The combined oral contraceptive (OC) is the most common method of birth control used by women in the United States.1 Despite its potentially high efficacy, effectiveness in typical use is much lower.2 The need for daily compliance is a factor in the lower-than-expected efficacy of combined OCs.3,4 In the third month of use, up to 50% of women miss three or more pills per cycle.3
In 2002, two products became available to U.S. consumers that deliver the same types of hormones found in combined OCs without requiring daily use. The contraceptive vaginal ring (NuvaRing, Organon USA, Inc., Roseland, NJ) and the contraceptive patch (OrthoEvra, Ortho Pharmaceuticals, Raritan, NJ) were developed as a means of potentially improving compliance and satisfaction as compared with combined OCs. Although published studies directly compare clinical outcomes with use of the contraceptive ring and combined OCs5–8 and the contraceptive patch and combined OCs,9,10 the contraceptive ring and patch have never been directly compared.
Clinicians need a better understanding of how these methods compare to provide accurate counseling for women interested in a nondaily combined hormonal contraceptive. Currently, no objective information exists to advise satisfied combined OC users whether they would be happier with the contraceptive ring or patch. We conducted this trial to evaluate these choices for women who are content with combined OCs but who are considering a change to a nondaily delivery system. We designed this trial to compare continuation rates after 3 months of use of the contraceptive ring and the contraceptive patch. Continuation rates tend to indicate method satisfaction, and satisfaction after the third cycle with the contraceptive ring predicts a low risk of subsequent discontinuation.11 This study also compared adverse effects, the incidence of bacterial vaginosis, and direct measures of acceptability between the two groups.
MATERIALS AND METHODS
The study was reviewed and approved by the institutional review boards of each of the 10 participating institutions: Boston University, Columbia University, Eastern Virginia Medical School, Johns Hopkins University, Oregon Health and Science University, University of California, Los Angeles, University of Chicago, University of Pennsylvania, University of Pittsburgh, and University of Wisconsin. The University of Pittsburgh served as the sponsoring institution; all protocol changes were submitted and approved by the University of Pittsburgh Institutional Review Board before submission to other institutions. All participants were informed and provided written consent before participation in the study.
Potential participants were recruited locally at each site using advertisements and referrals from local health care providers. We enrolled healthy women with no past experience using the contraceptive ring or patch who were currently using combined OCs or had stopped combined OCs within 3 months of the enrollment date. All participants stated they had been content with combined OCs and would otherwise continue using their product. Women who had recently discontinued did so for reasons unrelated to satisfaction (eg, did not obtain a refill). Women agreed to avoid intravaginal products unless prescribed by an investigator. If heterosexually active, participants needed to have a single partner for 3 months before and while participating in the study and agreed to avoid anal intercourse and use the study method as their primary contraceptive. Exclusion criteria included contraindication to combined hormonal contraception (see box), allergy to a component of the study contraceptives, use of drugs that increase cytochrome P450 activity (see box) within the preceding 2 months, recent pregnancy, planning a pregnancy in the next 6 months, or use of injectable contraception within 6 months. Women weighing more than 90 kg were ineligible, as were women with a recent (within 3 months) diagnosis of gonorrhea, Chlamydia, or trichomonas vaginalis.
After completing informed consent, we obtained demographic information, a medical history, blood pressure, and height and weight and performed breast and pelvic examinations. Women with symptomatic vaginal discharge were evaluated with vaginal pH, wet preparation, and potassium hydroxide preparation. Participants returned for their enrollment visit after results of gonorrhea and Chlamydia testing were known. At enrollment, each participant completed a preenrollment visual analogue scale (VAS) assessment of satisfaction with her current combined OC. The VAS used a 100-mm line with the left and right sides of the 100-mm line labeled “worst birth control method” and “best birth control method,” respectively. In a similar fashion, participants were asked which nondaily method they preferred to use with the left and right sides of the line labeled “contraceptive patch” and “contraceptive ring,” respectively.
Blood pressure and weight were measured, and a vaginal smear was performed for Gram stain evaluation. The participants were randomly assigned to use either the contraceptive ring or patch for four continuous cycles beginning with the next menses. Research staff at the study site opened the next sequentially numbered opaque envelope that contained the product assignment. An individual unassociated with the clinical portion of the study prepared the envelopes. Randomization sequence was created using Stata 9.0 (StataCorp., College Station, TX) statistical software and was stratified by center with a 1:1 allocation using random block sizes of 2, 4, and 6. Participants received four cycles of their assigned study medication, a copy of the U.S. Food and Drug Administration (FDA)–approved patient information, and written initiation instructions. Participants assigned to the ring were to insert it within the first 5 days of the onset of menses. Participants using the contraceptive patch were to apply the first patch within 24 hours of the onset of menses.
Follow-up included two telephone calls and one visit. Participants contacted the research office by telephone within five days after their menses following the enrollment visit to confirm method initiation. Participants were contacted during the second week of the second study cycle of study medication to verify continued product use. Participants returned during the first week of the fourth cycle (or sooner if she required or requested early discontinuation). At this final visit, we measured blood pressure and weight and obtained a vaginal smear for Gram stain evaluation. Each participant completed a poststudy VAS assessment of how much they liked their assigned study product. The left and right sides of the 100-mm line labeled “very little” and “very much,” respectively. In a similar fashion, participants were asked which method they preferred to use with the left and right sides of the line labeled “contraceptive patch” and “contraceptive ring,” respectively.
Participants completed an acceptability questionnaire modified from a validated questionnaire in women using the contraceptive ring.12 We developed two versions of the questionnaire—one for ring users and one for patch users. Those questions from the original questionnaire that referred to the contraceptive ring were changed to reflect the patch (for example, expulsion questions were changed to reflect adherence of the patch). Each participant was asked before final contraceptive counseling whether she desired to continue using her assigned method after she completed her current cycle. All telephone calls between the researchers and study participants were recorded in the participant's chart, with careful documentation of the reason for and duration of the telephone call.
In November 2005, the FDA changed the package labeling for the contraceptive patch and issued a warning regarding serum estrogen levels in the contraceptive patch as compared with combined OCs. Participants enrolled in the study, regardless of study product assignment, were given standard, written information about this warning.
The primary aim of this study was to assess continuation rates after three cycles of study medications. Estimates of these rates are unavailable in the literature. A 12-month efficacy study (which enrolled both switchers and new starters) demonstrated a 3-month continuation rate of 80% for the contraceptive ring.13 We assumed that a study of short duration that enrolled women who were content with oral contraception would yield a higher continuation rate of 90%. Similar data were unavailable for the patch. A 3-month study of new starters of the contraceptive patch found that only 60% of participants planned to continue the contraceptive patch after the study14; thus, we hypothesized a lower continuation rate for the patch. Based on continuation rates of 90% with the contraceptive ring and 80% with the contraceptive patch, 436 total participants were required when using a two-tailed .05 significance level and 80% power. We assumed that up to 15% of women would not have final outcome data for the primary analysis and, therefore, increased the total participants recruited to 500 women.
Our primary analysis group included all enrolled participants who started their assigned treatment and attended a discontinuation or final follow-up visit. Analyses were repeated using an intent-to-treat population of all randomly assigned participants in which those participants without follow-up were considered to have discontinued product use. Statistical analyses were performed using SPSS 14.0.1 statistical software (SPSS Inc., Chicago, IL), and all tests were evaluated at the .05 significance level. Baseline demographic data were compared according to treatment group using Student t test, Fisher exact test, or χ2 test, as appropriate. The status of participants at the end of the study (ie, completed study, lost to follow-up, or discontinued) was compared by treatment group using χ2 tests stratified by study site. The continuation rates at three cycles and the intended continuation rate after four cycles were compared using Fisher exact test. We developed logistic regression models to determine whether product assignment was independently associated with continuation outcomes after adjusting for the effects of potential confounding variables that were identified by univariable analyses. Statistical evaluation of product assignment in these models was based on the Wald χ2 test statistic. To control the type I error for multiple analyses of the primary endpoint, Holm's step-down procedure was used to adjust the P values associated with the continuation outcomes.15
To minimize the intrusiveness of the study, information on bleeding and adverse effects were not collected using a diary; such information was derived from the poststudy questionnaires. Acceptability, as measured by visual analog scales, was compared using the Mann-Whitney U test. Acceptability measured by categorical questions from the acceptability questionnaires was evaluated by χ2 analysis.
Nugent's scores were categorized as normal flora (score=0–3), intermediate flora (score=4–6), and bacterial vaginosis flora (score=7–10). Between-group differences in the categorized Nugent score at each visit and the acquisition and loss of bacterial vaginosis flora by the final visit was evaluated using the χ2 test. McNemar's test was used for evaluating paired changes in flora within each group. Because adverse effects, acceptability, and Nugent's scores were secondary endpoints, these P values were not adjusted for multiplicity.
Between June 2005 and September 2006, we enrolled 500 participants (Fig. 1). Twenty-one women were excluded from the analyses, including 12 women lost to follow-up, six women who withdrew consent after randomization, and three women who never started their assigned treatment. The 479 women in primary analysis population exceeded the sample size estimate of 436 women. Both groups were similar in demographic characteristics (Table 1). No pregnancies were reported in either group.
The number of women who completed three cycles of use of their assigned product in the per protocol population was 228 of 241 (94.6%, 95% CI 91.0–97.1%) and 210 of 238 (88.2%, 95% CI 83.4–92.0%) for ring and patch users, respectively (P=.03). The completion rates were similar for the intent-to-treat population as well (91.6% and 83.7%, respectively, P=.03) and did not vary by study center. The most common reasons for discontinuation among the 13 women who discontinued the ring were discomfort with use (n=4) and adverse effects (n=4). The most common reasons for discontinuation among the 28 women who discontinued the patch were adverse effects (n=10), skin irritation (n=6), and problems with adherence (n=5).
Of the women who completed three cycles of their assigned product, 171 of 241 (71.0%, 95% CI 64.8–76.6%) of ring users and 63 of 238 (26.5%, 95% CI 21.0–32.6%) patch users planned to continue use of their method (P<.001). The 70 women who did not wish to continue the ring opted for a combined OC (n=52), the contraceptive patch (n=5), another method (n=12), or no method (n=1). The 175 women who did not wish to continue the patch opted for a combined OC (n=99), the contraceptive vaginal ring (n=61), or another method (n=15). Only five women who used the contraceptive patch stated they were influenced by the FDA labeling change; all completed use but opted to discontinue after the study because of the warnings.
When adjusting for the possible confounding effects of education, age, and whether the participant was actively using combined OCs at the time of enrollment, patch users were twice as likely to discontinue product use by the third cycle (odds ratio 2.3, 95% CI 1.1–4.6) and seven times more likely to state that they did not wish to continue using the study product at study exit (odds ratio 7.1, 95% CI 4.7–10.7) compared with ring users. These findings did not change significantly from the unadjusted odds ratios.
Adverse effects are presented in Table 2. Women switching from the pill to the patch were significantly more likely than women switching to the ring to experience longer periods, increased dysmenorrhea, frequent nausea, frequent mood swings, and frequent skin rash and less likely to experience frequent vaginal discharge. These differences affected acceptability, because more women stated they were satisfied with the contraceptive ring than the contraceptive patch (Tables 3 and 4). Acceptability based on VAS assessments (Table 4) indicates that ring users preferred the ring to the combined OC (P<.001), and patch users preferred the combined OC to the patch (P<.001).
Women noted use problems with both products (Table 5). The patch fell off at least once during any 3-week use period in 108 (46.0%, 95% CI 39.5–52.6) women and the ring was expelled at least once during any 3-week use period in 49 (20.4%, 95% CI 15.5–26.1) women (P<.001). Overall, the patch was either removed or fell off during a 3-week use period in 117 of 235 (49.8%, 95% CI 43.2–56.4) users. For comparison, 100 of 240 (41.7%, 95% CI 35.4–48.2) women had the ring out of the vagina at least once during a 3-week use period (P=.08). Reasons cited by patch users (individuals could have more than one reason) included removal for discomfort (n=18, 7.7%), removal for bleeding or spotting (n=5, 2.1%), fell off during sex (n=12, 5.1%), and fell off at times unrelated to sex (n=100, 42.6%). Reasons (individuals could have more than one reason) for ring users included removal for discomfort (n=10, 4.2%), removal for bleeding or spotting (n=3, 1.3%), removal because of interference with sex (n=32, 13.3%), fell out during sex (n=30, 12.5%), and fell out at times unrelated to sex (n=25, 10.4%).
Women using the contraceptive patch were more likely to call the study site with a problem than women using the contraceptive ring. Overall, 66 ring users made 115 calls and 90 patch users made 179 calls (P=.03 for number of women making calls). The total time for these calls was similar, approximately 3 minutes each. Calls regarding adverse effects were similar: 62 calls from 40 ring users and 62 calls from 29 patch users (P=.2). Calls for problems with adherence or retention of the method were more common among patch users (97 calls from 52 participants) than ring users (45 calls from 23 participants, P<.001 for number of women making calls). Twenty-one women (8.7%) using the contraceptive ring made 24 interim visits, whereas 11 women (4.6%) using the patch made one interim visit each. Combining early discontinuation and interim visits, 34 (14.1%) ring users and 38 (16.0%) patch users came back to the office during the first 3 months of use (P=.6).
Baseline and follow-up Nugent scores (Table 6) did not vary between groups after three consecutive cycles of use. There was no difference in the percentage of women using the vaginal ring improving from bacterial vaginosis to normal or intermediate flora as compared with the number worsening from normal or intermediate flora to bacterial vaginosis (3% compared with 5%, P=.53). These changes were significantly different for women using the contraceptive patch (2% improving compared with 8% worsening, P=.01). Most of these women were asymptomatic.
Advances in hormonal contraceptive delivery systems have led to the development of nondaily methods, which provide the potential for increased compliance. Health care providers need to understand how these newer delivery systems fit into the contraceptive landscape. Although combined OCs are the most commonly used reversible form of contraception in the United States,1 continuation rates are only 40% at 6 months of use.16 The two alternative delivery systems for combined hormonal contraceptives had not been evaluated in a single randomized trial, making it difficult for clinicians to compare and contrast these options for their patients. We found that women who are happy with combined OCs but are interested in switching to a nondaily method were much more satisfied with a contraceptive vaginal ring than a patch. Most (71.0%) women who used the ring planned to continue that method after the study, whereas most (73.5%) women who used the patch did not. In our population, 21.6% of ring users and 41.6% of patch users opted to return to their oral contraceptive after using their assigned nondaily method.
Consistent with these continuation rates, women reported being more satisfied with the ring than the patch and were more likely to recommend the ring to their friends. This higher acceptability was reported by women despite findings that the ring was more likely to be bothersome during sex to them or their partners. Patch users reported significantly more systemic adverse effects. Surprisingly, some women experienced an increase in menstrual duration or pain; both of these outcomes were significantly more likely when women switched to the patch as compared with the ring. The only adverse effect (Table 2) that was more common in ring users was an increase in vaginal discharge. This discharge was not associated with a higher likelihood of bacterial vaginosis.
We think that the patch detachment and ring expulsion rates we report likely reflect the experiences of women in typical use. Audet et al9 reported that a total of 4.6% of all patches were replaced for either complete (1.8%) or partial (2.8%) detachment. We did not specifically ask our participants how many patches detached. If we assume that, over the three study cycles, only one of the nine patches detached, approximately 5% of all patches would have detached in our participant population. This estimate is similar to that reported by Audet et al.9 Ring expulsion rates are generally not reported, because most studies include this information within an overall category of general compliance. A single study of 1,130 Dutch women reported that 14% of users removed the ring at least once for intercourse during the first 3 months of use, 18% removed the ring incidentally for reasons other than intercourse, and 6% experienced expulsion.17 Our rates of removal and expulsion were much higher, which we hypothesize could be due to potential differences in the populations studied or the definitions used. If the user felt the ring had fallen out, we counted the event as an expulsion regardless of whether the expulsion was partial or complete. The prior study may have only counted complete expulsions. Regardless, the high ring removal and expulsion rates did not result in poor acceptability. Although some women and providers worry about ring expulsion and discomfort, especially with intercourse, these data are reassuring.
We expected that adverse effects and use problems would result in telephone calls to our study sites. Victor and Fink18 previously reported in a prospective study of 208 women followed for three cycles of use of their desired product that significantly more women using the contraceptive patch called with problems (43%) compared with women using a combined OC (27%) or ring (15%). The authors stated that most telephone calls were for adverse effects. We also found more calls among contraceptive patch users in our larger randomized trial, but the number of calls for adverse effects was similar for each method in our population. The major difference in calls was related to adherence of the patch or retention of the ring. Our results, combined with those of Victor and Frank,18 confirm that women using the contraceptive patch will call the office with more complaints than those using the contraceptive ring.
Our findings differ significantly from randomized trials reporting that women find the ring and the patch superior or equal to the pill in terms of acceptability.5–10 Prior trials typically randomly assigned women to a pill or a new, otherwise unavailable method. Women may enroll in such trials because they are unhappy with their current birth control method, which, most commonly, was a combined OC. Our study demonstrates that women without such a bias are unlikely to be happy with a contraceptive patch and will likely be happier with a contraceptive ring. A major strength of our study was that, unlike clinical trials to establish efficacy, we followed women in a manner that more closely reflects clinical practice. Women did not use daily diaries and returned for only a single follow-up visit after 3 months. A weakness of our study, given the finding of superior acceptability of the ring, is that we did not specifically ask women why they preferred it to the combined OC. Additionally, we only assessed adverse effects at the final study visit. This retrospective method might not accurately reflect the true incidence of adverse effects, but it does capture the overall perception of adverse-effect burden considered by women.
We had an unexpected modification during the trial in relation to the FDA revision of contraceptive patch information. It is likely that these warnings affected enrollment in a manner that we cannot accurately measure; however, only five participants who enrolled in the study stated that their decision not to continue with the method was influenced in any way by the FDA actions. In other words, even although we required and ensured that all participants receive the FDA warning, this information apparently did not factor into the decision to complete the study or continue with the patch. A smaller single-center study of 59 new start patch users that pre-dated the FDA warning found 91.5% complete three cycles of use, but only 57% planned to continue beyond the fourth cycle.14 Our planned rate of continuation of 26.5% is much lower and could represent more realistic findings across multiple centers. Alternatively, this lower rate may represent an effect of the FDA warning that participants did not reveal to the investigators. Regardless, weekly transdermal contraception remains an important option for women seeking to avoid unintended pregnancy, and providers should not uniformly discount the potential benefits of this particular method for some of their patients.19 Because the patch remains an important contraceptive option for women, the results of our study are useful for individuals switching from their current method.
This randomized comparison of women switching from a combined OC to a contraceptive patch or ring clearly demonstrates that women are more satisfied and have fewer adverse effects and problems related to product use with the ring than the patch. Additionally, contraceptive ring users require less provider time for telephone consultations than patch users. Women who are content with their combined OC and have never used a patch or ring are more likely to be happy with the ring than the patch. These findings do not imply that all women using a combined OC should be counseled to switch to a ring. However, the information from this study can help health care providers counsel women who desire a nondaily combined hormonal contraceptive method.
1. Mosher WD, Martinez GM, Chandra A, Abma JC, Willson SJ. Use of contraception and use of family planning services in the United States: 1982-2002. Adv Data 2004;(350):1–36.
2. Trussell J. Contraceptive failure in the United States. Contraception 2004;70:89–96.
3. Potter L, Oakley D, de Leon-Wong E, Canamar R. Measuring compliance among oral contraceptive users. Fam Plann Perspect 1996;28:154–8.
4. Rosenberg MJ, Waugh MS, Burnhill MS. Compliance, counseling and satisfaction with oral contraceptives: a prospective evaluation. Fam Plann Perspect 1998;30:89–92.
5. Oddsson K, Leifels-Fischer B, de Melo NR, Wiel-Masson D, Benedetto C, Verhoeven CH, et al. Efficacy and safety of a contraceptive vaginal ring (NuvaRing) compared with a combined oral contraceptive: a 1-year randomized trial. Contraception 2005;71:176–82.
6. Bjarnadottir RI, Tuppurainen M, Killick SR. Comparison of cycle control with a combined contraceptive vaginal ring and oral levonorgestrel/ethinyl estradiol. Am J Obstet Gynecol 2002;186:389–95.
7. Ahrendt HJ, Nisand I, Bastianelli C, Gomez MA, Gemzell-Danielsson Urdl W, et al. Efficacy, acceptability and tolerability of the combined contraceptive ring, NuvaRing, compared with an oral contraceptive containing 30 microg of ethinyl estradiol and 3 mg of drospirenone. Contraception 2006;74:451–7.
8. Veres S, Miller L, Burington B. A comparison between the vaginal ring and oral contraceptives. Obstet Gynecol 2004;104:555–63.
9. Audet MC, Moreau M, Koltun WD, Waldbaum AS, Shangold G, Fisher AC, et al. Evaluation of contraceptive efficacy and cycle control of a transdermal contraceptive patch vs an oral contraceptive: a randomized controlled trial. JAMA 2001;285:2347–54.
10. Urdl W, Apter D, Alperstein A, Koll P, Schonian S, Bringer J, et al. Contraceptive efficacy, compliance and beyond: factors related to satisfaction with once-weekly transdermal compared with oral contraception. Eur J Obstet Gynecol Reprod Biol 2005;121:202–10.
11. Szarewski A. High acceptability and satisfaction with NuvaRing use. Eur J Contracept Reprod Health Care 2002;7 suppl:31–6.
12. Novak A, de la Loge C, Abetz L. Development and validation of an acceptability and satisfaction questionnaire for a contraceptive vaginal ring, NuvaRing. Pharmacoeconomics 2004;22:245–56.
13. 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.
14. Murthy AS, Creinin MD, Harwood B, Schreiber CA. Same-day initiation of the transdermal hormonal delivery system (contraceptive patch) versus traditional initiation methods. Contraception 2005;72:333–6.
15. Holm S. A simple sequentially rejective multiple test procedure. Scand J Stat 1979;6:65–70.
16. Westhoff C, Heartwell S, Edwards S, Zieman M, Cushman L, Robilotto C, et al. Initiation of oral contraceptives using a quick start compared with a conventional start: a randomized controlled trial. Obstet Gynecol 2007;109:1270–6.
17. Roumen FJ, op ten Berg MM, Hoomans EH. The combined contraceptive vaginal ring (Nuvaring): first experience in daily clinical practice in the Netherlands. Eur J Contracept Reprod Health Care 2006;11:14–22.
18. Victor I, Fink RA. Comparing patient telephone callback rates for different hormonal birth control delivery systems. Am J Ther 2006;13:507–12.
19. Burkman R. Transdermal hormonal contraception: benefits and risks. Am J Obstet Gynecol 2007;197:134.e1–6.
© 2008 The American College of Obstetricians and Gynecologists
Figure. No caption available.