C trachomatis and N gonorrhoeae are major causes of maternal and neonatal morbidity and mortality in developing countries. To identify characteristics predictive of cervical infection, we examined pregnant women attending an antenatal clinic in Nairobi, Kenya. C trachomatis was isolated from 14/178 (8%), and N gonorrhoeae from 17 (10%); cervical infection with either pathogen was present in 28 (16%). Two characteristics were independently predictive of cervical infection by logistic regression analysis; the presence of either endocervical mucopus or induced endocervical bleeding, relative risk 4.2 (95% confidence interval (CI) 2.2 to 8.0) and having more than one sexual partner during pregnancy, relative risk 3.3 (95% CI 1.4 to 7.6). A screening programme for cervical infection which tested women with one or both risk markers would have a sensitivity of 68% (95% CI 51 to 85%) and a positive predictive value of 0.35 (95% CI 0.22 to 0.47). In countries where resources are limited, diagnostic algorithms incorporating clinical signs and behavioural characteristics may be useful in identifying pregnant women at high risk of cervical infection.
PIP: Chlamydia trachomatis and Neisseria gonorrhoea are major causes of maternal and neonatal morbidity in developing countries. 178 pregnant women attending an antenatal clinic in Nairobi, Kenya, were therefore examined in the attempt to identify characteristics predictive of cervical infection. Chlamydia trachomatis was isolated from 14 (8%) and Neisseria gonorrhoea from 17 (10%); cervical infection with either pathogen was present in 28 (16%). Logistic regression analysis found that the presence of either endocervical mucopus or induced endocervical bleeding and having more than 1 sex partner during pregnancy were independently predictive of cervical infection. A screening program for cervical infection which tested women with 1 or both risk markers wold have a sensitivity of 68% and a positive predictive value of 0.35. The authors conclude that countries with limited resources may use diagnostic algorithms and incorporate clinical signs and behavioral characteristics to help identify pregnant women at high risk of cervical infection.