection in the prior 2 years. The study population was recruited using RDS, a chain-referral strategy for recruiting hard-to-reach populations [18, 19]. RDS was initiated in each site by 2 or 3 “seeds,” and the target recruitment was 1000 persons per site. For this analysis, we included only persons with a confirmed HIV-positive antibody test result. Correlates of HIV infection in these groups have been described elsewhere [21, 22].
Study Procedures
After giving verbal consent, participants provided a fingerprint image that was converted to a unique hexadecimal code by proprietary software to prevent duplicate enrollment. Participants completed an interviewer-administered survey, which captured information on demographics, network characteristics, risk behavior, HIV testing and treatment, and substance use. Participants underwent rapid HIV testing with pre- and posttest counseling and were given 2 hologram-labeled referral coupons to give to 2 members of their network. Samples were shipped to a central laboratory in Chennai.
Laboratory Methods
HIV infection was diagnosed using 3 rapid tests according to Indian national guidelines [25]. Tests included Alere Determine HIV-1/2 (Alere Medical); First Response HIV card test 1–2.0 (PMC Medical India); and Signal Flow Through HIV 1 + 2 Spot/Immunodot Test Kit (Span Diagnostics). In HIV-infected participants, CD4 was estimated using the FlowCARE Pan-leukogate CD4 (CD45–fluorescein isothiocyanate/CD4-phycoerythrin) assay (Beckman Coulter) and HIV-1-RNA was measured with RealTime HIV-1 assay with a limit of detection of 150 copies/mL (Abbott Laboratories). All results were provided to participants.
Measures
HIV care continuum outcomes, except for viral suppression, were based on self-reported data. Participants had to have achieved the prior step to be eligible for each subsequent step. HIV-infected participants who self-reported either a previous positive HIV test result or being informed by a healthcare provider that he/she was HIV-infected were considered aware of their status. Linkage to care was defined as having ever seen a medical practitioner for HIV. Pre-ART retention was defined as an appointment with an HIV medical practitioner in the prior 6 months or a history of ART (assuming that those who started ART were successfully retained before ART initiation). Current ART was defined as ART use in the prior 30 days.
Statistical Analyses
All samples satisfied RDS process measures (number of waves, homophily, and equilibrium) [21, 22]. Data from seeds were excluded. The RDS-II estimator (Volz-Heckathorn), which weights for network effects, was used to calculate site-specific estimates; weights incorporated self-reported network size (number of MSM/PWID seen in the prior 30 days) [26]. For overall proportions, a composite weight was calculated separately for MSM and PWID, which, in addition to network size, accounted for either the number of men aged 15–59 years [27] (assuming similar proportion of MSM across all cities) or the number of PWID in each city derived from state-level data [28]. Outcomes are presented as the median of site-level values.
Correlates of awareness of HIV-positive status were identified using multilevel logistic regression models with a random intercept for each site and scaled RDS-II weights. Covariates included sociodemographic factors, markers of general healthcare access, and lifetime risk behaviors. Models were constructed separately for MSM and PWID. Variables to represent the number of integrated counseling and testing centers (ICTCs), ART centers, and the proportion of MSM- or PWID-focused targeted interventions in each state were created based on data from the National AIDS Control Organisation, India [15–17]. Separate multivariable models were constructed incorporating these variables and region because of collinearity. As sensitivity analyses, regression analyses were performed unweighted (Supplementary Table 2) and using RDS-I weights (Salganik-Heckathorn estimator) [29]; inferences remained unchanged. All statistical analyses were performed using RDS Analyst software, version 1.0 (http://hpmrg.org) and Stata software, version 12.0 (StataCorp).
Ethical Clearances
This study was approved by the Johns Hopkins and Y.R. Gaitonde Centre for AIDS Research and Education institutional review boards.
RESULTS
12 022 MSM and 14 481 PWID were recruited from September 2012 to December 2013; a total of 1146 (7%) of MSM and 2906 (21%) of PWID were HIV-infected (Figure 1). Site-level characteristics of HIV-infected MSM and PWID are provided in Table 1.
Table 1.
Median Site-Level Characteristics of Human Immunodeficiency Virus-Infected Men Who Have Sex With Men and Persons Who Inject Drugs Across 27 Sites in India (n = 4052)a
| Characteristic | MSM (n = 1146; 12 Sites) |
PWID (n = 2906; 15 Sites) |
||
|---|---|---|---|---|
| Median | Site Range | Median | Site Range | |
| Age, y | 33 | 27–40 | 30 | 27–40 |
| Sex, % | ||||
| Male | 100 | 100–100 | 94.5 | 57.6–100 |
| Female | 4.9 | 0–42.4 | ||
| Marital status, % | ||||
| Never married | 39.5 | 10.6–55.2 | 27.9 | 11.9–70.5 |
| Currently married | 54.4 | 39.0–84.5 | 43.1 | 21.4–76.0 |
| Widowed, divorced, or separated | 5.3 | 0.8–20.1 | 8.0 | 1.2–46.2 |
| Educational level, % | ||||
| Primary school or less | 32.7 | 13.5–70.3 | 33.6 | 9.5–80.6 |
| Secondary school | 45.2 | 15.9–86.5 | 44.1 | 17.3–69.7 |
| High school and above | 17.3 | 0–32.7 | 12.8 | 2.0–37.5 |
| Personal monthly income, $ | 130 | 81–162 | 162 | 97–372 |
| Living situation, % | ||||
| Living with spouse or children | 56.5 | 32.1–69.7 | 55.4 | 6.6–80.5 |
| Living with extended family | 58.1 | 40.4–75.8 | 62.5 | 16.8–87.7 |
| Homeless | 0.7 | 0–27.1 | 0 | 0–68.3 |
| Incarcerated in prior 6 mo | 2.3 | 0–16.4 | 9.4 | 1.2–31.5 |
| Household size, No. | 4 | 2–6 | 4 | 2–5 |
| Network size, No.b | 11 | 5–73 | 15 | 5–30 |
| History of tuberculosis, % | 8.9 | 1.0–45.7 | 13.9 | 1.9–62.2 |
| History of STI, % | 17.6 | 1.7–47.5 | NA | NA |
| Alcohol use (AUDIT), % | ||||
| None or limited use | 70.0 | 55.2–87.3 | 79.3 | 45.2–86.3 |
| Harmful or hazardous use | 17.6 | 7.4–34.1 | 11.8 | 3.3–21.7 |
| Alcohol dependence | 9.6 | 1.7–35.3 | 11.2 | 3.2–34.4 |
| Ever injected drugs, % | 0.5 | 0–51.0 | 100 | 100–100 |
| Injected drugs in prior 6 mo, % | 0.5 | 0–48.4 | 84.9 | 36.7–100 |
| History of needle exchange, % | 0 | 0–14.7 | 50.1 | 4.8–86.1 |
| History of opioid agonist therapy, % | 0 | 0–12.1 | 26.3 | 0–68.7 |
| Median CD4 cell count | 416 | 239–466 | 332 | 251–475 |
| Median log10 HIV RNA, copies/mL | 4.2 | 1.9–4.8 | 4.1 | 1.9–4.8 |
| ICTCs per city, No. | 27 | 5–101 | 7 | 3–17 |
| ART centers per city, No. | 3 | 1–9 | 1 | 1–12 |
| MSM-focused targeted interventions in state, No. | 11 | 5–31 | … | … |
| Targeted interventions in state that were MSM focused, % | 13 | 3–23 | … | … |
| PWID-focused targeted interventions in state, No. | … | … | 23 | 3–55 |
| Targeted interventions in state that were PWID focused, % | … | … | 36 | 2–74 |
Abbreviations: ART, antiretroviral therapy; AUDIT, Alcohol Use Disorders Identification Test; HIV, human immunodeficiency virus; ICTCs, integrated counseling and testing centers; MSM, men who have sex with men; NA, not applicable; PWID, persons who inject drugs; STI, sexually transmitted infection.
a All site-level characteristics were weighted using respondent-driven sampling (RDS) II weights, except for network size and city- and state-level variables [26].
b Network size is the number of MSM/PWID the participant reported seeing in the prior 30 days.
HIV Care Continuum
Among 1146 HIV-infected MSM, a median of 30% were aware of their HIV-positive status, 23% were linked to care, 22% were retained before ART, 16% had started ART, 16% were currently receiving ART, and 10% had suppressed viral loads (Figure 2; Table 2). Awareness in MSM ranged from 0% (Lucknow) to 90% (Chennai) and viral suppression from 0% (Lucknow and Mangalore) to 58% (Madurai).
Figure 2.
Human immunodeficiency virus (HIV) care continuum for 1146 HIV-infected men who have sex with men (MSM) and 2906 HIV-infected persons who inject drugs (PWID) across 27 sites (26 cities) in India. Numbers represent medians of the respondent-driven sampling (RDS) II–weighted site-specific percentages, along with the 10th and 90th percentiles [26]. Of the 911 PWID who had been linked to care and had not started antiretroviral therapy (ART), 90.7% were eligible for ART, and 9.3% were not; of the 446 MSM, 86.9% were eligible for ART, and 13.1% were not.
Table 2.
Human Immunodeficiency Virus Care Continuum Outcomes by Sitea
| Site | HIV infected, % | Diagnosed |
Linked |
Pre-ART Retention |
Initiated ART |
Current ART |
Suppressed VL |
||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Crude, % | Weighted, % | Crude, % | Weighted, % | Crude, % | Weighted, % | Crude, % | Weighted, % | Crude, % | Weighted, % | Crude, % | Weighted, % | ||
| MSM | |||||||||||||
| North | |||||||||||||
| Bhopal | 61 | 34.4 | 23.4 | 23.0 | 17.0 | 21.3 | 16.0 | 9.8 | 11.6 | 8.2 | 10.6 | 6.6 | 10.0 |
| Lucknow | 40 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| New Delhi | 80 | 26.3 | 10.3 | 21.3 | 7.3 | 20.0 | 6.7 | 15.0 | 4.8 | 15.0 | 4.8 | 13.8 | 4.7 |
| Tamil Nadu | |||||||||||||
| Chennai | 66 | 92.4 | 89.9 | 90.9 | 89.6 | 89.4 | 89.4 | 74.2 | 76.1 | 65.2 | 69.4 | 54.6 | 57.0 |
| Madurai | 130 | 82.3 | 76.9 | 80.8 | 76.2 | 78.5 | 74.2 | 73.1 | 71.8 | 70.8 | 70.2 | 58.5 | 58.4 |
| Coimbatore | 159 | 34.0 | 31.5 | 34.0 | 31.5 | 32.1 | 29.9 | 23.3 | 24.3 | 23.3 | 24.3 | 20.1 | 21.5 |
| Karnataka | |||||||||||||
| Bangalore | 91 | 41.8 | 28.5 | 31.9 | 21.0 | 26.7 | 19.8 | 17.6 | 8.0 | 17.6 | 8.0 | 15.4 | 6.1 |
| Mangalore | 20 | 10.0 | 2.5 | 10.0 | 2.5 | 5.0 | 1.4 | 5.0 | 1.4 | 5.0 | 1.4 | 0 | 0 |
| Belgaum | 60 | 38.3 | 42.6 | 31.7 | 32.1 | 30.0 | 32.1 | 26.7 | 29.8 | 25.0 | 28.7 | 11.7 | 10.2 |
| Andhra Pradesh | |||||||||||||
| Hyderabad | 188 | 27.1 | 19.3 | 24.5 | 16.3 | 23.4 | 16.0 | 18.6 | 13.2 | 18.6 | 13.2 | 14.9 | 10.7 |
| Visakhapatnam | 116 | 62.9 | 63.3 | 51.7 | 49.6 | 47.4 | 44.5 | 44.8 | 42.9 | 44.8 | 42.9 | 38.8 | 35.4 |
| Vijayawada | 135 | 38.5 | 34.8 | 29.6 | 24.9 | 25.9 | 23.5 | 20.7 | 18.1 | 20.0 | 18.0 | 6.7 | 7.1 |
| PWID | |||||||||||||
| Northeast | |||||||||||||
| Dimapur | 197 | 78.2 | 81.6 | 72.6 | 75.6 | 70.1 | 71.4 | 59.9 | 58.1 | 55.3 | 53.9 | 46.7 | 45.2 |
| Gangtok | 103 | 92.2 | 92.8 | 88.4 | 85.1 | 68.9 | 67.7 | 48.5 | 43.7 | 43.7 | 39.6 | 34.0 | 26.0 |
| Lunglei | 102 | 69.6 | 60.1 | 52.0 | 48.1 | 48.0 | 42.2 | 44.1 | 40.6 | 40.2 | 35.0 | 28.4 | 25.2 |
| Aizawl | 306 | 68.0 | 61.8 | 44.1 | 40.6 | 38.6 | 35.5 | 21.9 | 20.2 | 17.0 | 18.3 | 12.4 | 14.9 |
| Churachandpur | 279 | 39.8 | 35.9 | 33.0 | 32.5 | 30.8 | 32.1 | 24.7 | 29.4 | 23.7 | 29.1 | 21.2 | 24.5 |
| Imphal | 319 | 43.3 | 38.8 | 39.5 | 36.3 | 32.3 | 27.8 | 23.8 | 22.6 | 21.9 | 20.1 | 17.2 | 15.4 |
| Moreh | 198 | 59.1 | 66.9 | 43.4 | 54.6 | 42.9 | 54.3 | 37.4 | 47.4 | 36.9 | 47.3 | 30.8 | 42.3 |
| North | |||||||||||||
| Amritsar | 226 | 38.9 | 40.8 | 12.8 | 13.0 | 11.1 | 11.8 | 10.2 | 11.4 | 8.0 | 4.6 | 6.2 | 2.9 |
| Chandigarh | 121 | 17.4 | 19.6 | 7.4 | 6.1 | 6.6 | 4.7 | 5.0 | 3.8 | 4.1 | 2.0 | 2.5 | 1.7 |
| Ludhiana | 228 | 26.8 | 25.5 | 20.6 | 20.6 | 12.3 | 13.1 | 7.5 | 8.9 | 7.0 | 7.5 | 4.0 | 5.7 |
| New Delhi | 166 | 18.7 | 17.8 | 6.6 | 2.8 | 4.8 | 2.1 | 1.8 | 0.9 | 1.2 | 0.4 | 0 | 0 |
| West/central | |||||||||||||
| Bilaspur | 153 | 18.3 | 16.6 | 3.9 | 2.9 | 3.9 | 2.9 | 2.6 | 2.2 | 0.7 | 0.7 | 0 | 0 |
| Bhubaneshwar | 61 | 72.1 | 70.2 | 63.9 | 65.3 | 63.9 | 65.3 | 52.5 | 53.9 | 52.5 | 53.9 | 42.6 | 47.2 |
| Kanpur | 353 | 2.8 | 2.4 | 1.7 | 1.5 | 1.1 | 1.3 | 0.3 | 0.2 | 0 | 0 | 0 | 0 |
| Mumbai | 94 | 50.0 | 54.4 | 40.4 | 43.2 | 28.7 | 30.7 | 11.7 | 14.7 | 11.7 | 14.7 | 8.5 | 11.1 |
Abbreviations: ART, antiretroviral therapy; HIV, human immunodeficiency virus; MSM, men who have sex with men; PWID, persons who inject drugs; VL, viral load.
a Only those who achieved the prior step along the continuum were considered for achievement of the next step. Crude values represent percentages of all HIV-infected persons achieving each outcome; weighted proportions were calculated using respondent-driven sampling (RDS) II weights.
Among 2906 HIV-infected PWID, a median of 41% were aware of their HIV-positive status, 36% were linked to care, 31% were retained before ART, 20% initiated ART, 18% were currently receiving ART, and 15% had suppressed viral loads. Awareness in PWID ranged from 2% (Kanpur) to 93% (Gangtok) and viral suppression from 0% (Kanpur, New Delhi, and Bilaspur) to 47% (Bhubaneswar).
Awareness of HIV-Positive Status Among MSM
Of 643 MSM not aware of their HIV-positive status, 357 (51%) reported being tested previously. Of these 357, 167 (43%) were tested >1 year earlier, and 25 (14%) reported never receiving their results. Of 286 never tested, primary reasons for never being tested were low perceived risk (48%), not knowing where to go (24%), no time (14%), and being afraid of a positive diagnosis (8%). Among 503 aware of their HIV-positive status, primary reasons for testing were wanting to know one's status (44%), symptoms (18%), physician referral (9%), routine testing (7%), and sex work (7%). Of those tested, 78% were tested at a government ICTC.
Among HIV-infected MSM, factors positively associated with awareness of HIV-positive status in multivariable analysis included older age (odds ratio [OR] per 10 years, 2.11), being married (OR, 1.85), being widowed, divorced, or separated (vs never married; OR, 2.73), having more lifetime male sexual partners (OR for ≥300 vs ≤5 partners, 2.42), history of a sexually transmitted infection (OR, 3.49), and history of tuberculosis (OR, 26.5) (P < .05 for all; Table 3). Persons with harmful or hazardous alcohol use (OR, 0.22) or alcohol dependence (OR, 0.09) were significantly less likely to be aware of their HIV-positive status (P < .001 for both). Finally, persons living in the southern state of Tamil Nadu (OR, 31.9) were significantly more likely to be aware than those in North/Central India. In separate multivariable models, the number of ICTCs and ART centers were significantly associated with higher awareness (OR, 1.22 [P < .001] and 1.47 [P = .03], respectively); the proportion of MSM-focused targeted interventions was not significantly associated with awareness.
Table 3.
Correlates of Awareness of Human Immunodeficiency Virus-Positive Status Among Men Who Have Sex With Men and Persons Who Inject Drugs Across 27 Sites in India (n = 4052)a
| Correlate | MSM (n = 1146) |
PWID (n = 2906) |
||||||
|---|---|---|---|---|---|---|---|---|
| Unadjusted |
Adjusted |
Unadjusted |
Adjusted |
|||||
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | |
| Age (per 10 y) | 2.66 | 2.17–3.26 | 2.11 | 1.63–2.74 | 1.50 | 1.26–1.77 | 1.26 | 0.99–1.61 |
| Sex | ||||||||
| Male | … | … | … | … | Reference | … | Reference | … |
| Female | … | … | … | … | 2.88 | 1.69–4.91 | 4.33 | 2.36–7.96 |
| Sexual identity | ||||||||
| Panthi (penetrative) | Reference | … | … | … | … | … | … | … |
| Kothi (receptive) | 1.86 | 0.75–4.60 | … | … | … | … | … | … |
| Double-decker (both) | 0.86 | 0.26–2.89 | … | … | … | … | … | … |
| Gay/MSM | 1.03 | 0.25–4.22 | … | … | … | … | … | … |
| Bisexual | 0.92 | 0.30–2.86 | … | … | … | … | … | … |
| Marital status | ||||||||
| Never married | Reference | … | Reference | … | Reference | … | … | … |
| Currently married or living with partner | 2.89 | 1.79–4.65 | 1.85 | 1.06–3.23 | 1.40 | 0.90–2.18 | … | … |
| Widowed, divorced, or long-term partner | 3.72 | 2.03–6.80 | 2.73 | 1.80–4.14 | 1.34 | 1.02–1.76 | … | … |
| Educational level | ||||||||
| Primary school or less | Reference | … | … | … | Reference | … | Reference | … |
| Secondary school | 1.04 | 0.54–2.00 | … | … | 1.22 | 0.90–1.65 | 1.24 | 0.90–1.71 |
| High school graduate | 0.57 | 0.24–1.34 | … | … | 1.50 | 1.03–2.21 | 1.59 | 1.07–2.35 |
| Personal monthly income, $ | ||||||||
| <50 | Reference | … | … | Reference | … | Reference | … | |
| 50–115 | 0.59 | 0.32–1.09 | … | … | 0.57 | 0.40–0.80 | 0.66 | 0.47–0.93 |
| >115 | 0.46 | 0.23–0.85 | … | … | 0.69 | 0.47–1.01 | 0.73 | 0.50–1.07 |
| Homeless | 0.48 | 0.14–1.65 | … | … | 0.48 | 0.14–1.65 | … | … |
| Incarcerated in prior 6 mo | 0.54 | 0.18–1.60 | … | … | 0.50 | 0.26–0.97 | … | … |
| Ever injected drugs | 0.24 | 0.04–1.35 | … | … | … | … | … | … |
| Time since 1st injection (per 10 y) | … | … | … | … | 1.49 | 1.22–1.83 | 1.32 | 1.04–1.66 |
| Ever shared needle or syringe | 0.28 | 0.08–0.94 | … | … | 1.47 | 1.06–2.03 | 1.56 | 1.06–2.30 |
| Injection frequency in prior 6 mo | ||||||||
| None | Reference | … | … | … | Reference | … | Reference | … |
| Less than daily | 0.85 | 0.16–4.40 | … | … | 0.62 | 0.38–1.02 | 0.72 | 0.43–1.21 |
| Daily | 0.07 | 0.01–0.59 | … | … | 0.40 | 0.27–0.61 | 0.46 | 0.27–0.79 |
| Alcohol use (AUDIT) | ||||||||
| None or limited use | Reference | … | Reference | … | Reference | … | … | … |
| Harmful or hazardous use | 0.30 | 0.14–0.68 | 0.22 | 0.11–0.42 | 0.73 | 0.49–1.09 | … | … |
| Alcohol dependence | 0.14 | 0.06–0.35 | 0.09 | 0.03–0.30 | 0.60 | 0.30–1.19 | … | … |
| No. of lifetime sexual partners (per 10 partners) | 1.01 | 1.01–1.02 | … | … | 0.99 | 0.97–1.02 | … | … |
| Lifetime male sexual partners, No. | ||||||||
| 1–5 | Reference | … | Reference | … | … | … | … | … |
| 6–15 | 0.50 | 0.26–0.96 | 0.57 | 0.31–1.08 | … | … | … | … |
| 16–60 | 0.64 | 0.35–1.17 | 0.50 | 0.25–0.99 | … | … | … | … |
| 61–300 | 1.33 | 0.75–2.34 | 1.44 | 0.81–2.55 | … | … | … | … |
| ≥300 | 2.01 | 1.05–3.87 | 2.42 | 1.13–5.16 | … | … | … | … |
| Disclosed MSM to anyone | 1.50 | 0.94–2.38 | … | … | … | … | … | … |
| Lifetime history of sex work | 1.75 | 1.23–2.49 | … | … | 0.85 | 0.33–2.17 | … | … |
| Lifetime history of STI | 3.41 | 2.07–5.62 | 3.49 | 1.93–6.32 | … | … | … | … |
| Any history of tuberculosis diagnosis | 16.11 | 5.33–48.7 | 26.5 | 7.21–97.1 | 2.88 | 1.74–4.76 | 2.53 | 1.52–4.20 |
| Ever visited NEP | 0.24 | 0.04–1.36 | … | … | 1.08 | 0.69–1.69 | … | … |
| Ever visited OAT | 0.11 | 0.02–0.80 | … | … | 2.03 | 1.44–2.85 | 2.28 | 1.59–3.27 |
| Region (PWID) | ||||||||
| North/central India | … | … | … | … | Reference | … | … | … |
| Northeastern India | … | … | … | … | 5.62 | 1.69–18.7 | 2.99 | 0.97–9.17 |
| Region (MSM) | ||||||||
| Central/north | Reference | … | Reference | … | … | … | … | … |
| Karnataka | 3.09 | 0.41–23.1 | 4.40 | 0.40–48.70 | … | … | … | … |
| Tamil Nadu | 26.0 | 2.97–228.7 | 31.9 | 2.34–433.6 | … | … | … | … |
| Andhra Pradesh | 6.76 | 1.01–45.0 | 8.77 | 0.91–84.6 | … | … | … | … |
| No. of ICTCs in city (per 5 centers)b | 1.20 | 1.06–1.37 | 1.22 | 1.06–1.40 | 0.80 | 0.47–1.36 | … | … |
| No. of ART centers in city (per 1 center)b | 1.43 | 1.00–2.04 | 1.47 | 1.04–2.09 | 1.02 | 0.91–1.40 | … | … |
| No. of MSM-focused TIs in state (per 1 TI) | 0.98 | 0.90–1.06 | … | … | … | … | … | … |
| % of TIs that were MSM focused (per 10%) | 0.93 | 0.30–2.85 | … | … | … | … | … | … |
| No. of PWID-focused TIs in state (per 1 TI) | … | … | … | … | 1.00 | 0.97–1.04 | … | … |
| % of TIs that were PWID-focused (per 10%) | … | … | … | … | 1.31 | 0.98–1.74 | … | … |
Abbreviations: ART, antiretroviral therapy; AUDIT, Alcohol Use Disorders Identification Test; CI, confidence interval; ICTCs, integrated counseling and testing centers; MSM, men who have sex with men; NEP, needle exchange program; OAT, opioid agonist therapy; OR, odds ratio; PWID, persons who inject drugs; STI, sexually transmitted infection; TIs, targeted intervention.
a Models incorporated scaled respondent-driven sampling (RDS) II weights; see Supplementary materials for unweighted model results.
b Separate multivariable models were constructed using these variables, given collinearity with the region variable.
Factors Associated With Awareness of HIV-Positive Status Among PWID
Of 1682 PWID not aware of their HIV-positive status, 693 (34%) reported being tested previously. Of these 693 individuals, 344 (52%) reported being tested >1 year earlier and 232 (31%) reported never receiving their results. Of 989 never tested, primary reasons for not being tested were low perceived risk (42%), not knowing where to go (25%), no time (15%), and being afraid of a positive diagnosis (12%). Among 1224 PWID aware of their HIV-positive status, primary reasons for getting testing were wanting to know one's status (55%), physician referral (16%), symptoms (8%), and being taken for a test by an outreach worker (7%). Most were tested at a government ICTC (74%).
Among HIV-infected PWID, in multivariable analysis, factors significantly associated with awareness of HIV-positive status included higher educational attainment (OR for high school graduate vs no education, 1.59), female sex (OR, 4.33), longer duration of injection drug use (OR per 10 years of injection history, 1.32), and ever sharing a needle (OR, 1.56) (P < .05 for all; Table 3). Persons with higher income (OR for monthly income $50–$115 vs <$50, 0.66) and those injecting daily (OR for daily injection vs no injection in past 6 months, 0.46) were significantly less likely to be aware. Persons living in Northeastern India were significantly more likely to be aware than those in other states (OR, 2.99). In univariable analysis, neither the number of ICTCs nor ART centers was associated with higher levels of awareness, but a having a larger proportion of PWID-focused targeted intervention organizations in the state was associated with higher awareness (OR, 1.3), though this association did not achieve statistical significance. None of these factors were significantly associated with awareness in multivariable analysis.
DISCUSSION
Heterosexual transmission accounts for most HIV infections in India. During the past decade, India has made substantial progress, with HIV prevalence among women attending antenatal clinics declining by approximately 50% [30]. However, similar declines have not been observed among MSM and PWID; failure to make inroads in these key populations may stymie UNAIDS targets in India and other resource-limited settings where they have received less attention than heterosexual epidemics. In this large multicity community-based sample of key populations from India, much needs to be done to achieve the ambitious UNAIDS 90-90-90 target. Less than half of the PWID and a third of the MSM were aware of their HIV-positive status. Furthermore, among those aware of their status only 52% of PWID and 68% of MSM were receiving sustained ART. It was encouraging, however, that 78% of MSM and 83% of PWID receiving ART were virologically suppressed. Although these proportions fall short of the UNAIDS target, they are comparable to those in the general population in India [8, 31] and high-income settings [11, 12], highlighting the fact that high viral suppression can be achieved in populations historically considered difficult to treat.
This report represents one of the first multisite characterizations of the HIV care continuum in PWID and MSM from any resource-limited setting. Our use of identical survey methods in both populations during a similar time frame provides unbiased estimates of the care continuum. Much of the prior data from resource-limited settings, including Africa and Asia, include only steps occurring after HIV diagnosis [9, 10, 32, 33]. A prior study evaluated the care continuum among 1057 HIV-positive persons at government ICTCs in the 6 high-prevalence Indian states [9] and found a comparatively higher proportion linked to care (74%). A cohort of 7701 HIV-positive persons in Andhra Pradesh similarly reported that 82% were linked to care, 70% were retained, 55% had started ART and 31% were virologically suppressed [8]. Both studies included primarily heterosexual populations.
Our data highlight the fact that MSM and PWID lag behind other HIV-infected persons in India with respect to linkage to care and other HIV care continuum outcomes and also reinforce the importance of community-based samples to measure the full spectrum of population-level engagement in HIV care; the majority of individuals identified as HIV infected in our sample would have been missed by approaches focusing only on those in care. More importantly, our data identify striking regional differences in the HIV care continuum that have not been described elsewhere, to our knowledge. For MSM, awareness of status was in part explained by regional variability in the numbers of ICTCs and ART centers while for PWID it was not. There were alarmingly low levels of engagement in HIV care in cities not located in states with high HIV prevalence where there are limited HIV services and even fewer services targeting MSM and/or PWID. For example, Uttar Pradesh the state with the largest population in India, has historically been considered a low–HIV prevalence state and consequently has fewer HIV testing and ART centers than similarly sized high–HIV prevalence states [15–17]. The lowest levels of awareness of HIV-positive status were observed in this state among both groups. By contrast, in cities where HIV programs have been implemented since the early years of the AIDS epidemic (eg, in Tamil Nadu and Manipur), the care continuum was comparable to that of MSM/PWID in high-income [11, 12] settings, if not better.
In all but 2 sites, the primary barrier to successful engagement was awareness of HIV-positive status. Awareness of HIV-positive status was better among those who had accessed other services, including tuberculosis treatment, opioid agonist therapy, and sexually transmitted infection treatment. Even associations observed with marriage (MSM) and female sex (PWID) probably reflect differential service access because women (and spouses) are typically tested for HIV in India in antenatal care. These data collectively reinforce the importance of integrating HIV testing across all services where vulnerable populations may come into contact with the healthcare system, which is currently not the norm in India. Moreover, testing in all sites should follow current government guidelines by offering same-day rapid, point-of-care testing so that individuals are not lost because of the requirement of a follow-up visit for receiving results [34].
Still, broad integration of HIV testing may not reach everyone, because some individuals do not come into contact with any services. Our data suggest that network-driven recruitment strategies such as RDS may be an efficient approach for reaching such persons. Each sample was recruited in approximately 3 months, and anecdotal reports from sites suggest that a large proportion of the participants recruited by RDS had not previously sought services at the organization. Improved access to testing and outreach alone will; however, not be sufficient to improve other outcomes such as linkage and ART initiation, which were lower than general population estimates in India [8, 9, 31]. To really improve outcomes across the entire care continuum, additional strategies that target other barriers will be needed, including integration of services, peer navigation and incentives [34, 35]. These will be most effective if combined with large-scale public policy changes that affect the broader social environment (eg, decriminalization of same-sex behavior).
Some of what we observed may reflect the differential availability of ICTCs and ART centers in different cities across India as well as the fact that for the most part HIV counseling and testing and ART are provided in geographically segregated centers [15–17]. Although all 448 ART centers in India have a colocated ICTC within the same campus, they are usually in large government hospitals, and most ICTCs (n = 4694) are geographically distinct from ART centers. Most participants reported being tested in a government ICTC, where the common practice is referral to an ART center. It was encouraging in the MSM sites that more widespread coverage of ICTC and ART centers seemed to be associated with more favorable outcomes; the lack of association in PWID sites may reflect that what is important for this group is having efforts specifically targeted at linking PWID to these services.
We were limited by the self-reported nature of the HIV care continuum outcomes with the exception of viral suppression. However, our surveys were translated into local languages and in each city, common terms for “HIV doctor” and ART were collected during a formative research. Although data were collected from a large number of sites, cities were purposely selected, and therefore estimates may not be nationally representative.
In conclusion, in this large community-based sample of MSM and PWID across India, the major barrier to successful engagement in HIV care was diagnosis. Several of our findings suggest that the UNAIDS 90-90-90 target could be achieved in these marginalized groups. First, viral suppression among those receiving ART was comparable to that in high-income settings. Second, sites with historically higher HIV prevalence and better access to government-led interventions had higher levels of awareness, suggesting that better engagement can be achieved with expanded HIV programming. Efforts need to be made to expand HIV testing, possibly through integration with other existing services, along with efforts to improve downstream outcomes of linkage to care and viral suppression.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online (http://cid.oxfordjournals.org). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.
Notes
Acknowledgments. We thank Jungen Yi for assistance with manuscript preparation and the National AIDS Control Organisation, India, all of our partner nongovernmental organizations throughout India, and the countless participants, without whom this research would not have been possible.
Author contributions. S. H. M., G. M. L., and S. S. S. were responsible for the design of the study, data collection, interpretation of data and drafting the manuscript. A. M. M. was responsible for data analysis. S. S., N. D., M. S. K. and D. D. C. contributed to the design of the study and provided critical feedback on the manuscript. A. K. S. led data collection. P. N. conducted laboratory testing. All authors read and approved the final manuscript.
Financial support. This work was supported by the National Institutes of Health (grants MH 89266 and DA 032059). Additional support was provided by the Johns Hopkins University Center for AIDS Research (grants P30 AI094189 and K24 DA035684).
Potential conflicts of interest. All authors: No reported conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.Kitahata MM, Gange SJ, Abraham AG et al. Effect of early versus deferred antiretroviral therapy for HIV on survival. N Engl J Med 2009; 360:1815–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Grinsztejn B, Hosseinipour MC, Ribaudo HJ et al. Effects of early versus delayed initiation of antiretroviral treatment on clinical outcomes of HIV-1 infection: results from the phase 3 HPTN 052 randomised controlled trial. Lancet Infect Dis 2014; 14:281–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Cohen MS, Chen YQ, McCauley M et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med 2011; 365:493–505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tanser F, Barnighausen T, Grapsa E, Zaidi J, Newell ML. High coverage of ART associated with decline in risk of HIV acquisition in rural KwaZulu-Natal, South Africa. Science 2013; 339:966–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Montaner JS, Lima VD, Barrios R et al. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study. Lancet 2010; 376:532–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.UNAIDS. 90-90-90: an ambitious treatment target to help end the AIDS epidemic. Available at: http://www.unaids.org/sites/default/files/media_asset/90-90-90_en_0.pdf Accessed 31 December 2014.
- 7.Malta M, Magnanini MM, Strathdee SA, Bastos FI. Adherence to antiretroviral therapy among HIV-infected drug users: a meta-analysis. AIDS Behav 2010; 14:731–47. [DOI] [PubMed] [Google Scholar]
- 8.Alvarez-Uria G, Pakam R, Midde M, Naik PK. Entry, retention, and virological suppression in an HIV cohort study in India: description of the cascade of care and implications for reducing HIV-related mortality in low- and middle-income countries. Interdiscip Perspect Infect Dis 2013; doi:10.1155/2013/384805. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sarna A, Sebastian M, Bachani D, Sogarwal R, Battala M. Pretreatment loss-to-follow-up after HIV diagnosis from 27 counseling and testing centers across India: findings from a cohort study. J Int Assoc Provid AIDS Care 2014; 13:223–31. [DOI] [PubMed] [Google Scholar]
- 10.Feelemyer J, Des Jarlais D, Arasteh K, Uuskula A. Adherence to antiretroviral medications among persons who inject drugs in transitional, low and middle income countries: an international systematic review. AIDS Behav 2015; 19:575–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Westergaard RP, Hess T, Astemborski J, Mehta SH, Kirk GD. Longitudinal changes in engagement in care and viral suppression for HIV-infected injection drug users. AIDS 2013; 27:2559–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Millett GA, Peterson JL, Flores SA et al. Comparisons of disparities and risks of HIV infection in black and other men who have sex with men in Canada, UK, and USA: a meta-analysis. Lancet 2012; 380:341–8. [DOI] [PubMed] [Google Scholar]
- 13.UNAIDS. Global report: UNAIDS report on the global AIDS epidemic. Available at: http://www.unaids.org/en/media/unaids/contentassets/documents/epidemiology/2013/gr2013/UNAIDS_Global_Report_2013_en.pdf Accessed 31 December 2014.
- 14.Gupte MD, Mehendale SM, Pandav CS, Paranjape RS, Kumar MS. HIV-1 prevalence in young adults in south India. Lancet 2006; 368:114–6. [DOI] [PubMed] [Google Scholar]
- 15.National AIDS Control Organisation. Annual report. Available at: http://www.naco.gov.in/NACO/Quick_Links/Publication/Annual_Report/NACO_Annual_Report/Annual_Report_NACO_2014-15_English/. Accessed 18 August 2015.
- 16.National AIDS Control Organisation. Details of ICTC centres. Available at: http://www.naco.gov.in/NACO/Directory_of_Services/Details_of_ICTC_Centres_/ Accessed 31 December 2014.
- 17.National AIDS Control Organisation. List of ART centres in India. Available at: http://www.naco.gov.in/upload/2014%20mslns/CST/List%20of%20ART%20Centres%20in%20India.pdf Accessed 18 August 2015.
- 18.Heckathorn DD. Respondent-driven sampling II: deriving valid population estimates from chain-referral samples of hidden populations. Soc Probl 2002; 49:11–34. [Google Scholar]
- 19.Heckathorn D. Respondent-driven sampling: a new approach to the study of hidden populations. Soc Probl 1997; 44:174–99. [Google Scholar]
- 20.Solomon SS, Lucas GM, Celentano DD, Sifakis F, Mehta SH. Beyond surveillance: a role for respondent-driven sampling in implementation science. Am J Epidemiol 2013; 178:260–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Solomon SS, Mehta SH, Srikrishanan AK et al. High HIV prevalence and incidence among men who have sex with men (MSM) across 12 cities in India. AIDS 2015; 29:723–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lucas G, Solomon SS, Srikrishanan A et al. High HIV burden among people who inject drugs in 15 Indian cities. AIDS 2015; 29:619–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.MacArthur GJ, Minozzi S, Martin N et al. Opiate substitution treatment and HIV transmission in people who inject drugs: systematic review and meta-analysis. BMJ 2012; 345:e5945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Gowing L, Farrell MF, Bornemann R, Sullivan LE, Ali R. Oral substitution treatment of injecting opioid users for prevention of HIV infection. Cochrane Database Syst Rev 2011; doi:10.1002/14651858.CD004145.pub4. [DOI] [PubMed] [Google Scholar]
- 25.National AIDS Control Organisation. Guidelines on HIV testing. Available at: http://naco.gov.in/upload/Policies & Guidelines/5-GUILDELINES FOR HIV TESTING.pdf Accessed 31 December 2014.
- 26.Volz E, Heckathorn DD. Probability based estimation theory for respondent driven sampling. J Off Stat 2008; 24:79–97. [Google Scholar]
- 27.Office of the Registrar General & Census Commissioner of India. 2011 Census district profiles. Available at: http://censusindia.gov.in/Tables_Published/Basic_Data_Sheet.aspx Accessed 8 May 2014.
- 28.National Institute of Medical Statistics and National Aids Control Organisation. Technical report: India HIV estimates. 2010. Available at: http://naco.gov.in/upload/Surveillance/Reports & Publication/Technical Report India HIV Estimates 2010.pdf Accessed 31 December 2014.
- 29.Salganik M, Heckathorn DD. Sampling and estimation in hidden populations using respondent-driven sampling. Soc Probl 2004; 49:11–34. [Google Scholar]
- 30.Arora P, Kumar R, Bhattacharya M, Nagelkerke NJ, Jha P. Trends in HIV incidence in India from 2000 to 2007. Lancet 2008; 372:289–90. [DOI] [PubMed] [Google Scholar]
- 31.Hingankar NK, Thorat SR, Deshpande A et al. Initial virologic response and HIV drug resistance among HIV-Infected individuals initiating first-line antiretroviral therapy at 2 clinics in Chennai and Mumbai, India. Clin Infect Dis 2012; 54(suppl 4):S348–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Boyles TH, Wilkinson LS, Leisegang R, Maartens G. Factors influencing retention in care after starting antiretroviral therapy in a rural South African programme. PLoS One 2011; 6:e19201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Clouse K, Pettifor AE, Maskew M et al. Patient retention from HIV diagnosis through one year on antiretroviral therapy at a primary health care clinic in Johannesburg, South Africa. J Acquir Immune Defic Syndr 2013; 62:e39–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Meyer JP, Althoff AL, Altice FL. Optimizing care for HIV-infected people who use drugs: evidence-based approaches to overcoming healthcare disparities. Clin Infect Dis 2013; 57:1309–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Solomon SS, Srikrishnan AK, Vasudevan CK et al. Voucher incentives improve linkage to and retention in care among HIV-infected drug users in Chennai, India. Clin Infect Dis 2014; 59:589–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.