Shimoni, Noa'a MD, MPH; Davis, Anne MD, MPH; Ramos, Maria Elena MA; Rosario, Linette MD; Westhoff, Carolyn MD, MSc
Women undergoing abortion are often at risk for repeat pregnancy and are likely to benefit from immediate initiation of highly effective contraceptive methods such as the intrauterine device (IUD). Use of IUDs has increased in the United States,1 and recent studies demonstrate that women undergoing suction abortion accept immediate IUD insertion when available.2,3 Intrauterine device perforation and expulsion rates immediately after first trimester suction abortion are low and comparable to interval insertion.3–6 A recent observational study reported a low expulsion rate of 4.1% when women received IUDs an average of 8 days after medical abortion.7
Medical abortion now comprises 22% of all abortions under 9 weeks gestation in the United States.8 Insertion of an IUD on the day complete abortion is confirmed may be ideal. Ovulation resumes as early as 8 days after medical abortion,9 some couples reinitiate sexual activity soon after the abortion is complete,10 and many women cannot return for delayed insertion appointments. In one study (n=53), only 32% of women choosing the IUD as their postabortion method received one by 6 months; the others did not return because of “time needed for an additional visit.”11 In a larger randomized trial of immediate compared with delayed IUD insertion after first trimester suction abortion (n=259), 106 women were assigned to delayed insertion; 42% of these women (n=43) never returned.12 Whether timing of IUD insertion immediately after medical abortion affects subsequent utilization, bleeding, or removal compared with delayed insertion remains unknown.
The goal of our study was to compare “immediate” copper IUD insertion 1 week after medical abortion to “delayed” copper IUD insertion 4–6 weeks later. We hypothesized that more women offered immediate copper IUD insertion would undergo insertion and would be using the IUD at 6 months compared with those offered delayed insertion. We also sought to compare rates of expulsion, removal, subsequent pregnancy, bleeding, and insertion pain in the immediate and delayed groups.
MATERIALS AND METHODS
This study was conducted at Columbia University Medical Center from July 2008 to October 2009 with approval from the local Institutional Review Board. Healthy English-speaking or Spanish-speaking women seeking medical abortion up to 63 days from the last menstrual period were eligible to enroll if they desired a copper IUD (ParaGard, DuraMed Pharmaceuticals, Pomona, NY) for contraception for at least 6 months. Additional inclusion criteria were a working telephone number and plans to be in the area for the next 6 months. We excluded women with contraindications to IUD use such as a documented cervical gonorrhea or Chlamydia infection in the past 3 months, a known bleeding diathesis, serum hemoglobin less than 10 g/dL, or an untreated high-grade squamous intraepithelial lesion.
All women undergoing medical abortion received routine contraceptive counseling and those selecting the copper IUD were offered participation in the study. All participants provided written informed research consent. We screened participants for gonorrhea and Chlamydia infection unless screening results were negative in the past 3 months. The investigator or research assistant enrolled participants and administered a baseline questionnaire to obtain demographic, contraceptive, and pregnancy history.
We specified gestational age based on transvaginal ultrasonography and first day of last menses. We administered 200 mg of oral mifepristone in the office and participants placed 800 micrograms of misoprostol in the buccal mucosa at home 24 to 48 hours later. Supportive medicines included promethazine, ibuprofen, and acetaminophen with codeine. All participants received a prescription for doxycycline 100 mg orally twice daily for 1 week as infection prophylaxis. Each participant received a 4-week diary to note bleeding, spotting, or cramping beginning with mifepristone administration.
Participants returned 1 week after mifepristone administration per usual clinic protocol. Participants who confirmed interest in the copper IUD were sequentially randomized by the investigator or research assistant using opaque numbered envelopes to either immediate or delayed insertion of the IUD in an allocation ratio of 5:6, respectively. We decided to randomize more women into the delayed group in anticipation of greater loss to follow-up. After randomization, we performed a transvaginal ultrasound examination to ascertain if the abortion was complete (defined as absence of the gestational sac). Randomization occurred before ultrasound examination to prevent selective exclusions based on ultrasound findings. Those randomized to the immediate group underwent IUD insertion during the same visit; those randomized to the delayed group received an appointment to return for insertion 4–6 weeks after mifepristone administration. In cases of a retained sac (without evidence of growth), we provided a repeat dose of misoprostol or vacuum aspiration according to patient preference. Continuing pregnancies were managed with vacuum aspiration. We also planned a priori to defer insertion for endometrial stripes thicker than 3 cm. Participants requiring these further interventions remained in the study and IUDs were inserted after the abortions were complete.
Participants received 800 mg of ibuprofen before insertion and we performed a transvaginal ultrasound examination after insertion to document IUD location in both groups. Insertion pain was rated on a 10-cm visual analog scale twice: first before speculum insertion (baseline) and then again after the IUD was inserted. Pain in excess of baseline was attributed to IUD insertion.
We offered delayed group participants interim contraception. If a participant did not return as scheduled, we contacted her and offered another insertion appointment as soon as possible within the 6-month study period.
We scheduled all participants for an IUD check 6 to 8 weeks after placement, which included a transvaginal sonogram. We defined expulsion as either movement of the entire IUD into the vagina or presence of any part of the IUD in the cervical canal on physical or ultrasound examination. We removed expelled IUDs and offered reinsertion.
We contacted participants 3 months after enrollment to ascertain contraceptive use. An exit interview was scheduled 6 months after enrollment; that visit included a transvaginal sonogram and satisfaction questionnaire. We conducted the exit interview by telephone if participants could not return or were no longer using the IUD. Participants received up to $60 for their participation. All visits occurred in the outpatient setting. Study investigators were not blinded to study arm.
We decided a priori that a 20% difference in IUD utilization between the two groups at 6 months would be clinically meaningful, and we assumed 70% of participants in the delayed group and 90% in the immediate group would be using the IUD at 6 months.13–15 These two assumptions, together with α of 0.05 and 80% power, yielded a total sample size needed of 142. We overenrolled to accommodate for loss to follow-up. A staff member not involved with enrollment used a random number table to generate the allocation sequence. We performed statistical analyses using Microsoft Access and SAS 9.2. We compared categorical and continuous variables using χ2, Fisher exact test, Student t test, or Wilcoxon-Mann-Whitney U test, as appropriate. We used intention-to-treat analysis for IUD utilization at 6 months and per-protocol analysis for expulsions and removals. We fitted a logistic model to compare bleeding in the two groups. Missing bleeding data were excluded.
We enrolled 204 women. We excluded 48 participants for failure to return for the randomization visit (n=27), Chlamydia infection (n=4), anemia (n=1), and change of decision (n=16). We randomized 156 participants, 71 to immediate and 85 to delayed IUD insertion (Fig. 1). Table 1 shows the two groups had comparable baseline characteristics. Nine percent of participants (n=14) were lost to follow-up. Three participants still using the IUD at 6 months completed exit interviews by telephone.
Sixty-nine of 71 participants randomized to the immediate group (97%) underwent IUD insertion during the study. Two participants in the immediate group declined IUD insertion after allocation. In contrast, 65 of 85 randomized to the delayed group (76%) underwent IUD insertion (P<.001).
Use of IUDs was greater in the immediate group but not statistically different; 69% (n=49 of 71) of immediate group participants compared with 60% (n=51 of 85) of delayed group participants were still using the IUD at 6 months (P=.24).
Of participants randomized to the immediate group, 56 underwent IUD insertion within 8 days of mifepristone administration, eight between days 9 and 14 of mifepristone administration, and five more than 14 days after mifepristone administration. Two immediate group participants delayed insertion beyond 8 days for management of a retained sac; both participants had the IUD placed at the time of a suction aspiration. No other insertions were delayed for ultrasound findings; other delays resulted from late follow-up (n=10) and deferred IUD insertion (n=1). In the delayed group, 60 participants had the IUD inserted within 42 days, and five delayed insertion further. Four delayed group participants presented with retained sacs at follow-up. Management occurred with misoprostol or aspiration.
Expulsion of the IUD was similar in the two groups: 12% (n=8) in the immediate and 11% (n=7) in the delayed group (P=.88). Eight participants elected to have another IUD inserted after expulsion, four in each group. Of these reinsertions, two in the immediate and one in the delayed group expelled the IUD again. Three expulsions in the immediate group and four in the delayed group were diagnosed by sonogram only.
More IUD removals occurred in the immediate group; however, this difference did not reach statistical significance. Ten participants (14%) in the immediate group and five participants (8%) in the delayed group requested removal during follow-up (P=.21). In both groups, the most commonly cited reasons for removal were bleeding and pain. Removals occurred throughout the study in both groups. Number of bleeding days reported on the 28-day bleeding diary did not predict removal; participants requesting removal reported a median of 17 days (range 10–27) of bleeding or spotting, whereas those continuing to use the IUD reported 20 days (range 6–28).
Four pregnancies occurred during the study, all in participants randomized to delayed insertion who did not return for IUD insertion, whereas no pregnancies occurred in the immediate group (P=.09). Two participants continued their pregnancies, one terminated, and one had a spontaneous abortion.
Participants returned 105 bleeding diaries, 45 (63%) from immediate and 60 (71%) from delayed group participants, all initiated on the day of mifepristone administration and continued for 28 days. Immediate group participants reported a median of 20 days (range 8–28) of bleeding or spotting during the 28 diary days; delayed group participants reported a median of 19 days (range 6–28; P=.15). Immediate and delayed group participants reported cramping for a median of 5 days (range 0–24) and 4 days (range 0–19) days, respectively (P=.14).
Figure 2 compares the proportion of women reporting bleeding or spotting in each group over time. A logistic regression model did not find bleeding or spotting differences in the two groups (P=.11).
Pain recorded on a visual analog scale during IUD insertion increased a median of 10 mm (range −44 to 93) from baseline in the immediate group and 13 mm (range −43 to 80) in the delayed group (P=.30). There was no difference in pain before or during insertion between the two groups (data not shown).
We identified no cases of serious infection, uterine perforation, hemorrhage, or anemia requiring IUD removal or transfusion. One IUD was inserted in a participant before review of results identifying a Chlamydia infection. She was treated and the IUD was left in place. One ectopic pregnancy was identified, treated successfully with methotrexate, and the IUD was subsequently inserted. Both participants were included in the analysis. One delayed group participant received a levonorgestrel intrauterine system. She was included in the analysis.
This randomized trial examined timing of IUD insertion after medical abortion. Intrauterine copper contraceptive uptake was significantly greater in women offered early insertion after medical abortion, with 97% receiving the IUD compared with 76% in the delayed group. We attribute this high level of uptake to easy IUD access incorporated into the medical abortion follow-up visit.
More women randomized to immediate insertion used the IUD at 6 months than those randomized to the delayed group; however, this modest difference did not reach statistical significance. Our study environment may have contributed to the smaller than expected difference between the two groups. In this study, we provided close follow-up, flexible and prompt appointments, and free IUDs for participants. This seamless access likely resulted in a higher rate of insertion and use than would occur outside a research setting, especially for women delaying IUD insertion.
Early IUD insertion did not increase expulsions, which were similar in the two groups. Our expulsion rate is higher than reported in recent studies of IUDs inserted after suction abortion. When we exclude the eight expulsions diagnosed by sonogram only, our expulsion rate is similar to expulsion rates reported elsewhere.3,16
The rate of IUD removal was greater among the immediate group, although this difference did not reach statistical significance. Bleeding and pain were common reasons for removal in both groups and removals occurred throughout the study in both groups. The reason for this modest removal difference is unclear.
The four pregnancies in the study occurred among the 20 delayed group participants who did not return for IUD insertion. This underscores the importance of early initiation of effective contraception in women after abortion.
Reported bleeding during the first 28 days after mifepristone was similar in the two groups; early IUD insertion did not increase the duration of bleeding. Prolonged bleeding or spotting is common after medical abortion; one study reported an average of 24 days, which is consistent with our results.17 Emphasizing the possibility of prolonged bleeding after medical abortion may improve IUD retention in women selecting early placement. Clinicians should be reassured that early insertion is safe and will not significantly worsen bleeding or pain after medical abortion.
Our study has limitations. We provided the intrauterine copper contraceptive to participants; thus, our results may not be generalized to the levonorgestrel intrauterine system. This study was designed to detect a 20% difference in IUD use at 6 months. The observed difference at 6 months was only 9% and was not statistically significant. Differences in IUD insertion were substantial; however, both groups had less IUD continuation than we expected. A larger study may have demonstrated greater IUD utilization after early insertion, as seen in a recent study of IUD insertion after uterine aspiration.3 Our study did not have sufficient power to detect differences in secondary outcomes. Because we only collected bleeding diaries for 28 days, we cannot assess whether later bleeding contributed to the removals requested by women in both groups.
Our data support that offering IUD insertion at the medical abortion follow-up visit results in more IUD insertions and is not associated with adverse outcomes such as expulsion or more bleeding. We conclude immediate intrauterine copper contraceptive insertion after medical abortion should be offered routinely to patients.
1. Mosher W, Jones J. Use of contraception in the United States: 1982–2008. National Center for Health Statistics. Vital Health Stat. 2010;23:29.
2. Goodman S, Hendlish SK, Benedict C, Reeves MF, Pera-Floyd M, Foster-Rosales A. Increasing intrauterine contraception use by reducing barriers to post-abortal and interval insertion. Contraception 2008;78:136–42.
3. Bednarek PH, Creinin MD, Reeves MF, Cwiak C, Espey E, Jensen JT, et al.. Immediate versus delayed IUD insertion after uterine aspiration. N Engl J Med 2011;364:2208–17.
4. Drey EA, Reeves MF, Ogawa DD, Sokoloff A, Darney PD, Steinauer JE. Insertion of intrauterine contraceptives immediately following first- and second-trimester abortions. Contraception 2009;79:397–402.
5. Grimes DA, Lopez LM, Schulz KF, Stanwood NL. Immediate postabortal insertion of intrauterine copper contraceptives. The Cochrane Database of Systematic Reviews 2010, Issue 6. Art. No.: CD001777. DOI: 10.1002/14651858.CD001777.pub3.
6. Stanwood NL, Grimes DA, Schulz KF. Insertion of an intrauterine contraceptive device after induced or spontaneous abortion: a review of the evidence. Malden (MA): Blackwell Science; 2001. p. 1168–73.
7. Betstadt SJ, Turok DK, Kapp N, Feng KT, Borgatta L. Intrauterine device insertion after medical abortion. Contraception 2011;83:517–21.
8. Jones R, Zolna MR, Henshaw SK, Finer LB. Abortion in the United States: incidence and access to services, 2005. Perspect Sex Reprod Health 2008;40:6–16.
9. Sober S, Ratcliffe S, Creinin MD, Schreiber C. Ovulation resumption after medical abortion with mifepristone and misoprostol. Contraception 2010;82:207.
10. Creinin MD, Bednarek PH, Reeves MF, Cwiak C, Espey E. Resumption of sexual activity after first trimester suction aspiration. Contraception 2010;82:188.
11. Stanek AM, Bednarek PH, Nichols MD, Jensen JT, Edelman AB. Barriers associated with the failure to return for intrauterine device insertion following first-trimester abortion. Contraception 2009;79:216–220.
12. Gillett PG, Lee NH, Yuzpe AA, Cerskus I. A comparison of the efficacy and acceptability of the Copper-7 intrauterine device following immediate or delayed insertion after first-trimester therapeutic abortion. Fertil Steril 1980;34:121–4.
13. Sivin I, Tatum HJ. Four years of experience with the TCu 380A intrauterine contraceptive device. Fertil Steril 1981;36:159–63.
14. A randomized multicentre trial of the Multiload 375 and TCu380A lUDs in parous women: three-year results. UNDP/UNFPA/WHO/World Bank, Special Programme of Research, Development and Research Training in Human Reproduction: IUD Research Group [published erratum appears in Contraception 1994;50:397]. Contraception 1994;49:543–9.
15. Van Kets HE, Van der Pas H, Delbarge W, Thiery M. A randomized comparative study of the TCu380A and Cu-Safe 300 IUDs. Adv Contracept 1995;11:123–9.
16. Fox MC, Oat-Judge J, Severson K, Jamshidi RM, Singh RH, McDonald-Mosley R, et al.. Immediate placement of intrauterine devices after first and second trimester pregnancy termination. Contraception 2011;83:34–40.
17. Davis A, Westhoff C, De Nonno L. Bleeding patterns after early abortion with mifepristone and misoprostol or manual vacuum aspiration. J Am Med Womens Assoc 2000;55(Suppl 3):141–4.
© 2011 by The American College of Obstetricians and Gynecologists.