Clinical algorithms for the screening of women for gonococcal and chlamydial infection -- evaluation of pregnant women and prostitutes in Zaire.

During March-August 1990 in Kinshasa Zaire data were collected on 1160 pregnant women attending 4 prenatal clinics and 1222 prostitutes attending a womens health clinic so researchers could identify predictors of gonococcal and chlamydial infections to evaluate the diagnostic validity of WHO-recommended algorithms used as tool for gonococcal and chlamydial infections in high and low-risk populations in Africa and to assess a non-hierarchical decision score system based on risk factors and signs or symptoms which may be used as a tool for case finding in these populations. 66% of pregnant women were in a monogamous relationship. The mean number of sexual partners per week among the prostitutes was 8. Less than 10% of the women in both groups used condoms. Prostitutes had a significantly higher prevalence of gonorrhea and chlamydia infection than the pregnant women (23.4% vs. 1.6% and 13% vs. 5.2% respectively). They were also more likely to have gonococcal and/or chlamydial infection (31% vs. 6.5%). 35% of the prostitutes tested positive for HIV infection compared to 5% of pregnant women. No single sign or symptom of gonorrhea and/or chlamydia had an acceptable level of sensitivity and specificity combined. Young age unmarried and more than 1 sexual partner in the last year were more predictive of infection than were symptoms or signs. The sensitivity of the score driven system was much higher than that of the hierarchical algorithms. These findings suggest that the risk assessment profile incorporated with the algorithms significantly improves the sensitivity in both pregnant women and prostitutes. The risk assessment profile is valid for pregnant women however.