Sexually transmitted infections (STI) constitute the main health risk among sexually active adolescents 1, as confirmed by data from developed countries where statistical systems are well-organized. In the United States and European countries the infection rates of Chlamydia trachomatis, Neisseria gonorrhoeae and human papillomavirus are highest among adolescents and young adults up to 24 years old 23. In addition, it is well recognized that STI may increase the likelihood of human immunodeficiency virus (HIV) transmission as much as 20-fold 4. In developing countries, including Brazil, although available data are scarce, these trends were confirmed 56.

Sexually active adolescents should be adequately targeted for prevention and treatment of STI by early screening strategies 7. However, in many developing countries, especially those in Africa 8, where the acquired immune deficiency syndrome (AIDS) epidemic has reached catastrophic proportions, lack of adequate diagnostic resources jeopardize this strategy. The World Health Organization (WHO) has proposed a syndromic approach for screening and treating STI, which uses algorithms favoring specific symptoms such as urethral discharge, the presence of ulcers and abnormal vaginal discharge 9. The first two have proven to be efficacious for treatment purposes 101112. However, the algorithm abnormal vaginal discharge, which has been validated in several studies, is currently criticized, especially because the most prevalent STI are frequently asymptomatic 1314. In addition, this algorithm has never been validated in Brazil in an exclusively adolescent population; such validation is also rare worldwide 1516. The aim of the present study was to validate the diagnosis of cervicitis due to C. trachomatis and N. gonorrhoeae using the WHO/Ministry of Health risk score and gynecological examination in a population-based sample of female adolescents in Brazil.

In order to achieve the necessary sample size, 1539 female adolescents belonging to the randomly selected households were invited for a medical visit at their nearest health unit. From August 2002 to September 2003, 914 (60.0%) adolescents aged 15 to 19 years attended the health units and were eligible for the study. The main reason for not responding was a change of address (N = 335), while 166 were not eligible due to age criteria, 89 were pregnant and 35 had other reasons.

Among the 914 adolescents who fulfilled the inclusion criteria and gave informed consent 472 (51.6%) were sexually active and 427 (90.5%) of these had a gynecological examination performed, including the collection of material for the diagnosis of a chlamydial and gonococcal infection. Only four teenagers refused to participate in the study and 41 were excluded because they were menstruating or had used oral medications or vaginal topical antibiotics in the last 15 days.

Sociodemographic data indicated a uniform distribution across ages (mean age = 17.2 ± 1.3 years), and revealed that an important proportion of these teenagers were already married or living with partners (32.1%). Most of them had between 5 and 8 years of schooling, had poorly educated mothers and low family income (Table 1)

This study was a community-based representative sample of the female adolescent population, aged 15 to 19 years old, residents of one of the most socially deprived regions of the city of Goiânia. The prevalence of infection, particularly with C. trachomatis, was very high.

It is well known that adolescents present the highest prevalence levels of these infections in most countries 1718. Moreover, the fact that this population is one of low family income and lower level of education further increases the risk of STI 1920.

In Brazil, recent data indicate that the fastest growing rate of HIV infection is among adolescents aged 15 to 19 years-old 21. Thus, the control of STI is imperative, not only because of their intrinsic complications, especially in relation to chlamydia and gonococcus infection among female adolescents, but also due to their facilitating action in relation to HIV infection 22232425.

STI screening programs in many developed and developing countries, including Brazil, are not fully implemented. In the United States, for example, the Centers for Disease Control recommend annual screening of chlamydia infection among sexually active female adolescents and young adults up to 25 years of age, but this program is not widely followed 2627.

Despite the national and international recognition of the Brazilian National AIDS Program, Brazil is lagging behind other countries in regard to the detection and treatment of other STI, including secondary prevention. Only recently, a more clear priority has been given to implementing public health policies for the control of STI, including chlamydia and gonococcus infections. The current choice for STI treatment adopted by the Ministry of Health is the use of a syndromic approach for diagnosing and treating cervicitis, despite it being known that these infections are asymptomatic in up to 80% of cases 28.

Our results indicate that the use of the risk score for cervicitis screening and diagnosis among female adolescents is not adequate due to the large proportion of asymptomatic infections, most of which are positive for chlamydia. The low sensitivity and low PPV of the risk score should be of major public health concern. In addition, the risk score does not appear to improve the management of cervical infections, and should not be used as a screening tool or a diagnostic test among asymptomatic or poorly symptomatic women 2930. Components of the gynecological examination assessed for validation of a cervicitis diagnosis proposed by the Ministry of Health also had a poor performance.

The prevalence of infection by C. trachomatis was alarmingly high. The risk score proposed by the Ministry of Health and the findings of gynecological examinations are not adequate for the diagnosis and treatment of cervicitis due to chlamydia and gonococcus infection in adolescents. Therefore, other strategies to control these infections among adolescents are urgently required. Among such strategies, emphasis must include investment in primary prevention through sexual education, which should include the family, school, health services and sexual partners. The 'reinvention' of sexual education is also a fundamental requirement, since the currently used methods, while achieving increased awareness among young people, especially regarding HIV infection, has not achieved the desirable and essential level of a change in behavior 31.

As far as secondary prevention is concerned, the development of rapid and less costly tests for chlamydia and gonococcus diagnosis in resource-poor public health services are urgently needed. Studies regarding the prevalence of chlamydia and gonococcus infection should be encouraged in other regions of Brazil so that improved clinical-epidemiological diagnosis of this highly relevant public health problem occurs at a national level, similar to developed countries.

AIDS: acquired immune deficiency syndrome; CI: confidence interval; HIV: human immunodeficiency virus; PCR: polymerase chain reaction; PPV: positive predictive value; STI: sexually transmitted infections; WHO: World Health Organization

The authors declare that they have no competing interests.

EMBG and MFCA were responsible for the study design, data collection and data analyses, and drafted the manuscript. MDCG contributed to the study design and was responsible for the epidemiological aspects, and helped to draft the manuscript. MASV collaborated in the elaboration of the instruments for collecting sociodemographic and sexual behavior data and participated in the fieldwork. NMB collaborated in the design of the study regarding aspects related to the syndromic approach. MSSS and MSDG participated in data collection and the realization of gynecological examinations. LESD and RLMC realized the laboratory examinations and collaborated in the discussion of the results.