Incidence of pelvic inflammatory disease in clinical trials with Cu-7 (intrauterine copper contraceptive) : a statistical analysis
O'Brien, F.B.; Stewart, W.C.; Sturtevant, F.M.
Contraception 27(2): 111-122
1983
ISSN/ISBN: 0010-7824 PMID: 6851553 DOI: 10.1016/0010-7824(83)90082-3Document Number: 207267
In December 1970 a large-scale, multicenter, clinical trial of the Cu-7 IUD was initiated in the US, the UK, Western Europe, Scandinavia, Japan, Southeast Asia, Australia, New Zealand, Latin America. Women were excluded if they had a history of gonorrhea, pelvic inflammatory disease (PID) in the past 3 months, repeated PID, or postpartum endometritis or infected abortion in the past 3 months. This ultimately provided a cohort of 16,444 women with a Cu-7, yielding 363,658 woman months of experience. Data were collected following 1st insertion of the Cu-7 only. For this analysis, attention was restricted to the results of the US trials. This cohort was represented by 10,760 women, principally white and urban, who entered the study and accumulated a total exposure of 242,169 woman months by closure in 1981. The investigators' clinical records were reviewed for terms included within the definition of PID as used herein: PID, salpingitis, oophoritis, tubo-ovarian abscess, adnexal tenderness, endometritis, and uterine tenderness. The incidence of PID, in terms of 1000 woman years (TWY) of use of the Cu-7 was determined for the 1st 6 months of the 1st year and for all consecutive periods of 1-year over the entire 4 years that were analyzed. Although the clinical trial progressed considerably longer than these 4 years, the data became too few for meaningful analysis. For the 1st 48 months of use of the Cu-7 there were 111 reports classified as PID for the 3911 nulliparous subjects and 254 for the 6849 parous subjects. The overall incidence of PID among nulliparous subjects was 15.5 cases/TWY. Among parous subjects, it was 19.5 cases/TWY. For both parity groups combined, it was 18.1 cases/TWY. There were an additional 6 cases among the 1030 subjects using the Cu-7 for more than 48 months, which are not included in the statistical analysis. Incidence of PID generally decreased as the duration of use increased. The incidence of PID among all subjects generally decreased as the age increased. Thus, the incidence of 27.7 cases/TWY for the under-20 subjects of all duration declined to 3.1 cases/TWY for the over-39 age group. With each successive aging of 5 years, the incidence tended to decrease about 4 cases/TWY. The age group standardized incidence for nulliparous subjects of all durations was less than that for parous subjects. The difference was -7.4 cases/TWY; the rate ratio was 0.62.