Dear Editor,

In the June 2007 issue of Sexually Transmitted Infections, Dunkle and Jewkes make a much needed case for increased attention to the role of gender-based violence (GBV) in fueling the HIV pandemic. In their editorial, Dunkle and Jewkes assert that "social ideals of asculinity," including, inter alia, heterosexual success with women in the context of an entrenched gender hierarchy and the ability to control women, is an underlying cause of the observed association between violence perpetration and sexual risk taking among men. Referring to an article published in the same issue of Sexually Transmitted Infections regarding gender-based violence and STI risk in Bangladesh, they find mounting evidence for national generalizability across a range of cultural settings "supporting the link between perpetration of GBV and STI/HIV risk". Lastly, they recommend replicating gender transformative interventions with men, such as the Stepping Stones program, in order to address the underlying causes of gendered HIV risk, namely repressive gender norms and socioeconomic inequality that serve to disempower women vis-á-vis men. Their commentary is welcome, especially in a policy climate that is increasingly turning towards biomedical and non-gender transformative solutions to HIV, such as male circumcision to stem the tide of a continually mounting epidemic.

While agreeing with the spirit of their editorial, I question the logic that they use to derive their recommendations and propose an alternative method for analyzing the relationship among social ideals of masculinity, gender-based violence and HIV/STI transmission using an ecologic and comparative framework to explain population-level variations in GBV and HIV risk.

Dunkle and Jewkes accurately point to the growing body of evidence that demonstrates unequivocally that women who experience GBV and gender inequitable attitudes are at greater risk of HIV at the individual level. However, despite this association at the individual level, at the ecologic level it does not necessarily hold that countries or regions with high rates of GBV/gender inequality experience heightened rates of HIV/STIs. For instance, although the WHO multi-country study on GBV has found that Bangladesh has among the highest rates of GBV of the countries studied¹ (findings reinforced in the Silverman et al study), Bangladesh as of 2005 has a negligible national HIV prevalence rate of <0.1% (see table 1). Evidence suggests that Muslim societies, though often endorsing highly repressive gender norms, have extremely low rates of HIV/STIs.² In fact, it may be the repressive sexual culture of these countries/regions that is protective against HIV, limiting opportunities for marital infidelity and sexual liberty.

Likewise, the WHO multi-country study found Ethiopia’s and Peru’s prevalence of intimate partner violence to be the highest of the ten countries under study respectively; yet, adult HIV prevalence was estimated at 4.4% in Ethiopia in 2003 and 0.6% in Peru as of 2005. Countries with higher national HIV prevalence rates, including Namibia (19.6% in 2005) and Tanzania (6.5% in 2005), exhibited relatively lower rates of GBV (see table 1). Further, whereas HIV prevalence rates tend to be higher in urban areas,³ according to the WHO multi-country study results, GBV is more common in rural (provincial) settings while urban residence appears to have a protective effect. In essence, Dunkle and Jewkes’ comment commits the atomistic fallacy, or incorrectly generalizing an individual-level association to the group level.⁴ In order to assess the contribution of GBV to HIV at the ecologic level, studies must move from thinking about the replication of findings in different settings to a comparative, population-level framework. The disciplines of sociology, political science and economics routinely use cross-national data to test hypotheses about variations in social phenomena across countries.⁵ In his classic piece “Sick Individuals and Sick Populations”, epidemiologist Geoffrey Rose (1985) reminds us that the question,“‘Why do individuals have hypertension?’ is a quite different question from ‘why do some populations have much hypertension, whilst in others it is rare?’”.⁶ In order to address this second question, Rose argues that “what distinguishes the two groups is nothing to do with the characteristics of individuals, it is rather a shift of the whole distribution- a mass influence acting on the population as a whole.” In public health, cross-national comparisons can shed light on a host of important public health questions, including those related to GBV and HIV. To reframe Rose’s question, ‘Why are individuals who experience GBV more likely to contract HIV?’ is a very different question from ‘Why do some countries experience higher rates of GBV and how does this affect population rates of HIV infection?’

In the case of HIV, it remains somewhat of a mystery how individual HIV risk factors link up to population risk. Ultimately, due to the limited number of transmission routes, some variation in behavior between populations, coupled with a tipping point and possible biomedical transmissible cofactors must be responsible for the heightened rate of HIV in certain populations. More evidence is needed to explain under what conditions gender inequitable beliefs and behaviors link up with population-level GBV and HIV risk.

Furthermore, interventions aimed at transforming a group level characteristic, such as repressive gender norms, should focus on interventions that address the ecologic or structural level.⁷ There is intuitively a contradiction inherent in using individual-level interventions to induce broad, socially transformative gender equality. Social change does not proceed one individual at a time, but rather when, in Rose’s words, there is a “mass influence causing a shift in the whole distribution”. In order to understand what that mass influence is, and to intervene at the population level, one must know what constitute the population-level risk factors. More research must be conducted on why HIV prevalence rates vary so greatly within and between countries despite similar risk factors, such as high rates of GBV. Differences in collection and reporting of national HIV sero-prevalence estimates previously hampered comparative research of cross-national HIV prevalence. However, new data sources, notably Demographic and Health Surveys (DHS) with linked HIV testing results are an extraordinarily rich, new data source that has not yet been fully exploited to address questions regarding the social-structural and population-level determinants of HIV infection and GBV. Further, comparable national probability samples such as those utilized in the DHS allow for national estimates of GBV that can facilitate cross-national and intra-national comparisons of ecologic-level risk factors.

This comment is in solidarity with Dunkle and Jewkes’ contention that “broad socially transformative programs that promote gender equality and discourage perpetration of gender-based violence and are needed to combat the global HIV pandemic”. It is clear in countries with highly generalized HIV epidemics that reducing gender inequality and sexually empowering women are crucial to stem the spread of HIV. However, the modes by which to generate social transformation must stem from an understanding of the macro-level causes of differing social phenomena across populations. Dunkle and Jewkes point to microbicides as a gender-sensitive technology, but one that fails to address the “underlying social constructions of gender”. To this I would add male circumcision, an anti-HIV intervention increasingly gaining in prominence, which also fails to transform the broader constructions of gender in society and arguably has the potential to actually reinforce entrenched gender norms and male power. In keeping with recent literature, GBV should be addressed as a public health problem and human rights concern in its own right⁸ and not solely in relation to its instrumental association with reducing HIV and STI prevalence. Public health interventions that reinforce gender inequality and gendered power relations should be eschewed in favor of gender transformative interventions that proceed at the population level.

References

1. World Health Organization. Summary Report, WHO multi-country study on women’s health and domestic violence against women: initial results on prevalence, health outcomes and women’s responses, 2005. Retrieved June 25, 2007, from http://www.who.int/gender/violence/who_multicountry_study/en/index.html. See also, Garcia-Moreno C, Jansen H, Ellsberg M. Prevalence of intimate partner violence: findings from the WHO multi-country study on women’s health and domestic violence, Lancet 2006; 368:1260-1269

2. Gray PB. HIV and Islam: is HIV prevalence lower among Muslims? Soc Sci Med 2004;58:1751–1756.

3. Shelton JD, Cassell MM, Adetunji J. Is poverty or wealth at the root of HIV? Lancet 2005; 366(9491): 1057.

4. Leyland AH, Groenewegen PP. Multilevel modeling and public health policy, Scandinavian Journal of Public Health 2003; 31(4):267-274.

5. e.g., Dogan M, Pelassy D. How to compare nations: strategies in comparative politics. New Jersey: Chatham House Publishers, 1984.

6. Rose G. Sick individuals and sick populations, Int J Epidemiol 2001;30:427-432.

7. Blankenship KM, Friedman SR, Dworkin S, et al. Structural Interventions: Concepts, Challenges and Opportunities for Research, Journal of Urban Health 2006;83(1):59-72.

8. Brundtland GH. Violence, health, and human rights: toward a shared agenda for prevention, Health and Human Rights: An International Journal 2003;6(2): 11-13.

Stephanie Dellicour 1,2,3 , Florence Diemo4, Kayla Laserson2,3, Feiko ter Kuile1, Meghna Desai2,3 1. Child and Reproductive Health, Liverpool School of Tropical Medicine, UK 2. Kenya Medical Research Institute (KEMRI), Center for Global Health Research (CGHR), Kisumu, Kenya 3. Centers for Disease Control and Prevention (CDC), Atlanta, USA 4. Rarieda District Medical Officer for Health (DMOH), Nyanza Province, Kenya

We read with great interest the article by Otieno-Nyunya and colleagues1. The authors report low prevalence of syphilis 1.8% (95% CI 1.5% to 2.1%) as detected through the Kenya AIDS Indicator Survey (KAIS) for 2007. The authors suggest these findings indicate that elimination of syphilis is a possibility in Kenya. They also show that syphilis testing is common among women attending antenatal care (ANC) clinics and that recent estimates suggest that prevalence of syphilis among pregnant women has declined to less than 1%. Despite this success, we wish to highlight that there may still be pockets of high syphilis prevalence in Kenya.

As part of an ongoing observational study among pregnant women attending ANC in Asembo, Rarieda District, western Kenya, we observed an unexpected high prevalence of syphilis of 11.5% (32/282). The study participants are women living in 33 villages within a 5km radius of the study health facility coming for routine antenatal care, representing the general population in this area (i.e. not selected according to any health related criteria). Antenatal profile laboratory tests are performed by the facility laboratory (supplied by the Kenya Medical Supplies Agency (KEMSA) or Mission for Essential Drugs and Supplies (MEDS)) and are available free of charge for study participants. Overall 282 first ANC clients have been tested for syphilis during the period between February and September 2011. Different tests have been used during that period depending on the supplies provided as follows: VDRL (from Euromedi Equipment LTD, UK) was used from February to beginning of May 2011 with an associated syphilis prevalence of 8% (11/143), treponemal only rapid test (Eurostrip from Euromedi Equipment LTD, UK) from May to July 2011 with a prevalence of 17% (17/101) and another treponemal only rapid test (from Acon Laboratories, USA) from August to September 2011 with a prevalence of 11% (4/38). Note that the latter two treponemal rapid tests show higher prevalence as these reflect both past and current infections. The prevalence of HIV infection in this population of pregnant women is 28% by rapid test (Determine from Alere Medical Company, Japan and Bioline Standard Diagnostics Inc., South Korea, with Uni-Gold Trinity Biotech PLC, Ireland as the tie-breaker).

All the 32 women testing positive for syphilis were asymptomatic and many did not get the prescribed treatment due to costs (one dose of Benzathine Penicillin and Erythromycin cost Ksh 300 equivalent to ~USD 3); 13 reported no treatment and 11 reported incomplete treatment out of 27 with follow up data. Untreated pregnancies can result in adverse outcomes due to syphilis, such as spontaneous abortion or stillbirth, neonatal death, low birth weight/premature birth or congenital infection of the newborn in up to 80% of cases.23 Furthermore due to the stigma associated with sexually transmitted infections (STIs), even if pregnant women get treated, their partners most often will not.

Similar high prevalence has since been reported at the district level. Collated data from monthly health facility reports in the district show 23% prevalence of syphilis for 2009/2010 and 12% for 2010/2011 (DMOH personal communication). These high rates remained unreported as syphilis is no longer a notifiable disease in Kenya since 2009. Furthermore, the prominence of HIV indicators at district and provincial level reporting may also be partly responsible for other STIs being under-reported and overlooked. The prevalence of syphilis, and of STIs in general, is higher in antenatal clients compared to the general population as these represent the sexually active group. Whereas, a few previous studies have shown the prevalence of syphilis among pregnant women in Kenya to range from 3% to 4%,3 4 Otieno-Nyunya et al report a syphilis prevalence of less than 1% in recently pregnant women from the KAIS survey and noted the limitation from this small sub-sample (which included only 1 positive pregnant woman).

The high rate observed in Rarieda district highlights the fact that there might be hot spots which will need to be managed in order to achieve elimination. Screening pregnant women for syphilis provides an important means to monitor population prevalence and of identifying pockets of high syphilis prevalence. In 1940s, western European countries introduced antenatal syphilis screening and management programmes as part of the strategy for syphilis elimination.5 However implementation of antenatal screening for syphilis in low income countries is often poor due to irregular procurement of tests and the additional cost of the syphilis test incurred by ANC clients. Otieno-Nyunya et al report that antenatal syphilis screening is common in Kenya. However, there is considerable variation in the uptake of syphilis testing. In Rarieda district, less than 10% of first ANC clients are tested for syphilis as reported by the health facilities in 2009 and 2010. New simple and user-friendly, point-of -care rapid diagnostic tests for syphilis provide an opportunity to scale- up antenatal syphilis screening even in facilities without laboratory capacity.

The low country-wide population prevalence detected by KAIS is encouraging, but should be interpreted carefully. It will be important to enhance antenatal syphilis screening both for sentinel surveillance and to reduce the incidence of adverse pregnancy outcomes attributable to syphilis as well as congenital syphilis. Maternal syphilis screening and treatment is among the most cost-effective public health interventions. 6 The effect of the antenatal syphilis screening program could be improved through earlier antenatal attendance, as well as expansion to cover free syphilis treatment and explicit efforts to encourage partner notification and treatment. Operational barriers including procurement inconsistencies with syphilis test types and stock outs of both tests and treatment should be addressed. Additionally, more efforts are needed to increase public and health professional awareness of the potential serious consequences of syphilis. All these efforts will be required for disease control and eventually elimination. More emphasis should also be given to HIV uninfected ANC clients who don't currently benefit from free care and treatment. As argued previously, there should be stronger integration between syphilis screening and HIV-PMTCT programmes which have higher level of funding and political commitment. 7 The Kenyan Division of Reproductive Health would benefit from additional support to strengthen national STI program and policies for the antenatal population.

References 1. Otieno-Nyunya B, Bennett E, Bunnell R, Dadabhai S, Gichangi AA, Mugo N, et al. Epidemiology of syphilis in Kenya: results from a nationally representative serological survey. Sex Transm Infect 2011;87(6):521-25. 2. Kamb ML, Newman LM, Riley PL, Mark J, Hawkes SJ, Malik T, et al. A road map for the global elimination of congenital syphilis. Obstet Gynecol Int 2010;2010. 3. Temmerman M, Gichangi P, Fonck K, Apers L, Claeys P, Van Renterghem L, et al. Effect of a syphilis control programme on pregnancy outcome in Nairobi, Kenya. Sex Transm Infect 2000;76(2):117-21. 4. Buve A, Weiss HA, Laga M, Van Dyck E, Musonda R, Zekeng L, et al. The epidemiology of gonorrhoea, chlamydial infection and syphilis in four African cities. AIDS 2001;15 Suppl 4:S79-88. 5. Deperthes BD, Meheus A, O'Reilly K, Broutet N. Maternal and congenital syphilis programmes: case studies in Bolivia, Kenya and South Africa. Bull World Health Organ 2004;82(6):410-6. 6. Vickerman P, Peeling RW, Terris-Prestholt F, Changalucha J, Mabey D, Watson-Jones D, et al. Modelling the cost-effectiveness of introducing rapid syphilis tests into an antenatal syphilis screening programme in Mwanza, Tanzania. Sex Transm Infect 2006;82 Suppl 5:v38-43. 7. Watson-Jones D, Oliff M, Terris-Prestholt F, Changalucha J, Gumodoka B, Mayaud P, et al. Antenatal syphilis screening in sub-Saharan Africa: lessons learned from Tanzania. Trop Med Int Health 2005;10(9):934-43.

Conflict of Interest:

None declared

STDs were commonly known as venereal diseases : Veneris is the Latin genitive form of the name Venus , the Roman goddess of love. Social disease was another euphemism. Public health officials originally introduced the term sexually transmitted infection , which clinicians are increasingly using alongside the term sexually transmitted disease in order to distinguish it from the former. According to Ethiopian AIDS Resource Center FAQ - Are sexually transmitted infections (STIs) different from sexually transmitted diseases (STDs)? , "Sometimes the terms STI and STD are used interchangeably. This can be confusing and not always accurate, so it helps first to understand the difference between infection and disease. Infection simply means that a germ, virus , bacteria , or parasite that can cause disease or sickness if present inside a person's body. An infected person does not necessarily have any symptoms or signs that the virus or bacteria is actually hurting his or her body; they do not necessarily feel sick. A disease means that the infection is actually causing the infected person to feel sick, or to notice something is wrong. For this reason, the term STI which refers to infection with any germ that can cause an STD, even if the infected person has no symptoms is a much broader term than STD. " The distinction being made, however, is closer to that between a colonization and an infection , rather than between an infection and a disease . Specifically, the term STD refers only to infections that are causing symptoms . Because most of the time people do not know that they are infected with an STD until they start showing symptoms of disease , most people use the term STD, even though the term STI is also appropriate in many cases. Moreover, the term Sexually Transmissible Disease is sometimes used since it is less restrictive in consideration of other factors or means of transmission. For instance, meningitis is transmissible by means of sexual contact but is not labeled as an STI because sexual contact is not the primary vector for the pathogens that cause meningitis. This discrepancy is addressed by the probability of infection by means other than sexual contact . In general, an STI is an infection that has a negligible probability of transmission by means other than sexual contact, but has a realistic means of transmission by sexual contact (more sophisticated means blood transfusion , sharing of hypodermic needles are not taken into account). Thus, one may presume that, if a person is infected with an STI, e.g., chlamydia , gonorrhea , genital herpes , it was transmitted to him/her by means of sexual contact. The English language has short words for two of the most common: "pox" ( syphilis ) and "the clap" ( gonorrhea ).

Conflict of Interest:

None declared

On Cuba we notify between 40 and 50 thousand cases annually only of Sifilis, Blenorragia and Condylomatum Acuminatum, with tendencies to the increment for years. In order to get modification of the tendencies and reduction of the transmission of these diseases, we take promotional actions of health directed to establish habits and sexual safe or minor- risk conducts, but that only we have impact in the long term, because it is well known that it does not prove to be easy to modify these behavioral aspects of the life of people. These educational and promotional actions from health are common for all the Sexually Transmitted Diseases, including the HIV for that reason the actions played inside a specific program benefit the prevention in the group. With this method we exercise a better control and we raise knowledge and we supply of tools that they avoid the propagation of these diseases.

Conflict of Interest:

None declared