Though controversy exists about the role of treating sexually transmitted infections (STI) as part of HIV prevention programming, in resource-limited countries with a concentrated HIV epidemic and where STI services are limited but where STI have a high prevalence, improved STI treatment can potentially reduce the incidence of HIV.^1–5 Additionally, STI themselves are a public health concern and cause significant morbidity, mortality and economic costs.⁶ In countries with a generalised HIV epidemic and among population with high-risk behaviours, STI control can be cost-saving.¹ The World Health Organization (WHO) estimated that there are more than 340 million new cases of STIs globally among adults aged 15–49 years⁷ with 50 million annually in the Americas.⁸ With this renewed emphasis on the importance of STI control in the era of HIV, prompt diagnosis and treatment for STI are crucial.

More than 75% of the STI burden is in the developing world,⁷ where laboratory diagnoses of STI are not always available or are costly. Recognising this challenge, the WHO developed guidelines for STI diagnosis and treatment in resource-constrained settings using syndromic management.⁹ With the syndromic management approach, easily recognised signs and consistent groups of symptoms are identified as syndromes, which are then assigned treatment regimens that will address the majority of, or the most serious, organisms responsible for producing the respective syndrome.⁹ Syndromic management relies on individuals recognising STI signs and symptoms, seeking healthcare and reporting symptoms to healthcare professionals. Numerous studies have been conducted to assess the effectiveness of STI syndromic management;^10–15 however, data outside a healthcare setting have shown conflicting results. In a HIV voluntary testing and counselling setting, syndromic management can lead to many missed diagnosis and 3.2% overtreatment, while in a pharmacy setting the majority of symptomatic individuals were seropositive for an STI.^16,17 Additionally, efficacy of these guidelines can vary by gender and disease type^10–15 as some STI can be asymptomatic^18,19 in which case syndromic management would have limited application and can lead to missed diagnoses.

The HIV prevalence in the general population in El Salvador is less than 1%. However, recent data have shown a high rate of STI and HIV among high-risk populations in El Salvador.^20–22 Among men who have sex with men (MSM) in San Salvador, HIV prevalence was 14.5%, Herpes simplex virus-2 (HSV-2) was 49.8%, syphilis was 14.1% and other STIs were 11.1%.²⁰ Among people living with HIV/AIDS (PLWHA) in San Salvador, STI prevalence was 18.6%.^21,23 According to a study among high-risk populations conducted in five Central American countries, MSM and female sex workers (FSWs) in El Salvador had the highest rates of HIV (15.3% and 3.2%, respectively), syphilis (15% in FSWs), gonorrhoea (10.7% in FSWs), chlamydia (23.3% in FSWs) and HSV-2 (56.5% and 95.7%, respectively) seroprevalence compared with other countries (other prevalence data not specifically stated).²⁴ Given the high prevalence of STI in these populations, there is an urgent need to accurately identify and promptly treat STI in El Salvador.

The objectives of this analysis were to determine if self-reported STI symptoms were an accurate measure for STI positivity by comparing them to laboratory diagnoses for three high-risk populations: FSWs, MSM and PLWHA, and if PLWHA participants differentially reported STI compared to MSM and FSWs.

From March to September 2008, the El Salvador Ministry of Health conducted an integrated behavioural and biologic survey among MSM, FSW and PLWHA in San Salvador, Santa Ana, San Miguel and Sonsonate using methods described in detail elsewhere.^20,22,23 PLWHA were recruited from eight public hospitals from three cities that treat approximately 80% of all individuals with HIV infection in El Salvador. Participants were identified and sampled consecutively through HIV care and treatment programmes, HIV well-being programmes and HIV support groups. MSM and FSWs were recruited from March to September 2008 using respondent-driven sampling (RDS)^25,26 in San Salvador, San Miguel and Sonsonate. Study sites were chosen based on high HIV care reporting, estimated STI prevalence rates and convenience and accessibility to the target populations. The protocol was approved by the US Centers for Disease Control and Prevention and the Ethics Committee of Clinical Research of the National Hospital Rosales in El Salvador.

After written informed consent was obtained, participants completed both an audio computer-assisted survey instrument (ACASI) and face-to-face interview addressing demographic background, sexual behaviours, STI history and service utilisation. The face-to-face interview was conducted by trained study staff. STI symptoms questions included asking whether the participant experienced any of the following symptoms in the last 12 months: vaginal or penile discharge, dysuria, genital or anal ulcers or wounds, genital itching, bad odour and genital or anal warts. If a participant was not familiar with these terms, the study staff defined these symptoms for participants. Women were also asked if they had any lower abdominal pain in the last 12 months. Following the interview, individuals were provided STI counselling and biological specimens (blood, urine, vaginal swabs for women and anal swabs for men) were collected by a trained laboratory technician. Individuals could refuse to participate in any component of the study. Syphilis testing was conducted using the rapid plasma reagent (RPR) test (Macro-Vue RPR; Becton, Dickinson and Company, USA) and confirmed by the Treponema pallidum agglutination (TPPA) particles test (Fujirebio Diagnostics, USA). Syphilis positivity was defined as serum-positive results for both RPR (at one in eight dilutions) and TPPA tests. HSV-2 testing was conducted using HerpeSelect^® (Herpes-Select; Focus Diagnostics, USA). HSV-2 seropositive was defined as having an index value of >1.1. Urine samples, vaginal swabs for women and anal swabs for men were tested using a multiplex real time polymerase chain reaction (PCR, Applied Biosystems GeneAmp; Life Technologies, USA) for Neisseria gonorrhoeae (GC), Chlamydia trachomatis (CT), Mycoplasma genitalium (MG) and Trichomonas vaginalis (TV). For diagnosing bacterial vaginosis (BV) a Gram stain test was performed. After pre-test counselling, HIV testing for MSM and FSWs was performed at the study site using two rapid HIV tests (Determine^® HIV-1/2; Abbott, USA and OraQuick^® Advance Rapid HIV-1 Antibody Test; Orasure Technologies, USA). HIV results and post-test counselling were provided after 30 min. Serology tests were conducted at the Central Laboratory in El Salvador, and PCR and Gram stain tests were conducted at Gorgas Memorial Laboratory in Panama. PCR quality control was provided by the Centers for Disease Control Laboratory in Atlanta, USA. HIV-positive samples were sent to the National Reference Laboratory for confirmatory enzyme-linked immunosorbent assay (ELISA) testing. Participants were provided with a voucher with their study identification number to access their results at the study site after two weeks. If any test was positive, the participant was offered free treatment for diagnosed STI and given a voucher for free partner STI testing and partner treatment as needed. If a participant reported STI symptoms, treatment was provided based on El Salvador’s syndromic management guidelines. HIV-positive individuals were referred to the closest HIV clinic and encouraged to refer their partners for HIV testing at the nearest HIV testing centre.

The high prevalence of laboratory-diagnosed STI in these high-risk populations is concerning, especially among individuals infected with HIV, given the association between STI and HIV transmission.²⁷ Though high rates of STI positivity were observed in these populations, the majority of men and high percentage of women living with HIV were asymptomatic. These findings are consistent with previous studies^11,19,28 and suggest that self-reported symptoms in the last 12 months may not be reliable for detecting STI. Outreach programmes for high-risk populations should thus consider an integrated approach of better recognition of STI signs and symptoms recognition, mobile-team outreach, presumptive treatment, condom promotion, routine or periodic STI screening and risk factors associated with STI including partner STI status.^29–31

Similar to previous results,^{14,15,19,32,33} the sensitivity, specificity and PPV analyses showed that STI screening, referral and treatment based solely on self-recognised and reported symptoms may not be an effective strategy for the high-risk populations in this study. Among female PLWHA, syndromic screening would have missed the majority of the GC or CT infections and about half of BV, TV or MG cases. Among all men, almost 80% of genital inflammatory infections would have been missed. These results were similar to those seen among high-risk men in previous studies.^19,34 Though self-report of genital ulcerative infections had a high specificity, the majority of ulcerative cases in this study population would be missed if STI screening and referral was dependent on self-recognising and reporting of STI symptoms in the last 12 months. The addition of a physical examination, including a speculum examination for women and anoscopy for men, would likely improve these measures.¹⁴

Sensitivity, specificity and PPV results for self-reported genital inflammatory infections were different when comparing men and women. This may be attributable to more men reporting no STI symptoms. These data suggest that routine screening may be useful even in the absence of self-reported symptoms in high-risk populations.^29–31 In contrast, sensitivity of self-recognition and report of genital inflammatory infections were higher among FSWs. The discrepancy seen in FSWs may be influenced by current programmes for FSWs in El Salvador that encourage FSWs to receive free HIV screening through the use of mobile units.³⁵ In 2007, the Ministry of Health published guidelines for physicians who work with FSWs regarding STI symptom recognition, risk factors, management and STI education for FSWs.³⁶ These interventions may be increasing STI awareness and symptom recognition by FSWs and may account for the increased sensitivity, specificity and PPV seen in this study within this population.

Our study had a number of limitations. The study was designed to assess behavioural risk factors. Further research should evaluate if similar results are found in a clinical setting. Questions regarding STI symptoms were included in the face-to-face questionnaire; thus, the data have been subject to underreporting due to social desirability bias. Respondents were asked about occurrence of symptoms over the 12 months prior to the study; however, STI can remain asymptomatic in individuals, especially for anorectal STI, for longer than 12 months. Therefore, both recall bias and potentially asymptomatic STI, could lead to an underreporting of STI symptoms. In addition, not all participants in the survey responded to STI questions and agreed to STI testing. Therefore, our findings may not be generalisable to the population as a whole. Finally, the study is cross-sectional by design, and although the questionnaire asked about symptoms over the past 12 months, laboratory tests identified current infections and there were no data on whether participants sought treatment for the symptoms they reported. This would likely underestimate specificity and PPV; however, to what degree our findings may be affected is unknown.

To date, this is the first study to validate self-reported STI symptoms among FSWs, MSM and PLWHA in Central America. Our findings suggest that relying on self-reported STI symptoms may be a poor indicator of infection among high-risk groups in El Salvador as many infections are asymptomatic and can lead to missed diagnoses. In order to minimize missed diagnoses, especially in the setting of minimal STI laboratory capability and to reach high-risk populations, El Salvador should consider using creative methods including conducting risk assessment, STI symptom screening and physical examinations at research study sites. Additionally, encouraging referrals to clinics at study sites may aid in early identification and treatment of STI.³⁷ The high prevalence of STI positivity noted in this study, their associated morbidity and the high cost of treating sequelae of STI,³⁸ may justify the potential of overtreatment of patients using syndromic management. Further, because there is an increased risk of HIV transmission with STI and that there is minimal STI laboratory testing capability in El Salvador, syndromic management may be an acceptable and the only alternative for high-risk populations in low-resource settings especially for symptomatic individuals.³⁹ Nevertheless, the evolution of resistant GC reported in North America and recently in South America and the Caribbean, can lead to the development of resistance strains in El Salvador as well.^40,41 If possible, El Salvador should consider increasing their laboratory diagnostic capabilities to facilitate prompt STI diagnosis, appropriate treatment and to minimize development of resistant strains of STI due to the potential of misdiagnosis with syndromic management. With limited laboratory resources, to increase referrals to STI clinics and prompt diagnosis, El Salvador should consider increasing efforts to provide more education regarding STI, improving symptom recognition and risk factors for STI, and perhaps routine or periodic screening of high-risk individuals.