Objective
We sought to examine the short-term (3- and 6-month), self-reported bleeding and cramping patterns with intrauterine devices (IUDs) and the contraceptive implant, and the association of these symptoms with method satisfaction.
Study Design
We analyzed 3- and 6-month survey data from IUD and implant users in the Contraceptive CHOICE Project, a prospective cohort study. Participants who received a long-acting reversible contraceptive (LARC) method (levonorgestrel-releasing intrauterine system [LNG-IUS], copper IUD, or the etonogestrel implant) and completed their 3- and 6-month surveys were included. Univariable and multivariable analyses were performed to examine the association of bleeding and cramping patterns with short-term satisfaction.
Results
Our analytic sample included 5011 Contraceptive CHOICE Project participants: 3001 LNG-IUS users, 826 copper IUD users, and 1184 implant users. At 3 months, >65% of LNG-IUS and implant users reported no change or decreased cramping, while 63% of copper IUD users reported increased menstrual cramping. Lighter bleeding was reported by 67% of LNG-IUS users, 58% of implant users, and 8% of copper IUD users. Satisfaction of all LARC methods was high (≥90%). LARC users with increased menstrual cramping (relative risk adjusted [RR adj ], 0.78; 95% confidence interval [CI], 0.72–0.85), heavier bleeding (RR adj , 0.83; 95% CI, 0.76–0.92), and increased bleeding frequency (RR adj , 0.73; 95% CI, 0.67–0.80) were less likely to report being very satisfied at 6 months.
Conclusion
Regardless of the LARC method, satisfaction at 3 and 6 months is very high. Changes in self-reported bleeding and cramping are associated with short-term LARC satisfaction.
Long-acting reversible contraception (LARC) methods such as the levonorgestrel-releasing intrauterine system (LNG-IUS), copper intrauterine device (IUD), and the subdermal etonogestrel implant are safe and highly efficacious in preventing pregnancy, with failure rates similar to sterilization. LARC methods are more effective than other reversible contraceptives because they require little effort or maintenance on the part of the user after insertion. Although these methods are used widely in parts of Europe and Asia, only 8.5% of US girls and women ages 15-44 years who use contraception reported using an IUD or implant in 2009.
LARC users in several studies have cited increased bleeding, irregular bleeding, and cramping as the most common causes for method discontinuation within the first 2 years of use. Short-term changes in bleeding and cramping patterns associated with LARC use have been reported in several studies in Europe, Asia, and Latin America. However, limited data are available regarding the short-term experience of LARC users in the United States. Thus, the purpose of this analysis is to examine the short-term (3- and 6-month) bleeding and cramping patterns with IUDs and the contraceptive implant, and the association of these symptoms with method satisfaction. We hypothesize that decreased amount and frequency of bleeding and a reduction in cramping compared to the previous survey time point will be associated with higher levels of method satisfaction.
Materials and Methods
The Contraceptive CHOICE Project (CHOICE) is a prospective cohort study conducted in St. Louis, MO. All contraceptive methods in CHOICE are provided at no cost for 2-3 years. This analysis is focused on LARC users and examines bleeding and cramping patterns, and level of satisfaction with the contraceptive method reported during the 3- and 6-month telephone surveys. The protocol for recruitment, enrollment, and follow-up of participants is detailed by Secura et al and was approved by the Human Research Protection Office at Washington University School of Medicine. The methods of CHOICE and this specific subanalysis are outlined briefly below.
CHOICE enrolled girls and women between 14-45 years of age who lived or sought clinical services in the greater St. Louis area and who were interested in a new method of birth control. To be eligible a woman must have been sexually active with a male partner or anticipated sexual activity within 6 months of enrollment, and did not desire pregnancy within a year of enrollment. Women who had undergone a sterilization procedure or hysterectomy were not eligible.
Women interested in enrolling in CHOICE were provided with a brief scripted introduction to the LARC methods, and then screened for eligibility by a trained staff member. Screening was conducted in person at various recruitment sites or via telephone. Eligible participants were enrolled during an in-person, 2-hour enrollment session where additional standardized contraceptive counseling about all reversible methods was provided. Written informed consent was obtained from each participant prior to enrollment. LARC methods were inserted at the enrollment site, or at the participant’s clinic, or private practice. Women who did not receive a LARC method on the day of enrollment were provided with a bridge method until pregnancy was ruled out definitively, and LARC insertion was deemed clinically appropriate. Demographic information, baseline bleeding patterns, and reproductive history were recorded during the enrollment interview. At baseline, women were asked about the number of bleeding and spotting days each month; whether their menses were regular or irregular; and the average cycle length from one menses to the next. Follow-up interviews were conducted at 3 and 6 months, and then every 6 months using a standardized telephone survey. At follow-up, participants were asked if their bleeding and cramping were the same, better, or worse than 3 months earlier.
This analysis examines the symptoms and satisfaction reported by CHOICE LARC users 3 and 6 months after enrollment. Thus, only CHOICE participants who had a LNG-IUS, copper IUD, or implant inserted prior to the 3-month interview were included in this analysis. Duration of LARC use at the time of the 3-month and 6-month interviews was calculated as the difference in days between the follow-up interview dates and the date of method insertion. For the 3-month analysis, only those participants who started their method 2-4 months (60-120 days) from the date of 3-month survey were included. Similarly, for the 6-month analysis, only those women who started the method 5-7 months (150-210 days) from the date of 6-month survey were included. Data from women who discontinued their LARC method due to a pregnancy or desire to get pregnant were excluded from the satisfaction analysis. During the 3-month telephone survey, participants were asked to report changes in the amount and frequency of bleeding, and changes in cramping since insertion of their LARC method. At the 6-month interview, participants were asked to report changes in bleeding and cramping compared to what they experienced at time of 3 month survey. For our reported results, “heavier” or “lighter” bleeding refers to the volume of flow and “increased” or “decreased” frequency of bleeding refers to the number of days of bleeding.
Finally, at the 3- and 6-month interviews, participants rated their satisfaction with their LARC method as “very satisfied,” “somewhat satisfied,” or “not satisfied.” Participants who discontinued their LARC method before the 3-month or 6-month follow-up interview were coded as “not satisfied” in the analyses except those who discontinued to conceive. For the analyses, we grouped “satisfied” or “very satisfied” together as “satisfied.”
To compare the baseline demographic characteristics and bleeding patterns among our study participants, we used χ 2 or Fisher exact test for categorical variables, and analysis of variance for continuous, normally distributed variables. Poisson regression with robust variance was conducted to evaluate associations between bleeding and cramping patterns and satisfaction. This approach provides an unbiased estimate of relative risk when the outcome is common (>10%). All factors significantly associated with satisfaction in the univariable analysis were further evaluated for effect modification and potential confounding. Effect modifications were identified if the corresponding interaction terms were statistically significant. Confounders were identified if there was a >10% change in the relative risk estimate with the potential factor in the model compared to the risk estimate without the potential factor in the model. All the confounders were included in the final multivariable model to evaluate the association between satisfaction and method used, change in bleeding amount, change in bleeding frequency, and change in cramping. Stata 11 (StataCorp, College Station, TX) was used for all statistical analyses. The statistical significance level was set at .05.
Results
The Figure shows the inclusion criteria and analytic sample. Among 9256 participants enrolled in CHOICE, 75% desired a LARC method. Of participants, 93% (n = 6415) received their LARC method; 1044 reported method use at the 3-month interview that was <2 months (n = 850) or >4 months (n = 194) and were therefore excluded from this analysis, leaving 5001 participants in the 3-month analysis. The mean duration of use at the 3-month survey was 86 days. Among 4850 continuers at 3-month survey, 247 did not finish the 6-month survey and another 359 women reported method use at the 6-month interview that was <5 months (n = 99) or >7 months (n = 260); thus, 4235 participants were included in the 6-month analyses. The mean duration of use at the 6-month survey was 179 days. We excluded 3 women from the 3-month analyses and 6 women from the 6-month analyses who discontinued their LARC method to conceive. Women who discontinued at 3 months for other reasons were coded as “not satisfied” for that analysis (n = 151) but were not included in the analysis at 6 months. All other LARC users who discontinued for other reasons (n = 313) by 3 or 6 months were included in the analyses and were coded as “not satisfied.”
The baseline demographic and reproductive characteristics of the participants stratified by LARC method are shown in Table 1 . The mean age was 25.6 years. Implant users were significantly younger than IUD users, and were more likely to be black, single, nulliparous, and to have less than a 12th-grade education. Prior to method insertion, implant users were more likely to have shorter cycles (<21 days) and irregular cycles. LNG-IUS users were more likely to have heavier menstrual bleeding and more commonly experienced dysmenorrhea. Copper IUD users were less likely to report baseline menstrual cramping compared to women who chose the hormonal LARC methods.
| Demographic | All LARC (n = 5011) | LNG-IUS (n = 3001) | Copper IUD (n = 826) | Implant (n = 1184) | P value | ||||
|---|---|---|---|---|---|---|---|---|---|
| Age (y), mean (SD) | 25.6 | (6.0) | 26.0 | (5.8) | 27.9 | (6.3) | 23.0 | (5.7) | < .001 |
| Age category (y), n (%) | < .001 | ||||||||
| <18 | 274 | (5.5) | 75 | (2.5) | 10 | (1.2) | 189 | (16.0) | |
| 18-20 | 713 | (14.2) | 378 | (12.6) | 67 | (8.1) | 268 | (22.6) | |
| 21-25 | 1859 | (37.1) | 1173 | (39.1) | 264 | (32.0) | 422 | (35.6) | |
| 26-30 | 1160 | (23.1) | 757 | (25.2) | 230 | (27.8) | 173 | (14.6) | |
| >30 | 1005 | (20.1) | 618 | (20.6) | 255 | (30.9) | 132 | (11.1) | |
| Race, n (%) | < .001 | ||||||||
| Black | 2442 | (48.7) | 1419 | (47.3) | 298 | (36.1) | 725 | (61.2) | |
| White | 2186 | (43.6) | 1376 | (45.9) | 454 | (55.0) | 356 | (30.1) | |
| Other | 383 | (7.6) | 206 | (6.9) | 74 | (9.0) | 103 | (8.7) | |
| Ethnicity, n (%) | < .001 | ||||||||
| Hispanic | 265 | (5.3) | 132 | (4.4) | 46 | (5.6) | 87 | (7.3) | |
| Marital status, n (%) | < .001 | ||||||||
| Single/never married | 2897 | (57.8) | 1714 | (57.1) | 383 | (46.4) | 800 | (67.6) | |
| Married/living with partner | 1738 | (34.7) | 1063 | (35.4) | 350 | (42.4) | 325 | (27.5) | |
| Separated/divorced/widowed | 373 | (7.4) | 223 | (7.4) | 92 | (11.2) | 58 | (4.9) | |
| Education level, n (%) | < .001 | ||||||||
| <12th grade | 1727 | (34.5) | 906 | (30.2) | 198 | (24.0) | 623 | (52.7) | |
| Some college | 2074 | (41.4) | 1316 | (43.9) | 346 | (41.9) | 412 | (34.8) | |
| ≥College degree | 1208 | (24.1) | 778 | (25.9) | 282 | (34.1) | 148 | (12.5) | |
| Gravidity, n (%) | < .001 | ||||||||
| 0 | 1331 | (26.6) | 724 | (24.1) | 231 | (28.0) | 376 | (31.8) | |
| 1 | 1038 | (20.7) | 598 | (19.9) | 122 | (14.8) | 318 | (26.9) | |
| 2 | 946 | (18.9) | 621 | (20.7) | 148 | (17.9) | 177 | (14.9) | |
| ≥3 | 1696 | (33.8) | 1058 | (35.3) | 325 | (39.3) | 313 | (26.4) | |
| Parity, n (%) | < .001 | ||||||||
| 0 | 2124 | (42.4) | 1196 | (39.9) | 348 | (42.1) | 580 | (49.0) | |
| 1 | 1285 | (25.6) | 810 | (27.0) | 169 | (20.5) | 306 | (25.8) | |
| 2 | 976 | (19.5) | 630 | (21.0) | 170 | (20.6) | 176 | (14.9) | |
| ≥3 | 626 | (12.5) | 365 | (12.2) | 139 | (16.8) | 122 | (10.3) | |
| Baseline BMI, n (%) | .06 | ||||||||
| <18.5 | 110 | (2.2) | 64 | (2.2) | 17 | (21) | 29 | (2.5) | |
| 18.5-24.9 | 1880 | (37.9) | 1093 | (36.8) | 352 | (43.0) | 435 | (37.0) | |
| 25-29.9 | 1311 | (26.4) | 810 | (27.3) | 197 | (24.1) | 304 | (25.9) | |
| ≥30 | 1662 | (33.5) | 1001 | (33.7) | 253 | (30.9) | 408 | (34.7) | |
| Any history of STI, excluding BV and PID, n (%) | .87 | ||||||||
| No | 2989 | (59.6) | 1798 | (59.9) | 492 | (59.6) | 699 | (59.0) | |
| Yes | 2022 | (40.4) | 1203 | (40.1) | 334 | (40.4) | 485 | (41.0) | |
| Smoking status, n (%) | < .001 | ||||||||
| Never smoker | 2653 | (53.1) | 1543 | (51.6) | 403 | (48.8) | 707 | (59.9) | |
| Previous smoker | 1189 | (23.8) | 744 | (24.9) | 220 | (26.7) | 225 | (19.1) | |
| Current smoker | 1155 | (23.1) | 704 | (23.5) | 202 | (24.5) | 249 | (21.1) | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
