ASSOCIATION BETWEEN REPRODUCTIVE-TRACT INFECTIONS AND CERVICAL INFLAMMATORY EPITHELIAL CHANGES

Objective: To determine the association, if any, between different rep roductive tract infections and cervical inflammatory epithelial change s. Study Design: Clinical, colposcopic, cytologic, and microbiologic s creening for reproductive tract infections was conducted. Subjects mer e screened for Chlamydia trachornatis, Neisseria. gonorrhoeae, Trichom onas vaginalis, bacterial vaginosis, yeast vaginitis, human papillomav irus, and serology for herpes simplex viruses, syphilis, and human imm unodeficiency virus-1 and -2. The subjects were 257 women who visited a maternal and child health center between January 1992 and December 1 993. Results: Cervical cytology revealed inflammatory epithelial chang es in 207 women (80.5%), the highest proportion in the published serie s. Inflammatory epithelial changes were significantly associated with the number of reproductive tract infections. Risk increased up to 72.6 fold when women were infected with two or more agents. The positive p redictive value of inflammatory epithelial changes for any infection w as 88.4%, whereas the negative predictive value for any infection was 63.4%. In univariate analysis, the cervical inflammatory epithelial ch anges were significantly associated with infections such as chlamydia (28.0%), human papillomavirus (56.5%), Trichomonas vaginalis (18.3%), bacterial vaginosis (38.2%), and herpes simplex virus as evidenced by the presence of immunoglobulin M antibodies (28%). Multivariate analys is, however, revealed an independent association of inflammatory epith elial changes with chlamydia (odds ratio, 21.3; 95% confidence interva l, 2.6, 181.3), human papillomavirus (odds ratio, 13.5; 95% confidence interval, 4.5, 39.6), and bacterial vaginosis (odds ratio, 22.6; 95% confidence interval, 2.9, 181) only. Conclusion: Cervical inflammatory epithelial changes are significantly associated with reproductive tra ct infection. Cervical inflammatory epithelial changes predict correct ly to an extent of 81% infections with chlamydia, bacterial vaginosis, and human papillomavirus infection. A negative smear, on the other ha nd, rules out these infections to an extent of 83%.