Dr EM Godfrey, Department of Family Medicine, University of Illinois College of Medicine, Chicago, IL 60612, USA. Email firstname.lastname@example.org
Please cite this paper as: Wu S, Godfrey E, Wojdyla D, Dong J, Cong J, Wang C, von Hertzen H. Copper T380A intrauterine device for emergency contraception: a prospective, multicentre, cohort clinical trial. BJOG 2010;117:1205–1210.
Objective To determine the effectiveness of the Copper T380A (CuT380A) intrauterine device (IUD) as an emergency contraceptive in women.
Setting Eighteen family planning clinics in China.
Sample A cohort of 1963 women, aged 18–44 years, requesting emergency contraception within 120 hours of unprotected sexual intercourse.
Methods Women requesting emergency contraception were followed at 1, 3 and 12 months after the insertion of CuT380A.
Main outcome measures Efficacy of CuT380A as emergency contraception and for up to 12 months of postinsertion use. Insertion complication rates, reported side-effects and continuation rates at 12 months were also recorded.
Results No pregnancies occurred prior to or at the first follow-up visit, making CuT380A 100% effective as emergency contraception in this study. The pregnancy rate over the 12-month period was 0.23 per 100 women. In all, 29 (1.5%) women experienced a difficult IUD insertion process, requiring local anaesthesia or prophylactic antibiotics. No uterine perforations occurred. The main side-effects were increased menstrual bleeding and menstrual disturbances. The 12-month postinsertion continuation rate was 94.0 per 100 woman-years.
Conclusions CuT380A is a safe and effective method for emergency contraception. The advantages of CuT380A include its ability to provide effective, long-term contraception.
The copper intrauterine device (IUD), inserted shortly after unprotected sexual intercourse, is an alternative emergency contraceptive (EC) method, first reported by Lippes et al.1 in 1976. Available data to date show that the postcoital insertion of a copper IUD is highly effective when used as EC, offering greater protection than hormonal EC pills against pregnancy in noncomparative trials.2 Further, studies suggest that the proportion of pregnancies with EC pills increases with treatment delay,3 whereas the copper IUD for EC remains just as effective when placed up to 5 days after unprotected sexual intercourse.4 In addition, the copper IUD continues to be effective for up to 20 years,5 is the most highly rated compared with other methods6 and may provide noncontraceptive benefits, such as protection against endometrial cancer.7
Previously, IUD insertion for long-term use was restricted to those women who had at least one child as a result of concerns about reduced fertility after its use. Evidence has emerged, however, showing no negative effect on fertility following IUD removal.8,9 As a result, medical eligibility guidelines no longer necessitate that IUD candidates be parous.10–12 In addition, opportunities for IUD insertion have expanded. IUD insertion can occur any time during a woman’s menstrual cycle if it is certain that she is not pregnant.13
In long-term comparative studies of various IUDs organised by the World Health Organization (WHO), the Copper T380A IUD (CuT380A) was most effective.14 It is also one of the least expensive IUDs, and was therefore selected for use in this study. The objectives of this study were threefold: (1) to investigate the efficacy of CuT380A as an EC among parous and nulliparous women up to 12 months after insertion; (2) to determine the complications and reported side-effects up to 12 months postinsertion; and (3) to investigate the continuation rate for up to 12 months of use. There was no comparison group in this study.
This prospective multicentre study was conducted in 18 family planning clinics in China. Eligible participants were healthy women, aged 18–44 years, with regular menstrual cycles between 24 and 42 days’ duration and a known date of the last menstrual period (LMP), who requested EC within 120 hours (5 days) of unprotected coitus. Women also needed to have had at least one spontaneous cycle of normal length after discontinued hormonal contraception or a recent delivery, miscarriage or induced abortion, a negative urine pregnancy test and a desire to use the CuT380A IUD as a long-term contraceptive method.
Reasons for exclusion included suspected or confirmed pregnancy, any episode of pelvic inflammatory disease or pelvic abscess in the 12 months preceding trial admission, sexually transmitted infection within the past 6 months, any evidence of sexually transmitted infection on clinical or laboratory examination during initial examination, multiple sexual partners, known or suspected genital tract malignancy, cervical or uterine malformations, vaginal bleeding of unknown aetiology, multiple uterine fibroids associated with previous menstrual anomalies, and clinical or laboratory evidence of anaemia (haemoglobin less than 9 g/l). In addition, women who were unsure about the date of their LMP (LMP needed for pregnancy risk assessment), were currently taking oral contraceptive pills or had used spermicide, condom (even in cases of slippage or breakage) or other barrier method during their last sexual intercourse were excluded from the study.
The initial visit included the recording of information regarding the date of the initial examination, age, parity and medication use; date of onset of LMP, usual cycle length and expected date of onset of next menstruation; date and clock time of unprotected coitus; and contraceptive history, including the use of contraception during the previous and current cycle. A pelvic examination was performed to determine whether there was any evidence of cervical or uterine infection. In addition, the following tests were performed: high-sensitivity urine pregnancy test [Lan Men Yun Zhi; human chorionic gonadotrophin (hCG), 25 iu/l], haemoglobin and microscopy examination for Trichomonas, bacterial vaginosis and candidal vaginitis, for which the results were known immediately. A Papanicolaou smear was also performed at the initial examination; the results, however, were reported 2 weeks after the initial examination, and women were contacted immediately if the results were abnormal.
All women who satisfied the admission criteria gave written informed consent and were subsequently enrolled into the study for the insertion of CuT380A. Following insertion, women received diary cards to record vaginal bleeding and spotting, and any side-effects. The date of the first follow-up visit was noted on the card. Women were asked to come for follow-up visits at the end of the first and third menstruation, as well as 12 months after CuT380A insertion. Women were also instructed to return for unscheduled visits if they had any problems. The clinics participating in this trial were open from Monday to Saturday, and acute care clinics were open in the evenings and at weekends, should an unscheduled visit be necessary.
The IUD used in this trial was CuT380A, produced by Siping Medical Apparatus and Instruments Factory, Jilin Province, China. The copper wire on the device is wound on the vertical stem, and two copper sleeves are placed on the horizontal arms of the T-shaped plastic frame to provide a total copper surface area of 380 mm2. The device is 36 mm in length and 32 mm in width. All devices were placed in presterilised packages. The United States Food and Drug Administration has approved the CuT380A IUD.
CuT380A has been shown to be highly effective; therefore, it was unlikely that the study would be discontinued because of device ineffectiveness. Stopping rules included rates higher than those reported in the literature: a perforation rate of more than three per 1000 insertions, or an infection rate within 3 weeks of insertion greater than 3%.15,16
Women could discontinue the study early at any time. Investigators discontinued any participation in cases of pelvic inflammatory disease, perforation, CuT380A displacement or expulsion, pregnancy or high-grade cytological findings on the Papanicolaou screening test. In cases of pain, irregular bleeding or heavy menstruation, the woman and the physician investigator could decide whether or not to continue the trial. Nonmedical reasons for discontinuation included failure to rely solely on CuT380A as a method of contraception, desire to become pregnant, a move out of the study area and loss to follow-up.
The primary outcome measure was the efficacy of CuT380A as EC. This was determined by evaluating the unintended pregnancy rates confirmed by a positive high-sensitivity urine pregnancy test and/or by vaginal ultrasound at the first follow-up visit. Secondary outcome measures included insertion difficulties and complications, infection rates, continuation rates up to 12 months after insertion, reasons for removal prior to the first menstrual period after insertion, expulsion rates and number of removals for medical or nonmedical reasons during the 12-month follow-up.
The data were collected on standard precoded duplicate forms at the admission, follow-up and final visits at each centre. The forms were sent to the National Research Institute for Family Planning (NRIFP) in Beijing, which was the coordinating centre responsible for the data management of the trial. Progress reports were sent regularly to the United Nations Development Programme/United Nations Population Fund/World Health Organization (UNDP/UNPFA/WHO)/World Bank Special Programme of Research, Development and Research Training in Human Reproduction, WHO, Geneva, Switzerland.
The sample size was based on the estimated failure rates of CuT380A as EC. In one review, the failure rate for copper IUDs for EC was 0.20%.17 As this study did not have a control group, we estimated our sample size by using the equation, ss = (z)2 × (p) × 1 – (p)/CI2, where (p) is the true failure rate. Assuming a 95% confidence interval (CI), normal distribution and an approximate precision of 0.2%, we calculated the sample size to be 1916. Additional participants were enrolled to account for a probable 3% attrition.
We analysed the data obtained at the first follow-up visit to measure the unintended pregnancy rates. The failure rate was calculated in terms of the raw pregnancy rate (number of pregnancies per 100 women treated) and in terms of the standardised rates for the number of pregnancies that would have been expected if no treatment had been given. The expected number of pregnancies was estimated by multiplying the number of women having unprotected sexual intercourse on their menstrual cycle day by the probability of conception on that cycle day. The date of ovulation was estimated by subtracting 14 days from the date of onset of the next menstrual period. We used the estimated conception probabilities by cycle day calculated by Trussell et al.18 Continuation rates were computed using Kaplan–Meier estimates and compared using the log-rank test.19
A total of 1963 women participated in the study between July 1997 and January 2000 in 18 family planning institutions or clinics in different provinces throughout China. There were 1459 women (74.3%) with 12-month outcome information. We tracked outcomes according to parity: parous, if a woman had had at least one child; nulliparous, if a woman had never given birth after 20 weeks of gestation. Because the outcomes of those lost to follow-up were not known, they were excluded from the analysis. The flow of study participants is shown in Figure 1.
The baseline characteristics of the participants based on parity are shown in Table 1. More than 95% of the participants were parous, and the majority had had at least one prior termination of pregnancy. The median age of the participants was 28 years (range, 18–46 years) and a minority (3.1%) had used EC before this trial.
Table 1. Baseline demographics and clinical characteristics of the women enrolled in the trial
Participant characteristics and obstetric history
Total (n = 1963)
Parous (n = 1868)
Nulliparous (n = 95)
Data are the mean ± standard deviation (SD) or n (%).
Age (years) (mean ± SD)
29.0 ± 4.5
29.2 ± 4.5
24.8 ± 3.4
Age (years) (median, quartiles)
Previous induced abortion
Previous use of emergency contraception
A total of 1360 women (69.3%) had CuT380A inserted within 72 hours after the first act of unprotected sexual intercourse. A majority of women (1476; 75.2%) had unprotected sexual intercourse only once prior to CuT380A insertion, 318 women (16.2%) had intercourse twice, and 169 women (8.6%) had intercourse three times or more prior to CuT380A insertion.
In all, 29 (1.5%) women encountered a difficult CuT380A insertion process. Two nulliparous women and one parous woman were given local anaesthesia; 333 (17%) women were given prophylactic antibiotic according to the individual clinical guidelines from different participating centres. Fourteen per cent of nulliparous women and 5.6% of parous women required cervical dilatation. There were no uterine perforations during the study.
Efficacy of emergency contraception
CuT380A EC efficacy analysis included 1893 women, because we excluded 32 women (1.6%) who discontinued CuT380A and 38 women (1.9%) who were lost to follow-up prior to the 1-month follow-up visit. No pregnancies occurred prior to or at the first follow-up visit, making CuT380A 100% effective as EC in this study.
12-month CuT380A IUD continuation rate
The 12-month CuT380A continuation rate was based on intention-to-treat analysis, with 1707 woman-years (20 484 woman-months), including 111 women who discontinued the trial for various reasons (Table 2). The 12-month discontinuation rate was 6.5 (95% CI, 5.32–7.83) per 100 woman-years. The probability of continuation at 1 month was 98.5% overall, and was similar for both parous (98.4%) and nulliparous (98.9%) women. At 12 months after insertion, the probability of continuation was 94.0% overall: 94.3% for parous women and 88.2% for nulliparous women (P = 0.0266).
Table 2. Reasons for discontinuation of Copper T380A within 12-month follow-up period
Total (n = 1963)
Parous (n = 1868)
Nulliparous (n = 95)
Pregnancy with IUD in situ
Removal for pain
Removal for bleeding
Pelvic inflammatory disease
Removal for other medical reasons
Removal, desire for further pregnancy
Removal for other personal reasons
The number of women who discontinued CuT380A prior to the 12-month follow-up period was 111 (5.7%); 10 of these women were nulliparous and 101 were parous. The most common reasons for removal were bleeding (36.0%) and partial expulsion (34.2%). More nulliparous women than parous women complained of pain. There were no reported infections during the study trial.
A total of four complete and 38 partial expulsions occurred during the 12-month trial period. All of the complete expulsions occurred in women who were parous. The majority of the partial expulsions (34) occurred among women who were parous; four occurred among women who were nulliparous.
The 12-month pregnancy rate was calculated as the number of pregnancies that occurred during the 12-month period, divided by the number of woman-years of follow-up. Four pregnancies with the CuT380A in situ occurred amongst the 1707 woman-years of follow-up, making the 12-month pregnancy rate 0.23 per 100 women. All pregnancies were intrauterine and occurred in parous women between 60 and 360 days after insertion of CuT380A.
The most common complaints reported by women with CuT380A were increased menstrual bleeding and menstrual disturbances. The number of women reporting unwanted side-effects decreased at each progressive follow-up visit throughout the duration of the study.
This study is the first prospective multicentre cohort trial to report the efficacy of the CuT380A IUD as EC. In addition, this study reports the continuation rates, reasons for removal and side-effects up to 12 months after postcoital insertion. Subsequent to the description by Lippes et al.1 of the postcoital use of the Copper-T in 1976, smaller studies have described its efficacy.20 Our findings show that CuT380A is very effective for EC; there were no failures within the first month after CuT380A placement among 1893 women who continued it for 31 days after insertion. Compared with a more recent postcoital IUD study using the Multiload Cu-375 SL IUD in 1013 participants by Zhou and Xiao,4 which showed a failure rate of 0.2%, CuT380A appears to be more effective as EC. The present study had a similar loss-to-follow-up rate as that mentioned by Zhou and Xiao4 (1.4%). Our findings are particularly promising in the light of recent reviews indicating that oral hormonal EC, when implemented within a population, increases use, but does not reduce overall unintended pregnancy rates in the population.21,22 The postcoital CuT380A IUD, however, reduces EC failure when there are further acts of unprotected sexual intercourse in the short term, and contributes further to the reduction of unintended pregnancies as a long-term contraceptive once in place.
One unique factor of this study is the reporting of 1459 women (74.3%) with 12-month outcome information after CuT380A postcoital insertion. Our overall 12-month discontinuation rate is consistent with or better than that in other large multicentre studies of women requesting long-term contraception with the CuT380A IUD, and suggests that, particularly for the circumstance of postcoital insertion, the acceptability of the method is high.14,23 The greater acceptability in this study may be attributed to the recruitment criterion of women desiring to use the IUD as long-term contraception. In addition, some evidence suggests that Chinese women, relative to women from other regions, are less likely to discontinue participation in IUD studies.24 Four pregnancies (0.23 per 100 women) occurred among only parous women during the 12-month observation period. This 12-month failure rate is comparatively low, and could be attributed to CuT380A itself, which has been reported to be the most effective copper-containing IUD.25
Although this study demonstrates the efficacy of CuT380A as EC, it is not without limitations. This was an observational study, and thus participants were not randomised to an alternative postcoital method. Although the Multiload Cu-375 SL copper IUD is also an effective postcoital EC, these data were not available at the time that our study was conducted.4 Second, we did not perform any statistical tests between the parous and nulliparous populations because there were significantly more parous women enrolled in this study, making statistical tests between the groups difficult. This report therefore provides a general sense of how nulliparous women tolerate CuT380A as a long-term contraceptive following postcoital insertion relative to parous women.
Based on this study, CuT380A appears to be an effective and safe EC method. We found CuT380A to be highly acceptable to the majority of both parous and nulliparous women who desired this method as a long-term contraceptive.
Disclosure of interest
SW has received the following from the World Health Organization: support-in-kind, such as writing, administrative support and travel/accommodation expenses covered or reimbursed, and is a member of the Steering Committee on IUD Research, Human Reproduction Programme, for the World Health Organization. EMG is a consultant to the World Health Organization in Switzerland, and has received honoraria from Merck, USA. The other authors have no potential conflicts of interest. The authors alone were responsible for the content and writing of the manuscript.
Contribution to authorship
SW and HvH conceived the idea for this project. EMG wrote the paper and assisted with data analysis. SW and CW designed and initiated the original study. JD and JC were study investigators and processed and analysed the data. DW also analysed the data. HvH coordinated the study through the World Health Organization. All authors read and approved the final manuscript.
Details of ethics approval
Ethics approval was given by the Institutional Review Board of the National Research Institute for Family Planning, Beijing, China, on 4 April 1996. Institutional review boards at each of the 18 participating centres and the World Health Organization Secretariat Committee on Research Involving Human Subjects also gave ethics approval (Protocol no: 96506). Written informed consent was obtained from all women participating in the study.
This study was funded by the UNDP/UNPFA/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction. The donors and sponsors of the study had no role in the study design, data collection, data analysis, data interpretation, writing of the report or the decision to submit the paper for publication.
The authors would like to acknowledge the following investigators for their contribution to this study: (1) Wang Suzhen, Family Planning Research Institute of Tong-ji Medical University, Wuhan, Hubei, China; (2) Cheng Linan, International Peace MCH Hospital, China Welfare Institute, Shanghai, China; (3) Wang Hong, Institute of Family Planning, Shenyang, Liaoning, China; (4) XuYufang, Family Planning Research Institute of Guangdong,Guangzhou, Guangdong, China; (5) Lei Zhenwu, Sichuan Reproductive Health College, Chengdu, Sichuan, China; (6) Shu Lizhen, Chongqing Family Planning Scientific Research Institute, Chongqing, China; (7) Zhao Guifen, Jilin Family Planning Research Institute, Changchun, Jilin, China; (8) Xu Xiao, Henan Institute of Planned Parenthood Research, Zhengzhou, Henan, China; (9) Yuan Xinhua, Jiangsu Family Planning Institute, Nanjing, Jiangsu, China; (10) Song Si, Shanghai Institute of Planning Parenthood Research, Shanghai, China; (11) Cheng Weiyu, Tianjin Municipal Research Institute for Family Planning, Tianjin, China; (12) Zhang Wanping, Hebei Family Planning Research Institute, Shijiazhuang, Hebei, China; (13) Zhao Dongxiao, Shanxi Family Planning Research Institute, Taiyuan, Shanxi, China; (14) Zhong Chunli, Guizhou Research and Technical Guiding Institute of Family Planning, Guiyang, Guizhou, China.