The transition from physiologic ovulatory menstrual cycles to a lack of follicles with little estradiol surrounding menopause is when childbearing capability ends. Although fecundity (the average monthly probability of conceiving) begins to decrease around age 32, it declines by 50% at age 43. There is also a declining fertility of the male partner after age 50. In spite of this, most women are still able to conceive, and are therefore “at risk” for conception during perimenopause. With advancing age there is generally a reduction in the frequency of intercourse; however, new relationships are common and contraception is needed.
It is easy for clinicians and their patients to forget to have a discussion about contraception. Such discussions are, however, essential; regardless of prior desire for pregnancy, an unplanned pregnancy is frequently unwanted, pregnancy is more risky andmedical co-morbidities aremore frequent during perimenopause.
Although no contraceptivemethod is contraindicated because of age, per se, some specific methods are better suited for women in their late reproductive years. Factors affecting choice include desire for (or fear of) permanent sterilization, frequency of intercourse, need for protection from sexually transmitted diseases (STDs), a desire for non-contraceptive health benefits, and/or the presence of medical conditions affecting choice. Insurance reimbursement issues and unwarranted fears ormisinformation about highly effective, reversible contraceptive methods are important related issues.
Female sterilization is the most common method of contraception used by perimenopausal women in the US.1 Sterilization is highly effective, the procedure can often be performed on an outpatient basis, and “adverse” effects are generally limited to short-term postoperative pain and a small surgical scar.
Surgical risks are directly related to surgical or anesthesia complications, which are more closely related to body habitus and general health than to age.2 Tubal sterilization has a higher morbidity rate than other methods.3
The risk of pregnancy after sterilization is related to age and the type of procedure performed. The rate of method-dependent failure ranges from 0.75% to 3.65%;2 if tubal sterilization is chosen, the patient should be informed that a resultant pregnancy may be ectopic. Women should also be reminded that sterilization (regardless of the method) provides no inherent protection from STDs.
Although it is commonly believed that sterilization leads to menstrual cycle irregularity, the best evidence suggests that the procedure does not causemenstrual problems.4 There is some suggestion that there may be a reduced risk of ovarian cancer after sterilization,5 and regret after sterilization is much less common among perimenopausal women than it is among younger women.
In addition to the factors described above, decision-making regarding
contraception used by perimenopausal women in the US.1
female sterilization should also include discussion of vasectomy when a woman has a single committed partner. Vasectomy is highly efficacious; the failure rate is 1/20006 and the associatedmorbidity rate is quite low. The relative cost of the procedures may also be a consideration for some patients.
First introduced in Europe, hysteroscopic tubal insertion (without an abdominal scar) is now common in the US. At this point in time, the considerable skill required to perform this procedure and the cost of the device render this method less suitable for widespread use worldwide.
With contraception efficacy equal to that obtained with female sterilization and vasectomy, the additional benefit of reduced menstrual flow makes insertion of an intrauterine device (IUD) an attractive choice for many perimenopausal women.
The copper-bearing IUD is considered an optimum method for older women without menstrual problems who need a safe, effective, long-term method of contraception that is also reversible. Because the associated cost is encountered at the front-end, the copper-bearing IUD is extremely cost-effective if used for the 10-year duration allowed. Heavier menses can occur after insertion of this device, and nonsteroidal anti-inflammatory agents effectively reduce cramping and the amount of bleeding.7
The low rate of failure after insertion of a levonorgestrel IUD is similar to that after tubal sterilization and vasectomy.8 This progestin- containing IUD is effective for 5 years and has the advantage of eventual amenorrhea in the majority of users; according to the package insert the rate of eventual amenorrhea is 20% (Mirena; Berlex package insert; available at: http://berlex.bayerhealthcare.com/ html/products/p/wp-content/themes/twentytwelve/images/mirena_patient_ insert.pdf).
Both the copper-bearing and the progestin IUDs can be associated with spotting between menstrual periods. Each IUD requires insertion by a clinician, but little in the way of maintenance thereafter. The progestin IUDmay provide the specific additional benefit of delivering therapeutic levels of a progestin to the endometrial cavity, averting the hyperplastic effects of unopposed estrogen during a perimenopausal anovulatory cycle, and potentially decreasing dysmenorrhea. Thismay be quite beneficial for patients with rectovaginal endometriosis without adnexal masses.9
The levonorgestrel systemis considered an excellent alternative to hysterectomy and/or uterine ablation for women with menorrhagia.10 When used formenorrhagia control it is associated with reduction of menses for 7 years or longer,11 but contraceptive efficacy is tested and licensed for only 5 years. For perimenopausal women who require hormonal therapy, the progestin IUD offers protection against endometrial hyperplasia and endometrial cancer. In women taking tamoxifen, it has also been used for protection against endometrial proliferation and associated endometrial cancer risk.11 World Health Organization (WHO) category III information suggests that it should not be used in patients currently afflicted with breast cancer; insertion may, however, be considered after 5 years of remission.11
The well-established risks of IUD use are infection within 20 days of insertion, perforation at the time of insertion, and ectopic pregnancy in the unlikely event of contraceptive failure. A common (off-label) practice among women over 40 is to retain a copper IUDuntil 18months after the last menses.12 In women over age 45, a common off-label practice is to keep the IUD in until contraception is no longer needed, since low pregnancy rates have been reported when the progestinonly IUD is used after 7 years. Clinicians should keep good records of their discussions with patients in such instances.
Regardless of which IUD is selected, the clinician should provide thorough counseling, preferably involving the male partner.
Combined oral contraceptives (COCs) are highly effective in preventing pregnancy, primarily by inhibiting ovulation. Failure rates range from 0.2 to 0.4 (in established adherent users) to 3 or more per 100 woman years.11
Only about 11%of women in the US between the ages of 40 and 44, and 4% of 45- to 50-year-old women use COCs.13 This is likely because women in these age groups tend to associate OCs with serious health risks. Low-estrogen-dose COCs may be prescribed to older, healthy non-smoking women up to the age of menopause, provided there is periodic monitoring to identify new risk factors for disease (such as diabetes, smoking, risk factors for venous thromboembolism [VTE], breast cancer, etc.). Little outcome data exist for this practice, however.
COCs are contraindicated in women over age 35 with risk factors for arterial or venous disease; established risk factors include excess weight, hypertension, dyslipidemia, diabetes and cigarette smoking. Hypertension (even if controlled) and migraine headaches are among the other contraindications to the use of any estrogen-containing contraceptive by perimenopausal women.
Approximately 1%of COC users become clinically hypertensive and this rate increases with age.11 Blood pressure that is repeatedly higher than 140/90 mm/Hg (WHO category III) is a cut-point at which the risk is thought to exceed the benefit, and blood pressure may rise with age, independent of COC use. Blood pressure should be carefully monitored before and after starting the COC; a common practice is to measure blood pressure after 3 months or sooner, and subsequently at intervals from 6 months to yearly.
These hormonal therapies increase the risk for VTE. Although the incidence of VTE is very low in reproductive-age women, VTE risk increases with age and body mass index.14,15 The risk of stroke appears to be increased in women with migraines, and the use of estrogencontaining contraceptives in these women further increases the risk.16 Risk factors for venous disease also include excess weight, immobility and a family history of a young firstdegree relative with early venous disease. The risk for venous disease is increased with COC use. This risk is apparently highest within the first severalmonths of use, and decreases with longer duration of COC use. The increased risk for venous disease may be related to previously undiagnosed thrombophilia. The absolute risk of idiopathic venous thromboembolism is small (15- 25 per 100,000 woman years).11 In healthy nonsmoking women the individual risk for arterial disease is very small with low-dose COCs.
Table 1 lists contraindications (relative and absolute) to COC use, and is intended to help determine relative versus absolute contraindications for women of this age. A systematic review of literature (up to 200217) addresses 10 controversial issues related to COC use in women with special conditions. Because of the rarity of events, case-control methodology was often used to infer added risk with OC use. WHO reports that users of COCs who have migraine and hypertension appear to be at increased risk of stroke relative to nonusers with migraine. The evidence for the other eight was either insufficient to enable conclusions or identified no excess risk.
It has been estimated that at age 20, women with migraine who are also taking COCs have an absolute stroke risk of 10 per 100,000, with most of this risk (8 in 10 cases) attributable to the presence of both migraine and COC use.17 By age 40, the absolute risk rises to 100 in 100,000 women, with an attributable risk of 80 per 100,000. As we age our vascular systems are more prone to injury, and the excess risk with OC use is thought to be due to changes in clotting and nonclotting factors in the liver with a slight shift toward thrombosis (Table 1).
Long-term use of COCs is associated with less benign breast disease (fibroadenoma and fibrocystic breast disease).11 After 35 years of age, the cumulative risk of breast cancer increases significantly with age. The COC is a weak co-factor for breast cancer (relative risk is 1.24 in users, as compared to neverusers). 11 After cessation the risk declines progressively, and disappears after 10 years.11 Not all epidemiologic studies agree on this as a risk factor; findings from case-control studies aremixed.Counseling should
weigh this risk against the diminished risk of ovarian, uterine, epithelial and, possibly, colon cancer. As with all contraceptives, risks and benefits should be reviewed with each individual. If breast cancer develops the COC should be discontinued. When breast cancer develops in COC users it is more often clinically less advanced than in non-users.11
COCs may act as a weak cofactor for cervical cancer. A systemic review suggests that the risk of cervical cancer gradually increases with increased duration of use.12 In the presence of human papillomavirus the risk increases after about 5 years,11 and women should be so advised.
The above risks notwithstanding, for healthy women over age 40 the benefits of COCs outweigh the risks, according toWHO.11 Because of reduced fertility and a diminishing need for a powerful contraceptive (as in premenopausal women), and because synthetic estrogens in the COC are prothrombotic, a common practice is to switch to an alternative method around age 51. Not all women will be menopausal at this age (90% will transition by age 53) and some may want to continue the pills to manage symptoms of perimenopause.
When prescribing COCs for perimenopausal women, some clinicians feel that ultra-low-dose estrogen OCs are the prudent choice. Use of traditional OC formulations (21 active tablets followed by 7 hormone-free days) with 20 μg of estrogen is associated with high rates of breakthrough bleeding and spotting. Newer formulations with 25 μg of estrogen and a triphasic regimen of the progestins desogestrel (Cyclessa) or norgestimate (Ortho Tricylen Lo)—both available in the US but not Canada— appear to achieve the low rates of unscheduled bleeding and spotting characteristic of higher-estrogen OCs. Accordingly, these 25-μg estrogen formulations may be a prudent choice for perimenopausal women. However, some women experience mood changes with changing hormones and prefer stable hormone levels. Some perimenopausal women may experience bothersome vasomotor symptoms or other symptoms during the hormone-free interval. Use of continuousOC tablets or formulations with shorter pill-free intervals often reduces such symptoms.
Two 20-μg estrogen COCs formulated with the progestin norethindrone acetate (Lo Estrin-24 Fe) or drospirenone (Yaz) are based on a 24/4 regimen (24 active pills followed by 4 placebo days) rather than the traditional 21/7 cycle. These newer formulations (both available in the US but not Canada) may have better bleeding profiles than those of other formulations containing 20 μg of estrogen. This has not, however, been tested in perimenopausal women.
Drospirenone differs from other synthetic progestins in that it is derived from 17-spirolactone, an analogue of spironolactone. The estrogen-plus-drospirenone OC has similar efficacy and safety profiles to those of other low-dose OCs, but seems to offer improved tolerability with regard to weight gain, acne and mood changes. Drospirenone has anti-mineralcorticoid activity, including the potential for hyperkalemia in high-risk patients. Monitoring of appropriate patients would therefore be prudent.
Use of extended-OC formulations results in withdrawal bleeding less frequently than monthly. Two extended regimens (Seasonale, Seasonique; available in the US but not Canada) include 84 tablets that combine 30 μg estrogen and the progestin levonorgestrel. Scheduled withdrawal bleeding occurs every 3 months. Seasonique has eliminated the hormone-free interval by replacing the 7 placebo pills with 10 μg of ethinyl estradiol.
One newer COC (Lybrel; available in the US but not Canada) is designed to be taken 365 days per year, without a placebo phase or pill-free interval. Consisting of 90 μg levonorgestrel and 20 μg ethinyl estradiol, this continuous-formulation OC provides a steady low dose of hormones. Although no scheduled withdrawal bleeding will occur with this OC, bleeding and spotting are common.
Common side effects of COCs include nausea and breast tenderness, which tend to resolve as OC use continues. Use of COCs does not cause weight gain or headaches in the large majority of patients.
Perimenopausal use of COCs for contraception has been associated with many non-contraceptive benefits (Table 2); these benefits may apply to other forms of estrogencontaining contraceptives.
Contraceptives FDA-approved non-oral combination contraceptives include a vaginal ring (NuvaRing; available in the US and Canada) and a weekly patch (Ortho Evra, available in the US but not Canada), both worn for 3 of 4 weeks. The ring releases ethinyl estradiol and etonogestrel, and the patch delivers ethinyl estradiol and norelgestromin. Etonogestrel and norelgestromin are the biologically active metabolites of desogestrel and norgestimate, respectively, both of which are progestins used in commonly prescribed COCs.
As with COCs, use of either the ring or patch results in cyclical monthly withdrawal bleeding. Because these methods do not require daily pill-taking, compliance (and therefore contraceptive effectiveness) may be higher than with oral combination products. Although side effects and contraindications are, in general, similar to those seen with COCs, breast tenderness is more common with initial use of the patch than with OCs.
Labeling indicates that the patch wearer is exposed to approximately 60% more estrogen than that found in a typical OC containing 35 μg estrogen. As a result, labeling for the patch now includes a warning that the risk of VTE may be higher than with other combination contraceptives. One epidemiologic study found that VTE risk was similar to that of combination OCs, and reported the risk to be 2 times higher with the patch.18 Because age itself is an independent risk factor for VTE, some experts believe that the ring and OCs are more appropriate contraceptive choices for perimenopausal women than is the patch.14 (Readers are referred to the labeling for these products for more details.)
Progestin-only contraceptives can be used by some perimenopausal women for whom contraceptive doses of estrogen are contraindicated. They are, for example, an effective alternative for cigarette smokers over the age of 35, women with hypertension (controlled and uncontrolled), and those with a history of VTE. Contraindications include a history of hormonedependent cancer and unexplained abnormal uterine bleeding.
Progestin-only contraceptives are available as oral tablets, injections, sub-dermal implants and IUDs. They also offer non-contraceptive advantages, such as menstrual control and less risk for thrombosis. (See labeling for more examples).
Oral. Progestin-onlyOCs formulated with norethindrone or norgestrel provide effective contraception for perimenopausal women.Women using these OCs may experience irregular bleeding, spotting or, less commonly, amenorrhea.14 Many consumers refer to this type of OC as “the mini-pill” or “POP” (progestin-only pill).
Injectables.The progestinmedroxyprogesterone acetate (MPA) provides highly effective, long-term contraception when the hormone is administered suspended in a solution (“depot”).DepotMPA (DMPA) is available in the US and Canada (Depo-Provera). The dose, 150-mg
intramuscular (IM) injection into either the buttock or upper arm every 3 months, does not need to be adjusted according to body weight. Although DMPA contraception is reversible, return of fertility after cessation of therapy can be delayed 12 to 18 months.14 Some DMPA users report weight gain. Less common side effects include change in libido, depression and nervousness. Some consumers call DMPA “the birth control shot.”
Depo-subQ Provera 104 is the newer version, containing 31%less progestin (104 mg) than Depo- Provera IM. The dose is injected subcutaneously into the anterior thigh or abdomen every 3 months, using a smaller needle than that needed for IM administration, which may cause less pain. DeposubQ Provera 104 is available in the US, not Canada. It is also FDAapproved for the treatment of endometriosis-related pain.
Initially, irregular bleeding or spotting is common with use of the injectable progestin-only contraceptives. After four or more injections, at least half of users will experience amenorrhea.14Use of these products suppresses ovarian estradiol production, which can reversibly lower bone mineral density (BMD); labeling contains a black-box warning regarding BMD loss.14 No fracture increase has been documented, and BMD has been shown to be reversible after discontinuing these products. Adding calcium/vitamin D supplementation is important. Clinicians prescribing these injectables to perimenopausal women may wish to supplement (“add back”) oral or transdermal estrogen, using doses similar to those used in conventional postmenopausal hormone therapy. Excessive added estrogen carries thrombotic risk above that seen with the progestin-only product.
Implants. An etonogestrel subdermal contraceptive implant product (Implanon; available in the US but not Canada), consisting of one toothpick-sized polymer capsule implanted under the skin of the inner arm, offers perimenopausal women highly effective contraception for up to 3 years. Irregular bleeding or spotting or amenorrhea may occur during implant use. The levonorgestrel subdermal implant product (Norplant), consisting of six silicone rods offering contraception for up to 5 years, is no longer available in North America; however, some women may still be using it.
These progestin-only methods may be useful when synthetic estrogens should be avoided, such as in older women who smoke or who have side effects with COC use. The main disadvantage is menstrual irregularity.
Emergency contraception (EC) refers to contraceptive measures which, if taken after coitus,may prevent pregnancy. It may be appropriate in certain cases, such as after unprotected sexual intercourse, a condom accident, a partially or totally expelled IUD, or if one or more progestin-only contraceptive pill or several COC pills have been missed. EC is meant only for occasional use (when, as stated above, primary means of contraception fail), as other contraceptive methods are more reliable. The EC pill is often referred to as the “morningafter pill,” but this term is figurative since these products are licensed for use for up to 72 hours after sexual intercourse.14
Since EC methods act before post-fertilization implantation of the embryo, they are medically and legally considered forms of contraception, according to the International Federation of Gynecology and Obstetrics. EC pills are not to be confused with the “abortion pill” (RU486, mifepristone, Mifeprex).
Various oral estrogen-progestin contraceptives are FDA-approved in the US and government-approved in Canada for safe and effective EC. Most employ a dose of 120 μg ethinyl estradiol plus 0.60 mg levonorgestrel taken within 72 hours after unprotected intercourse, followed by another dose 12 hours later. Itmust be used within this timeframe as EC has no effect once a fertilized egg is implanted. Efficacy is approximately 75%.14Nausea and vomiting are common side effects.
The progestin-only method, approved in the US and Canada for EC, is more effective and better tolerated than the combined regimen. Marketed as Plan B, this method uses levonorgestrel in a dose of 1.5 mg, either as two 0.7- mg tablets taken 12 hours apart or, more recently, as a single dose. The original FDA labeling indicated that Plan B reduces the risk of pregnancy by 89%.14 Current labeling does not include this figure, but states that if it is taken within 72 hours after coitus, 7 of 8 women who would have become pregnant will not, and Plan B works even better if taken within the first 24 hours after coitus.14 Individuals older that 18 years can purchase Plan B over the counter in the US; otherwise, Plan B is available by prescription. In Canada, Plan B is the only government-approved product for EC.
EC is not indicated for a woman with a known or suspected pregnancy because it is not effective in women who are already pregnant. Inadvertent ingestion of EC during pregnancy is not known to be harmful to the woman or her fetus.14 Concomitant administration of St. John’s Wort and some enzymeinducing drugs (including certain anticonvulsants and the antibiotic Rifampin) may reduce the effectiveness of EC, and larger doses of steroids may be required.
Condoms and diaphragms with spermicides are appropriate for any age group and may be very appealing to a perimenopausal woman with infrequent sexual exposure.
Male condoms. Condoms are used frequently by women over age 40, and have the advantage of protecting against STDs. Efficacy in preventing pregnancy is 85%, and would probably be higher if studied in this specific age group since the anticipated compliance is higher than in younger age groups.11 It is important to note that if a woman assumes she probably can’t get pregnant anyway, consistent use may diminish.
Patients should be told to use watersoluble lubricants with condoms, and should be cautioned against the use of oil-based vaginal creams, which damage latex condoms. Patientsmay have concerns about condom use since some older men may experience loss of erection while using condoms. The eZ-on condom is a loose-fitting, lubricated plastic condom designed to overcome the interference with penile sensation during vaginal penetration.
Female condom. The female condom is a lubricated polyurethane sac with an outer rim at the introitus and a loose-fitting inner ring to facilitate insertion and to anchor it within the vagina.While used infrequently by women over age 40, it helps to protect against STDs, and pregnancy is reported to occur in 5 of 100 women when used correctly (http://www.plannedparenthood. org/health-topics/birth-control/ female-condom-4223.htm#effective).
Diaphragm. Many women view the diaphragm as safe, simple and convenient. It can be carried in a pocket or purse, is generally not felt by the user or her partner, and, of course, has no effect on a woman’s natural hormones. The diaphragm is immediately effective without any interruption of foreplay; it can be inserted hours ahead of time. If women always use a diaphragm as directed, the reported failure rate is approximately 6 in 100 women (http://www.plannedparenthood. org/health-topics/birth-control/ diaphragm-4224.htm#effective). Because age reduces fertility and older women are more likely to use the device as directed, the failure rate is estimated to be lower in this age group, although precise figures are not available.
Most spermicides that are used with the diaphragm contain nonoxynol- 9. Nonoxynol-9 is associated with certain risks; if used many times a day, or by people at risk for HIV, it may irritate tissue and increase the risk of HIV and other STDs.
Sponge.TheToday Sponge is available over the counter. Likemale and female condoms, the sponge can reduce the risk of STDs. Of 100 women who use the Today Sponge and who have given birth, 32 will become pregnant within the first year of typical use and 20 of 100 will become pregnant with “perfect use” (http://www.plannedparenthood. org/health-topics/birth-control/ birth-control-sponge-today-sponge- 4224.htm#effective). Sub-grouped data for perimenopausal women are not available.
Barrier methods are completely patient-controlled and do not require any additional care from a clinician (other than an initial diaphragm fitting). Side effects are rare (latex allergy, bladder irritation and decreased sensitivity in the male partner) and there are no reported health risks. Other than the main advantage of STD protection afforded by condoms, there are no non-contraceptive benefits associated with barrier methods; they do not contribute to menstrual cycle control.
The typicalmenstrual pattern of the perimenopausal woman is unpredictable. Cycles are usually shorter by 2 to 7 days, although longer cycles are almost as common. Irregular bleeding becomes the norm. The quality of bleeding also changes; it is usually heavier at first followed by lighter bleeding as menopause approaches. Short luteal phases are common. Women should be made aware that these menstrual patterns make ovulation prediction (or the rhythm method) much more difficult for women during the perimenopause transition, compared with earlier in life.
Many options are now available for the perimenopausal woman in need of contraception. This aspect of her care should not be overlooked. Individualization with a thorough risk/benefit discussion appears to be the key to continued compliance. Discussion of contraceptive choices provides a window of opportunity for prevention counseling and monitoring for contraceptive and noncontraceptive benefits. ■
Robert A. Wild, MD, PhD, MRobert A. Wild, MD, PhD, MPH, is Professor of Reproductive Endocrinology, Gynecology, Medicine, Clinical Epidemiology, Oklahoma University Health Sciences Center, Oklahoma City, OK. Dr. Wild reports no conflicts related to the content of this article.
This article includes discussion of offlabel use of medications/medical devices.
Received: September 7, 2007; Accepted: November 18, 2007.
References1. Centers for Disease Control and Prevention. Use of contraception and use of family planning services in the United States: 1982-2002: a fact sheet for advance data No. 350. Atlanta: Centers for Disease Control and Prevention, National Center for Health Statistics 2002:1-4.
2. Williams JK. Contraceptive needs of the perimenopausal woman. Obstet Gynecol Clin North Am 2002;29:575-88, ix.
3. Peterson HB, Jeng G, Folger SG, et al. U.S. Collaborative Review of Sterilization Working Group. The risk of menstrual abnormalities after tubal sterilization. N Engl J Med 2000;343:1681-87.
4. Hankinson SE, Hunter DJ, Colditz GA, et al. Tubal ligation, hysterectomy, and risk of ovarian cancer. A prospective study. JAMA 1993;15;270:2813-18.
5. Schwingl PJ, Guess HA. Safety and effectiveness of vasectomy. Fertil Steril 2000;73:923-36.
6. Hendrix NW, Chauhan SP, Morrison JC. Sterilization and its consequences. Obstet Gynecol Surv 1999; 54:766-77.
7. Sivin I, Stern J, Diaz J, et al. Two years of intrauterine contraception with levonorgestrel and with copper: a randomized comparison of the TCu 380Ag and levonorgestrel 20 mcg/day devices. Contraception 1987;35:245-55.
8. Grimes DA, Hubacher D, Lopez LM, Schulz KF. Nonsteroidal anti-inflammatory drugs for heavy bleeding or pain associated with intrauterine-device use. Cochrane Database Syst Rev 2006;(4):CD006034.
9. Fedele L, Bianchi S, Zanconato G, et al. Use of a levonorgestrel- releasing intrauterine device in the treatment of rectovaginal endometriosis. Fertil Steril 2001;75:485-8.
10. Kailas NA, Sifakis S, Koumantakis E. Contraception during perimenopause. Eur J Contracept Reprod Health Care 2005;10:19-25.
11. Kaunitz AM, Stern L, Doyle J, Etschmaie MR. Use of the levonorgestrel-IUD in the treatment of menorrhagia: improving patient outcomes while reducing the need for surgical management. Manag Care Interface 2007;20:47-50.
12. Bhathena RK, Guillebaud J. Contraception for the older woman: an update. Climacteric 2006;9:264-76. 13. Mosher WD, Martinez GM, Chandra A, et al. Use of contraception and use of family planning services in the United States: 1982-2002. Adv Data 2004;350:1-36.
14. North American Menopause Society. Menopause practice: a clinician’s guide. 3rd edition. North American Menopause Society, Cleveland, OH, 2007:201-28.
15. Tosetto A, Frezzato M, Rodeghiero F. Prevalence and risk factors of non-fatal venous thromboembolism in the active population of the VITA Project 14. J Thromb Haemost 2003;1:1724-29.
16. Brass LM. Estrogens and stroke: use of oral contraceptives and postmenopausal use of estrogen: current recommendations. Curr Treat Options Neurol 2004; 6:459-67.
17. Curtis KM, Chrisman CE, Peterson HB. Contraception for women in selected circumstances. Obstet Gynecol 2002;99:1100-12.
18. Burkman RT. Transdermal hormonal contraception: benefits and risks. Am J Obstet Gynecol 2007;197: 134.e1-6.