Most currently available combined oral contraceptives (OCs) are 28-day regimens with 21 hormone-containing pills, followed by either a pill-free week or 7 spacer or placebo pills to induce a regular monthly withdrawal pattern. This bleeding is not necessary for contraceptive efficacy, and shortening the hormone-free week further decreases ovarian activity,1 whereas extending the hormone-free week increases the risk of ovulation and could lead to method failure.2 The dogma that women using contraception must menstruate is changing,3 and reversible amenorrhea is becoming more acceptable to women.4 Many women using the OC manipulate their withdrawal bleeding, and women are interested in reducing this bleeding.5 By decreasing the number of hormone-free weeks, cycle-related symptoms such as dysmenorrhea and headaches can be reduced.6
Recently, an OC was approved with extended use of 84 days of active pill use followed by 7 days of placebo or spacer pill use for a withdrawal week every 3 months.7 Irregular bleeding was the leading reason for study discontinuation by women using this regimen.7 An increase in irregular bleeding is commonly reported with extended use of OCs.8–10 A short extension, with just one additional package, or the 49-day regimen may have less irregular bleeding10 than the longer extension to the 91-day cycle,7 but there has not been a randomized direct comparison with identical formulation. Paradoxically, as with continuous hormone replacement therapy,11 rates of amenorrhea appear to increase with continuous OC use,12,13 and amenorrhea may be desirable by some women.14
NuvaRing (NV Organon, Oss, The Netherlands) is a monthly combined contraceptive vaginal ring that releases a daily dose of 120 μg etonogestrel and 15μg ethinyl estradiol.15 Like most OCs, the contraceptive vaginal ring is used for 21 days, followed by a ring-free week, to induce a regular monthly withdrawal bleed. The vaginal ring continuously delivers contraceptive hormones to suppress ovulation16 without the interference of the gastrointestinal tract or hepatic first pass to produce uniform serum levels.17
The 28-day ring schedule has been demonstrated in the contraceptive trials18,19 to result in infrequent irregular bleeding, and recent trials demonstrate significantly less irregular bleeding with ring use compared with 30 μg ethinyl estradiol20 or 20 μg ethinyl estradiol21 OC use. What was not known is what bleeding pattern would result with extended ring cycles or whether skipping the ring-free week would produce irregular bleeding not typically seen with conventional 28-day cycles. We designed a trial to compare the bleeding patterns and tolerability of 28-day cyclic ring use schedule with the continuous use of the ring (364-day cycle) and 2 commonly used extended cycle regimens, the 49-day and the 91-day schedules.
MATERIALS AND METHODS
This open-label, randomized, comparative, multicenter trial was conducted in 10 European centers (Denmark, Finland, Germany, and Norway) and in 10 centers in the United States. The study was performed in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines, with local review board approval from each center and written informed consent obtained from each enrolled subject.
Women interested in using the contraceptive vaginal ring were recruited by means of flyers and/or newspaper advertisements, depending on the site. Women were eligible if premenopausal and aged 18 years or more, with regular menstrual cycles, defined as 24–35 days in length, with an allowable individual variation of plus or minus 3 days. Important exclusion criteria included contraindications to contraceptive steroids, current use of a contraceptive injection, significant current history of a gynecologic abnormality, breastfeeding, postpartum or postabortion within prior month, use of drugs that interfere with the metabolism of contraceptive steroids, and abnormal cervical smear or abnormal laboratory result at screening. Women deemed eligible received verbal and written instructions on use of the contraceptive vaginal ring (NuvaRing; NV Organon, Oss, The Netherlands). Women were advised in these materials that irregular bleeding was expected, but that the rates for this were unknown, and that determining these rates was the primary reason for the study. For the first cycle, the women followed the instructions in the package insert. The purpose of this run-in cycle was to document that bleeding would occur in the ring-free period and to minimize study dropouts after randomization.
Following the run-in cycle, subjects were randomized to 1 of 4 treatment arms: 28-day cycle arm (21 days with ring in place followed by 7 ring-free days), 49-day cycle arm (42 days with ring in place followed by 7 ring-free days), 91-day cycle arm (84 days with ring in place followed by 7 ring-free days), and 364-day cycle arm (continuous ring use for 357 days followed by a ring-free week at the end of the study year) (Fig. 1). The total days of treatment following the run-in cycle were 364 days for all arms, except the 49-day cycle, which lasted 343 days to best accommodate the exit visit timing. Subjects in the 28-day cycle arm used 13 rings (thirteen 1-ring cycles), subjects in the 49-day cycle arm used 14 rings (seven 2-ring cycles), subjects in the 91-day cycle arm used 16 rings (four 4-ring cycles), and subjects in the 364-day cycle arm used 17 rings (continuous use for the full study period). Each ring was scheduled to be used for only 21 days. Computer-generated tables of randomly permuted numbers, in blocks of either 4 or 8 women, with a 50% chance of either block size, were used to generate treatment groups with a 1:1:1:1 ratio per site. Assignment to treatment groups was performed by calling a centralized automated telephone system. There was no blinding after arm assignment. Verification of correct study allocation was ascertained.
Study assessments were conducted at scheduled visits at the time of initial screening, at baseline (during the second ring), after 3, 6, and 9 months, and after 1 year or premature discontinuation. At the screening visit, subjects provided medical and gynecologic history and underwent a physical and gynecologic examination, including cervical cytology. The physical and gynecologic examinations were repeated at the last study visit. In addition, clinical safety laboratory tests, including lipids and complete blood count, were performed at screening and at the end of study treatment. At all study visits, blood pressure and body weight were measured.
A transvaginal ultrasound examination for assessment of endometrial thickness and to check for the presence of fibroids and other abnormalities was performed at the baseline visit and repeated after 1 year. If the double layer endometrial thickness was 10 mm or greater, a biopsy was performed to exclude pathology. In addition, a cohort of patients (10 subjects per trial arm) from each treatment group gave specific consent and underwent endometrial biopsy at the beginning and at the end of treatment to assess the effects of the different regimens on the endometrium.
A urinary pregnancy test was performed by the subjects before the first ring insertion to exclude pregnancy, at each study visit, and if a pregnancy was suspected during the trial. The occurrence of adverse events and the use of concomitant medication were recorded throughout the trial. Any problems directly related to ring use, such as vaginal discomfort, and device-related events, such as coital problems, discomfort from sensation of a foreign body, and expulsion, were considered device-related adverse events. In the event of early study discontinuation, the data collection form required selection of a specific reason for exit, such as seeking pregnancy. Unacceptable bleeding was specifically queried for, to ensure the collection of data on this rationale for discontinuation at early exit. Hazard ratios of the discontinuation rates for each of the extended regimens compared with the regular regimen, P values, and 95% confidence intervals were calculated using Cox proportional hazards model.
The acceptability of the various ring regimens was assessed by use of a questionnaire, which was completed by the subjects after 6 months and after 1 year or upon early discontinuation from the trial. Subjects were asked about the level of satisfaction with their bleeding schedule, menstrual pain in comparison with previous method, and the presence or absence of specific self-assessed symptoms such as premenstrual syndrome, headaches, and pain with menses. Furthermore, subjects were asked about the occurrence of unscheduled bleeding, if this was a problem, if it improved with time, and what would be their preferred bleeding schedule.
All subjects were asked to document the number of hours of ring use during each of the treatment days and the dates of ring insertions and removals. A subject was compliant with the 21-days of ring use if the ring use period did not deviate by more than 48 hours from the scheduled 3 weeks and with the 7-day ring-free period if the ring-free period length did not deviate by more than 24 hours from the scheduled 1 week. Vaginal bleeding was recorded daily by each subject on diary cards. Vaginal bleeding was classified by the subject as either spotting (requiring ≪ 1 pad/tampon per day) or bleeding (requiring ≫ 2 pads/tampons per day). A posttreatment phone call was made 6 weeks later to ascertain contraceptive method following the study, and if the woman was not using hormonal contraception, then she was asked if her menses had returned to its normal pattern.
The days of bleeding and days of bleeding and spotting during the days of both ring use and scheduled ring-free days were tabulated to produce medians and percentiles for the different schedules across 91-day reference periods for the year of method use. The first cycle (run-in cycle) was not included in the reference period analysis because this cycle was before randomization, and all subjects were scheduled to have a withdrawal bleeding. All analyses were performed on the intent-to treat population (defined as all randomized women) unless otherwise stated. Statistical analyses were not performed on the satisfaction or other nonbleeding data outcomes because the study was only powered to measure bleeding and spotting day differences.
The primary objective of this study was to evaluate the efficacy of 3 different extended regimens of the ring in decreasing the bleeding and/or spotting days as compared with the standard 28-day cycle. The difference between the treatment groups in the percentage of bleeding and/or spotting days was compared using an analysis of variance model, with treatment group and study center as explanatory variables. The interaction effect of study center and regimen was not statistically significant (P = .93) and therefore was not included in the final model. The null hypothesis of the study was that there would be no difference between the extended and standard regimens in percentage of days with bleeding and/or spotting. The alternative hypothesis, that treatment with at least one of the extended regimens would reduce the total percentage of bleeding and/or spotting by at least 15% compared with the standard 28-day regimen was used as the basis for the sample-size calculation. With an expected percentage of bleeding and/or spotting days of 21.6% of days in the standard 28-day regimen group and a reduction of 15%, ie, at most, 18.3% bleeding and/or spotting days in at least one of the extended regimen groups, a standard deviation of 5.5%, a power of 80%, a 2-sided α level of 0.017 (Bonferroni correction for multiple comparisons), and an anticipated discontinuation rate of 40%, this resulted in the randomization of 100 women to each of the 4 treatment arms.
The study began May 2002 and data collection was complete February 2004. A total of 596 women were screened, and 561 women began the run-in cycle. Of these, 429 women were subsequently randomized to 1 of the 4 treatment regimens (Fig. 2). A total of 140 women (32.6%) discontinued the trial prematurely, and the reasons are listed in Figure 2. Across the U.S. sites, 247 women were randomized, with 161 (65.2%) completing the year, compared with 182 women randomized at European sites, with 128 (70.3%) completing the year. Individual sites experienced a range of study completion rates, from 44–89% for the U.S. sites and 50–86% for European sites. For the 91-day cycle and 364-day cycle, the discontinuation rates were similar and higher than for the 2 shorter regimens, and this difference was statistically significant (P = .017 and P = .005 for the 91-day and 364-day cycles compared with the 28-day cycle, respectively). The main difference in discontinuation rates between arms was for bleeding. Subject baseline demographic and clinical characteristics of the randomized women are shown in Table 1 and the medical reasons for early study discontinuation are further detailed in Table 2. There were no notable differences between the 4 treatment regimens.
Extended use of the ring did not reduce the combined bleeding and spotting days compared with the 28-day cycle use; hence, the primary endpoint, a reduction of 15% or greater over 28-day cycle bleeding and spotting days, was not met. The mean percentages of bleeding and/or spotting days were 17.6% (28-day), 15.5% (49-day), 20.9% (91-day), and 24.4% (364-day). Compared with the standard 28-day cycle, the 49-day cycle showed a mean decrease of 2% (95% confidence interval –6.5% to 2.5%) in total bleeding and spotting days (ie, a reduction of 11%), whereas the 91-day and 364-day cycle arms increased this by 3.5% and 7.1%, with 95% confidence intervals of –1.1% to 8/1% and 2.6% to 11.7%, respectively.
The median numbers of bleeding and spotting or bleedings days per 91-day reference period over the full treatment period following the run-in cycle for the 4 treatment groups are shown on Table 3. It is difficult to directly compare the 91-day reference periods for the 4 treatment groups because of the difference in the number of ring-use days and ring-free days per reference period (Fig. 1). However, one can see from the table that the median numbers of bleeding days decrease with fewer scheduled ring-free days, but this is accompanied with an increase in spotting days. Notably there is a wide variation between individuals; for example, some women experienced complete amenorrhea and yet others reported frequent spotting with the 364-day cycle. Table 4 compares the bleeding and spotting days during scheduled ring use when the woman would have expected no bleeding or spotting. As one can see, the 49-day cycle was very similar to the 28-day cycle, with only an occasional unscheduled bleeding or spotting day reported (Table 4).
One in-treatment pregnancy was reported for a woman randomized to the 91-day cycle who reported not realizing her ring had been lost after her ring-free week. Compliance with the 21-day ring-use period and 7-day ring-free period was high and comparable for all treatment regimens. Compliance with the 21-day ring period ranged from 88.6% to 98.9%, and compliance with the 7-day ring-free period ranged from 84.4% to 96.3%. Prolonged ring-free periods with more than 24 hours occurred in 2.5%, 1.7%, and 3.1% of the cycles for the 28-day, 49-day, and 91-day cycle arms, respectively.
The majority of women at their exit visits were satisfied with ring use, and the highest satisfaction was reported for the shorter cycles (Table 5). The frequency of experiencing unscheduled bleeding increased for longer regimens, from 16% for the 28-day cycle to 92% for women using the 364-days cycle, but for many of these women unscheduled bleeding was not considered a problem (Table 5). Many women specified a preference to decrease the number of periods per year, with some bias toward selecting the schedule of their study assignment. For example, 48% of women in the 49-day arm preferred that schedule compared to only 24% of women selecting that choice in the 364-day arm (Table 5). Although self-reported premenstrual syndrome, pre-existing menstrual pain, and migraine/recurrent headache symptoms did decrease, as measured by our limited questionnaire, we did not collect any additional medical history or prospective data to confirm the presence or absence of these diagnoses in our study population (Table 5). The questionnaire responses for women completing the entire year were similar to the responses of those exiting early (data not shown).
Tolerability of all treatment regimens was good. Medical reasons for early study termination and adverse events were comparable over the 4 treatment groups (Fig. 2 and Table 2). Overall, throughout the 1-year study period, 63.0 % of the women in the 28-day cycle, 64.5% in the 49-day cycle, 68.6% in the 91-day cycle, and 67.0% in the 364-day cycle reported an adverse event starting after randomization (ie, after the run-in cycle). Rates of adverse events such as breast tenderness and weight changes were sporadic and did not appear to occur more frequently in extended users. Slightly more subjects in the extended cycle arms experienced an adverse event that was considered to be drug-related by investigators (30.6% in the 28-day cycle versus 37.4%, 40.0%, and 35.8% in the 49-day cycle, 91-day cycle, and 364-day cycle, respectively). The distribution for specific adverse events such as headache, mood changes, and breast complaints were similar across the arms, although there were 7 women reporting a breast complaint who had been assigned to the 364-day cycle compared with none using the 28-day cycle regimen. But headache events were reported by 16, 21, 21, and 18 women assigned to the 28-day, 49-day, 84-day, and 364-day cycle arms, respectively.
Nine women reported serious adverse events during the study. Six of these events were deemed not related to study drug use and include an appendectomy and a hysterectomy for ovarian endometriosis and pelvic pain, hospitalizations for infectious diarrhea, diabetic ketoacidosis, mania, and an episode of abdominal and pelvic pain with a normal laparoscopy. Three of the serious events were considered to be possibly related to the study drug: a case of acute and chronic cholecystitis and cholelithiasis (49-day cycle), for which the subject underwent cholecystectomy during her ring-free week and then continued the study; one case of a uterine 4.7-cm leiomyoma at baseline, in which the subject later exited for hysterectomy (364-day cycle), presumably for bleeding, although her diary pages were lost; and one case of deep venous thrombosis (91-day cycle). This third woman was found on testing to have a genetic predisposition for thrombosis with elevated anticardiolipin immunoglobulin (Ig)G antibodies, elevated treatment anticardiolipin IgM antibodies, and protein S deficiency, and additionally, she had a grandmother with a history of thrombosis.
Physical and gynecological examinations revealed very few clinically relevant abnormalities and no apparent differences between the treatment groups. Clinically relevant abnormalities for hematology and biochemistry values rarely occurred, and there were no significant differences between the treatment groups in these parameters, including blood pressure, hemoglobin, body weight, or endometrial thickness at baseline or exit (Table 6).
There was nothing from demographic or clinical data that correlated with a risk of exiting the study early for unacceptable bleeding. We randomized 46 women (10.7% of total randomized population) with a body mass index of 30 or more, and 17 of these women discontinued early, distributed equally over the 4 arms. Specifically current smoking, heavy smoking, body mass index, older age, recent pregnancy, and recent contraceptive-type history were all no different between women completing the study and those exiting early for bleeding (data not shown). Women exiting the study typically continued with hormonal contraception and reported returning to normal menstrual patterns. Of women completing the poststudy telephone interview, use of the ring was continued by 50 (50%), 50 (53%), 40 (44%), and 40 (44%) women, and women switching to the OC numbered 8 (8%), 11 (12%), 15 (16%), and 14 (16%), in the 28-day, 49-day, 91-day, and 364-day arms, respectively, which supports the finding that many women chose the ring at study exit.
We have shown extension of the contraceptive vaginal ring use schedule is feasible and will result in fewer overall days of bleeding, but for many women there can be an increase in unscheduled spotting days compared with the predictable cycle of the 28-day regimen. The hypothesis that a 15% reduction in total bleeding and spotting days would be seen with extended ring use was not met, and it is likely that this was too optimistic compared with the standard regimen. The frequency of unscheduled bleeding was lowest with the shortest extension—the 49-day cycle assignment. Typically, the unscheduled bleeding requires no protection or only one pad or tampon used for the day and is, by definition, spotting only. With cycle extension, particularly with continuous use (364-day cycle), we measured bleeding variability not seen with shorter regimens, underscoring that an individual woman’s experience can range from amenorrhea to unacceptable bleeding.
Acceptability of a contraceptive is a major factor influencing whether women continue to use it. In our study, many women found the ring acceptable and were satisfied with ring use and continued to use the ring following this study. Even women reporting unscheduled bleeding did not always feel that it was a problem and reported desiring a reduction in scheduled bleeding. However, more women discontinued the study for unacceptable bleeding with the longer extended cycles. Perhaps, if these women had been given a more realistic expectation of the amount of irregular bleeding or spotting, it may have been more acceptable. Expectation of irregular bleeding is likely to lead to better acceptance of the bleeding when it does occur. We also did not recruit women seeking a reduction in bleeding, and it is possible that women with conditions worsened by the menses, such as dysmenorrhea or endometriosis, may have more tolerance for irregular bleeding to attain menstrual suppression for these symptoms.
With extended use, one will inherently and by definition get an additional week of hormone exposure per 28-day cycle, and without this hormone-free week the exogenous hormone levels do not drop. The third week of use of the OC and the patch are associated with higher ethinyl estradiol levels compared with the first week, unlike the ring with a steady state of release in which the levels are remarkably stable.16
In an earlier OC study, there were only minor changes between baseline and 1 year in 70-day OC cycle users on lipid parameters and hemostatic factors, but there was a trend toward an increase of sex hormone binding globulin levels compared with women with 28-day cycle use.23 Safety parameters reported with an earlier extended OC study measured a small but significant increase in weight in the extended cycle users.9 In the more recent 91-day OC cycle study, the authors did not share the data regarding weight and only state “there were no clinically meaningful changes.”7 The results of our study (Table 6) are reassuring although there is a small trend in the 364-day cycle subjects toward an increase in weight. This could be spurious and the result of our small numbers, or it is possible the extra week of progestin exposure could facilitate weight gain. It is interesting to speculate whether women would tolerate some change in weight to have amenorrhea. Certainly the long-term use of depo-medroxyprogesterone acetate injection has been associated with both of these conditions—weight gain and amenorrhea.
The pelvic ultrasound examination and endometrial histology results from our study were reassuring, and with ring schedule extension there was no evidence of endometrial hyperplasia (Table 6). The irregular bleeding also did not impact hemoglobin. We did not detect a difference in reported bleeding days between OC switchers and new hormonal contraceptive starts, possibly because the run-in cycle to demonstrate a functioning endometrium before randomization may have eliminated any differences because all endometria were equally exposed to the ring before randomization.
Various extended OC regimens have been studied, and there is no evidence that one is superior to another. One could argue that any cyclic use would always require a scheduled reminder, so for simplicity, a continuous regimen may be preferable. In our study we looked at several extended ring regimens, and it is not possible to rate one over the others. Women who are willing to tolerate some spotting may choose a longer extension to gain more menstrual suppression or to relieve cycle-related symptoms, whereas women who are seeking a more predictable bleeding pattern may want to stay with a shorter regimen. In summary, we compared 4 different schedules for ring use and withdrawal bleeding, and they were all tolerable, effective, and apparently safe. More work should be done to find solutions to ameliorate the irregular bleeding with the longer extended cycles. This study reaffirms that cycle control with the 28-day regimen is excellent. Yet many women may be willing to trade regularity for a reduction in the scheduled menstrual period. Women are individuals, and the selection of the most appropriate regimen will depend upon each woman’s wishes and expectations.
1.Sullivan H, Furniss H, Spona J, Elstein M. Effect of 21-day and 24-day oral contraceptive regimens containing gestodene (60 μg) and ethinyl estradiol (15 μg) on ovarian activity. Fertil Steril 1999;72:115–20.
2.Pierson RA, Archer DF, Moreau M, Shangold GA, Fisher AC, Creasy GW. Ortho Evra/Evra versus oral contraceptives: follicular development and ovulation in normal cycles and after an intentional dosing error. Fert Steril 2003;80:34–42.
3.Thomas SL, Ellertson C. Nuisance or natural and healthy: should menstruation be optional for women? Lancet 2000;355:922–4.
4.Glasier AF, Smith KB, van der Spuy ZM, Ho PC, Cheng L, Dada K, et al. Amenorrhea associated with contraception: an international study on acceptability. Contraception 2003;67:1–8.
5.den Tonkelaar I, Oddens BJ. Preferred frequency and characteristics of menstrual bleeding in relation to reproductive status, oral contraceptive use, and hormone replacement therapy use. Contraception 1999;59:357–62.
6.Sulak PJ, Cressman BE, Waldrop E, Holleman S, Kuehl TJ. Extending the duration of active oral contraceptive pills to manage hormone withdrawal symptoms. Obstet Gynecol 1997;89:179–83.
7.Anderson FD, Hait H. A multicenter, randomized study of an extended cycle oral contraceptive [published erratum appears in Contraception 2004;69:175]. Contraception 2003;68:89–96.
8.Loudon NB, Foxwell M, Potts DM, Guild AL, Short RV. Acceptability of an oral contraceptive that reduces the frequency of menstruation: the tri-cycle pill regimen. Br Med J 1977;2:487–90.
9.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.
10.Miller L, Notter K. Menstrual reduction with extended use of combination oral contraceptive pills: randomized controlled trial. Obstet Gynecol 2001;98:771–8.
11.Sturdee DW, Ulrich LG, Barlow DH, Wells M, Campbell MD, Vessey MP, et al. The endometrial response to sequential and continuous combined oestrogen-progesterone replacement therapy [published erratum appears in BJOG 2001;108:335]. BJOG 2000;107:1392–400.
12.Coutinho EM, O’Dwyer E, Barbosa IC, Gu ZP, Shaaban MM, Aboul-Oyoon M, et al. Comparative study on intermittent versus continuous use of a contraceptive pill administered by vaginal route. Contraception 1995;51:355–8.
13.Miller L, Hughes JP. Continuous combination oral contraceptive pills to eliminate withdrawal bleeding: a randomized trial. Obstet Gynecol 2003;101:653–61.
14.Vercellini P, Frontino G, De Giorgi O, Pietropaolo G, Pasin R, Crosignani PG. Continuous use of an oral contraceptive for endometriosis associated recurrent dysmenorrhea that does not respond to a cyclic pill regimen. Fertil Steril 2003;80:560–3.
15.Timmer CJ, Mulders TMT. Pharmacokinetics of etonogestrel and ethinyl estradiol released from a combined contraceptive vaginal ring. Clin Pharmacokinet 2000;39:233–42.
16.Alexander NJ, Baker E, Kaptein M, Karck U, Miller L, Zampaglione E. Why consider vaginal drug administration? Fertil Steril 2004;82:1–12.
17.Mulders TM, Dieben TO. Use of the novel combined contraceptive vaginal ring NuvaRing for ovulation inhibition. Fertil Steril 2001;75:865–70.
18.Dieben TO, Roumen FJ, Apter D. Efficacy, cycle control and user acceptability of a novel combined contraceptive vaginal ring. Obstet Gynecol 2002;100:585–93.
19.Novak A, de la Loge C, Abetz L, van der Meulen EA. The combined contraceptive vaginal ring, NuvaRing: an international study of user acceptability. Contraception 2003;67:187–94.
20.Oddsson K, Leifles-Fischer B, Wiel-Masson D, de Melo NR, Benedetto C, Verhoeven CH, et al. Superior cycle control with a contraceptive vaginal ring compared with an oral contraceptive containing 30 μg ethinylestradiol and 150 μg levonorgestrel: a randomized trial. Hum Reprod 2005;20:557–62.
21.Veres S, Miller L, Burington B. A comparison between the vaginal ring and oral contraceptives. Obstet Gynecol 2004;104:555–63.
22.Tuppurainen M, Klimscheffskij R, Venhola M, Dieben TO. The combined contraceptive vaginal ring (NuvaRing) and lipid metabolism: a comparative study. Contraception 2004;69:389–94.
23.Cachrimanidou AC, Hellberg D, Nilsson S, von Schoulz B, Crona N, Siegbahn A. Hemostasis profile and lipid metabolism with long-interval use of a desogestrel-containing oral contraceptive. Contraception 1994;50:153–65.
Figure. No caption available.