There is currently an unmet need for transdermal contraceptive patch formulations with reduced ethinyl estradiol doses for improved safety profile with the potential to improve compliance in comparison with oral contraceptives. To date, the only contraceptive patch approved for use in the United States is the norelgestromin and ethinyl estradiol transdermal system. However, this patch is associated with 60% greater total drug exposure to ethinyl estradiol compared with a 35-microgram ethinyl estradiol pill1 and possibly an increased risk of venous thromboembolism compared with a combination oral contraceptive.2,3 Additionally, use of a transdermal contraceptive delivery system (Patch) in clinical trials has been associated with greater compliance than combination oral contraceptives.4
An investigational contraceptive patch with levonorgestrel and a lower dose of ethinyl estradiol has been evaluated in phase 1 and 2 studies, which demonstrated that this low-dose combination patch is well tolerated and maintains ethinyl estradiol plasma levels within the range of those reported for low-dose combination oral contraceptives.5,6 The objective of this phase 3 study was to evaluate contraceptive efficacy, safety, participant compliance (using participant diaries and plasma drug levels), and incidence of unscheduled bleeding of the low-dose levonorgestrel and ethinyl estradiol contraceptive patch compared with a marketed low-dose combination oral contraceptive.
PATIENTS AND METHODS
The study enrolled sexually active women aged 17–40 years with regular menstrual cycles (24–35 days). Smokers younger than 35 years of age, women of any body mass index (BMI, calculated as weight (kg)/[height (m)]2), and women with well-controlled hypertension or diabetes mellitus without vascular disease were eligible. All participants were in good general health, confirmed by medical history, physical and gynecologic examinations, clinical chemistry, and liver function values, and were appropriate candidates for combination estrogen–progestin contraception. New users of hormonal contraceptives were women with no exposure to hormonal contraceptives within 6 months of study start; current users were participants who switched to study drugs within 7 days of previous active hormonal contraceptives; and recent users were those who had used a hormonal contraceptive within the previous 6 months (but not within 7 days of study drugs). Study protocol was approved by an institutional review board (IRB) before recruitment. The IRB was New England IRB, Newton, Massachusetts, except for three sites using an IRB of their choice (Schulman Associates, Cincinnati, Ohio; Crescent City IRB, New Orleans, Louisiana; Western IRB, Olympia, Washington). All participants provided written informed consent.
This was an open-label, randomized, parallel-group, multicenter study conducted in the United States from August 2010 to November 2011 comparing contraceptive efficacy and safety of the levonorgestrel and ethinyl estradiol contraceptive patch (Patch) with a combination oral contraceptive (Pill) containing 100 micrograms levonorgestrel and 20 micrograms ethinyl estradiol (Fig. 1). Participants were randomized in a 3:1 (Patch:Pill) ratio. Randomization was further stratified by hormonal contraceptive user status (new users or current or recent users) and BMI (33% of participants were obese [BMI 30 or greater]; 50% of obese participants had BMIs of 35 or greater). After consent, participants were randomized to treatment groups using an interactive voice response system. Women initially randomized to the Patch group continued on the Patch for 13 cycles, whereas those initially randomized to the Pill group were treated for six cycles and crossed over to Patch treatment for an additional 7 cycles. A treatment cycle was defined as a 28-day period: 21 days on treatment (consecutive three 7-day patch wear or 21 days of active pill-taking) followed by 7 days off treatment.
The Patch contains levonorgestrel and ethinyl estradiol (manufactured by Bayer AG) in an active matrix core (15-cm2 area) surrounded by a perimeter adhesive system (total area 26 cm2). The Patch has been shown to provide levonorgestrel and ethinyl estradiol systemic exposure comparable to a combination oral contraceptive containing 120 micrograms levonorgestrel and 30 micrograms ethinyl estradiol.5,6 Each cycle, participants selected their preferred site of Patch application: abdomen, buttock, or upper torso (excluding breasts). The Pill was provided as a calendar pack of tablets for 28-day dosing regimens; the last seven pills contained no steroids. Each active Pill tablet contained 100 micrograms levonorgestrel and 20 micrograms ethinyl estradiol.
Participants used daily diary cards to record information detailing drug administration (cycle days when the Patch was worn, applied, or removed or active pill ingested as well as number of active pills taken during each particular day) and presence of uterine bleeding. All active participants made clinic visits during cycles 2, 4, 6, 9, and 13 for diary review of bleeding patterns, adverse events, and use of concomitant medications, and measurement of vital signs.
Contraceptive efficacy was evaluated using the Pearl Index7 (number of in-treatment pregnancies per 100 woman-years of exposure) based on pregnancies classified as having occurred while on the drug, defined as pregnancies with date of conception after initiation of Patch or Pill and up to and including 14 days after last Patch removal or last active Pill. Pregnancy rate was determined for the intention-to-treat (ITT) population, which included all cycles during which participants were exposed to the study drug (Patch or Pill). As a result of the high number of laboratory-verified noncompliant participants in both groups (9–13%), pregnancy rates were also calculated for the data set excluding these participants. The first six cycles of therapy were used for primary efficacy assessments of the Patch compared with the Pill. The Pearl Index was also calculated for all Patch cycles (from women randomized to the Patch for 13 cycles and women continuing with the Patch for seven cycles after completion of Pill for six cycles).
Participants received urine pregnancy test kits to be used when they did not experience withdrawal bleeding before the start of a new treatment cycle. At the clinic visit; randomization; cycles 2, 4, 6, 9, and 13; and 14 days after the last Patch removal or after the last active Pill ingestion, serum β-human chorionic gonadotropin (β-hCG) measurement (Physician Reference Laboratory Immunoassay using ADVIA) was performed. If urine or serum β-hCG was positive, the study drug was discontinued. Women with a positive pregnancy test had the pregnancy confirmed and (if applicable) date of conception estimated using information based on the following hierarchy: 1) ultrasonography; 2) estimation of gestational age based on pelvic examination, abdominal examination, or both or pregnancy outcome; 3) date of first positive pregnancy test; 4) daily diary information (eg, absence of withdrawal bleeding, participant's complaints); 5) serum β-hCG level; and 6) qualitative urine β-hCG determination.
Treatment compliance was assessed through participant diaries and measurements of plasma drug levels (levonorgestrel and ethinyl estradiol). Self-reported compliance metrics were based on participant diaries; a perfectly compliant cycle was defined as 21 days of Patch wear without missed days or 21 days of active Pill-taking without days of missed Pills.
Blood samples were collected for plasma drug level determinations for all participants at cycles 2 (days 8–11), 6 (days 15–17), and 13 (days 15–17). Blood sampling was scheduled to occur on any day after application of the last scheduled Patch during the cycle (Patch group) or 1–1.5 hours after the active Pill was ingested (Pill group). After venipuncture, plasma was separated by centrifugation and samples were stored at −70°C. Levonorgestrel and ethinyl estradiol plasma concentrations were determined by validated liquid chromatography assay with tandem mass spectrometric detection at the end of the study.8 Values below minimal quantification level (less than 50 pg/mL for levonorgestrel; less than 2 pg/mL for ethinyl estradiol) defined participants as laboratory-verified-noncompliant. Any participant with an undetectable drug level at any of the time points (cycle 2, 6, or 13) was classified as “laboratory-verified-noncompliant.”
Uterine bleeding was defined as any bloody discharge requiring at least one tampon or sanitary pad per day; spotting was defined as bloody discharge requiring only panty liners or no protection. Unscheduled bleeding, spotting episodes, or both were defined as any number of days during the drug-taking interval with bleeding or spotting episodes preceded and followed by at least two bleeding-free days. All participants who took study medication and provided diary cards with relevant bleeding and drug-taking information were included in the cycle control population.
Safety was evaluated through adverse events, discontinuation information, vital signs, physical and gynecologic examinations, and laboratory test results (including lipid panel). Complete physical and pelvic examinations were performed at baseline, cycle 6, and final visit, 14 days after removal of the last Patch, or intake of the last Pill. Pap smears were conducted at baseline and final clinic visit, 14 days after the last active dose of study drug. The safety population included all participants who received the study drug.
All statistical analyses were performed using SAS 9.2 or later. Two-sample t test was used for treatment comparisons of continuous variables and χ2 test was used for treatment comparisons of categorical variables. All statistical tests were two-sided with a significance level of .05.
Sample size for this study was based on U.S. Food and Drug Administration requirements of 10,000 28-day treatment cycles of exposure to the Patch with at least 200 women completing 1 year of therapy. The one-arm, crossover study design (from Pill to Patch after six cycles) was adopted for this purpose. Assuming a 25% dropout rate by the end of cycle 6 and a 40% dropout rate by end of the year of therapy, the study was predicted to generate an adequate number of cycles of exposure to the Patch.
A total of 1,504 women were randomized (n=1,129 Patch; n=375 Pill) and 1,387 participants (n=1,043 Patch; n=344 Pill) were included in the safety population (all women exposed to study drugs) for the comparative part of the study (first six cycles of therapy). A comparable percentage of safety-evaluable participants discontinued prematurely in the Patch group (385/1,043 [36.9%]) compared with the Pill group (95/344 [27.6%]; Fig. 2) during cycles 1–6.
Baseline demographics were similar between the Patch-only and Pill groups that were used for comparison during cycles 1–6 (Table 1) as well as for the group that included switchers from Pill to Patch after six cycles (data not shown); relatively minor treatment differences were not statistically significant. Racial and ethnic minorities were well represented, accounting for 43.1% and 44.5% of participants in Patch and Pill groups, respectively. Of those who were obese, 54.5% had BMIs of 35 or greater in each treatment group. More than half of participants were new users of hormonal contraceptives. Less than 10% were older than 35 years of age and 16–17% were smokers. Most women chose to place the Patch on the abdomen (40%) or buttock (48%); 12% placed the Patch on the upper torso.
Contraceptive efficacy was compared during cycles 1–6. During these cycles, 17 (0.3% of cycles) and five (0.3% of cycles) pregnancies occurred in the Patch (4,962 cycles) and Pill (1,618 cycles) groups, respectively. Contraceptive efficacy was similar in both groups (Table 2). For cycles 1–6, Pearl Index was 4.45 for the Patch and 4.02 for the Pill. When laboratory-verified-noncompliant participants were excluded from the ITT population, Pearl Index decreased to 2.82 for the Patch and 3.80 for the Pill. For all calculated Pearl Indices, between-group treatment differences were not statistically significant (Table 2) (Efficacy data are based on the data included in a new drug application under review by the FDA. During the review, the FDA suggested that 1 of 6 pretreatment pregnancies had a date of conception after drug exposure and that 3 of 4 post-treatment pregnancies occurred within the 14-day post-treatment window. In addition, the FDA review suggested that 4 pregnancies of 17 with insufficient evidence of pregnancy did have sufficient evidence of pregnancy. Each of these 8 pregnancies were in the Patch group, with 2 of these pregnancies occurring during the comparator part of the study (cycles 1–6). In the Patch group, 53% of pregnancies (21/40) occurred in women who were found to have undetectable drug levels; in the Pill group, 80% of pregnancies (4/5) occurred in women with undetectable drug levels.
Pearl Indices for the Patch group were higher in new users than in current or recent users for both the ITT and the ITT that excluded laboratory-verified-noncompliant participants (Table 3). Pearl Indices for the Patch were 6.32 and 4.18 for new users and 2.81 and 2.02 for current or recent users of hormonal contraceptives in the ITT and ITT groups excluding noncompliant participants, respectively. The difference in Pearl Indices for the ITT population reached a statistically significant level (P=.033).
Pearl Indices for the Patch were 4.58 in obese participants (ITT group excluding noncompliant participants 4.63) and 4.40 in nonobese participants (ITT group excluding noncompliant participants 2.15). No pregnancies were noted among obese women in the Pill group; Pearl Indices in nonobese women in the Pill group were similar to the Patch group (ITT 5.59; ITT group excluding noncompliant participants 5.26; Table 2).
Self-reported perfect compliance assessed by participant diary was 91.6% in the Patch group compared with 79.8% in the Pill group (P<.001). However, proportions of participants considered noncompliant based on absence of detectable levonorgestrel and ethinyl estradiol levels in cycles 2 or 6 were not statistically significantly different for the Patch group (9.9 and 11%) and the Pill group (8.8 and 12.6%). Rate of laboratory-verified noncompliance was three times higher among new users (21.8% Patch, 23.5% Pill) than among experienced (current or recent) users of hormonal contraceptives (7.2% Patch, 7.5% Pill; P<.001).
Mean plasma levonorgestrel and ethinyl estradiol concentrations were determined for all participants. Among participants with detectable levels, mean (standard deviation) levonorgestrel concentrations in the Patch group were 1,202 (1,003) pg/mL in cycle 2, 1,753 (1,505) pg/mL in cycle 6, and 1,590 (1,502) pg/mL in cycle 13; in the Pill group, levels 1–1.5 hours after pill ingestion were 2,386 (1,987) pg/mL in cycle 2 and 2,611 (1,933) in cycle 6. Between-treatment differences were statistically significant at both cycles 2 and 6 (P<.001). Mean (standard deviation) plasma ethinyl estradiol concentrations in the Patch group were 30.5 (23.3) pg/mL in cycle 2, 36.7 (30.2) pg/mL in cycle 6, and 31 (27.4) pg/mL in cycle 13; in the Pill group, 1–1.5 hours after pill ingestion, they were 35.4 (31.4) pg/mL and 40.1 (51.7) pg/mL in cycles 2 and 6, respectively. The between-treatment difference was statistically significant only at cycle 2 (P<.01).
The percentage of women with unscheduled bleeding or spotting episodes was similar in both groups during cycles 3 and 5 (Patch compared with Pill: 25.4 compared with 23.2% in cycle 3; 22.1 compared with 21.9% in cycle 5). The percentage of women with unscheduled bleeding episodes only was also similar between both groups (Patch compared with Pill: 3.5 compared with 5.5% in cycle 3; 2.6 compared with 3.2% in cycle 5). For all noted parameters, treatment differences were not statistically significant. During cycles 1–6, 13 (1.3%) women in the Patch group and four (1.2%) in the Pill group discontinued the study as a result of bleeding problems (difference not statistically significant). During cycles 7–13, one (0.1%) additional participant discontinued Patch treatment as a result of menorrhagia.
Safety of the two treatments was evaluated by overall incidence of treatment-emergent adverse events. Treatment-emergent adverse events were comparable between both groups in cycles 1–6 (Table 4). For cycles 1–6, 47.5% (495/1,043) of Patch users and 47.4% (163/344) of Pill users experienced a treatment-emergent adverse event; 21.8% (227/1,043) and 16% (55/344), respectively, experienced a drug-related treatment-emergent adverse event. The majority of treatment-emergent adverse events in both groups were mild or moderate; severe treatment-emergent adverse events were experienced by 4.7% (49/1,043) and 2.6% (9/344) of participants in the Patch and Pill groups, respectively. In the Patch group, over 13 cycles, few individual treatment-emergent adverse events had incidence rates 2% or greater; the incidence exceeded 5% only for nasopharyngitis. Treatment differences in overall incidence of treatment-emergent adverse events and incidence of individual treatment-emergent adverse events were not statistically significant (Table 4). Incidence of serious adverse events was low and a similar percentage of women in the Patch (14/1,043 [1.3%]) and Pill (4/344 [1.2%]) groups experienced serious adverse events. Three (0.3%) in the Patch group (uncontrollable nausea and vomiting, drug [antihistamine] overdose, thrombosis in the left subclavian vein) and one (0.3%) in the Pill group (liver problem lacking diagnostic confirmation) were considered possibly related to the study drug. The subclavian thrombosis occurred in an active weight lifter. For cycles 1–6, incidence of study discontinuation resulting from treatment-emergent adverse events was 9.2% (96/1,043) for the Patch group and 4.4% (15/344) for the Pill group. The difference in discontinuation rates resulting from treatment-emergent adverse events was statistically significant (P<.01). Higher discontinuation rates in the Patch group are mainly explained by the incidence of treatment-emergent adverse events related to application site reactions (3.2%). In the Patch group over 13 cycles, incidence of study discontinuations resulting from treatment-emergent adverse events was 10.8% (138/1,273). Incidences of treatment-emergent adverse events commonly associated with estrogen (eg, headache, nausea, vomiting, breast pain) were less than 5% of participants with similar incidences among users of hormonal contraceptives in the Patch and Pill groups (Table 4).
Participant-reported percentage of patches falling off, as reported by women in daily diaries, was 3% (2.1, 3.5, and 3.7% of patches applied to the abdomen, buttock, and upper torso, respectively).
This trial showed that contraceptive efficacy, cycle control, safety, and self-reported and verified compliance with use of the low-dose Patch and a low-dose Pill were similar. The Patch was well tolerated with excellent adhesion and associated with a low rate of treatment-emergent adverse events, including those related to wearability.
With regard to demographics, this clinical trial overenrolled minorities and participant BMI distribution was representative of the U.S. population of women 18–40 years of age.9 Previous contraceptive trials excluded women with BMIs of 32 or greater. 10,11 Immediate switchers from other hormonal contraceptives are less likely to experience contraceptive failure compared with new users.12 In this study, only 17.9% of participants in the Patch group were immediate switchers from other hormonal contraceptives compared with 44–63% of participants in other trials of new hormonal contraceptives. 13–15
Noncompliance with hormonal contraceptives contributes to typical user failure rates.16 This trial found similar rates of noncompliance based on plasma drug levels as those previously reported for women using combination oral contraceptives (17%).17
The high Pearl Index in this study is consistent with observations of higher Pearl Indices in recent combination oral contraceptive clinical trials compared with older studies.12 The Pearl Index in this study appears to reflect a lack of participant compliance (eg, “user failure” as opposed to “method [Pill or Patch] failure,” which reflects contraceptive failures when the contraceptive is administered according to dosing instructions) and suggests that participants in this clinical trial should be considered “typical users” of a hormonal contraceptive. Low socioeconomic status represents a predictor of contraceptive noncompliance12,17 and obesity and poverty are highly correlated in the United States.18 Accordingly, the high prevalence in this clinical trial of minorities, obese women, and women who have not recently used hormonal contraceptives may explain the high rates of contraceptive noncompliance observed. Ideally, future clinical trials assessing efficacy of oral and transdermal contraceptives should consider using laboratory assessments of participants' compliance.
This trial found that the Patch was associated with a Pearl Index of 2.82 in compliant women over six cycles of use and had contraceptive efficacy similar to that of the comparator low-dose Pill. We found that the rate of contraceptive failure among women treated with the Patch was similar in obese and normal-weight women in contrast to findings from the pivotal trials of the norelgestromin and ethinyl estradiol patch, which observed higher failure rates among heavier women.19 Similar failure rates among obese and nonobese women are consistent with results from several previous studies of women using combination oral contraceptives.17,20–23
Unscheduled bleeding and spotting with the Patch was comparable to that of the Pill and improved with time. The low-dose combination Patch was safe and well tolerated, consistent with data from initial clinical trials with the Patch.5,6 There was generally a lower incidence of treatment-emergent adverse events compared with data for the currently marketed higher estrogen exposure ethinyl estradiol and norelgestromin contraceptive patch (Ortho Evra® (norelgestromin/ethinyl estradiol transdermal system) prescribing information. Titusville (NJ): Janssen Pharmaceuticals, Inc; 2013.).2,13,24 Thrombosis represents the principal health risk associated with combination estrogen–progestin contraceptives with venous thromboembolism more common than arterial thrombosis.25 One case of subclavian vein thrombosis considered possibly related to Patch treatment occurred in a participant in this study who was an active weightlifter. Repetitive, strenuous upper extremity activities may be associated with subclavian vein thrombosis.26 Risk for venous thromboembolism is associated with ethinyl estradiol dose and is substantially higher in obese compared with nonobese women, both with and without combination oral contraceptive use.27,28 The association of the ethinyl estradiol and norelgestromin patch with a higher risk of venous thromboembolism than a combination oral contraceptive is uncertain with conflicting epidemiologic findings.3,29 Although our trial did not have the statistical power to assess venous thromboembolism risk, a patch that is associated with substantially lower ethinyl estradiol exposure than that of the norelgestromin and ethinyl estradiol patch and similar to that of a 30-microgram ethinyl estradiol combination oral contraceptive might be a better option for women (both obese and nonobese).
The Patch had good adhesion properties. Fall-off rates were similar to those reported for Ortho Evra.13 The Patch also reported low incidences of skin reaction.
1. Devineni D, Skee D, Vaccaro N, Massarella J, Janssens L, LaGuardia KD, et al.. Pharmacokinetics and pharmacodynamics of a transdermal contraceptive patch and an oral contraceptive. J Clin Pharmacol 2007;47:497–509.
2. van den Heuvel MW, van Bragt AJ, Alnabawy AK, Kaptein MC. Comparison of ethinylestradiol pharmacokinetics in three hormonal contraceptive formulations: the vaginal ring, the transdermal patch and an oral contraceptive. Contraception 2005;72:168–74.
3. Cole JA, Norman H, Doherty M, Walker AM. Venous thromboembolism, myocardial infarction, and stroke among transdermal contraceptive system users. Obstet Gynecol 2007;109:339–46.
4. Archer DF, Cullins V, Creasy GW, Fisher AC. The impact of improved compliance with a weekly contraceptive transdermal system (Ortho Evra) on contraceptive efficacy. Contraception 2004;69:189–95.
5. Archer DF, Stanczyk FZ, Rubin A, Foegh M. Ethinyl estradiol and levonorgestrel pharmacokinetics with a low-dose transdermal contraceptive delivery system, AG200-15: a randomized controlled trial. Contraception 2012;85:595–601.
6. Stanczyk FZ, Rubin A, Flood L, Foegh M. Pharmacokinetics, tolerability and cycle control of three transdermal contraceptive delivery systems containing different doses of ethinyl estradiol and levonorgestrel. Horm Mol Biol Clin Invest 2011;6:231–40.
7. Pearl R. Factors in human fertility and their statistical evaluation. Lancet 1933;222:607–11.
8. Licea-Perez H, Wang S, Bowen CL, Yang E. A semi-automated 96-well plate method for the simultaneous determination of oral contraceptives concentrations in human plasma using ultra performance liquid chromatography coupled with tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 2007;852:69–76.
9. Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. JAMA 2012;307:491–7.
10. Grimes DA, Shields WC. Family planning for obese women: challenges and opportunities. Contraception 2005;72:1–4.
11. Society of Family Planning, Higginbotham S. Contraceptive considerations in obese women: release date 1 September 2009, SFP Guideline 20091. Contraception 2009;80:583–90.
12. Trussell J, Portman D. The creeping pearl: why has the rate of contraceptive failure increased in clinical trials of combined hormonal contraceptive pills? Contraception 2013;88:604–10.
13. Audet MC, Moreau M, Koltun WD, Waldbaum AS, Shangold G, Fisher AC, et al.; ORTHO EVRA/EVRA 004 Study Group. 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.
14. Hampton RM, Short M, Bieber E, Bouchard C, Ayotte N, Shangold G, et al.. Comparison of a novel norgestimate/ethinyl estradiol oral contraceptive (Ortho Tri-Cyclen Lo) with the oral contraceptive Loestrin Fe 1/20. Contraception 2001;63:289–95.
15. Anderson FD, Hait H. A multicenter, randomized study of an extended cycle oral contraceptive. Contraception 2003;68:89–96.
16. Trussel J. Contraceptive failure in the United States. Contraception 2011;83:397–404.
17. Westhoff CL, Torgal AH, Mayeda ER, Stanczyk FZ, Lerner JP, Benn EK, et al.. Ovarian suppression in normal-weight and obese women during oral contraceptive use: a randomized controlled trial. Obstet Gynecol 2010;116:275–83.
18. Bleich SN, Thorpe RJ Jr, Sharif-Harris H, Fesahazion R, Laveist TA. Social context explains race disparities in obesity among women. J Epidemiol Community Health 2010;64:465–9.
19. Zieman M, Guillebaud J, Weisberg E, Shangold GA, Fisher AC, Creasy GW. Contraceptive efficacy and cycle control with the Ortho Evra/Evra transdermal system: the analysis of pooled data. Fertil Steril 2002;77(suppl 2):S13–8.
20. Kaneshiro B, Edelman A, Carlson N, Nichols M, Jensen J. The relationship between body mass index and unintended pregnancy: results from the 2002 National Survey of Family Growth. Contraception 2008;77:234–8.
21. Westhoff CL, Hait HI, Reape KZ. Body weight does not impact pregnancy rates during use of a low-dose extended-regimen 91-day oral contraceptive. Contraception 2012;85:235–9.
22. Foegh M, Archer DF, Stanczyk FZ, Rubin A, Mishell DR Jr. Ovarian activity in obese and nonobese women treated with three transdermal contraceptive patches delivering three different doses of ethinyl estradiol and levonorgestrel. Contraception 2013;87:201–11.
23. Lopez LM, Grimes DA, Chen M, Otterness C, Westhoff C, Edelman A, et al.. Hormonal contraceptives for contraception in overweight or obese women. The Cochrane Database of Systematic Reviews 2013, Issue 4. Art. No.: CD008452. DOI: 10.1002/14651858.CD008452.pub3.
24. Wan GJ, Barnowski CE, Ambegaonkar BM, Bolge SC, McDonnell DD. Treatment satisfaction with a transdermal contraceptive patch or oral contraceptives. Contraception 2007;75:281–4.
25. Petitti DB. Hormonal contraceptives and arterial thrombosis—not risk-free but safe enough. N Engl J Med 2012;366:2316–8.
26. Joffe HV, Goldhaber SZ. Upper-extremity deep vein thrombosis. Circulation 2002;106:1874–80.
27. Pomp ER, le Cessie S, Rosendaal FR, Doggen CJ. Risk of venous thrombosis: obesity and its joint effect with oral contraceptive use and prothrombotic mutations. Br J Haematol 2007;139:289–96.
28. Meade TW, Greenberg G, Thompson SG. Progestogens and cardiovascular reactions associated with oral contraceptives and a comparison of the safety of 50- and 30-microgram oestrogen preparations. Br Med J 1980;280:1157–61.
29. Sidney S, Cheetham TC, Connell FA, Ouellet-Hellstrom R, Graham DJ, Davis D, et al.. Recent combined hormonal contraceptives (CHCs) and the risk of thromboembolism and other cardiovascular events in new users. Contraception 2013;87:93–100.