Background and objectives: In developing countries, simple and cheap procedures for the diagnosis of sexually transmitted diseases (STDs) are urgently needed, especially for screening purposes in high risk groups.
Goals: To evaluate the sensitivity and specificity of a screening algorithm for STDs among 364 female sex workers in Bénin, in comparison with reference laboratory tests.
Study design: The algorithm relied on the following criteria, which were evaluated in sequence: the presence of endocervical mucopus on visual inspection of the cervix, a positive swab test, or a microscopic examination of vaginal fluid showing more than 10 polymorphonuclear cells per field. The algorithm diagnosed an infection if any one of these criteria was fulfilled. True infectious status was determined by the combined results of culture for Neisseria gonorrhoeae, enzyme immunoassay for Chlamydia trachomatis, and polymerase chain reaction assays for both infections.
Results: Gonococcal or chlamydial infection was diagnosed in 39.8% of the study population according to the reference tests. The algorithm had a sensitivity of 57.9% and a specificity of 61.2%. In the presence of Candida sp or Trichomonas vaginalis, specificity decreased to 39.1%, but sensitivity increased to 67.5%.
Conclusions: These results underscore the limitations of simple, nonlaboratory diagnostic tools for screening STDs in high-risk groups in developing countries. Further research is needed to increase the validity--especially the sensitivity--of these algorithms.