Improvement of clinical algorithms for the diagnosis of Neisseria gonorrhoeae and Chlamydia trachomatis by the use of Gram-stained smears among female sex workers in Accra, Ghana.

Background: Screening for cervical infection is difficult in developing countries. Screening strategies must be improved for high-risk women, such as female sex workers. Goal: To evaluate the sensitivity and specificity of screening algorithms for cervical infection pathogens among female sex workers in Accra, Ghana. Study Design: A cross-sectional study among female sex workers was conducted. Each woman underwent an interview and a clinical examination. Biologic samples were obtained for the diagnosis of HIV, syphilis, bacterial vaginosis, yeast infection, Trichomonas vaginalis, Neisseria gonorrhoeae, and Chlamydia trachomatis infection. Signs and symptoms associated with cervicitis agents were identified. Algorithms for the diagnosis of cervical infection were tested by computer simulations. Results: The following prevalences were observed: HIV, 76.6%; N gonorrhoeae, 33.7%; C trachomatis, 10.1%; candidiasis, 24.4%; T vaginalis, 31.4%; bacterial vaginosis, 2.3%; serologic syphilis, 4.6%; and genital ulcers on clinical examination, 10.6%. The best performance of algorithms were reached when using a combination of clinical signs and a search for gram-negative diplococci on cervical smears (sensitivity, 64.4%; specificity, 80.0%). Conclusions: In the algorithms, examination of Gram-stained genital smears in female sex workers without clinical signs of cervicitis improved sensitivity without altering specificity for the diagnosis of cervical infection.