Oral contraceptive practice guidelines

ORIGINAL RESEARCH ARTICLE

Oral Contraceptive Practice Guidelines Ronald T. Burkman and Lee P. Shulman

Introduction

O

ral contraceptives (OC) are currently being used by 65 million women worldwide, approximately 6% of all women of reproductive age.1 One-fourth of all women of reproductive age in the United States use OCs, which are now more widely used in the United States than any other method, including sterilization.2 Their widespread use throughout the world for several decades suggests that both medical professionals and the lay public believe that the benefits of OCs outweigh potential side effects.1 Greater awareness of the metabolic effects of OCs and the role of smoking and hypertension in the development of cardiovascular disease has led to more selective prescribing and a substantial reformulation of the steroidal components.3 A variety of formulations are available, but all combination OCs contain two components: an estrogen, which maintains the endometrium and prevents breakthrough bleeding, and a progestin with sufficient activity to suppress pituitary gonadotropin secretion. Each of the steroids also has additional functions. The progestin component causes changes in the cervical mucus and the endometrium that enhance the antifertility effects of OCs should ovulation occur4 and the estrogen acts synergistically with the progestin to inhibit the pituitary.4 Since OCs were first introduced, both hormonal components have been substantially reduced in amount and reformulated. Prior to 1992, OCs in the United States contained one of two estrogens, either ethinyl estradiol (EE) or mestranol. Since then, all low-dose OCs (ie, preparations containing ,50 mg) have contained EE exclusively. Low-dose formulations are now the standard, accounting for almost all OC prescriptions. Prior to 1992, OCs in the United States contained one of four progestins: norethindrone; norgestrel and its active isomer, levonorgestrel; norethindrone acetate; or ethynodiol diacetate. All of these mediumDepartment of Obstetrics and Gynecology, Baystate Medical Center, Springfield, Massachusetts; and Department of Obstetrics and Gynecology, University of Tennessee at Memphis, Memphis, Tennessee Name and address for correspondence: Ronald T. Burkman, Chairman, Department of Obstetrics and Gynecology, Baystate Medical Center, 765 Chestnut St., Springfield, MA 01109

© 1998 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010

and high-androgenic progestins were less selective than later formulations, and they were associated with a variety of adverse androgenic effects. They had a negative effect on carbohydrate and lipid metabolism3,5 and often produced acne,6 hirsutism,7 and weight gain.8,9 In 1992, low-androgenic progestins, norgestimate and desogestrel, were introduced in the United States. Because the affinity of these progestins for progesterone receptors is higher than for androgen receptors, their activity is more selective. Consequently, androgenic side effects associated with OCs containing these progestins are less common compared with the side effects of OCs containing high doses of the more androgenic progestins.10 –14 All currently available OCs are highly effective for preventing pregnancy, but variations in their hormonal components are clinically important. Variations in progestin formulations, for example, modulate the overall metabolic and clinical effect of a given combination OC.3 Variations in estrogen dose among formulations containing ,50 mg EE may also affect cycle control. However, all low-estrogen dose formulations are safe for use by nonsmoking women without major cardiovascular risk factors such as uncontrolled hypertension. To assist clinicians in selecting the most appropriate OC for their patients, a panel of experts in gynecology and contraception has developed a set of practice guidelines for OC selection. These physicians, who are members of the editorial board of Dialogues in Contraception, reviewed the guidelines with a panel of 70 visiting faculty members.15,16 Together, they reviewed the clinical effects of the variations in OC formulations on safety, cycle control, side effects, and noncontraceptive benefits—the factors which, collectively, provide the basis for selecting one OC over another for a particular patient. (It must be recognized that these guidelines were developed within the context of countries with advanced medical systems and low maternal mortalities. In less developed countries, with high rates of maternal morality and limited health care resources, the use of any form of oral contraceptive with limited screening and follow-up may be quite appropriate.) ISSN 0010-7824/98/$19.00 PII S0010-7824(98)00081-X

36S

Burkman and Shulman

Table 1. US Preventive Services Task Force Literature Classification I

Evidence obtained from at least one properly designed, randomized controlled trial II-1 Evidence obtained from well-designed controlled trials without randomization II-2 Evidence obtained from well-designed cohort or case-control analytic studies, preferably from more than one center or research group II-3 Evidence obtained from multiple time series with or without the intervention III Opinions of respected authorities, based on clinical experience, descriptive studies, or reports of expert committees

Practice Guidelines Clinical practice guidelines, which are being employed increasingly in both fee-for-service and managed care settings, are used to improve the effectiveness and efficiency of diagnosis and treatment of a variety of medical conditions. By clarifying the potential benefits, risks, and costs of available interventions, such guidelines help to improve care, enhance patient satisfaction, and facilitate medical education.17 With regard to contraception, guidelines can assist providers in evaluating each woman’s needs and expectations, educating patients, monitoring use, enhancing the successful use of a contraceptive method, and changing methods when indicated. Current methods for developing practice guidelines include informal consensus development and a number of progressively more analytic approaches. The most explicit of these strategies involves mathematical modeling to estimate the probability of specific benefits, harms, and costs of potential interventions.17 Most efforts at guideline development are, however, substantially less formal than this. With regard to OC selection, there are at present no guidelines based on formal analytic methods, and the dearth of level I evidence presents a challenge to their development. Nevertheless, clinicians seek advice on how to differentiate between the many and varied OC formulations currently on the market. The panel has attempted to meet this need by developing evidencebased guidelines that reasonably interpret the available literature and clinical experience while seeking to avoid unintentional bias or misinterpretation. Every effort was made to document the sources appropriately, distinguishing between empiric evidence and expert opinion (Table 1). Estrogen Dose Recommendations The effect of OC formulations containing #35 mg EE on cardiovascular risk was reviewed in the earlier article in this supplement.18

Contraception 1998;58:35S– 43S

However, a growing body of evidence suggests that particular formulations may be preferable for identifiable patient subgroups for reasons other than cardiovascular safety. For example, 30- to 35-mg EE formulations may be preferred for nonsmoking perimenopausal women, for reasons that are discussed later in this article. On the other hand, there are circumstances in which the use of 20-mg EE formulations may be appropriate. Women with persistent estrogenic sensitivity may experience nausea or other side effects that can be a source of dissatisfaction and discontinuance. These women may find their symptoms are reduced if they switch to a formulation with a lower estrogen dose. Although they are no longer widely used, even the 50-mg EE formulations have a useful role in some circumstances. As discussed in a later section, this formulation may be appropriate for women with medical problems that require the use of hepaticenzyme–inducing drugs.19 Overview of Clinical Recommendations Regarding Progestin Component Most OC prescribed in the United States today contain progestins with low or medium androgenic activity. These formulations are less likely to produce adverse androgenic side effects than are OC formulated with high androgenic activity. Although several studies have suggested that desogestrel-containing formulations may increase the risk of venous thromboembolism (VTE),20 –24 most experts think that this effect is unlikely to be the result of a causal relation. As summarized in the earlier article,18 it is widely believed that these results are biologically implausible1,20,23,24 and that the findings are due to prescribing bias and other confounding factors.1,23,26,27 The evidence does not support any changes in current prescribing patterns at this time, and no change is advocated by the US Food and Drug Administration (FDA).28

General Guidelines for Nonsmoking Women New Starts Based on a comparative analysis of safety, cycle control, and noncontraceptive benefits, low-androgenic OCs containing #35 mg EE are the panel’s recommended first choice for most new starts, including adolescents, postpartum mothers, and nonsmoking perimenopausal women. Continuing Use, Restarting, or Switching OC Formulations If a woman is doing very well on a low-dose combination OC, change is rarely indicated. However, there

Contraception 1998;58:35S– 43S

are some situations in which it may be appropriate to switch to a different formulation. For example, if an older woman has been using a 50-mg EE formulation for many years, every effort should be made to switch her to a less thrombogenic formulation containing #35 mg EE. Similarly, if a woman is experiencing an exacerbation of an existing condition, such as acne, and her current OC contains a progestin with relatively high androgenic activity, she would probably benefit from a switch to a formulation containing a low-androgenic progestin. If a woman is resuming OC use after an interval without use, she should be given the same formulation she used previously, even if it is not the one that might theoretically be preferred, provided it contains ,50 mg EE and she was satisfied with it. When consideration is given to switching to an alternative formulation in the hope of reducing or eliminating a troublesome side effect, the change generally should be made only after giving the current OC a 3- to 6-month trial, because side effects may resolve in that time. Equally important, when a woman is switched to an alternative OC, it should be, initially, to a formulation containing a different progestin rather than to one containing a different dose of the same progestin. Certain side effects are clearly related to the androgenicity of a particular progestin, and it is reasonable to anticipate improvements after changing to a less androgenic formulation. However, it is not presently understood why women with side effects apparently unrelated to androgenicity (for example, headache) do better on one type of progestin than with another. Adolescents It may be helpful to consider in detail the noncontraceptive benefits of OC use to adolescent women. Like older women, most are more familiar with the potential risks of OC use than they are with these documented benefits. Adolescents are often less interested than older women in preventing gynecologic cancer and bone loss and are much more interested in cycle control, acne management, and reduction of dysmenorrhea and menorrhagia. When these issues are important, a 30- to 35-mg EE formulation may be especially beneficial. Postpartum Women Nonlactating women may safely begin using combination OCs about 2 to 3 weeks after delivery.29,30 This approach provides women with effective contraception prior to the earliest ovulation at 4 weeks postpartum. Lactating women who regularly breastfeed at ap-

OC Practice Guidelines

37S

proximately 4-h intervals without supplemental feeding will not become pregnant for at least 3 months. The World Health Organization recommends that combination OCs should be avoided during this period, since estrogen decreases the quantity and duration of lactation.31 If an exclusively breastfeeding woman wishes to use oral contraception during the postpartum period, progestin-only pills can be prescribed. Use of these agents during lactation causes no adverse effects on breastfeeding performance or on the health, growth, or development of the infant.31 When supplemental feeding is introduced, the woman should be switched to combination OCs, as these are more effective. Formulations containing #35 mg EE and progestins with low or medium androgenic activity are the first choice for most postpartum women.

Perimenopausal Women Healthy, nonsmoking women can use low-dose OCs throughout the perimenopausal years. Ovarian function declines over several years, and women in transition may ovulate sporadically. Therefore, perimenopausal women are potentially fertile and may be at risk for unintended pregnancy. OCs provide safe and effective contraception for women in this population and cause regular episodes of uterine bleeding. Two studies have now demonstrated that 30- to 35-mg EE formulations are more effective in preventing breakthrough bleeding and spotting than are 20-mg EE OCs with the same dose of progestin.32,33 Therefore, during the period of ovarian decline, the use of OCs containing 30 or 35 mg EE may be superior to lower-dose formulations with respect to maintaining or restoring cycle control, which is important for many perimenopausal women. Evidence from several studies suggests that the use of OCs in premenopausal women may stabilize or even increase bone mass.34,35 Eight of 12 published studies have shown that women using OC have greater bone mass than nonusers.35 The protective effects of OCs may be dependent on estrogen dose and duration of use. The greatest increase in bone mineral density (BMD) is seen in women who have used OCs for at least 10 years.36 Finally, 30- to 35-mg EE formulations have been shown to provide protection against endometrial and ovarian cancers.37,38 The degree of reduction of risk of developing these gynecologic cancers is directly related to the duration of OC use.37– 40 The effect of lowering the estrogen dose to 20 mg on protection against these malignancies is unknown. Data based on the use of doses of $35 mg suggest that protection

38S

Burkman and Shulman

against endometrial cancer may be independent of estrogen dose.37 Cancer prevention, protection against bone loss, and improved cycle control are all of more interest to perimenopausal women than to younger women, and these important noncontraceptive benefits should be explained to older women who are selecting a contraceptive method.

Guidelines for Women Who Smoke Women Aged 35 Years and Younger Because an increased risk of myocardial infarction (MI) and stroke among women who smoke and use OCs is largely confined to those who are over age 35,41,42 there is no reason that younger smokers should not use low-dose OCs. Current epidemiologic data provide no indication that the various formulations containing 20 to 35 mg EE carry different cardiovascular risks in smokers versus nonsmokers. Therefore, this panel recommends that smokers under age 35 should be permitted to use any of the currently available low-dose formulations. Cigarette smoking in OC users increases the risk for breakthrough bleeding43 and bone loss.44 – 46 No data are available comparing the relative benefits of the different EE doses with respect to breakthrough bleeding among women who smoke, but among all women, including smokers, the incidence of breakthrough bleeding is higher with 20-mg formulations than with 30- to 35-mg formulations.32,33,47

Women Over Age 35 This panel recommends limiting the contraceptive options of smokers over the age of 35 to nonestrogen or nonhormonal methods. This approach is consistent with the FDA-approved warnings for combination OCs that any formulation containing estrogen should not be used in women over the age of 35 who smoke.

Guidelines for Women With Medical Problems Women with underlying medical problems present a number of contraceptive and obstetric challenges. Both perinatal and maternal risks in women with medical problems are higher than among healthy women. Highly effective contraceptive methods are required to avoid putting women with medical disorders at risk for unintended pregnancy. Determining which reversible method of contraception to use requires careful consideration.

Contraception 1998;58:35S– 43S

Hepatic Enzyme-Inducing Drugs Decrease OC Efficacy The use of pharmaceutical agents that increase the hepatic metabolism of sex steroids by hepatic enzyme induction, such as antiepileptic drugs, may reduce compliance with low-dose OCs. Hepatic enzyme induction can decrease ethinyl estradiol levels, resulting in irregular bleeding and lack of continuation of the methods.19 For women who use such agents, it may be advisable to prescribe OCs containing 50 mg EE to maintain the endometrium.19,49

Certain Antiepileptic Drugs Are Potent Hepatic Enzyme Inducers Several antiepileptic drugs (AED) are potent hepatic enzyme inducers, and women who use AED may be at risk for breakthrough bleeding if they are using formulations containing ,50 mg EE. Epilepsy affects approximately 1% of the population, manifesting itself early in life and persisting through adulthood.19 Failure of contraception among women using AEDs is a serious issue. The AEDs are teratogenic, and pregnancy has potential complications for both mother and infant, making it very important to maintain good contraceptive control in this population.19 Because of this potential for teratogenicity, the need for continued use of AEDs should be re-examined critically in women of childbearing age, regardless of their intention to use contraceptive drugs.48 Of the AEDs, those most likely to cause hepatic enzyme induction are phenobarbital, phenytoin, primidone, and carbamazepine.48 Other AEDs, such as valproic acid, do not cause hepatic enzyme induction. Other contraceptive options may be preferable to high-estrogen dose OCs for women with epilepsy. Both the intrauterine device (IUD) and depot medroxyprogesterone acetate (DMPA) are reasonable alternatives. The latter has the added benefit of reducing seizure frequency.49 If a woman who has been using OCs containing 50 mg EE or more discontinues hepatic enzyme-inducing AEDs, she should change to a formulation containing #35 mg EE.48

Other Hepatic Enzyme-inducing Drugs A small number of other hepatic enzyme-inducing medications may also place women at risk for contraceptive failure. Some antituberculosis agents (rifampin)50,51 and antifungals (griseofulvin)52 may affect liver metabolism and reduce the effectiveness of OCs. It has been suggested that women taking rifampin for prolonged periods should not use OCs.52

Contraception 1998;58:35S– 43S

OC Efficacy Unaffected by Most Antibiotics, Minor Tranquilizers, and Tricyclic Antidepressants Despite anecdotal evidence, there is no established proof that antibiotics other than rifampin decrease the efficacy of OCs.50 There is little evidence that ampicillin, metronidazole, quinolone, and tetracycline affect OC efficacy.50 Valproate and the benzodiazepines have not been associated with contraceptive pill failure, which is consistent with the fact that neither drug is an enzyme inducer.19,48 The action of diazepam, chlordiazepoxide, tricyclic antidepressants, and theophylline may be potentiated by OCs.53–55 Type I and Type II Diabetes Mellitus/ Gestational Diabetes Pregnancy carries appreciable risks for women with diabetes. Diabetic women should, therefore, become pregnant only by plan, and they need qualified advice on which contraceptive method is best for them. It is important for these women to use a safe and reliable contraceptive method, and, for many diabetic women, OCs are a very good option. There is no evidence that OCs use worsens the course of type I or type II diabetes mellitus.56,57 OC may be used by young diabetic women as long as they are free of retinopathy, nephropathy, hypertension, or other vascular complications. Either the IUD or a progestinonly method is an appropriate contraceptive choice for diabetic women with vascular complications. Among women with a history of gestational diabetes, the incidence of frank diabetes developing after pregnancy is no higher in women using low-dose OCs than in women using nonhormonal contraceptive methods.58,59 In more than half of women with a prior history of gestational diabetes, diabetes mellitus will eventually develop within 10 to 20 years after delivery, regardless of the contraceptive method used.59 Metabolic status in these women must be supervised closely regardless of OCs use. Systemic Lupus Erythematosus (SLE) Effective contraception is also very important for women with systemic lupus erythematosus (SLE) because pregnancy is associated with disease exacerbations. However, SLE constitutes a relative contraindication to use of estrogen-containing OCs because OCs may exacerbate SLE in some patients.60 Moreover, women with SLE—particularly those with anticardiolipin antibodies—are at increased risk for vascular disease, and the risk of deep venous thrombosis may be further increased by the use of estrogencontaining OCs.61,62 For these reasons, vascular disease associated with lupus represents an absolute contraindication to estrogen-containing OCs.

OC Practice Guidelines

39S

Recommended methods of reversible contraception for women of reproductive age who have SLE include the IUD, a progestin-only pill, DMPA, or implants. Cardiovascular Disease Mitral valve prolapse affects approximately 15 million Americans and is more common among women than men.63 Nonsmoking women with asymptomatic mitral valve prolapse (that is, with no mitral regurgitation) may safely use low-dose OCs.63 OCs are contraindicated only if there is marginal cardiac reserve or a condition that predisposes to thrombosis (for example, atrial fibrillation or clotting factor abnormalities).63 When OCs are contraindicated, IUDs may be a good alternative. The Copper T 380A can be used in women with mitral valve prolapse, provided appropriate antibiotics are used at the time of insertion64; however, the manufacturer’s prescribing recommendations proscribe use of the hormone-releasing IUDs for women with valvular heart disease.65 Among women with chronic hypertension, OCs should be used only by nonsmokers whose blood pressure is well controlled (#140/90 mm Hg), and who are otherwise healthy.63 Frequent monitoring of blood pressure is extremely important, as even OC with low estrogen dosages can cause an increase in blood pressure.63 OCs are contraindicated for women with uncontrolled hypertension, but the IUD or progestin-only OCs methods can be used.63 OCs do not alter the risk of developing hypertension in women with a history of pregnancy-induced hypertension who are now normotensive.63 Any OCs containing ,50 mg EE may be used by normotensive women with a history of pregnancy-induced hypertension.63 As long as blood pressure has returned to normal within 4 – 6 weeks of delivery, OC use may be initiated at that time. Combination OCs are contraindicated for women with a personal history of idiopathic VTE and for women with congestive heart failure, coronary artery disease, angina, or MI. Headache and Migraine OCs may safely be used by women who suffer from nonvascular (tension) headaches, the most common type of headaches. They are caused primarily by muscle contractions and few, if any, interactions with exogenous sex steroids have been reported.19 Tension headaches are not related to OC use19 and can be treated with over-the-counter analgesics. A history of migraine headache is of concern when considering OC use because of the vascular changes associated with these headaches.19

40S

Burkman and Shulman

Migraine headache is more often unilateral than tension headache and is more commonly associated with nausea and vomiting.19 It is estimated that 8.7 million women in the US have mild to moderate migraine headaches,19 about 20% of all women of reproductive age. Analysis of the limited and conflicting literature suggests that the severity and frequency of migraine headache increase in 15%–50% of established migraineurs taking OC.19 However, as many as one-third of migraineurs report improvement with OC use.19 Classic migraine is characterized by well-defined symptoms that precede or accompany the headache. The symptoms are usually visual, consisting most often of scotomata, but unilateral paresthesia, aphasia, and other symptoms of neurologic dysfunction may also occur.19 Individuals with classic migraine are probably best advised to use some form of contraception other than OCs, because of the thrombophilic effect of OCs.19 Low-estrogen dose OCs can be used by women with common migraine (without focal symptoms) as no increased risk of stroke in this group of women who have used OC has been reported.19 If headache severity or frequency increases with combination OC use in women with a history of migraine headaches, an alternative method should be considered. Women who have focal neurologic symptoms after taking OCs should be advised to discontinue use immediately.19 Although some women report that OC use causes headaches or exacerbates them, headaches are also common among women not taking OCs. If a woman using a combination OC with a high- or mediumandrogenic progestin develops headaches during the active pill-taking portion of the cycle, switching her to an OC with a low-androgenic formulation may be beneficial. If there is no improvement, or if headache worsens, another method of birth control should be considered. If headaches consistently occur during the pill-free interval, the woman may have an estrogen-withdrawal headache. In this situation, use of a daily, continuous, combined OC to avoid cyclicity can be helpful.

History of Benign Breast Disease and Family History of Breast Cancer Anxiety regarding breast cancer continues to be a major concern among women considering OC use. Thus, it is important to provide women with accurate information on this topic. There is now persuasive evidence (level II-2) that neither a family history of breast cancer nor a personal history of benign breast disease increases the risk of breast cancer in OC users

Contraception 1998;58:35S– 43S

compared with nonusers.66 – 68 Therefore, women with these background factors can be reassured that OC use is not contraindicated. More than a decade ago, investigators analyzed case-control data from the Cancer and Steroid Hormone (CASH) study of the Centers for Disease Control and Prevention to establish the relative risk of breast cancer in various subgroups of women and in women who used specific OC formulations.67 Data from 4711 women with newly diagnosed breast cancer and 4676 community controls were analyzed. The findings of the CASH study investigators were confirmed by later reports attesting to the overall safety of OC, and the study also demonstrated that neither family history of breast cancer nor a history of surgery for benign breast disease increased risk of developing breast cancer among OC users compared with nonusers of the same age.67 Overall risk of breast cancer was not increased by OC use and was unrelated to OC type and formulation, dose and type of estrogen and progestin, and duration of use.67 Subsequently, the Collaborative Group on Hormonal Factors in Breast Cancer collected and reanalyzed worldwide epidemiologic evidence on the relation between breast cancer risk and use of hormonal contraceptives.68 This study incorporated data of 54 studies on 53,297 women with breast cancer and 100,239 women without. Data were obtained on several potential risk factors, including sociodemographic factors, use of hormonal contraceptives and hormone replacement therapy, family history of breast cancer, height, weight, age at menarche, reproductive history, menopausal status, age at menopause, gynecologic surgery, and alcohol consumption. The data confirmed that risk of breast cancer was not increased in OC users with a family history of breast cancer. The data indicated there was an increased incidence of diagnosis of breast cancer, but it occurred only in current and recent users and may have been due to detection bias and improved surveillance of OC users. A personal history (instead of a family history) of breast cancer is currently regarded as a contraindication to use of all hormonal contraceptive methods, including OCs. For women with a personal history of breast cancer, the Copper T 380A IUD is an effective reversible contraceptive method that can be used.

Conclusions and Recommendations Risks and benefits associated with OC use for healthy, nonsmoking women should be balanced against those of other contraceptives and against those associated with unintended pregnancy and abortion. An increased absolute risk of serious disease

Contraception 1998;58:35S– 43S

associated with the use of all formulations containing ,50 mg EE is extremely small and limited to VTE. Furthermore, the incidence of persistent side effects is low, and noncontraceptive benefits are substantial. Formulations containing #35 mg EE should remain the first choice for most new starts, including adolescents and nonsmoking perimenopausal women. There are special subpopulations of women who are candidates for a particular estrogen formulation (20, 30 to 35, or 50 mg EE). Use of other methods—the IUD, progestin-only pills, DMPA, and implants—may be appropriate when contraceptive doses of estrogen are contraindicated (as, for example, in women over age 35 who smoke or have uncontrolled hypertension).

Observation on Evidence-Based Practice Guidelines Evidence-based results are now being used to establish practice guidelines. These link recommendations directly to the supporting evidence, in accordance with the criteria shown in Table 1. Evidence-based approaches primarily use only level I or level II evidence.17,69 The evidence-based strategy has increased the scientific rigor of practice guidelines development, but it is not a panacea. Only a small proportion of current interventions has been validated through clinical trials. In the absence of acceptable evidence, some clinicians say this approach cannot yield useful recommendations, and neutral recommendations—neither for nor against the procedure— can result.17 In an effort to develop useful recommendations for practitioners, many guideline development panels, including this one, have adopted a methodology that combines expert opinion in assessing clinical benefits and harm with a detailed review of relevant literature that relies mainly on studies that meet level I or level II standards. Guidelines are documented so that it is clear which recommendations are based on scientific evidence and which are based on expert opinion. The process involves a formal assessment of scientific evidence, panel meetings, and open forum sessions that seek broad input on relevant analytic issues.17

Levels of Evidence in OC Research Little of the evidence with respect to contraceptive formulations achieves level I status (ie, evidence obtained from at least one properly designed, randomized controlled trial). The most persuasive evidence available on cardiovascular safety and cancer risk is based on observational studies and achieves level II-2 status (i.e., well-designed cohort or case-control ana-

OC Practice Guidelines

41S

lytic studies, preferably from more than one center or research group). Level II-2 evidence is available on risk of MI,41,70 ischemic and hemorrhagic stroke,42,71,72 and VTE.20 –25,73–75 Level II-2 evidence is also available with respect to the relationships of OC use to risk of breast cancer,67,68 ovarian cancer,38 and endometrial cancer.37 Lacking a more conclusive empiric base, recommendations regarding OC use for subgroups of women, particularly women with medical conditions, are often based on level III evidence (ie, opinions of respective authorities, based on clinical experience, descriptive studies, or reports of expert committees). Many of the recommendations of this panel are of this type.

References 1. Lidegaard O, Milsom I. Oral contraceptives and thrombotic diseases: impact of new epidemiological studies (editorial). Contraception 1996;53:135–9. 2. Ortho Pharmaceutical Corp. 1997 Annual Birth Control Study. Raritan, NJ: Ortho Pharmaceutical Corp, 1996. 3. Godsland IF, Crook D. Update on the metabolic effects of steroidal contraceptives and their relationship to cardiovascular disease risk. Am J Obstet Gynecol 1994; 170:1528 –36. 4. Guillebaud J. The Pill and Other Hormones for Contraception. 4th ed. Oxford: Oxford University Press, 1991: 46. 5. Godsland IF, Walton C, Felton C, Proudler A, Patel A, Wynn V. Insulin resistance, secretion, and metabolism in users of oral contraceptives. J Clin Endocrinal Metab 1992;74:64 –70. 6. Palatsi R, Hirvensalo E, Liukko P, et al. Serum total and unbound testosterone and sex hormone binding globulin (SHBG) in female acne patients treated with two different oral contraceptives. Acta Derm Venereol 1984; 64:517–23. 7. Bilotta P, Favilli S. Clinical evaluation of a monophasic ethinylestradiol/desogestrel-containing oral contraceptive. Arzneim Forsch/Drug Res 1988;38:932– 4. 8. Rekers H. Multicenter trial of a monophasic oral contraceptive containing ethinyl estradiol and desogestrel. Acta Obstet Gynecol Scand 1988;67:171– 4. 9. Runnebaum B, Grunwald K, Rabe T. The efficacy and tolerability of norgestimate/ethinyl estradiol (25 mg of norgestimate/35 mg of ethinyl estradiol): results of an open, multicenter study of 59,701 women. Am J Obstet Gynecol 1992;166:1963– 8. 10. Phillips A, Demarest K, Hahn DW, Wong F, McGuire JL. Progestational and androgenic receptor binding affinities and in vivo activities of norgestimate and other progestins. Contraception 1990;41:399 – 410. 11. Van der Vange N, Blankenstein MA, Kloosterboer HJ, Haspels AA, Thijssen JHH. Effects of seven low-dose combined oral contraceptives on sex hormone binding globulin, corticosteroid binding globulin, total and free testosterone. Contraception 1990;41:345–52. 12. Chapdelaine A, Desmorais J-L, Deeman RJ. Clinical evidence of the minimal androgenic activity of norgestimate. Int J Fertil 1989;34:347–52. 13. Jung-Hoffman C, Heidt F, Kuhl H. Effect of two oral

42S

14.

15. 16. 17. 18. 19. 20.

21.

22.

23.

24.

25.

26. 27. 28. 29.

30. 31.

Burkman and Shulman

contraceptives containing 30 mg ethinylestradiol and 75 mg gestodene or 150 mg desogestrel upon various hormonal parameters. Contraception 1988;38:593– 603. Mall-Haefeli M, Werner-Zodrow I, Huber PR. Biochemical and clinical results of the new micropill Cilest. In: Keller PI, ed. Aktuelle Aspekte der Hormonalen Kontrazeption (in German with English abstract). Basel: S. Karger AG, 1991:55– 66. Practice guidelines for OC selection. Dialogues in Contraception 1996;4(7). Practice guidelines for OC selection: update. Dialogues in Contraception 1997;5(4). Woolf SH. Practice guidelines, a new reality in medicine. II: Methods of developing guidelines. Arch Intern Med 1992;152:946 –52. Ory HW. Cardiovascular safety of oral contraceptives. Contraception 1998;58:9S–13S. Mattson RH, Rebar RW. Contraceptive methods for women with neurologic disorders. Am J Obstet Gynecol 1993;168:2027–32. World Health Organization (WHO) Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Effect of different progestagens in low oestrogen oral contraceptives on venous thromboembolic disease. Lancet 1995;346:1582– 8. Jick H, Jick SS, Gurewich V, Myers MW, Vasilakis C. Risk of idiopathic cardiovascular death and nonfatal venous thromboembolism in women using oral contraceptives with differing progestagen components. Lancet 1995;346:1589 –93. Spitzer WO, Lewis MA, Heinemann LAJ, et al. Third generation oral contraceptives and risk of venous thromboembolic disorders: an international case-control study. Br Med J 1996;312:83– 8. Lewis MA, Heinemann LAJ, MacRae KD, et al. The increased risk of venous thromboembolism and the use of third generation progestagens: role of bias in observational research. Contraception 1996;54:5–13. Bloemenkamp KWM, Rosendaal FR, Helmerhorst FM, Bu¨ller HR, Vandenbroucke JP. Enhancement by factor V Leiden mutation of risk of deep-vein thrombosis associated with oral contraceptives containing a thirdgeneration progestagen. Lancet 1995;346:1593– 6. Farmer RDT, Lawrenson RA, Thompson CR, Kennedy JG, Hambleton IR. Population-based study of risk of venous thromboembolism associated with various oral contraceptives. Lancet 1997;349:83– 8. Jamin C, de Mouzon J. Selective prescribing of third generation oral contraceptives (OC). Contraception 1996;54:55– 6. Farmer RDT, Lawrenson R. Utilization patterns of oral contraceptives in UK general practice. Contraception 1996;53:211–5. Oral contraceptives and risk of blood clots. FDA Talk Paper. Rockville, MD: US Food and Drug Administration; November 14, 1995; T95–T61. Gray RH, Campbell OM, Zacur HA, Labbok MH, MacRae SL. Postpartum return of ovarian activity in nonbreastfeeding women monitored by urinary assays. J Clin Endocrinol Metab 1987;64:645–50. Campbell OMR, Gray RH. Characteristics and determinants of postpartum ovarian function in women in the United States. Am J Obstet Gynecol 1993;169:55– 60. World Health Organization (WHO) Task Force on Oral Contraceptives, Special Programme of Research, Devel-

Contraception 1998;58:35S– 43S

32.

33.

34. 35. 36. 37.

38.

39. 40. 41.

42.

43. 44.

45. 46. 47.

48.

opment, and Research Training in Human Reproduction. Effects of hormonal contraceptives on breast milk composition and infant growth. Stud Fam Plann 1988; 19:361–9. Åkerlund M, Røde A, Westergaard J. Comparative profiles of reliability, cycle control and side effects of two oral contraceptive formulations containing 150 mg desogestrel and either 30 mg or 20 mg ethinyl oestradiol. Br J Obstet Gynaecol 1993;100:832– 8. Endikrat J, Mu¨ller U, Du¨sterberg B. A twelve-month comparative clinical investigation of two low-dose oral contraceptives containing 20 mg ethinylestradiol/75 mg gestodene and 30 mg ethinylestradiol/75 mg gestodene, with respect to efficacy, cycle control, and tolerance. Contraception 1997;55:131–7. Corson SL. Oral contraceptives for the prevention of osteoporosis. J Reprod Med 1993;38:1015–20. DeCherney A. Bone-sparing properties of oral contraceptives. Am J Obstet Gynecol 1996;174:15–20. Kleerekoper M, Brienza RS, Schultz LR, et al. Oral contraceptive use may protect against low bone mass. Arch Intern Med 1991;151:1971– 6. Cancer and Steroid Hormone (CASH) Study of the Centers for Disease Control and the National Institute of Child Health and Human Development. Combination oral contraceptive use and the risk of endometrial cancer. JAMA 1987;257:796 – 800. Cancer and Steroid Hormone (CASH) Study of the Centers for Disease Control and the National Institute of Child Health and Human Development. The reduction in risk of ovarian cancer associated with oral contraceptive use. N Engl J Med 1987;316:650 –5. Schlesselman JJ. Cancer of the breast and reproductive tract in relation to use of oral contraceptives. Contraception 1989;40:1–38. Schlesselman JJ. Net effect of oral contraceptive use on the risk of cancer in women in the United States. Obstet Gynecol 1995;85:793– 801. Croft P, Hannaford PC. Risk factors for acute myocardial infarction in women: evidence from the Royal College of General Practitioners’ oral contraception study. Br Med J 1989;298:165– 8. World Health Organization (WHO) Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Haemorrhagic stroke, overall stroke risk, and combined oral contraceptives: results of an international, multicentre, case-control study. Lancet 1996; 348:505–10. Rosenberg MJ, Waugh MS, Stevens CM. Smoking and cycle control among oral contraceptive users. Am J Obstet Gynecol 1996;174:628 –32. Hollenbach KA, Barrett-Connor E, Edelstein SL, Holbrook T. Cigarette smoking and bone mineral density in older men and women. Am J Public Health 1993;83: 1265–70. Johnston JD. Smokers have less dense bones and fewer teeth. J R Soc Health 1994;114:265–9. Hopper JL, Seeman E. The bone density of female twins discordant for tobacco use. N Engl J Med 1994;330:387– 92. Appel TB, Arman KA, Birdsall C, et al. A comparison of a new graduated estrogen formulation with three constant-dosed oral contraceptives. Contraception 1987; 35:523–32. Mattson RH, Cramer JA, Darney PD, Naftolin F. Use of

Contraception 1998;58:35S– 43S

49. 50. 51. 52. 53.

54. 55.

56.

57.

58.

59. 60.

61. 62.

oral contraceptives by women with epilepsy. JAMA 1986:256:238 – 40. Mattson RH, Cramer JA, Caldwell BV, Siconalfi BC. Treatment of seizures with medroxyprogesterone acetate: preliminary report. Neurology 1984;34:1255– 8. Helms SE, Bredle DL, Zojic J, Jarjoura D, Brodell RT, Krishnarao I. Oral contraceptive failure rates and oral antibiotics. J Am Acad Dermatol 1997;36:705–10. Stockley I. Interactions with oral contraceptives. J Pharm 1976;216:140 –3. Shenfield GM. Oral contraceptives: are drug interactions of clinical significance? Drug Safety 1993;9:21–37. Abernethy DR, Greenblatt DJ, Divoll M, Arendt R, Ochs HR, Shader RI. Impairment of diazepam metabolism by low-dose estrogen-containing oral contraceptive steroids. N Engl J Med 1982;306:791–2. Baciewicz AM. Oral contraceptive drug interactions. Ther Drug Monit 1985;7:26 –35. Tornatore KM, Kanarkowski R, McCarthy TL, Gardner MJ, Yurchak AM, Jusko WJ. Effect of chronic oral contraceptive steroids on theophylline disposition. Eur J Clin Pharmacol 1982;23:129 –34. Garg SK, Chase HP, Marshall G, Hoops SL, Holmes DL, Jackson WE. Oral contraceptives and renal and retinal complications in young women with insulin-dependent diabetes mellitus. JAMA 1994;271:1099 –102. Peterson KR, Skouby SO, Sidelmann J, Mølsted-Pedersen L, Jespersen J. Effects of contraceptive steroids on cardiovascular risk factors in women with insulindependent diabetes mellitus. Am J Obstet Gynecol 1994;171:400 –5. Kjos SL, Shoupe D, Douyan S, et al. Effect of low-dose oral contraceptives on carbohydrate and lipid metabolism in women with recent gestational diabetes: results of a controlled, randomized, prospective study. Am J Obstet Gynecol 1990;163:1822–7. Kjos SL. Contraception in diabetic women. Obstet Gynecol Clin N Am 1996;23:243–58. Jungers P, Dougados M, Pe´lissier C, et al. Influence of oral contraceptive therapy on the activity of systemic lupus erythematosus. Arthritis Rheum 1982;25:618 – 23. Julkunen HA. Oral contraceptives in systemic lupus erythematosus: side effects and influence on the activity of SLE. Scan J Rheumatol 1991;20:427–33. Julkunen HA, Kaaja R, Friman C. Contraceptive practice in women with systemic lupus erythematosus. Br J Rheumatol 1993;32:227–30.

OC Practice Guidelines

43S

63. Sullivan JM, Lobo RA. Considerations for contraception in women with cardiovascular disorders. Am J Obstet Gynecol 1993;168:2006 –11. 64. ParaGardt Prescribing Information. Raritan, NJ: Ortho Pharmaceutical Corp, January 1996. 65. Progestasertt System Prescribing Information. Palo Alto, CA: Alza Pharmaceutical, April 1996. 66. Malone KE, Daling JR, Weiss NS. Oral contraceptives in relation to breast cancer. Epidemiol Rev 1993;15:80 – 97. 67. Cancer and Steroid Hormone (CASH) Study of the Centers for Disease Control and the National Institute of Child Health and Human Development. Oral contraceptive use and the risk of breast cancer. N Engl J Med 1986;315:405–11. 68. Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and hormonal contraceptives: collaborative reanalysis of individual data on 53,297 women with breast cancer and 100,239 women without breast cancer from 54 epidemiological studies. Lancet 1996;347:1713–27. 69. Eddy DM. Clinical Decision Making: From Theory to Practice: A Collection of Essays from JAMA. Boston: Jones and Bartlett, 1996. 70. Jick H, Jick SS, Myers MW, Vasilakis C. Risk of acute myocardial infarction and low-dose combined oral contraceptives (letter). Lancet 1996;347:627– 8. 71. Petitti DB, Sidney S, Bernstein A, Wolf S, Quesenberry C, Ziel HK. Stroke in users of low-dose oral contraceptives. N Engl J Med 1996;335:8 –15. 72. World Health Organization (WHO) Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Ischaemic stroke and combined oral contraceptives: results of an international, multicentre, casecontrol study. Lancet 1996;348:498 –505. 73. Farmer RDT, Preston TD. The risk of venous thromboembolism associated with low oestrogen oral contraceptives. J Obstet Gynaecol 1995;15:195–200. 74. Gerstman BB, Piper JM, Tomita DK, Ferguson WJ, Stadel BV, Lundin FE. Oral contraceptive estrogen dose and the risk of deep venous thromboembolic disease. Am J Epidemiol 1991;133:32–7. 75. World Health Organization (WHO) Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Venous thromboembolic disease and combined oral contraceptives: results of international multicentre case– control study. Lancet 1995;346:1575– 82.