This prospective non-inferiority randomized controlled study will recruit women at the time of an abortion at gestational age of less than 12 weeks (first trimester). Consenting women choosing to use an IUD post-abortion will be randomized to type of device, and women choosing other post-abortion contraceptive methods, or no method of contraception, will be enrolled in a non-intervention, observational arm. Primary outcome will be rate of expulsion at 1 year. Secondary outcomes will include rates for pregnancy; complications such as infection or perforation; continuation of, and satisfaction with, the chosen contraceptive method as determined at 1, 2, 3, 4 and 5 years; and expulsion rate for each type of copper IUD at 2, 3, 4 and 5 years.
Sample-size justification for comparing expulsion rates for the NovaT and FlexiT
Our aim in this non-inferiority study is to provide evidence that the expulsion rate for the Flexi-T does not exceed that of the Nova-T by more than 10%. The rationale for this choice rests on the assumption that the detection rate for IUD expulsions is at least 85%
, which implies that the loss in efficacy of a device due to expulsions is less than 0.15 times its expulsion rate, and correspondingly, that any difference in efficacy between devices is less than 0.15 times the difference in expulsion rates. Thus differences of 10% in expulsion rates translate into differences in efficacy of 1.5% or less.
We calculated power calculations for non-inferiority testing following the approach described by Blackwelder
 using the on-line calculator provided by S. Patten
Our null hypothesis is: H0: FlexiT expulsion rate ≥ NovaT expulsion rate + 10%.
This formula requires specification of the problem in terms of success proportions (that is, those not experiencing expulsion by 1 year). We will assume the rate of those not experiencing expulsion will be 96% for the Nova-T
[17–19] and 92% for the Flexi-T. Thus, we require a sample size of 192 per group to attain 80% power at α = 0.05 to detect a difference in expulsion rates exceeding 10%. We will enroll a total of 400 women, a sufficient number to allow for changes to intention to leave the province and thus loss to follow-up using the health administrative databases. This recruitment will ensure the final sample is at least the 384 women required to test the null hypothesis that FlexiT has a significantly inferior (higher) expulsion rate compared with the NovaT.
Thus, our recruitment will include 400 women who choose to have IUDs and consent to be randomized to either the Flexi-T (200 women) or the Nova-T (200 women), and a further 200 who decide to not to use intrauterine contraception, who will be enrolled as a non-intervention comparator group. Therefore, a total of 600 women will be recruited.
The study clinic (see ‘Recruitment facility’, below) provides abortions to about 700 women meeting the entry criteria each year, and based on our previous research in this setting
 we expect 40% of eligible women to choose an IUD, and 80% of these to enroll in IUD groups. Target enrollment is anticipated at 22 months.
All women at participating study sites who have completed informed consent for an abortion at less than 12 weeks gestational age (as determined by ultrasonography) and who are residents of the Canadian province of British Columbia (BC) enrolled in the universal provincial medical services plan are eligible to participate. Additionally, those who have chosen to use an IUD for post-abortion contraception are eligible to be randomized to the device groups. The study has no minimum age criteria for enrollment.
Women are not eligible if they intend to move from BC within the next year, if they intend to conceive within 1 year, or if they are currently enrolled in another clinical trial. The contraindications to the use of an IUD are also exclusion criteria.
Uterine cavity anomalies causing distortion of the endometrial canal including fibroids of more than 50 mm, excluding repaired uterine septum.
Current untreated pelvic inflammatory disease, cervicitis, or lower genital tract infection.
Undiagnosed abnormal uterine bleeding.
Known uterine or cervical malignancy or cervical dysplasia.
Established immunodeficiency (HIV positivity is not an exclusion unless immunodeficiency is present).
Acute malignancies affecting blood or leukemias.
Recent trophoblastic disease with raised levels of human chorionic gonatotropin.
Post-randomization exclusion factors for women in the randomized groups include perforation or excessive bleeding at the time of their abortion, failure to undergo the abortion procedure for any reason, or a uterine anomaly detected at the time of the abortion procedure. These exclusions are designed to be only those that, in real-life clinical practice, would preclude a woman from being able to use an IUD as an immediate method of contraception.
Recruitment for this study will be undertaken at a teaching-hospital outpatient abortion service in the Canadian province of British Columbia (BC).
All women presenting at the study clinic for an abortion of a pregnancy before 12 weeks gestational age will receive information via the research study web page
 at the time they book their appointment, and a study information brochure upon check-in. All women will undergo a post-abortion contraception information session with a trained clinic nurse who is not part of the research staff, and the women will decide upon their contraceptive method of choice before they are asked if they wish to participate in the study. Interested potential participants will then be referred for an information session with a trained women’s health research assistant (RA). The RA will explain all study procedures, answer all questions, and complete the informed consent process with women before the time of their abortion, including establishing preferred personal contact methods to ensure follow-up. Enrolled women will complete an intake version of the contraception satisfaction questionnaire (CSQ) on site. We have previously described the development and use of the CSQ
. Briefly, this is a detailed instrument eliciting the effectiveness of, degree of satisfaction with, and adverse effects experienced for the range of contraceptive methods available in Canada. The CSQ has been adapted from a published instrument
[25, 26] for use in post-abortion studies. It has been translated into the three most common non-English languages in our study population, and piloted for accuracy and cultural sensitivity. This trial has been registered at ClinicalTrials.Gov (Identifier: NCT01174225).
Participants choosing a copper IUD will be randomized to receive either the Flexi-T or the Nova-T. Stratified (that is, separate) randomization lists in sequential blocks of sizes of eight will be generated for parous and nulliparous women. Allocation will be performed remotely using a Canadian on-line randomization service (http://www.randomize.net
), to maintain strict concealment. Staff will be kept unaware of the randomization design and of stratification factors involved, to mitigate the potential for unblinding of individual blocks. For women choosing an IUD, the study device determined by the randomization will be inserted by their surgeon immediately after the abortion and before the worman leaves the procedure or operating room. Blinding of the surgeon or the woman receiving the IUD is not possible, as the IUDs are distinct in appearance. Even once in situ, these two devices have a differing number and consistency of strings palpable at the cervical os. Following standard clinic protocol, all women will undergo PCR testing for chlamydia and gonorrhea before their abortion, and receive 2 g metronidazole as a single observed dose as prophylaxis against postoperative infection
. Women deemed to be at higher risk of a sexually transmitted infection (STI)
 and those with positive PCR results will receive 1 g of azithromycin as well. A post-placement ultrasound image of the IUD in situ will be recorded.
Participants choosing no contraception and all other methods of contraception will receive normal instructions from the facility physician for implementation of their contraceptive method, undergo the same STI testing and prophylaxis protocol, and complete an intake CSQ before leaving the abortion facility.
Enrolled participants will complete the subsequent CSQ either by mail or as a web-based questionnaire at 3, 6, and 12 months post-abortion. User-friendly internet-based and paper-based formats are available for each of the four most common languages spoken in our study population.
All recruited women will receive reimbursement for their time in the form of a gift certificate at enrollment and for each completed CSQ submitted. In addition, women choosing to be randomized to an IUD arm will receive their device free of charge.
Primary outcome determination will be through data obtained from the provincial government health administrative databases and from the CSQ to determine the expulsion rate for each group at 1 year. Permission will be obtained from participants to use personal health numbers (the unique identifiers used in these databases) and date of birth to search these databases, detailing all subsequent care received in BC within the subsequent year (including IUD removals and re-insertions, and prescriptions for alternate forms of birth control; information on subsequent pregnancies such as miscarriages, abortions, or deliveries; and any hospital re-admissions, surgery or prescriptions dispensed for antibiotics, which may indicate possible complications). The CSQ will collect data on expulsion of IUD, effectiveness of and satisfaction with the contraceptive method and insertion timing assigned, any removal of the IUD, any change to contraceptive method or intention to conceive, any interval pregnancies and their outcomes, and any adverse events. Using both the databases and the CSQ methods for outcome determination over a period of 5 years (the device life for the IUDs) we will determine, with a close to perfect follow-up rate, the intention to treat (ITT) effectiveness of post-abortion contraception in Canada.
This study has received institutional review board approval from the University of British Columbia-Children’s and Women’s Research Ethics Board (H10-00798) and the Interior Health Authority Research Ethics Board (2010-034).
Primary outcome will be examined as expulsion rate at 1 year for women randomized to receive the FlexiT device compared with women randomized to a NovaT. Secondary outcomes will be examined annually over the 5-year period using an ITT framework for both randomized and non-intervention groups, and will include: pregnancy rate, rate for continuation of method, adverse events (such as infection or perforation), outcomes of those participants who were chlamydia-positive at the time of abortion, and satisfaction with contraceptive method chosen and expulsion rate for years 2 to 5.
Because our aim is to examine expulsion rates and establish the non-inferiority (within a margin of 10%) of the Flexi-T compared with the NovaT, the primary analysis will be based on an upper (one-sided) 95% confidence limit for the difference in expulsion rates (FlexiT rate minus NovaT rate). Non-inferiority will be inferred if the confidence limit value is less than 10%. This simple approach is valid so long as no systematic difference in follow-up occurs between groups. As a check, we will also conduct analysis to account for partial follow-up. Because our outcome definition is essentially composite and the relevant risk periods differ by components, Kaplan-Meier estimates for each component event will be determined, and composite estimates will be obtained by summing the estimated cumulative event rates (calculated as 1 minus the survival function) at the time points indicated in our operational definitions. Confidence intervals around the difference in these estimates will be calculated using the bootstrap method.
Rates for all secondary outcomes will be calculated in an ITT framework for events occurring within 1year and within successive years following the abortion, and for pregnancy events, those associated with a conception within a year or successive years. Outcomes will be determined using the CSQ, direct access to clinical follow-up visit records, and billing and procedure coding data from the administrative health-system databases. Multivariate logistic regression will be used to examine demographic, socioeconomic, and obstetric factors in relationship to primary and secondary objectives.
The quantitative data from the CSQ will be analyzed using descriptive and correlational statistics. The CSQ contains several scales, providing composite scores that can be used to indicate differences in the secondary outcomes. Open ended questions will be analyzed through content analysis, focusing on key topics.