Jacobson, Jerry O. PhD*; Creswell, Jacob MPH†; Guardado, Maria Elena MD, MPH‡; Lee, Janet C. MPH§; Isabel Nieto, Ana MD¶; Paz-Bailey, Gabriela MD, MSc, PhD§
Positive prevention interventions (PIs), which target persons with diagnosed HIV infection, have demonstrated effectiveness in supporting the adoption of safer sexual behaviors in resource-limited settings in the context of both counseling and testing1 as well as care and treatment.2,3 Scale-up of antiretroviral therapy (ART) brings people with HIV (PWH) in contact with diagnostic and treatment services and makes access to positive prevention more important to reduce the risk of transmission of primary drug-resistant HIV strains to HIV-negative partners, and the risk of acquiring other sexually transmitted infections (STIs) and dual or recombinant HIV infection.4–6 At the same time, rapid ART scale-up can exacerbate existing health systems resource constraints, posing a challenge to prevention implementation.7,8 Yet, there is scarce information on actual exposure to positive prevention components in resource-limited settings where ART has scaled up rapidly.
In El Salvador, ART coverage of individuals eligible for treatment (presence of HIV-related symptoms or CD4<350 cells/mm3) increased rapidly from 5% in 2001 to 96% in 2008, with 7104 patients attending 16 specialized, public HIV clinics.9 El Salvador accounts for just 15% of the Central American region's estimated number of PWH but 34% of patients on ART.5,9 HIV infection prevalence is greatest in men who have sex with men (MSM), at 11%, the highest prevalence among MSM in the region.10,11 Prevention among MSM is thus of particular concern. However, MSM also face important challenges in accessing health services due to stigma and discrimination by health providers,12 a recognized barrier to implementation of positive prevention globally.13,14
Recent World Health Organization guidance for resource-limited settings recommends that prevention components at health services include risk-behavior screening, family and partner counseling and testing, STI screening and management, condom promotion and provision, safer sex and risk reduction counseling, and psychosocial support.15 We estimate coverage levels of posttest counseling (POC), subsequent exposure to HIV PIs, and STI management among MSM, heterosexual men (HM) and women with diagnosed HIV infection and attending public hospitals or HIV support groups in the 3 largest cities in El Salvador. Our analysis centers on individuals diagnosed at least 12 months before the study to allow time for exposure to interventions posterior to testing.
MATERIALS AND METHODS
Data are from a larger study of men and women aged 18 years or older with previous confirmed diagnosis of HIV infection, who reported having anal or vaginal sex in the past 12 months, recruited between March and September of 2008 using consecutive sampling from the HIV service areas at the country's 2 major public hospitals, Rosales and Zacamil. Participants were also recruited from HIV peer support groups at these and 6 other public hospitals providing ART in El Salvador's largest cities: San Salvador, Santa Ana, and San Miguel. The Ministry of Health estimates public hospitals account for 80% of patients on ART. Participants completed face-to-face and audio computer-assisted interviews (ACASI) based on standardized survey instruments16 on demographics, past 12-month risk behaviors, HIV diagnosis, care and treatment services, and perceived HIV-related stigma and discrimination. Physical examinations and STI testing were also provided, including vaginal swabs (female participants) and urine (male) samples to identify Chlamydia trachomatis, Neisseria gonorrhoeae, Trichomonas vaginalis, and Mycoplasma genitalium, using polymerase chain reaction. Blood samples were tested for syphilis, based on reactive rapid plasma reagin and positive Treponema particle-agglutination assay (TPPA), and for herpes simplex virus type 2 (HSV-2), with HerpeSelect (Focus Technologies, Cypress, CA). Active syphilis was defined as a rapid plasma reagin titer ≥1:8 and a reactive TPPA result. Rosales National Hospital's ethics committee approved the study protocol. We examine study participants who reported being diagnosed with HIV at least 12 months before the survey interview. Additional details on data collection are available in the main study publication.17
Primary measures of positive prevention components were having received HIV POC (based on a question item on whether counseling was received before testing, upon receiving the result, at both or neither of these times), subsequent exposure to HIV PIs (defined broadly as having received counseling and education during regularly scheduled medical visits at the HIV clinic, having participated in information or educational activities on HIV/AIDS during the past year, or having attended counseling sessions for health reasons since HIV diagnosis), treatment-seeking for STIs (defined for participants reporting any STI symptoms in the past 12 months as having sought help for those symptoms at a health clinic or hospital) and completing prescribed treatment for STIs (for participants who reported receiving a medical prescription for past-year symptoms). Participants were not asked about experiencing “STI symptoms” per se, but whether they had experienced each of a list of specific signs and symptoms (i.e., ulcers, sores, swelling, odors, pain, etc.) using local terms.
HIV-related discrimination in health settings was defined as having ever experienced “any kind of mistreatment at health services” or “poor health services or denial of treatments or medicines” due to being a person living with HIV/AIDS. Discrimination in other contexts was defined as reporting isolation or abandonment by friends or family, rejection by neighbors or others, threats, abuse, insults, violence, denial of access to housing or religious ceremonies, denial of a promotion at work, loss of employment due to HIV infection, or undesired disclosure of HIV infection. Similar measures have been used previously in studies of PWH in the Central American region.18
Male participants who reported any of their last 3 sexual partners as male, or who self-identified as (1) homosexual or gay, (2) bisexual, or (3) transvestite, transsexual, or transgender, were defined as MSM. Other male respondents were defined as heterosexual.
Subgroup (MSM, women, and HM) differences were explored using F-tests for categorical variables and Kruskal-Wallis tests for continuous variables. Differences were assessed on demographic variables, years since HIV diagnosis, sexual and drug transmission risk variables, current STI, whether currently on ART, reason for HIV testing and test location, stigma and discrimination measures, and measures of exposure to prevention components.
When a difference among the 3 subgroups was detected, we then tested for differences among subgroup pairs. For brevity, the pairwise findings are described in the text and the global differences appear in the tables. Additionally, Mann-Whitney 2-sample tests were used to examine whether POC and PI coverage changed over time by assessing median differences in (untransformed and log-transformed) time since diagnosis.
Multivariate logistic models were developed to identify factors independently associated with POC and PI. Variables with a significant unadjusted odds ratio (P ≤ 0.1) relative to the respective outcome were included as covariates while subgroup indicators were retained in all models. Likelihood-ratio tests were used to compare alternate model specifications. Interactions between significant main effects were also considered.
Of the 1133 individuals approached to participate in the study meeting eligibility criteria, 72% (N = 811) agreed to participate. Twenty percent (N = 159) were from HIV support groups. 698 (86.1%) participants had been diagnosed at least a year before the study, but sexual orientation could not be determined for 9 of these. The final sample of 689 participants included 110 MSM, 237 HM, and 342 women (Table 1).
Median age of the sample was 35 and ranged from 18 to 73 years. HM were significantly older than MSM and women. Completion of primary education was highest among MSM (89.1%) compared with HM (68.4%) and women (53.5%) (all P < 0.001). Most participants were single (53.9%) with MSM least likely to be married or living with a partner (14.7%) (P < 0.001). Overall, participants had been diagnosed HIV infection for 4.0 years at the median, although HM had been diagnosed for 3.0 years (P = 0.05). Two participants who reported being diagnosed >20 years ago (in 1983 and 1984) were MSM who resided in the United States at the time of diagnosis.
Sexual Behavior, Substance Use, and STIs
Over the 12 months before the study, all participants had at least 1 sex partner, in accordance with study eligibility criteria. Most reported multiple partners (76.4%), many reported at least 1 casual partner (31.4%), whereas few reported multiple casual partners (13.0%).
Indicators of potential transmission risk varied by subgroup. Overall, 21.6% reported unprotected sex during the most recent intercourse with a casual partner and 28.7% had had unprotected sex with at least 1 noncommercial partner in the past 12 months. Regarding unprotected sex at last intercourse with a casual partner, HM had moderately lower rates compared with other subgroups combined, but pairwise comparisons were nonsignificant. Women were more likely (68.5%) and MSM were least likely (44.1%) to report multiple stable sex partners (P < 0.001); both were different than the estimate for HM (59.7%) (both P < 0.05). In contrast, MSM were much more likely to report multiple casual sex partners (35.2%) versus either HM (14.9) or women (4.5) (both P < 0.001); the latter difference between HM and women was also significant (P < 0.01). Ten percent of MSM had received money for sex in the past year compared with 3.2% among women (P < 0.004) and just 0.9% of HM (P < 0.001). Both HM (13.6%) and MSM (9.1%) were more likely to have paid for sex than women (0.6%) (both P < 0.001). MSM had twice the rates of past-month alcohol use (20.0%) and past-year drug use (10.5%) compared to HM (11.0% and 3.8%, respectively) and 4 times those in women (4.7% and 1.2%, respectively) (all P < 0.05), although drug and alcohol use were low overall.
HSV-2 infection was high in all subgroups (>78%) but greater in MSM and women compared with HM (P < 0.01). Women were more likely to have a treatable STIs (syphilis, Chlamydia, gonorrhea, trichomoniasis, or M. genitalium) (24.9%) compared with HM (10.6%) (P < 0.001) and to a lesser extent compared with MSM (15.9%) (P = 0.053). MSM were more likely to have syphilis infection than the other subgroups (P < 0.001), whereas women were most likely to report experiencing STI symptoms (50.3% compared with 25%–30% among HM and MSM; P < 0.001).
HIV Diagnosis and Services
Most PWH in the sample had been diagnosed at hospitals or public clinics (77.3%). However, MSM were more likely to be tested at private clinics or laboratories (20.9%) (P < 0.001). MSM were also more likely to have self-initiated their test (44.5%), whereas testing by HM was more often initiated by a health provider due to existing illness (46.4%) (P < 0.001). Overall, just 37.0% had voluntarily sought HIV testing. In all subgroups, diagnosis resulting from being asked or required to be tested by a school or employer was common (10.9% to 32.6%), although illegal in El Salvador.19 Nearly all participants (98.5%) were regularly receiving HIV care or treatment at a hospital and most were on ART (74.5%).
HIV-Related Stigma and Discrimination
Mistreatment, abuse, denial of services or social activities, or other forms of HIV-related discrimination and stigma were more likely to be experienced in family, community, workplace, school, or religious contexts (27.6%) than at health facilities (10.6%; P < 0.001). Women were more likely to report discrimination at facilities (13.2%) and in other contexts (33.4%) compared with HM (7.6% and 21.1%, respectively) (P < 0.05); levels of discrimination experienced by MSM (9.1% and 23.6%, respectively) were not significantly different than those reported by other subgroups.
Positive Prevention Components
Counseling and PIs.
Only 22.8% of participants had received POC (Table 2) with no significant differences among subgroups. During the time period from diagnosis to the survey interview (median 4 years), only about one-third (31.4%) of participants reported receiving any kind of additional HIV PI, including HIV education, HIV-related information or participation in counseling sessions; this was higher among women (37.7%) compared with HM (22.8%) (P < 0.001), with no differences compared with MSM (30.0%). Perceived access to information and advice at the point of HIV care was also limited across subgroups; only 15% to 30% of participants said they had access to information, advice, and education for their illness (agreement on the four measures >70%, Cohen's Kappa = 0.39, P < 0.001). Yet, most did not perceive communication problems with care providers per se (72.5%). Differences in median time since diagnosis by PI and POC were nonsignificant (P > 0.1).
Of 267 participants who reported experiencing STI symptoms in the past 12 months, 44.1% had sought medical attention at a health facility. Of those who did seek care and were prescribed medications (n = 70), nearly all (91.4%) completed the prescription. There were no significant differences across subgroups.
In multivariate models, exposure to POC was negatively associated with multiple sex partners in the past year, self-initiated HIV testing, and female sex (Table 3). Subsequent exposure to PIs was associated with current treatable STIs, self-initiated HIV testing, and female sex (Table 3). Both POC and PI were also associated with experienced HIV-related discrimination outside of the health services context. Correlations between independent variables were at most 0.25, indicating the absence of multicolinearity problems. No interactions proved significant. Analysis was repeated excluding M. genitalium, given the controversy regarding its clinical significance in women, as it was the most prevalent of the treatable STIs (12.9% in the combined sample and 17.3% among women), although evidence is accumulating regarding its association with HIV risk.20–22 Excluding M. genitalium eliminated the bi- (P = 0.8) and multivariate (P = 0.5) relationships of treatable STIs with PI; however, other findings were unchanged.
Positive prevention has been forwarded as an essential prevention strategy since the first initiatives to expand ART access globally.8,15,23 This study provides evidence that in El Salvador, relatively few of the individuals with HIV who are in contact with HIV care and treatment programs at public hospitals are exposed to positive prevention components, or more broadly, to any information, advice, education, or counseling with HIV prevention content either immediately following testing or after a year following diagnosis. Furthermore, we found no evidence that coverage of POC or subsequent PIs has increased over time. Providers of care to PWH should be trained in providing prevention messages and proper supervision should ensure this happens. Methods of delivering messages and behavioral change strategies should be gender-specific and appropriately tailored to risk behavior profile. Our data also evidence low levels of seeking treatment for STI symptoms, which suggests that individuals with diagnosed HIV infection may have limited understanding of the importance and resources available for STI diagnosis and treatment. STI coinfection with HIV can result in increased concentration of HIV virus in both genital tract and plasma and increased probability of sexual transmission of both infections.24,25
Levels of transmission risk are important in evaluating the need for positive prevention. Unsafe sexual practices in the larger sample were explored previously.17 This analysis documented in greater detail risk behaviors among sexually active individuals a year or more following diagnosis. Most participants reported multiple sex partners, 3 in 10 reported casual partners and about 3 in 10 reported inconsistent condom use in the past year. Past-month alcohol and past-year drug use were low overall although MSM reported higher levels of both. More than three-fourths of participants had HSV-2 infection and 1 in 5 had a treatable STI. Women had higher prevalence of most of the STIs examined, whereas MSM had higher levels of syphilis infection. However, this study likely underestimated the prevalence of Chlamydia and gonorrhea among HIV-positive MSM, as polymerase chain reaction testing was only done in urine and no oral or rectal screening was conducted. Our findings of continuing risk behaviors and STIs, in most cases among a sizable minority of PWH, and of higher prevalence of risk behaviors among MSM with HIV, are consistent with other recent studies in Central and South America.18,26,27 However, declines in risk behavior over time have been reported elsewhere.28,29
Our data also indicate that PWH who are required to be tested for HIV by their employer or educational institution are significantly less likely to receive any kind of counseling or education later on. Measures are therefore urgently needed to ensure compliance with laws prohibiting such practices19 and to facilitate linkage to care and support services when mandatory testing occurs.
Treatable STI in this study was related to exposure to HIV prevention following diagnosis. This relationship did not persist when excluding M. genitalium from the analysis and, as this study was cross-sectional, the relationship could indicate an appropriate tendency to target prevention to those at elevated transmission risk, or conversely, a failure of prevention to reduce subsequent risk behavior. However, we failed to detect a similar relationship between prevention and HSV-2 infection, which is associated with HIV transmission,25,30 or with risk behaviors (condom use, sexual partnerships, alcohol and drug use, and MSM behavior). Thus, in our view, STI and risk behavior screening should be strengthened to improve prevention targeting in the context of HIV care and treatment, following World Health Organization guidance.15
In multivariate models, experience of HIV-related discrimination was more likely among subjects who had received posttesting counseling and subsequent prevention. Although the mechanisms are unclear, we believe the relationship may capture a tendency for individuals who experience discrimination to seek support in more tolerant settings in which HIV information is provided, such as support groups, or for these groups to sensitize individuals to discrimination and legal rights. In follow-on analysis of our sample, we found a positive relationship between discrimination and support group participation (P < 0.001).
Self-report data can lead to underreporting of stigmatized behaviors due to social desirability bias.31 Although likely mitigated by using audio computer-assisted interviews, such bias may still have led to misclassification of MSM as heterosexuals, underestimated risk behaviors and thus reduced power to detect associations with prevention use. Similarly, the study required recall of receiving services at any time following diagnosis (more than 20 years for 2 individuals) that may have introduced recall bias, although the extent of the bias could not be assessed with available data. As the study was cross-sectional, the reported associations between exposure to prevention and participants' characteristics cannot be interpreted as causal; underlying mechanisms should be investigated further. The relatively high nonresponse rate (28%) of the larger study, largely due to lack of time or interest, may make our findings less representative of busier individuals, possibly those who are employed or have greater family obligations. The study also cannot generalize to PWH who do not present at hospitals or belong to support groups, and therefore may have fewer opportunities to access prevention. Our broadly framed measures of counseling and education capture exposure to any kind of behavior-change prevention message and do not necessarily indicate participation in an intervention or intervention effectiveness. Similarly, our measure of treatment-seeking for STI symptoms is just one of the many inputs to an effective STI management program. All of these factors suggest, this study likely overestimates actual coverage of positive prevention and STI management.
People diagnosed with HIV are within easy reach of prevention efforts at the time of diagnosis and later on during care. The growing number of available interventions in health settings,32 their demonstrated effectiveness1 and cost-effectiveness,33 together with the low exposure to prevention presented here, suggest a great need to expand and intensify positive prevention in testing and care facilities in El Salvador.
1. Kennedy CE, Medley AM, Sweat MD, et al.. Behavioural interventions for HIV positive prevention in developing countries: A systematic review and meta-analysis. Bull World Health Organ 2010; 88:615–623.
2. Crepaz N, Lyles C, Wolitski R, et al.. Do prevention interventions reduce HIV risk behaviours among people living with HIV? A meta-analytic review of controlled trials. AIDS 2006; 20:143.
3. Johnson B, Carey M, Chaudoir S, et al.. Sexual risk reduction for persons living with HIV: Research synthesis of randomized controlled trials, 1993–2004. J Acquir Immune Defic Syndr 2006; 41:642.
4. Casado C, Pernas M, Alvaro T, et al.. Coinfection and superinfection in patients with long-term, nonprogressive HIV-1 disease. J Infect Dis 2007; 196:895–899.
5. World Health Organization, United Nations Children's Fund, Joint United Nations Programme on HIV/AIDS. Towards universal access: Scaling up priority HIV/AIDS interventions in the health sector. 2009.
6. Global HIV Prevention Working Group. Behavior change and HIV prevention: (Re)Considerations for the 21st century. 2008.
7. Van Damme W, Kober K, Laga M. The real challenges for scaling up ART in sub-Saharan Africa. AIDS 2006; 20:653.
8. Valdiserri R. International scale-up for antiretroviral treatment: Where does prevention fit? J Acquir Immune Defic Syndr 2004; 37:S138.
9. Ministerio de Salud Pública y Asistencia Social de El Salvador. Resumen de casos de VIH-sida: Notificación en El Salvador 2009; 1984–2008.
10. Guardado ME, Cresswell J, Paz-Bailey G. Encuesta Centroamericana de Vigilancia de Comportamientos Sexuales y Prevalencia de VIH/ITS en Poblaciones Vulnerables de El Salvador: Reporte de país: Programa Nacional de ITS/VIH/SIDA de El Salvador y Universdad del Valle de Guatemala 2010.
11. Calleja JM, Walker N, Cuchi P, Lazzari S, Ghys PD, Zacarias F. Status of the HIV/AIDS epidemic and methods to monitor it in the Latin America and Caribbean region. AIDS 2002; 16:S3–S12.
12. Cáceres C. HIV among gay and other men who have sex with men in Latin America and the Caribbean: A hidden epidemic? AIDS 2002; 16:S23.
13. Global HIV Prevention Working Group. HIV prevention in the era of expanded treatment access. 2004.
14. Castro A, Farmer P. Understanding and addressing AIDS-related stigma: from anthropological theory to clinical practice in Haiti. Am J Public Health 2005; 95:53.
15. World Health Organization. Essential Prevention and Care Interventions for Adults and Adolescents Living With HIV in Resource-limited Settings. Geneva, Switzerland: World Health Organization, 2008.
16. Family Health International. Behavioral surveillance surveys (BSS) guidelines for repeated behavioral surveys in populations at risk for HIV 2000.
17. Paz-Bailey G, Shah N, Cresswell J, et al.. Risk behaviors and STI prevalence among people with HIV in El Salvador. Open AIDS J In press.
18. Paz-Bailey G, Fernandez VI, Miranda SM, et al.. Unsafe sexual behaviors among HIV-positive men and women in Honduras: The role of discrimination, condom access, and gender. Sex Transm Infect 2012; 39:35–41.
19. Ministerio del Interior de El Salvador. Diario Oficial. , 2001;Tomo #353, No. 222:Decreto No. 40.
20. Laga M, Manoka A, Kivuvu M, et al.. Nonulcerative sexually transmitted diseases as risk factors for HIV-1 transmission in women: results from a cohort study. J Acquir Immune Defic Syndr 1993; 7:95–102.
21. Ghys P, Fransen K, Diallo MO, et al.. The associations between cervicovaginal HIV shedding, sexually transmitted diseases and immunosuppression in female sex workers in Abidjan, Cote d'Ivoire. J Acquir Immune Defic Syndr 1997; 11:F85–F93.
22. Haggerty CL. Evidence for a role of M. genitalium in pelvic inflammatory disease. Curr Opin Infect Dis 2008; 21:65–69.
23. Merson M, Dayton J, OíReilly K. Effectiveness of HIV prevention interventions in developing countries. AIDS 2000; 14:S68–S84.
24. Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy and practice: The contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect 1999; 75:3–17.
25. Paz-Bailey G, Sternberg M, Puren AJ, et al.. Improvement in healing and reduction in HIV shedding with episodic acyclovir therapy as part of syndromic management among men: A randomized, controlled trial. J Infect Dis 2009; 200:1039–1049.
26. Valverde EE, Cassetti I, Metsch LR, et al.. Sex risk practices among HIV-positive individuals in Buenos Aires, Argentina. AIDS Patient Care STDS 2009; 23:551–556.
27. Guimaraes MD, Grinsztejn B, Chin-Hong PV, et al.. Behavior surveillance: prevalence and factors associated with high-risk sexual behavior among HIV-infected men in Brazil in the post-HAART era. AIDS Behav 2008; 12:741–747.
28. Gorbach PM, Weiss RE, Jeffries R, et al.. Behaviors of recently HIV-infected men who have sex with men in the year postdiagnosis: Effects of drug use and partner types. J Acquir Immune Defic Syndr 2011; 56:176.
29. Kalichman SC, Cherry C, White D, et al.. Sexual HIV transmission and antiretroviral therapy: A prospective cohort study of behavioral risk factors among men and women living with HIV/AIDS. Ann Behav Med 2011: 42:1–9.
30. Paz-Bailey G, Ramaswamy M, Hawkes SJ, Geretti AM. Herpes simplex virus type 2: Epidemiology and management options in developing countries. Postgrad Med J 2008; 84:299–306.
31. Catania JA, Gibson DR, Chitwood DD, et al.. Methodological problems in AIDS behavioral research: Influences on measurement error and participation bias in studies of sexual behavior. Psychol Bull 1990; 108:339–362.
32. Collins CB Jr, Hearn KD, Whittier DN, et al.. Implementing packaged HIV-prevention interventions for HIV-positive individuals: Considerations for clinic-based and community-based interventions. Public Health Rep 2010; 125:55.
33. Marseille E, Shade SB, Myers J, et al.. The cost-effectiveness of HIV prevention interventions for HIV-infected patients seen in clinical settings. J Acquir Immune Defic Syndr 2011; 56:e87.