Bacterial Vaginosis and Risk of Pelvic Inflammatory Disease

Abstract

This multicenter study was conducted to investigate the association of pelvic inflammatory disease (PID) and bacterial vaginosis. Participants were recruited from women who were attending family planning, health, gynecology, and sexually transmitted disease (STD clinics in 5 medical centers. Eligible patients were women not seeking care for STD, but who were considered at high risk for acquiring STDs according to an algorithm that weighed age, race parity, number of sexual partners, habit of douching, and a history. The vaginal swabs were self-collected. Participating patients were instructed in the use of a cotton swab to collect their vaginal specimens. At intervals of 6 to 12 months, the self-obtained specimens were examined for the characteristics of bacterial vaginosis. A vaginal microflora gram stain score of 7 to 10 was considered bacterial vaginosis. Women who developed pelvic pain or who were positive for Neisseria gonorrhoeae or Chlamydia trachomatis underwent a clinical examination and endometrial biopsy for detection of PID. A diagnosis of PID required the presence of histologic endometritis and/or pelvic pain and tenderness accompanied by either a fever of 101° F or higher, sedimentary rate greater than 15 mm/hr, elevated white blood count, or leukorrhea, mucopus, N. gonorrhea, or C. trachomatis in the lower genital tract. There were 1179 patients included in the analysis. The average follow up was 4 years. At the initial examination, 428 women had normal vaginal flora (36%), 280 had intermediate flora (29%), and 471 had bacterial vaginosis (40%). The baseline diagnosis was not associated with the rate of detection of PID over the 4 years of follow up. Nor was the development of PID significantly associated with age, race, education, income, smoking, sex during menses, condom use, or a history of STD or PID. Analyses according to various subgroups of patients (younger/older women, black/white women, women with/without a history of PID, with/without baseline gonococcal or chlamydia genital infection) found that only women who had a baseline report of 2 or more sexual partners in the previous 2 months and who had a baseline diagnosis of bacterial vaginosis were significantly more likely to have PID. An absence of hydrogen peroxide-producing lactobacillus was not associated with PID, even among the various subgroups. A baseline diagnosis of G. vaginalis or Gram-negative rod growth above 4 had no association with PID except in the subgroup of women who reported 2 or more sexual partners in the previous 2 months. Women with baseline diagnoses of N. gonorrhea or C. trachomatis were more likely to have PID.