Genital ulcers: Etiology, clinical diagnosis, and associated human immunodeficiency virus infection in Kingston, Jamaica

Individuals presenting consecutively with genital ulcers in Kingston, Jamaica, underwent serological testing for human immunodeficiency virus (HIV) infection, chlamydial infection, and syphilis. Ulcer material was analyzed by multiplex polymerase chain reaction (M-PCR) analysis. DNA ti om herpes simplex virus (HSV), Haemophilus ducreyi, and Treponema pallidum was detected ia 158 (52.0%), 72 (23.7%), and 31 (10.2%) of 304 ulcer specimens. Of the 304 subjects, 67 (22%) were HIV-seropositive and 64 (21%) were T. pallidum-seroreactive. Granuloma inguinale was clinically diagnosed in nine (13.4%) of 67 ulcers negative by M-PCR analysis and in 12 (5.1%) of 237 ulcers positive by M-PCR analysis (P = .03). Lymphogranuloma venereum was clinically diagnosed in eight patients. Compared with M-PCR analysis, the sensitivity and specificity of a clinical diagnosis of syphilis, herpes, and chancroid were 67.7%, 53.8%, and 75% and 91.2%, 83.6%, and 75.4%, respectively, Reactive syphilis serology was 74% sensitive and 85% specific compared with M-PCR analysis. Reported contact with a prostitute in the preceding 3 months was associated with chancroid (P = .009), reactive syphilis serology (P = .011), and HIV infection (P = .007). The relatively poor accuracy of clinical and locally available laboratory diagnoses pleads for syndromic management of genital ulcers in Jamaica. Prevention efforts should be intensified.