Abstract

OBJECTIVE--To identify the risk factors for gonorrhoea, syphilis, and trichomonas infections among low risk women in Nairobi, Kenya. METHOD--In a cross-sectional study, 4,404 women attending two peri-urban family planning clinics between 1989 and 1991 were interviewed using a structured questionnaire and examined for signs of sexually transmitted disease (STD) infection. Cervical cultures for gonorrhoea, PAP smear (including microscopy for trichomonas), RPR and HIV testing were done. RESULTS--Positive cervical cultures for gonorrhoea were found in 3.2% of women, positive syphilis serology in 1.9%, and positive trichomonas microscopy in 5.2%. Genital ulcers were found in 1.9% of women. Although unmarried status and reporting more than one sex partner in the previous year were both significantly associated with each disease in the crude analysis, these associations were attenuated after controlling for each other and for other risk factors. The population attributable risks (PARs) for these factors were low (7-16%) owing to the high proportion of cases who were married and monogamous. The majority of women with microbiological evidence of infection had normal pelvic examinations. Clinical diagnostic algorithms for STDs in this population had a low sensitivity and positive predictive value. Nevertheless, a strong association between HIV seropositivity and STDs was observed. CONCLUSION--The low population attributable risks found in this study suggest that behaviour change messages directed to women, particularly if they are married have a low potential for preventing STDs. The poor performance of clinical diagnostic algorithms illustrates the desirability of testing these algorithms in a variety of populations and reinforces the need for low-cost methods of microbiologic diagnosis if populations with relatively low prevalences of these infections are to be included in programmes to diagnose and treat STDs.