Comparison of contraceptive ring and patch for the treatment of symptomatic endometriosis

Received

publishe

P.V. has no

to disclo

L.F. has

Supported

Grant F

Reprint req

Universi

20122

E-mail: p

FC

2150

Comparison of contraceptive ring and patch for thetreatment of symptomatic endometriosisPaolo Vercellini, M.D.,a Giussy Barbara, M.D.,a Edgardo Somigliana, M.D., Ph.D.,b

Stefano Bianchi, M.D.,c Annalisa Abbiati, M.D.,a and Luigi Fedele, M.D.a

a Istituto Ostetrico e Ginecologico Luigi Mangiagalli, Universit�a degli Studi di Milano; b Fondazione Ospedale Maggiore

Policlinico Mangiagalli e Regina Elena; and c Division of Obstetrics and Gynecology, Ospedale San Giuseppe-

Fatebenefratelli, Milan, Italy

Objective: To evaluate the efficacy and tolerability of a contraceptive vaginal ring and transdermal patch in thetreatment of endometriosis-associated pain.Design: Patient preference cohort study.Setting: Academic center.Patient(s): Two-hundred and seven women with recurrent moderate or severe pelvic pain after conservative sur-gery for symptomatic endometriosis.Intervention(s): Continuous, 12-month treatment with a vaginal ring supplying 15 mcg of ethinyl E and 120 mcg ofetonogestrel per day or a transdermal system delivering 20 mcg of ethinyl E and 150 mcg norelgestromin per day.Main Outcome Measure(s): Satisfaction with treatment.Result(s): One-hundred and twenty-three women preferred the ring, and 84 preferred the patch. Forty-four ringusers (36%) and 51 patch users (61%) withdrew. Thirty-six of 79 subjects (46%) in the ring group and 14 of 33(42%) in the patch group shifted from continuous to cyclic use because of irregular bleeding. Pain symptomswere reduced by both treatments, with the ring being more effective than the patch in patients with rectovaginallesions. According to an intention-to-treat analysis, 88 of 123 ring users (72%) and 40 of 84 patch users (48%)were satisfied with the treatment received.Conclusion(s): Patients who preferred the ring were significantly more likely to be satisfied and to comply withtreatment than those who chose the patch. Both systems were associated with poor bleeding control when usedcontinuously. (Fertil Steril� 2010;93:2150–61. �2010 by American Society for Reproductive Medicine.)

Key Words: Endometriosis, dysmenorrhea, dyspareunia, pelvic pain, estrogen-progestogen combination, patientpreference trial, satisfaction with treatment

Symptomatic endometriosis is suppressed but not cured byhormonal therapies, and pain recurrence is frequent afterthe withdrawal of drugs. Consequently, medications aresometimes required for prolonged periods. Because of theiradequate safety and efficacy, acceptable tolerability, and af-fordable costs, oral progestogen and estrogen-progestogencombinations (oral contraceptives [OCs]) have gained in-creasing popularity among clinicians and patients as long-term treatments. In addition to pain reduction, these factorscontribute substantially to overall patient satisfaction, whichis, in our opinion, the most important outcome of the thera-peutic strategy of this benign but chronic condition.

However, despite the potentially high efficacy of oral pro-gestogens and OCs, effectiveness in typical use is lower.

November 25, 2008; revised and accepted January 7, 2009;

d online March 27, 2009.

thing to disclose. G.B. has nothing to disclose. E.S. has nothing

se. S.B. has nothing to disclose. A.A. has nothing to disclose.

nothing to disclose.

in part by the University of Milan School of Medicine Research

IRST no. 12-01-5068118-00067.

uests: Paolo Vercellini, M.D., Clinica Ostetrica e Ginecologica I,

ty of Milan, Istituto Luigi Mangiagalli–Via Commenda, 12 -

Milan, Italy (TEL: þ390255032331; FAX: þ390250320252;

[email protected]).

ertility and Sterility� Vol. 93, No. 7, May 1, 2010opyright ª2010 American Society for Reproductive Medicine, P

Many patients regard the need for daily pill intake as a draw-back of oral medications. Inconvenience of the daily pill rou-tine in the context of womens’ social and affective lives maycause reduced compliance (1). As with all medications, ad-herence to the prescribed regimen is essential for the successof treatments for symptomatic endometriosis.

Two contraceptives are available, a monthly vaginal ringand a weekly transdermal patch, that deliver the same typesof hormone combinations used in OCs without requiringdaily use, thus providing greater convenience and the poten-tial for superior tolerance and compliance (2). The overallhigh satisfaction observed with transvaginal and transdermalcontraception (2, 3) may be regarded as an important prereq-uisite for lasting acceptability in patients with symptomaticendometriosis.

Beyond the convenience and simplicity of dosing sched-ules, transvaginal and transdermal drug delivery offer severalbiochemical and metabolic advantages. Continuous adminis-tration maintains constant plasma drug levels and eliminatesthe peaks and troughs associated with oral administration.Loss of bioavailability, owing to first-pass hepatic metabo-lism and enzymatic degradation in the gastrointestinal tractobserved in oral drug use, is prevented. Theoretically, thetherapeutic effect is achievable with lower doses of drugs,

0015-0282/10/$36.00ublished by Elsevier Inc. doi:10.1016/j.fertnstert.2009.01.071

mailto:[email protected]

with reduction of side effects associated with high plasmalevels. Finally, absorption is not influenced by episodes ofvomiting or diarrhea.

Given this background, we sought to compare the efficacyand tolerability of the vaginal ring with those of the transder-mal patch for the management of recurrent pelvic pain afterconservative surgery for endometriosis. Because the mostfrequent symptom in women with endometriosis is dysmen-orrhea (4, 5), both systems were prescribed continuously tosuppress menstruation.

The primary end point was the degree of patient satisfac-tion with treatment. Thus, the issues of patient preferenceand lifestyle must be considered when choosing different de-livery systems such as a vaginal ring and a transdermal patch.If those disappointed with the prescribed alternative tend todecide to withdraw, this would introduce a selection bias.Furthermore, motivational differences could affect compli-ance with the assigned therapy, particularly when the treat-ments under investigation differ dramatically. Motivationmay also introduce bias in the measurement of outcome, par-ticularly if these measures are subjectively assessed by thoseparticipating in the trial; this applies, for example, to treat-ment success, satisfaction, acceptability, and quality of lifefollowing management of symptomatic endometriosis.

In light of these considerations, we decided to conduct apatient preference trial. According to this study design, twogroups of recruits are generated in whom motivational factorsare optimized by allowing them to receive their preferredtreatment (6). The preference groups can be viewed as partof a nonrandomized prospective cohort study. A patient pref-erence study design has been adopted occasionally in gyne-cology (6), obstetrics (7, 8), and internal medicine (9).

MATERIALS AND METHODS

The manuscript was prepared according to the Strengtheningthe Reporting of Observational studies in Epidemiologyguidelines for reporting observational studies (10).

Design

This patient preference, prospective cohort study comparedtwo estrogen-progestogen combinations delivered by twodifferent systems, a vaginal ring and a transdermal patch,for the treatment of recurrent pelvic pain after conservativesurgery for symptomatic endometriosis. The investigationwas performed in an academic department specializing inthe study and management of endometriosis. The study wasapproved by the local institutional review board and wasnot supported by pharmaceutical companies. Informed con-sent was obtained from the subjects before any protocol-spe-cific procedures were completed.

Participants

At pretrial screening, demographic information and a medicalhistory were obtained. We considered 18- to 40-year-old

Fertility and Sterility�

women who were menstruating regularly and not wantingpregnancy, with a baseline body mass index of 18–27 kg/m2,and who had undergone laparoscopy or laparotomy for stageI to IV (11) symptomatic endometriosis in other hospitals inthe previous 12 months.

Patients in whom rectovaginal lesions were not excised

were also included. In our center, the diagnosis of rectovagi-

nal endometriosis is based on vaginal and rectal examination,

transvaginal and transrectal ultrasonography, and histological

demonstration of endometriosis in specimens taken during

previous surgery or during biopsy of the posterior fornix. In

these cases, the diagnostic workup includes kidney and

urinary tract ultrasonography and rectosigmoidoscopy.

Subjects with persistent pelvic pain of more than 6 months’duration and who were unwilling to undergo further short-term surgery were deemed eligible. Exclusion criteria were:obstructive uropathy or bowel stenosis; evidence of complexadnexal cysts or an ovarian endometrioma with a diametergreater than 3 cm at vaginal ultrasonography; therapies forendometriosis other than nonsteroidal antiinflammatorydrugs (NSAIDs) in the 3 months before study entry (6 monthsfor GnRH analogues); the typical contraindications for estro-gens and progestogens; the use of drugs that interfere withcontraceptive steroid metabolism; history of dermal hyper-sensitivity; allergy to components of the study medicationsor to NSAIDs; abnormal findings at breast examination andmammary ultrasound scan; presence of active cutaneous dis-orders; an abnormal cervical smear; any degree of genitalprolapse; a diagnosis of sexually transmitted diseases andvaginal infections; a diagnosis of concomitant pelvic inflam-matory disease, pelvic varices, or genital malformations atprevious surgery; known gastrointestinal, urologic, and or-thopedic diseases; psychiatric disturbances; history of drugor alcohol abuse; and unwillingness to tolerate menstrualchanges.

After an introductory explanation, the women wereinformed that medical therapies for endometriosis generallyinduce only temporary relief and are not expected to be defin-itively curative. In addition, when hormonal treatments are tobe continued for long periods, estrogen-progestogen combi-nations appear to be the compounds that most favorably bal-ance benefits, harm, and cost.

Standard and predefined verbal and written information onthe vaginal ring and the transdermal patch was provided, in-cluding detailed description of composition and modality ofuse. Patients were told that these systems deliver the samehormones as OCs without requiring daily assumption andthat their weekly or monthly use may provide greater conve-nience as well as superior tolerance and compliance com-pared with OCs. Participants were counseled regarding thecommon hormonal side effects associated with estrogen-pro-gestogen combinations. Specific mention was given to unto-ward effects unique to the ring (including expulsion andincreased vaginal secretions) and to the patch (includingskin irritation and system detachment).

2151

Women were also informed that, whereas encouragingevidence exists on the use of progestogens and OCs inpatients with endometriosis, no data were available to sup-port the use of the study drugs and systems as a medicaltreatment for this condition. Finally, the technical difficul-ties and the risks associated with radical repeat surgerywere described, especially to subjects with persistent recto-vaginal lesions.

Measurements

Each patient was asked to complete a questionnaire on thepresence and severity of dysmenorrhea, deep dyspareunia,and nonmenstrual pelvic pain graded using a 0- to 3-pointmultidimensional categorical rating scale modified fromthat devised by Biberoglu and Behrman (12), which definesdysmenorrhea according to loss of work efficiency andneed for bed rest, nonmenstrual pain according to various de-grees of discomfort and use of analgesics, and deep dyspareu-nia according to limitation of sexual activity.

The women were also requested to grade the severity of dys-menorrhea, deep dyspareunia, and nonmenstrual pelvic painusing a 100-mm visual analog scale, with the left extreme indi-cating the absence of pain and the right extreme indicating themost pain possible; a score of 1 to 50 was considered mild pain,51 to 80 moderate pain, and 81 to 100 severe pain. Cut-offpoints defining different categories of pain were chosen onthe basis of a previous correlation analysis with the Biberogluand Behrman multidimensional categorical rating scale (4, 12).Subjects were recruited if they had at least one moderate orsevere persistent symptom on both scales.

Cycle control was evaluated from information recordeddaily by the participants on diary cards. Irregular bleedingduring treatment was defined as spotting (scanty bleeding re-quiring % 1 pad or tampon per day); breakthrough bleeding(light or moderate bleeding requiring R 2 pads or tamponsper day); and metrorrhagia (more than normal menstruation).Any bleeding or spotting occurring during a ring- or patch-free week was classified as withdrawal bleeding.

Baseline fasting peripheral blood samples were obtained tomeasure concentrations of glucose, aspartate and alanineaminotransferases, total bilirubin, lactate dehydrogenase,alkaline phosphatase, total, high-, and low-density- lipopro-tein cholesterol, and triglycerides. Only women with normalvalues were considered for recruitment.

Treatment Assignment

The vaginal ring is a monthly contraceptive that supplies15 mg of ethinyl estradiol (EE) and 120 mg of etonogestrelper day in a low, steady state. Etonogestrel is the biologicallyactive metabolite of desogestrel. The ring is used intravagi-nally for 21 days, usually followed by a 7-day hormone-free interval. The need for compliance is restricted to theday of insertion and the day of removal (2).

2152 Vercellini et al. Contraceptive ring and patch for endo

The transdermal system marketed in Europe is a combinedhormonal contraceptive system containing 0.60 mg of EE and6.0 mg of the progestin 17-deacetylnorgestimate (the primaryactive metabolite of norgestimate). The 20-cm2 patch de-livers 20 mg of EE and 150 mg of norelgestromin per day tothe systemic circulation and is generally used as a cyclic reg-imen of three consecutive 7-day patches followed by 1 patch-free week (13).

After the pretrial screening was completed and thepatient’s understanding of the objectives of the study and oftreatment alternatives was assessed, eligible subjects were re-quested to express their preference of the two delivery sys-tems. Participants who chose the ring were to insert itwithin the first 3 days of onset of menses, whereas thosewho chose the patch were to apply the first transdermal sys-tem within 24 hours of the onset of menses. Subjects in thetransdermal system group were instructed to apply one patchper week to normal, healthy skin of the buttock, upper outerarm, lower abdomen, or upper torso (excluding the breast).

Women were invited to use the ring or the patch continu-ously. However, in case of prolonged spotting (R 7 days)or breakthrough bleeding, they were advised to suspend treat-ment and apply a new system after 1 week. Patches that had tobe removed or were accidentally partially or completely de-tached were replaced with a new patch, which was worn foran entire week.

Patients were allowed to take NSAIDs when needed (nap-roxen sodium, one 550-mg tablet twice a day unless contra-indicated). Compliance was assessed by checking personaldiaries and questionnaires at each visit.

End Point Examination

Every 3 months the patients underwent clinical assessment,vaginal and rectal examination, and transvaginal ultrasonog-raphy, and they were requested to complete the pain question-naire again. Pain during both erratic and regular withdrawalbleeding was considered as dysmenorrhea. In women whowere completely amenorrheic, dysmenorrhea was countedas zero on both scales. Bleeding pattern, analgesic use, andadverse events were also recorded. At the same time points,fasting peripheral blood samples were obtained to measurethe blood chemistries determined at baseline.

At the 12-month evaluation, the women rated the overall de-gree of satisfaction with their treatment on a five-category scale(very satisfied, satisfied, uncertain, dissatisfied, very dissatis-fied) by answering the following question: ‘‘Taking intoconsideration the variations in pain symptoms, in overall well-being and quality of life, the side effects experienced, if any,as well as the convenience of the system chosen, how wouldyou define the level of satisfaction with your treatment?’’

Statistical Analysis

In calculating the sample size required, the primary consider-ation was the rate of satisfied patients (very satisfied plus sat-isfied) after 12 months of therapy. No literature data are

metriosis Vol. 93, No. 7, May 1, 2010

available on the effect of the vaginal ring and transdermalpatch in women with symptomatic recurrent endometriosis.Based on our previous experience on estrogen-progestogencombinations used continuously (14, 15), we hypothesizedthat approximately 75% of the subjects treated with one ofthe two systems would be satisfied. We considered clinicallyrelevant a difference of 20% in satisfaction rate between thestudy groups. To have an 80% chance of detecting such a dif-ference at an overall significance level of 5%, 60 patients pergroup are required. Allowing for dropouts, the aim was torecruit 80 subjects in each group.

Data were archived using Excel 2003 (Microsoft, Red-mond, WA) and exported to SPSS 15.0 (SPSS, Inc., Chicago,IL) for analysis. Baseline patient characteristics were com-pared with the Fisher’s exact test. The statistical significanceof differences in patient satisfaction rates was evaluated usingthe c2 test with Yates’ correction, performing the analysis ac-cording to the intention-to-treat principle. Contingency tableswere also constructed to compare distributions of pain symp-tom frequency and severity, and different bleeding patterns inthe two groups. Variations in pelvic symptoms were assessedby computing pain scores on the verbal rating and visual an-alogue scales, then comparing differences between baselineand final mean scores in the two groups with the independentStudent’s t test. Within-group variations between basal and12-month values were evaluated with the Student’s t testfor paired data. When appropriate, 95% confidence intervals(CIs) were calculated for the observed differences. The inde-pendent Student’s t test was also adopted to compare analge-sic use. All statistical tests were two-sided. In order toestimate the effect of several covariates simultaneously,a multivariate analysis was performed by applying a linear re-gression model. Included in the equations were terms for age,previous pregnancies, body mass index, American Societyfor Reproductive Medicine classification stage, ovarian endo-metriotic cysts, rectovaginal lesions, pain symptoms severity,type of delivery system, and NSAID use. Probability valuesof less than 5% were considered statistically significant.

RESULTS

A total of 292 women evaluated at our endometriosis outpa-tient clinic during 2004–2007 were eligible for the study, but71 declined the suggested delivery systems and requestedother treatment methods and 14 were lost to further contact.One hundred and twenty-three of the remaining patients(59%) chose the vaginal ring and 84 (41%) chose the trans-dermal patch. The mean interval (�SD) between the last sur-gical intervention and the start of the study was 8.1 � 1.8months in the former group and 8.8� 1.4 months in the latter.

The baseline clinical characteristics of the women enrolledin the study are shown in Table 1. The distribution of the vari-ables considered was similar in the two groups. Fifty-nine(28%) patients had rectovaginal lesions. The reasons for thechoice between the two delivery systems are indicated inTable 2.


Tolerability

A total of 80 subjects (39%) discontinued the study and 15(7%) were lost to follow-up. Forty-four (36%) women inthe vaginal ring group and 51 (61%) in the transdermal patchgroup withdrew at different times during the study period forvarious reasons, including side effects and treatment ineffi-cacy, or were lost to follow-up (Fig. 1). The between-groupdifference in proportion of withdrawals was statistically sig-nificant (c2

1, 11.52; P % .001), as subjects who chose thetransdermal patch were 70% more likely to abandon the study(relative risk [RR], 1.70; 95% CI, 1.27–2.28).

The most frequent reason for withdrawal was adverse effects(Fig. 1). Excluding irregular bleeding episodes, side effectswere reported by 46/123 (37%) subjects who chose the vaginalring, and by 36/84 (43%) of those using the patch (P¼0.52;Table 3). Overall, 13 subjects (11%) in the former group and13 (15%) in the latter discontinued treatment because of sideeffects. As expected, vaginal discomfort was more commonin ring users, whereas nausea, vomiting, headache, breast ten-derness, and cutaneous reactions were more common in patchusers. Only 8 subjects using the ring and 4 using the patch re-ported weight increase. The mean gain (�SD) in these patientswas 4.7� 1.7 kg and 3.3� 1.5 kg, respectively. Decreased li-bido was infrequent in both groups. Ring expulsion was expe-rienced by three women (all with fixed uterine retroversion)and patch detachment by four. The second most frequent rea-son for withdrawal was the desire to use an OC. Noticeably,only three participants chose to change the type of deliverysystem (two shifted from the patch to the ring).

All of these subjects were included as failures in the eval-uation of satisfaction with treatment (intention-to-treat anal-ysis), the primary end-point of our study, but were excludedfrom all the other analyses (per protocol analysis) becausecomplete diaries and questionnaires were not available.

Bleeding Control

Irregular bleeding was frequent during continuous use of bothsystems and determined a change to cyclic use in a substantialproportion of participants (ring group, 33/87, 38% at 6-monthfollow-up, and 36/79, 46%, at 12-month follow-up; patchgroup, 14/47, 30%, and 14/33, 42%, respectively). In the re-maining women adhering to the continuous regimen, amenor-rhea was reported at 6- and 12-month evaluation by 21 (40%)and 16 (37%) participants using the ring, and by 11 (34%) and6 (11%) of those using the patch, respectively. Correspondingfigures for spotting were, respectively, 20 (38%) and 22 (51%)in the former group and 17 (52%) and 13 (68%) in the latter. Atthe same time points, breakthrough bleeding was experiencedby 12 (23%) and 5 (12%) ring users, and 5 (15%) and 4 (21%)patch users, respectively. Four women in the patch groupwithdrew because of poor bleeding control.

Effect on Pain

According to the verbal rating scale, at baseline dysmenorrheawas moderate to severe in 106 of 123 (86%) subjects in the

2153

TABLE 1Distribution of study patients according to age, parity, body mass index, endometriosis stage atprevious surgery, number of prior operations, presence of endometriotic cysts or rectovaginal lesions,and preferred treatment.a

Vaginal ring Transdermal patch

Variable n % n %

Age, y<30 28 23 18 21R30 95 77 66 78

Parity0 100 81 71 85R1 23 19 13 15

Body mass index, kg/m2

<21 70 57 47 5621–22 27 22 20 2423–24 12 10 6 7R25 14 11 11 13

Disease stagec

I 31 25 25 30II 30 24 15 18III 29 24 23 27IV 33 27 21 25

No. of previous operations1 92 75 65 772 26 21 17 20R3 5 4 2 3

Endometriotic cysts 39 32 23 27Rectovaginal endometriosis 38 31 21 25

a None of the between-group differences is statistically significant (Fisher’s exact test).bAccording to the revised American Fertility Society classification (11).

Vercellini. Contraceptive ring and patch for endometriosis. Fertil Steril 2010.

vaginal ring compared with 59 of 84 (70%) in the patch group(P¼0.018). In addition, dyspareunia was significantly moresevere in the ring group based on the visual analog scale (44� 35 versus 34 � 35 [mean � SD]; P¼0.047). Consideringall the subjects followed at 6 and 12 months independentlyof severity of pain at baseline, a substantial decrease was ob-served in all symptom scores on the visual analogue and ver-bal rating scales in both study arms. In the vaginal ring group,the mean dysmenorrhea visual analog score fell from 80� 16to 32� 28 at 6 months and to 22� 21 at 12 months (mean�SD). Corresponding figures in the patch group were, 77� 19,42 � 25, and 35 � 20 (mean � SD), respectively (between-group difference, P¼0.03 at 6 months, and P < 0.01 at 12months). Based on the verbal rating scale at 6 and 12 months,pain at menstruation was moderate to severe in 10 of 87 (11%)and 2 of 79 (3%) patients in the former group and in 7 of 47(15%) and 4 of 33 (12%) in the latter, respectively.

In vaginal ring users, mean visual analog scale dyspareuniascores were 28� 27 (mean� SD) at six months, and 22� 17


at the end of the study period. Corresponding figures in patchusers were 27 � 31, and 22 � 23, respectively. At the sametime points, pain at intercourse, as assessed by the verbal rat-ing scale, was moderate to severe in 8 (9%) and 4 (5%) sub-jects in the former group, and, respectively, 4 (12%) and4 (12%) in the latter. Variations in nonmenstrual pain fol-lowed a similar pattern with no between-group statisticallysignificant difference (data not shown).

Details of pain symptom variations separately for womenwithout and with rectovaginal lesions who completed thestudy (per protocol analysis) are shown in Tables 4 and 5,as well as in Figures 2 and 3, in which only subjects reportingthe symptom at baseline were included. In women with rec-tovaginal endometriosis, vaginal ring use was associated witha statistically significant greater reduction of dysmenorrheaon both scales compared with the patch (Table 5, Fig. 3).

In women using the ring, the mean number of days permonth (�SD) using naproxen sodium decreased from 4.0


TABLE 2Reported main reasons for patients’ choice ofthe preferred study delivery system.

Main reason for choice n %

Vaginal ringComfort and practicality 53 43Fear of patch detachment 23 19Greater discretion than the patch 19 15Greater hygiene than the patch 11 9Fear of cutaneous reaction 10 8Attending swimming pool or sauna 7 6

Transdermal patchDifficulty in using vaginal devices 33 39Practicality 27 32Psychological reasons (fear of losing

the ring or wearing a foreign body)19 23

Frequent vaginitis 5 6


� 2.0 at baseline to 1.5� 1.3 at six months and to 1.2� 2.2 at12 months. Corresponding figures in the patch group were 4.2� 2.9, 1.9 � 2.1, and 1.6 � 1.8, respectively, without statis-tically significant between-group differences.

Satisfaction with Treatment

At a 12-month assessment, 30 (24%) women in the vaginalring group were very satisfied with their treatment, 58(47%) were satisfied, 5 (4%) were uncertain, 24 (20%) weredissatisfied, and 6 (5%) were very dissatisfied. Correspondingfigures for the transdermal patch group were, respectively, 10

FIGURE 1

Flow-chart showing recruitment and women’s progress t



(12%), 30 (36%), 2 (2%), 32 (38%), and 10 (12%) (c2 fortrend, 14.61; P< .01). This analysis included all recruited pa-tients. Overall, 71% of subjects in the vaginal ring group weresatisfied or very satisfied after 12 months of treatment com-pared with 48% in the transdermal patch group (c2

1 ¼11.12; P< .001; RR, 1.50; 95% CI, 1.17–1.93). The satisfac-tion rate in participants without rectovaginal lesions was 68%(58/85) in the ring group and 44% (28/63) in the patch group.Corresponding figures in patients with rectovaginal endome-triosis were 79% (30/38) and 57% (12/21), respectively.

Some women were satisfied with their treatment even ifthey requested to shift from the ring or patch to an OC. Mul-tivariate analysis confirmed that the type of system chosenwas the only variable significantly associated with likelihoodof satisfaction with treatment (RR, 1.31; 95% CI, 1.15–1.49).

Safety

No major adverse event was recorded. Limited and nonsig-nificant within-group variations in fasting levels of total,high-density, and low-density lipoprotein cholesterol wereobserved in both groups (data not shown). No blood chemis-try abnormalities were detected in either treatment arm.

An increase in diameter of known endometriomas, appear-ance of new ones, and signs of follicular maturation were notdetected at ultrasonography scheduled during the study period.

DISCUSSION

Endometriosis is a chronic, recurrent disease and recentlyreported reoperation rates for pain relapse after conservativeprocedures are higher than 50% (16, 17). Prolonged medicaltreatment with progestogen and estrogen-progestogen

hrough the study.

2155

TABLE 3Side effects reported during the study period according to treatment group.a

Vaginal ring Transdermal patch

Side effectb n % n % P value

Weight gain 8c 6 4d 5 0.82Headache 7 6 15 18 0.01Vomiting 0 0 2 2 0.31Nausea 3 2 7 8 0.10Depression 7 6 4 5 0.76Decreased libido 5 4 4 5 0.80Breast tenderness 6 5 7 8 0.47Cutaneous reaction 0 0 4 5 0.05Bloating or swelling 12 10 3 3 0.15Vaginal discomfort 8 7 0 0 0.04Vaginal dryness 0 0 2 2 0.31Other 4 3 5 6 0.55

a Patients who withdrew because of side effects are included.b Some women reported more than one side effect.c Mean � SD weight gain, 4.7 � 1.7 kg.d Mean � SD weight gain, 3.3 � 1.5 kg.


combinations has been demonstrated to successfully suppressectopic implants, relieve pain symptoms, and improve health-related quality of life (14, 15, 18), thus offering a valuable al-

TABLE 4Grading of pain symptoms based on the verbal ratinendometriosis before and after 12 months of treatm

Vaginal ring(n [ 51)

Baseline 12 m

Symptom n % n

DysmenorrheaAbsent 0 0 24Mild 11 21 27Moderate 12 24 0Severe 28 55 0

DyspareuniaAbsent 23 45 33Mild 15 29 17Moderate 8 16 1Severe 5 10 0

Chronic pelvic painAbsent 12 24 30Mild 23 45 18Moderate 12 24 3Severe 4 7 0



ternative to serial surgery (19, 20, 21). To achieve these ben-efits, treatments must also be acceptable to promote long-term compliance.

g scale in patients without rectovaginalent according to preferred delivery system.

Transdermal patch(n [22)

o Baseline 12 mo

% n % n %

47 0 0 6 2753 7 32 16 730 5 23 0 00 10 45 0 0

65 12 55 16 7333 6 27 5 222 3 13 1 50 1 5 0 0

59 6 27 9 4135 7 32 11 506 6 27 2 90 3 13 0 0


TABLE 5Grading of pain symptoms based on the verbal rating scale in patients with rectovaginal endometriosisbefore and after 12 months of treatment according to preferred delivery system.

Vaginal ring(n [ 28)

Transdermal patch(n [ 11)

Baseline 12 mo Baseline 12 mo

Symptom n % n % n % n %

DysmenorrheaAbsent 0 0 18 64 0 0 4 36Mild 0 0 8 29 3 27 3 27Moderate 9 32 2 7 4 36 3 27Severe 19 68 0 0 4 36 1 9

Dyspareuniaa

Absent 4 15 6 22 4 36 3 27Mild 7 26 18 66 3 27 5 46Moderate 14 52 3 11 2 18 3 27Severe 2 7 0 0 2 18 0 0

Chronic pelvic painAbsent 5 18 17 61 4 36 5 46Mild 14 50 11 39 5 46 6 54Moderate 7 25 0 0 2 18 0 0Severe 2 7 0 0 0 0 0 0

a One woman in the vaginal ring group who reported no sexual activity was excluded.


Study Design

The nonrandom allocation of treatment alternatives consti-tutes a drawback of this study. However, before designingthe trial, we conducted structured interviews in a sample ofwomen who were representative of our patient populationwith disappointing results, as only a tiny minority of subjectswere willing to accept assignment to both delivery systems.

To our knowledge, the contraceptive ring and patch weredirectly compared in a single randomized controlled trial,conducted for three cycles only in asymptomatic OC users in-terested in nondaily combined hormonal contraception (2).Different entry criteria may explain the difficulties encoun-tered in our setting with regard to random allocation of con-traceptive methods in comparison with the study by Creininet al. (2). In that trial, potential participants were recruited us-ing advertisements and referrals from local health care pro-viders. Hence, only women already willing to accept bothmethods were preselected. Our patients were identified solelyon pain symptoms and were not informed of study details be-fore pretrial counseling. Moreover, our planned period of usewas rather long, thus potentially magnifying patient prefer-ence toward one of the two alternatives. Providing standardand predefined verbal and written information on both deliv-ery systems should have limited undue investigator influence.

Given the apparently insurmountable difficulties in settingup a formal randomized controlled trial, we opted for a more


pragmatic approach. A study based completely on patientpreference introduces a major selection bias, thus limiting in-terpretations of the observed results. In fact, preference armscomparisons have the potential limitations of observationalstudies. Nevertheless, it has been recognized that such a re-search environment may be more similar to ‘‘real world’’ con-ditions (22).

Based on our previous experience, we were convinced thatthe effect of different estrogen-progestogen combinations onpain would have been similar, the main outcome of interestbeing satisfaction with treatment. This latter outcome is theresult of several factors, including not only pain relief butalso tolerance and convenience, which contribute greatly tolong-term compliance. In this context, a patient preferencetrial seemed to us methodologically acceptable, even thoughtreatment alternatives were not allocated randomly.

The rationale for randomization is control of selection biasand elimination of confounding—that is, equal distribution offactors that may influence the outcome—thus ensuring inter-nal validity of the comparison between study groups.However, random allocation of treatment takes all elementsof choice away from the patient. Women involved in the med-ical decision-making process are likely to perform better thanthose who have treatments imposed upon them, especially ifthey are unhappy with their allocation and would have pre-ferred the alternative. Resentful demoralization may lead to

2157

FIGURE 3

Bar graphs showing mean pain symptom scores inpatients with rectovaginal endometriosis before andafter 12 months of treatment according to thepreferred system and pain scale: (A) visual analogscale and (B) verbal rating scale. Error bars representstandard deviation. Mean differences correspond tobaseline values minus 12-month values. *Thebetween-group difference in dysmenorrhea variationis statistically significant on both scales(independent Student’s t test, P < 0.01).


FIGURE 2

Bar graphs showing mean pain symptom scores inpatients without rectovaginal endometriosis beforeand after 12 months of treatment according to thepreferred system and pain scale: (A) visual analogscale and (B) verbal rating scale. Error bars representstandard deviation. Mean differences correspond tobaseline values minus 12-month values. None of thebetween-group differences is statistically significant(independent Student’s t test).


differential acceptance, compliance, and adherence (23);consequently, the outcome measure could be affected byhow satisfied patients are with their allocated medical inter-vention. This effect is largely dependent on the type of out-come being measured, with minimal or no effect for hardoutcomes (survival), but a potentially major influence if theoutcome is patient satisfaction. In the latter case, a compari-son between two groups of patients who have chosen theirtreatment, thereby optimizing their satisfaction, may repre-sent the maximum effect size of the intervention (24).

The main difference between an observational study com-paring two case series and a patient preference trial is the mo-dality of treatment allocation, by the clinicians in the formercase and by patient choice in the latter. Clinicians are prone torecruit different types of women to the two different arms,depending on their belief about the effectiveness of differentoptions. Thus, any difference in outcome cannot be assumed


to be due exclusively to a difference in efficacy of the inter-ventions. Allocation by the patients themselves does not pre-vent the potential for bias, but may offer some reassurancethat the results could be extrapolated to a wider group of sub-jects when women have such strong treatment preferences asto refuse randomization. However, the effect observed underthese conditions can be referred exclusively to patients whospecifically choose that treatment (25).

In addition, analyses of nonrandomized groups should betreated as observational studies, adjusting for known con-founding factors. In this regard, linear regression analysisidentified treatment choice as the sole factor influencingoutcome.

Despite the drawbacks of this particular cohort study, sev-eral trials including both randomized and nonrandomized


participants have demonstrated a substantial consistency inoutcome, regardless of whether the treatment had been ran-domly allocated or chosen (6–8). Moreover, contrary to theprevailing belief, two systematic literature reviews demon-strated that the findings of well-designed observational stud-ies and randomized controlled trials are remarkably similar(26, 27).

According to Brewin and Bradley (25) the pragmatic ques-tion of which treatment is more useful under optimal condi-tions of motivation should be considered as a legitimatealternative to the explanatory, but sometimes unrealistic,question of which is the best treatment per se. Moreover, ithas been observed that preference-based treatment allocationmay optimize cost-effectiveness of intervention (28).

Tolerability

The delivery systems evaluated were potentially promising,hypothetically representing better choices compared withoral medications. However, the observed findings do notseem to fully confirm the expectations. The proportion ofwithdrawals was much higher than that observed in previousstudies conducted in our department using oral progestogensand OCs (14, 15, 18). In particular, more than half of thewomen who chose the transdermal patch did not completethe 12-month study period. This was particularly surprising,because system assignment was based on patient preferencewith the specific objective of increasing acceptability andthe likelihood of treatment continuation. Moreover, manymore women preferred the ring to the patch, with an approx-imate ratio of 3:2. This explains the unbalanced numbers ofsubjects in the two study groups. In fact, we extended the re-cruiting phase until the preplanned number of patients wasreached in the patch group. Other authors observed that theproportion of women choosing the ring is almost doublethat choosing the patch (29).

In the study by Creinin et al. (2), after three cycles 71% ofring users but only 27% of patch users planned to continuetheir treatment method. Moreover, at study completion onlya minority of the subjects who did not wish to continue thepatch (61/175, 35%) opted for the ring, and just 5 of the 70(7%) women who did not wish to continue the ring optedfor the patch. These findings are in line with our pretrialobservations and confirm that women generally have markedpreferences in regard to such different contraceptive modali-ties, even when the planned period of use is brief.

The weight increase reported by a few ring and patch usersappears excessive. However, alternative causal explanationsfor this gain in addition to the medications used were notidentified.

The three ring expulsions were observed in patients withfixed uterine retroversion. According to an in vivo placementmagnetic resonance imaging study, the vaginal ring is locatedsuperior to the urogenital diaphragm, surrounding the cervix.The superior-most aspect of the ring lies behind (posterior to)


the cervix (30). However, rotation of the cervix toward thepubic symphysis facilitates displacement of the ring fromthe posterior to the anterior vaginal fornix, thus increasingthe risk of device expulsion.

Bleeding Control

A fundamental issue in medical treatment for endometriosisis limitation of irregular bleeding, an event particularly wor-rying and frustrating for the patient and frequently associatedwith pain (14, 18). Extended regimens of the vaginal ring upto one year of continuous use have already been evaluated(31, 32). Bleeding days were reduced with postponement ofthe withdrawal bleeding week and adverse events, as wellas several clinical and laboratory variables, were comparablewith the traditional 28-day cycle. Extended use of the trans-dermal system also resulted in fewer median bleeding days.Moreover, a greater incidence of amenorrhea, reduced fre-quency of headache episodes, and high subject satisfactionwere observed (13, 33).

There is a major difference between the extended regimenused in the present study and those assessed in the other trials(13, 31–33). In fact, we instructed the patients to continue thetreatment indefinitely and to interrupt for 1 week only whenprolonged spotting or breakthrough bleeding occurred. In ourexperience with continuous oral progestogens and OC use,this practice optimizes the extended regimen modality, avoid-ing unnecessary induced withdrawal bleedings. With this‘‘1 week off as required’’ modality, drug interruptions areconsidered only as a treatment for irregular and unscheduledbleeding episodes, thus rendering management usually lessproblematic than with fixed period regimens (14, 15, 18).

Despite this, both delivery systems were associated withsuboptimal results, as by the end of the study 46% of thewomen who chose the ring and 42% of those who chosethe patch changed from continuous to cyclic use because ofpoor bleeding control. Moreover, only a minority of thosewho persisted with the continuous use experienced amenor-rhea (37% with the ring and only 11% with the patch). Theseresults compare unfavorably with findings observed withcyproterone acetate (65% amenorrhea rate) (14), an estro-gen-progestogen combination (46%) (14), and particularlywith norethindrone acetate (72%) (18). Moreover, whereasin cases of irregular bleeding with oral medications womenmay just interrupt treatment for one week, with the ringand patch a system change is needed.

Effect on Pain

In subjects who continued the study, both delivery systemswere associated with a major reduction in pain symptoms,with the vaginal ring superior to the patch in regard to dys-menorrhea relief. Actually, subgroup analysis showed thatthis difference in effect is limited specifically to patientswith rectovaginal lesions. A considerable benefit wasobtained also in women experiencing pain during inter-course. We deemed it of interest to compare the two systems

2159

TABLE 6Cost of 12 months of continuous treatmentwith an estrogen-progestin combination ora progestin only: Italy, 2008.

Cost

Medication V $

Vaginal ring 233a 371Transdermal patch 220 a 349Low-dose monophasic OC 80–260b 126–412Norethisterone acetate

2.5 mg/d18c 28

a Cost potentially higher owing to the need for systemremoval/replacement when breakthrough bleedingoccurs.

b The least expensive OCs contain 30 mg of EE and are

in women with rectovaginal endometriosis because a greatereffect of the ring might be expected because of the higher lo-cal progestogen concentration obtainable from direct contactwith lesions of the posterior fornix. Unfortunately, the num-ber of subjects who completed the study in the patch groupwas inadequate to detect as statistically significant the differ-ence observed in dyspareunia scores.

The patients themselves completed pain diaries and ques-tionnaires for the assessment of dysmenorrhea, dyspareunia,and nonmenstrual pain. In addition, two scales were adminis-tered: a visual analog scale that mainly measured the subjec-tive aspect of symptoms and a multidimensional categoricalrating scale that measured the functional impairment causedby pelvic pain. However, only cautious conclusions can bedrawn on the effect of the two study systems on pain symp-toms, as the high number of withdrawals rendered interpreta-tion of data somewhat problematic.

partly reimbursed by the Italian National Health Sys-tem, with an overall yearly patient cost of V 18 - $ 28.

c Partly reimbursed by the Italian National Health Sys-tem with an overall yearly patient cost of V 4 - $ 6.


Satisfaction with Treatment

Based on an intention-to-treat analysis that included all with-drawals and dropouts as failures, the degree of satisfactionwith treatment at 12-month follow-up was significantlyhigher in the ring group than in the patch group. The propor-tion of 71% of satisfied subjects in the vaginal ring group is inline with previous findings on the use of progestogens or otherestrogen-progestogen combinations. In contrast, the propor-tion of 48% of satisfied women in the transdermal patch groupis lower than that usually observed by us (14, 15, 18). Thebetween-group difference in satisfaction rate was similar inparticipants without and with rectovaginal endometriosis(24% and 22%, respectively). Based on these differences,the correctness of the pretrial power calculation is confirmed.

Patient satisfaction was assessed by means of a closed, pre-defined, and written question that was answered by simplycircling or ticking one in a suitable but limited range of givenresponses. This should have guaranteed correctness in deter-mination of outcome grading. Moreover, our findings areconsistent with those of Creinin et al. (2), as in their random-ized controlled trial satisfaction with treatment was signifi-cantly higher in subjects allocated to the ring group (78%)than to the patch group (39%).

Safety

No major adverse event was observed in our study popula-tion. However, it has been reported that use of the transdermalpatch results in an increased overall estrogen exposure com-pared with the ring and low-dose OCs. Data from epidemio-logic studies evaluating the risk of adverse events, includingvenous thromboembolism, in patch and OC users are discor-dant (34, 35). Area under the curve calculations show thatexposure to EE with the patch is 60% higher comparedwith a 30-mg pill, and 250% higher compared with the ring(36). The exposure to EE with the ring was 3.4-fold lowerthan with the patch and 2.1-fold lower than with the OC. Inaddition, the ring produced less variation in hormone serum


levels than the patch or the OC. The elevated exposure toEE may explain the high incidence of nausea, vomiting,and breast-related symptoms frequently observed in patchusers in comparison with OC or ring users (2, 3, 37).

Conclusions

Women who chose a contraceptive vaginal ring for the treat-ment of recurrent symptomatic endometriosis were signifi-cantly more satisfied than those who chose a contraceptivetransdermal patch. Both systems were associated with subop-timal results in terms of patient adherence. The proportion ofparticipants who discontinued the study was particularly highin the patch group. Medical therapy of endometriosis with thetransdermal patch should be limited to highly motivatedwomen who refuse other treatment modalities.

Neither delivery systems guaranteed sufficient bleedingcontrol when used continuously, and our observationsdo not confirm the encouraging findings of previous trials(13, 31–33). During cyclic use, the effect of the vaginalring in terms of pain relief is probably similar to that ofOCs, and the choice should be based on patient preference.

The benefit of the vaginal ring in women with rectovaginallesions should be evaluated further, as the best results wereobserved specifically in this study group.

Finally, when considering prolonged therapies, economicaspects should be taken into adequate account. In this regard,both the ring and the patch appear less advantageous com-pared with OCs and oral progestogens (Table 6).


REFERENCES1. Bajos N, Leridon H, Goulard H, Oustry P, Job-Spira N, COCON group.

Contraception: from accessibility to efficiency. Hum Reprod 2003;18:

994–9.

2. Creinin MC, Meyn LA, Borgatta L, Barnhart K, Jensen J, Burke AE,

et al. Multicenter comparison of the contraceptive ring and patch. Obstet

Gynecol 2008;111:267–77.

3. Urdl W, Apter D, Alperstein A, Koll P, Sch€onian S, Bringer J, et al. Con-

traceptive efficacy, compliance and beyond: factors related to satisfac-

tion with once-weekly transdermal compared with oral contraception.

Eur J Obstet Gynecol Reprod Biol 2005;121:202–10.

4. Vercellini P, Trespidi L, De Giorgi O, Cortesi I, Parazzini F,

Crosignani PG. Endometriosis and pelvic pain: relation to disease stage

and localization. Fertil Steril 1996;65:299–304.

5. Vercellini P. Endometriosis: what a pain it is. Semin Reprod Endocrinol

1997;15:251–61.

6. Cooper KG, Grant AM, Garrat AM. The impact of using a partially rand-

omised patient preference design when evaluating alternative manage-

ments for heavy menstrual bleeding. BJOG 1997;104:1367–73.

7. Henshaw RC, Naji SA, Russell IT, Templeton AA. Comparison of med-

ical abortion with surgical vacuum aspiration: women’s preferences and

acceptability of treatment. BMJ 1993;307:714–7.

8. Nicolaides K, Brizot Mde L, Patel F, Snijders R. Comparison of cho-

rionic villus sampling and amniocentesis for fetal karyotyping at 10-13

weeks’ gestation. Lancet 1994;344:435–9.

9. Knight G, Boulton AJ, Drury J, Gamsu DS, Moses JL, Bradley C, et al. A

feasibility study of the use of continuous subcutaneous insulin infusion in

diabetic clinic: patients’ choice of treatment. Diabet Med 1984;1:267–72.

10. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC,

Vandenbroucke JP, STROBE Initiative. Strengthening the Reporting of

Observational Studies in Epidemiology (STROBE) statement: guidelines

for reporting observational studies. BMJ 2007;335:806–8.

11. Revised American Fertility Society classification of endometriosis:

1985. Fertil Steril 1985;43:351–2.

12. Biberoglu KO, Behrman SJ. Dosage aspects of danazol therapy in endo-

metriosis: short-term and long-term effectiveness. Am J Obstet Gynecol

1981;139:645–54.

13. Stewart FH, Kaunitz AM, LaGuardia KD, Karvois D, Fisher AC,

Friedman AJ. Extended use of transdermal norelgestromin/ethinylestra-

diol: a randomized trial. Obstet Gynecol 2005;105:1389–96.

14. Vercellini P, De Giorgi O, Mosconi P, Stellato G, Vicentini S,

Crosignani PG. Cyproterone acetate versus a continuous monophasic

oral contraceptive in the treatment of recurrent pelvic pain after conserva-

tive surgery for symptomatic endometriosis. Fertil Steril 2002;77:52–61.

15. Vercellini P, Frontino G, De Giorgi O, Pietropaolo G, Pasin R,

Crosignani PG. Continuous use of an oral contraceptive for endometri-

osis-associated recurrent dysmenorrhea that does not respond to a cyclic

pill regimen. Fertil Steril 2003;80:560–3.

16. Cheong Y, Tay P, Luk F, Gan HC, Li TC, Cooke I. Laparoscopic surgery

for endometriosis: how often do we need to reoperate? J Obstet Gynecol

2008;28:82–5.

17. Shakiba K, Bena JF, McGill KM, Minger J, Falcone T. Surgical treatment

of endometriosis: a 7-year follow up on the requirement for further sur-

gery. Obstet Gynecol 2008;111:1285–92.

18. Vercellini P, Pietropaolo G, De Giorgi O, Pasin R, Chiodini A,

Crosignani PG. Treatment of symptomatic rectovaginal endometriosis


with an estrogen-progestogen combination versus low-dose norethin-

drone acetate. Fertil Steril 2005;84:1375–87.

19. Vercellini P, Cortesi I, Crosignani PG. Progestins for symptomatic endo-

metriosis: a critical analysis of the evidence. Fertil Steril 1997;68:

393–401.

20. Vercellini P, Fedele L, Pietropaolo G, Frontino G, Somigliana E,

Crosignani PG. Progestogens for endometriosis: forward to the past.

Hum Reprod Update 2003;9:387–96.

21. Vercellini P, Somigliana E, Vigan�o P, Abbiati A, Daguati R,

Crosignani PG. Endometriosis: current and future medical therapies.

Best Pract Res Clin Obstet Gynaecol 2008;22:275–306.

22. Brocklehurst P. Partially randomised patient preference trials. BJOG

1997;104:1332–5.

23. Togerson D, Sibbald B. Understanding controlled trials: what is a patient

preference trial? BMJ 1998;316:360.

24. Schmoor C, Olschewski M, Schumacher M. Randomized and non-ran-

domized patients in clinical trials: experiences with comprehensive

cohort studies. Stat Med 1996;15:263–71.

25. Brewin CR, Bradley C. Patient preferences and randomised clinical tri-

als. BMJ 1989;299:313–5.

26. Concato J, Shah N, Horwitz I. Randomized, controlled trials, observa-

tional studies, and the hierarchy of research designs. N Eng J Med

2002;342:1887–92.

27. Benson K, Hartz AJ. A comparison of observational studies and random-

ized, controlled trials. N Eng J Med 2008;342:1878–86.

28. Sculpher M. The cost-effectiveness of preference-based treatment allo-

cation: the case of hysterectomy versus endometrial resection in the

treatment of menorrhagia. Health Econ 1998;7:129–42.

29. Victor I, Fink RA. Comparing patient telephone callback rates for differ-

ent hormonal birth control delivery systems. Am J Therapeutics 2006;13:

507–12.

30. Barnhart KT, Timbers K, Pretorius ES, Lin K, Shaunik A. In vivo assess-

ment of NuvaRing placement. Contraception 2005;72:196–9.

31. Barreiros FA, Aparecida C, Guazzelli F, De Ara�ujo FF, Barbosa R.

Bleeding patterns of women using extended regimens of the contracep-

tive vaginal ring. Contraception 2007;75:204–8.

32. Miller L, Verhoven CHJ, Hout J. Extended regimens of the contraceptive

vaginal ring. Obstet Gynecol 2005;106:473–82.

33. LaGuardia KD, Fisher AC, Bainbridge JD, LoCoco JM, Friedman AJ.

Suppression of estrogen-withdrawal headache with extended transder-

mal contraception. Fertil Steril 2005;83:1875–7.

34. Jick SS, Kaye JA, Russmann S, Jick H. Risk of nonfatal venous throm-

boembolism in women using a contraceptive transdermal patch and

oral contraceptives containing norgestimate and 35 ng of ethinyl estra-

diol. Contraception 2006;73:223–8.

35. Cole JA, Norman H, Doherty M, Walzer AM. Venous thromboembolism,

myocardial infarction, and stroke among transdermal contraceptive sys-

tem users. Obstet Gynecol 2007;109:339–46.

36. Van De Heuvel MW, Van Bragt AJ, Alnabawy AK, Kaptein MC. Com-

parison of ethinylestradiol pharmacokinetics in three hormonal contra-

ceptive formulations: the vaginal ring, the transdermal patch and oral

contraceptive. Contraception 2005;72:168–74.

37. Audet MC, Moreau M, Koltun WD, Waldbaun AS, Shangold G,

Fisher AC, et al. Evaluation of contraceptive efficacy and cycle control

of a transdermal contraceptive patch vs an oral contraceptive. A random-

ized controlled trial. JAMA 2001;285:2347–54.

2161

Documents

Comparison of contraceptive ring and patch for the treatment of symptomatic endometriosis