Combined oral contraceptives pills (OCPs) are a popular form of reversible contraception worldwide. Although perfect use failure rates are low, typical use failure rates are 5% to 7% or higher because of incorrect or sporadic use.1,2 It is estimated that more than one million unintended pregnancies are related to OCP use, misuse, or discontinuation.3 Discontinuation rates for OCPs are very high in developing countries, ranging from 17% in Zimbabwe to 60% in Paraguay.4
Alternate regimens of OCPs, in which the duration of the active OCP phase is more than 21 days or the placebo phase is less than 7 days, may reduce symptoms associated with the hormone-free interval, decreasing bleeding (and potentially anemia), enhancing acceptability, and improving continuation rates. Although most studies show that extended or continuous use of OCPs decreases overall menstrual blood loss by reducing the total number of bleeding days, some studies show an initial increase in unscheduled bleeding, which may lead women to discontinue use.5 However, breakthrough bleeding can be successfully managed by using a short hormone-free interval.6
Few data exist regarding the continuation or acceptability of extended or continuous use oral contraceptive regimens in women in under-resourced settings. We conducted this randomized trial to compare the effects of continuous use of OCPs (with short flexible hormone-free intervals) with the conventional 28-day cyclic regimen regarding OCP continuation rates and contraceptive effectiveness in women in the Dominican Republic. We also evaluated participant satisfaction, menstruation-associated symptoms, and effects on anemia, a substantial problem for many women in developing countries.
MATERIALS AND METHODS
We conducted this randomized trial at the Profamilia clinic in Santo Domingo, Dominican Republic from October 2008 to September 2010. Women requesting OCPs as their method of contraception (new users or those switching from nonhormonal methods) were given the opportunity to participate in the study provided they were between the ages of 18 and 30 years, had menstrual periods every 21 to 35 days, were willing to be randomly assigned to one of the study groups, and had negative urine pregnancy tests at enrollment. We excluded women who had contraindications to OCP use,7 had been pregnant in the previous 3 months or wished to become pregnant in the next 12 months, were breastfeeding or had breastfed in the previous 2 months, or had recently used hormonal contraceptives. Study staff and participants were not blinded to group. All participants provided written informed consent. Enrolled participants received free OCPs (lofemenal; 0.3 mg norgestrel and 0.03 mg ethinyl estradiol) donated by the U.S. Agency for International Development. The FHI 360 Protection of Human Subjects Committee and the Ethics Committee of Profamilia, Dominican Republic, approved the study.
Women presented for enrollment during their menses. To minimize bias, we enrolled, counseled, and followed-up women in both groups in an identical fashion and avoided mentioning that OCP continuation was the primary outcome measure. At the admission visit, we assessed eligibility, collected baseline demographic, menstrual, and contraceptive use data, and measured weight and blood pressure. We also assessed attitudes toward menstruation and its suppression using Spanish adaptations of previously validated questions.8,9 We similarly developed questions to address menstrual discomfort based on the Menstrual Distress Questionnaire.10
Women were then allocated equally into two groups using a computer-generated block randomization scheme concealed in sequentially numbered opaque sealed envelopes to either a 28-day cyclic regimen or continuous use (omitting placebos) of the same combined OCP for 12 months. Neither participants nor site staff was blinded to assigned regimen. Study staff provided women with four to five packs of OCPs, depending on their assigned regimen, and detailed use instructions, including what to do in case of missed OCPs. Participants in the continuous use regimen also received instructions that they could stop using OCPs for 3 days as a flexible short hormone-free interval to manage breakthrough bleeding if they experienced persistent bleeding or spotting for 7 days.
We asked participants to schedule five clinic visits (admission, 3-month follow-up, 6-month follow-up, 9-month follow-up, and 12-month follow-up). At each follow-up visit, we resupplied OCPs and reviewed OCP use, discontinuation, serious adverse events, and bleeding. At the 3-month and 9-month visits, we performed urine pregnancy tests when pregnancy was suspected. At the 6-month and 12-month visits, we performed urine pregnancy tests, assessed hemoglobin and hematocrit, reviewed side effects and serious adverse events, and administered structured acceptability questionnaires to all participants. Women were considered study completers at the 12-month visit or if pregnancy occurred. Women who discontinued OCPs use were not discontinued from study participation and remained in the study through the 12-month follow-up visit.
Statistical analyses were performed blinded to study group using SAS 9.3. The primary endpoint was time to OCP discontinuation; secondary endpoints were cumulative pregnancy probabilities, OCP continuation rates, acceptability, perception of bleeding and other side effects, and hemoglobin and hematocrit at 6 months and 12 months of follow-up.
The primary analysis population was the intent-to-treat population, for which we planned to include all women enrolled and randomized into the study. During the study we discovered that four women had enrolled twice; data from their second enrollment were excluded. We also defined a user population, a subset of the intent-to-treat population excluding women who had multiple enrollments, never used OCP, or who were found to have conceived before randomization, and a per-protocol population, which further excluded women who had violations of the study eligibility criteria.
For the intent-to-treat population, time in analysis was calculated as the day of OCP discontinuation minus the day of enrollment plus 1 day or time to the last follow-up visit. For the user population, if a participant became pregnant during the study, they were censored at their estimated date of fertilization. We compared the probabilities of OCP continuation and pregnancy over time using Kaplan–Meier estimates and log-rank tests. We evaluated associations between baseline and time-varying covariates and also menstrual attitudes regarding OCP discontinuation and pregnancy using Cox proportional hazards regression. We used exploratory factor analysis with orthogonal rotation to reduce a set of 22 menstrual attitude statements adapted from the following two different instruments: the Attitudes Towards Menstrual Suppression Instrument,9 which assessed interest in reduction or elimination of menses, including through contraceptive use; and the Menstrual Attitudes Questionnaire,8 which included menstruation as a debilitating, bothersome, or natural event. Some items assessed denial of menstrual symptoms and attitudes toward suppression of menses. The exploratory factor analysis produced three factors, with eigenvalues more than 1, expressing the following: desire for reduced menstruation (seven items, alpha=0.75); menstruation as problematic (four items, alpha=0.56); and menstruation as natural (five items, alpha=0.36). We only included the Desire for Reduced Menstruation factor in further analyses because of low reliability of the other two factors. Finally, we factor-analyzed four items, of which three were retained: the idea that OCPs are easy to use; instructions are easy to follow; and OCPs are an effective way to avoid pregnancy (three items, alpha=0.65).
We estimated that a sample size of 362 women would provide 80% power to detect a difference in continuation rates if the true proportions of participants continuing OCPs at 12 months were 65% and 50% in the continuous use and cyclic groups, respectively.
Of the 398 women screened for this study, 362 were randomized and 358 were included in the primary intent-to-treat analysis (Fig. 1). Of these women, 335 (93.6%) completed the study: 165 (92.7%) in the continuous use group and 170 (94.4%) in the cyclic group. The main reasons for study discontinuation were duplicate enrollment and personal reasons.
Participants had a mean age of 22.7 years; most of them were of mixed ethnic origin, unmarried but living with partner, nonsmokers, and of low educational status (12 years of schooling or less) (Table 1). Most participants were parous; a small proportion of participants reported history of unplanned pregnancy.
Forty-eight percent of participants were previous users of OCPs, although most of them (69.2%) had used them for less than 1 year. The most common reasons for previous OCP discontinuation were becoming abstinent (27.9%) and desire for pregnancy (18.6%).
Reported menstrual-related symptoms at baseline included abdominal pain or cramping (74.9%), backache (48.9%), moodiness (48.9%), breast tenderness (48.6%), bloating (36.3%), headaches (29.1%), and nausea (17.9%). At baseline, 90% of participants agreed with the following statements: “having a regular menstrual period is a sign of health” and “menstruation is part of what makes me a woman.” Nevertheless, considerable variation was evident in women's attitudes toward reduction or suppression of menses (Table 2). Although only one-quarter of women agreed that “stopping menstrual periods is a good idea,” two-thirds of women stated that they “would be interested in not having a period every month,” and more than half agreed that they “would be happy to use a birth control method that made my period stop for a certain amount of time.”
The OCP cumulative continuation probabilities at 6 months and 12 months of follow-up were relatively high for both study groups and did not differ significantly (Fig. 2) (P=.21). The OCP continuation probabilities were 86.1% (95% CI 81.0–91.3%) at 6 months and 77.6% (95% CI 70.8–84.4%) at 12 months for the continuous use group, and 81.3% (95% CI 75.5–87.1%) at 6 months and 71.7% (95% CI 64.1–79.3%) at 12 months for the cyclic group. Main reasons for OCP discontinuation in both groups were “running out of pills or forgetting” (44.9%) and side effects other than bleeding (20.2%). Only two women in the cyclic group and no women in the continuous use group discontinued OCPs because of bleeding issues.
In univariate analyses, factors associated with OCP discontinuation included cohabiting (hazard ratio [HR] 0.61, 95% CI 0.40–0.94) (Table 3), previous pregnancy (HR 0.53, 95% CI 0.31–0.91), and previous birth (HR 0.48, 95% CI 0.29–0.78). In multivariable analyses, factors associated with risk of OCP discontinuation were previous birth (adjusted HR 0.45, 95% CI 0.27–0.78) and desire for reduced menstruation (adjusted HR 1.36, 95% CI 1.06–1.76). Ease of use of OCPs was only marginally statistically significant (adjusted HR 0.76, 95% CI 0.55–1.05). Treatment group did not predict OCP discontinuation in either the bivariate (unadjusted HR 0.76, 95% CI 0.50–1.16) or the multivariable (adjusted HR 0.84, 95% CI 0.55–1.28) analyses.
A total of 58 women became pregnant during the study, 28 in the continuous use group and 30 in the cyclic group, and were included in the intent-to-treat population and analysis. In the intent-to-treat population, pregnancy rates over time were similar in both groups: 5.8% (95% CI 2.3–9.4%) at 6 months and 16.2% at 12 months (95% CI 10.2–22.1%) in the continuous use group and 8.5% (95% CI 4.3–12.6%) at 6 months and 17.4% (95% CI 11.1–23.6%) at 12 months in the cyclic group. Nineteen of these pregnancies occurred after OCP discontinuation and were excluded as endpoints from the per-protocol analysis, along with two pregnancies estimated to have been conceived before the first OCP intake (one in each group). Pregnancy rates in both groups were also similar in the per-protocol analysis (P=.21).
Treatment group was not associated with risk of pregnancy in either univariate (unadjusted HR for continuous OCP use 0.91, 95% CI 0.54–1.53) (Table 4) or multivariable analyses (HR 0.90, 95% CI 0.54–1.51). No other factor was significantly associated with pregnancy in either univariate or multivariable analysis.
Six serious adverse events were reported, including two in the continuous use group (appendicitis and cholelithiasis that were considered possibly related) and four in the cyclic group (pneumonia, femur fracture, gunshot injury, and abdominal pain that was considered possibly related).
We evaluated OCP adherence by self-report. Across all visits, 55.4% of women in each group reported that they had never missed any OCPs in the past month, whereas 26.1% of women in the continuous use group and 22.3% of women in the cyclic group ever reported missing three or more OCPs in the past month (P=.43). The main reason for missed OCPs was running out or forgetting, with no statistically significant difference between groups at the 6-month (P=.76) and 12-month (P=.74) visits.
We also evaluated participant perceptions of bleeding and other side effects at the 3-month, 6-month, 9-month, and 12-month visits. Women randomized to the cyclic group were more likely to report any bleeding during the study (P<.001) as well as regular bleeding (P<.001) and normal volume of bleeding (P<.001) compared with the continuous use group. Women in the continuous use group who experienced bleeding were more likely to report spotting. Most women found their bleeding patterns acceptable, but more women in the cyclic group perceived their bleeding patterns as acceptable at 3 months (99% compared with 93%; P=.02) and 6 months (98% compared with 89%; P<.01); this difference disappeared by 9 months and 12 months. When we compared OCP continuers compared with discontinuers, regardless of group, a higher proportion of OCP discontinuers (83%) than OCP continuers (76%) reported moderate to heavy menstrual cycles at the time of OCP discontinuation, and most OCP discontinuers reported some menstrual bleeding at each study visit before discontinuing. In contrast, 20–25% of women who continued using OCPs reported no menstrual bleeding at each study visit. In addition, although less than 10% of study participants reported heavy bleeding at any time point, discontinuers were approximately twice as likely as continuers to report heavy bleeding in the first 6 months.
Reported symptoms such as backache, bloating, breast tenderness, and abdominal pain or cramping decreased from baseline in both groups, whereas symptoms such as nausea, dizziness, and headaches increased in both groups after OCPs were started. Few significant differences emerged between the two study groups. Fewer women in the continuous use group reported backache at the 6-month visit (38.5% for the continuous use group compared with 52.0% for the cyclic group; P=.02) and breast tenderness at the 12-month time point (28.9% for the continuous use group compared with 42.3% for the cyclic group; P=.03). Finally, no significant differences were noted in 12-month hemoglobin and hematocrit levels for the continuous use group (mean hemoglobin, 12.2; hematocrit, 37.3) compared with the cylic group (mean hemoglobin, 12.3; hematocrit, 37.4; P=.36 and P=.64, respectively)
In this randomized trial of continuous OCP use compared with cyclic OCPs in the Dominican Republic, both regimens were effective and acceptable, with no differences in continuation and pregnancy rates or acceptability.
It is possible that the greater ovarian suppression with continuous use could lead to greater contraceptive effectiveness. Alternatively, greater acceptance of a continuous use regimen with flexible breaks could theoretically enhance continuation or adherence. However, pregnancy rates were relatively high in both groups, likely reflecting poor adherence. Although longer hormone-free intervals may result in greater follicular development, other mechanisms of action are likely still functioning.11 Few factors predicted OCP continuation or pregnancy. For both groups, the main reason for stopping OCPs was forgetting to take OCPs or running out of OCPs. Poor adherence may be partially attributable to the overall young age of study participants (mean age, 22.7 years) who may have been less likely to follow a daily OCP regimen. Almost 20% of women in both groups reported missing three or more OCPs in the past month. Although we had no biological measures of adherence, this is likely an underestimate.
Irregular bleeding is a common side effect of OCP use, especially at the beginning of treatment and during extended cycle treatment, and irregular bleeding is one of the most common reasons for discontinuing OCPs. In this study, although the continuous use regimen was associated with more irregular bleeding, acceptance of bleeding patterns improved over time in both groups. Importantly, no women in the continuous use group reported bleeding issues as the main reason for OCP discontinuation. Similar to other studies in Brazil and Mexico,12,13 continuous use was also associated with less bleeding overall; however, no difference existed in hematocrit over time or between groups.
Few studies have addressed the issue of continuous contraceptive use and other menstrual-related symptoms. Similar to other studies,14–16 the proportion of women reporting symptoms including backache, bloating, abdominal pain or cramping, and breast tenderness decreased over time in both groups, although most reported side effects did not differ by group.
Strengths of our study include the large sample size, the randomized design, and efforts to ensure that women's interactions with the clinic were as close to “real-world” conditions as possible. For example, women were not told that assessing continuation rates was the main outcome of the study to reduce the possibility that they would change their behavior as a result of their participation (the “Hawthorne effect”). We attempted to minimize any such effect by enrolling, counseling, and following-up women in both groups in an identical fashion. We also did not ask participants to complete daily diaries; we designed our questionnaires to avoid having participants respond in a way that they might think was socially appropriate or desirable, and we minimized the number of clinic encounters. We also had a very high retention rate during the study.
Another strength was that women using the continuous use regimen were allowed to use flexible hormone-free intervals when experiencing persistent bleeding or spotting. Recent trials have supported the use of such flexible “breaks” in response to bleeding.17,18 However, our data failed to demonstrate the benefit of this approach with regard to OCP continuation. We do not know how many women actually used flexible hormone-free intervals.
In conclusion, our study results support that the continuous use regimen is acceptable and safe and is as effective as a 28-day cyclic OCP regimen in our setting. Although our study did not demonstrate that continuous use increased OCP continuation over 12 months, it does add to the literature documenting the long-term satisfaction, continuation, and improvement of menstrual-related symptoms with a continuous use regimen.
For many women, OCPs are a preferred contraceptive. The OCPs must be offered with adequate counseling to increase adherence and reduce risk of pregnancy. For women in both resource-poor and resource-rich settings who choose to use OCPs and who prefer to eliminate or reduce menses, a continuous use regimen is a viable option and should be offered. Emphasizing the safety and health benefits of continuous OCP use during counseling may be especially important in cultures like the Dominican Republic, where some women perceive menstruation as a sign of health or fertility or are concerned about the safety of amenorrhea. When adequately counseled, women in our study were quite accepting of reduced or eliminated menses. Presenting women with the choice of continuous use as an alternative to cyclic use enables women to select a regimen that works best for their own lifestyle.
1. Cleland J, Ali MM. Reproductive consequences of contraceptive failure in 19 developing countries. Obstetrics Gynecol 2004;104:314–20.
2. Trussell J. Contraceptive failure in the United States. Contraception 2004;70:89–96.
3. Rosenberg M, Waugh MS. Causes and consequences of oral contraceptive noncompliance. Am J Obstetrics Gynecol 1999;180:276–9.
4. Ali MM, Cleland J. Oral contraceptive discontinuation and its aftermath in 19 developing countries. Contraception 2010;81:22–9.
5. Edelman AB, Gallo MF, Jensen JT, Nichols MD, Schulz KF, Grimes DA. Continuous or extended cycle vs. cyclic use of combined oral contraceptives for contraception. The Cochrane Database of Systematic of Reviews 2005, Issue 3. Art. No.: CD004695. doi: 10.1002/14651858.CD004695.pub2.
6. Coffee AL, Sulak PJ, Kuehl TJ. Long-term assessment of symptomatology and satisfaction of an extended oral contraceptive regimen. Contraception 2007;75:444–9.
7. World Health Organization. Medical eligibility criteria for contraceptive use. 3rd ed. Geneva (Switzerland): World Health Organization; 2004.
8. Brooks-Gunn J, Ruble DN. The menstrual attitude questionnaire. Psychosomatic Med 1980;42:503–12.
9. Andrist LC, Arias RD, Nucatola D, Kaunitz AM, Musselman BL, Reiter S, et al.. Women's and providers' attitudes toward menstrual suppression with extended use of oral contraceptives. Contraception 2004;70:359–63.
10. Moos RH. The development of a menstrual distress questionnaire. Psychosomatic Med 1968;30:853–67.
11. Steward R, Melamed A, Granat A, Mishell DR Jr. Comparison of cervical mucus of 24/4 vs. 21/7 combined oral contraceptives. Contraception 2012;86:710–5.
12. Machado RB, de Melo NR, Maia H Jr. Bleeding patterns and menstrual-related symptoms with the continuous use of a contraceptive combination of ethinylestradiol and drospirenone: a randomized study. Contraception 2010;81:215–22.
13. Bustillos-Alamilla E, Zepeda-Zaragoza J, Hernandez-Ruiz MA, Briones-Landa CH. Combined hormonal contraception in cycles artificially extended [Article in Spanish]. Ginecol Obstet Mexico 2010;78:37–45.
14. Cachrimanidou AC, Hellberg D, Nilsson S, Waldenstrom U, Olsson SE, Sikstrom B. Long-interval treatment regimen with a desogestrel-containing oral contraceptive. Contraception 1993;48:205–16.
15. Wiegratz I, Stahlberg S, Manthey T, Sanger N, Mittmann K, Lange E, et al.. Effect of extended-cycle regimen with an oral contraceptive containing 30 mcg ethinylestradiol and 2 mg dienogest on bleeding patterns, safety, acceptance and contraceptive efficacy. Contraception 2011;84:133–43.
16. Dmitrovic R, Kunselman AR, Legro RS. Continuous compared with cyclic oral contraceptives for the treatment of primary dysmenorrhea: a randomized controlled trial. Obstet Gynecol 2012;119:1143–50.
17. Jensen JT, Garie SG, Trummer D, Elliesen J. Bleeding profile of a flexible extended regimen of ethinylestradiol/drospirenone in US women: an open-label, three-arm, active-controlled, multicenter study. Contraception 2012;86:110–8.
18. Klipping C, Duijkers I, Fortier MP, Marr J, Trummer D, Elliesen J. Long-term tolerability of ethinylestradiol 20 mug/drospirenone 3 mg in a flexible extended regimen: results from a randomised, controlled, multicentre study. J Fam Plann Reprod Health Care 2012;38:84–93.