Cover Story: When a chronically ill patient needs contraceptives

October 1, 2003

While some hormonal contraceptives are clearly contraindicated in specific chronic disorders, others pose little or no threat for the same condition. To provide practical guidelines on the subject, two experts review the "WHO Medical Eligibility Criteria for Contraceptive Use."


Cover Story

When a chronically ill patient needs contraceptives

By Robert Hatcher, MD, MPH, and Carrie Cwiak, MD, MPH

While some hormonal contraceptives are clearly contraindicated in specific chronic disorders, others pose little or no threat for the same condition. To provide practical guidelines on the subject, two experts review the WHO Medical Eligibility Criteria for Contraceptive Use.

Most of the women you see in the office or clinic requesting contraceptives are young and healthy. For these patients, contraceptives carry little risk of complications. There are, however, those women with chronic medical conditions, many of whom still need contraception. In fact, these women may be more at risk for pregnancy-related disorders and death, and so their need for safe and effective contraceptives may be even greater.

To determine which contraceptives are contraindicated in these patients, clinicians have to pay close attention to their individual medical histories. At the same time, you'll also want to bear in mind that there are chronic illnesses that are not affected by hormonal or other methods.

In order to take advantage of the best evidence in this specialized area, the World Health Organization assembled an international group of clinicians and scientists to review the available research and devise comprehensive guidelines for the safe use of contraceptives. The resulting guidelines, which are not intended as a substitute for clinical judgment, were compiled into the WHO Medical Eligibility Criteria for Contraceptive Use, which can be easily accessed via the Internet (see Web site references).1

Where epidemiological data are available, the WHO criteria reflect this. Where there are insufficient or absent data, the criteria may present theoretical risks. Our goal here is to review some of the eligibility criteria, as well as the available literature for several chronic conditions.

The WHO document lists over 120 diseases and conditions and assigns a numeric classification of 1 to 4 for the use of 10 contraceptive methods in that circumstance. The categories are as follows:

1 – Can use the method. No restrictions on use.

2 – Can use the method. Advantages generally outweigh theoretical or proven risks. If the method is chosen, more than usual follow-up may be needed.

3 – Should not use the method unless clinician makes the clinical judgment that the patient can safely use it. Theoretical or proven risks usually outweigh the advantages of the method. Method of last choice, for which regular monitoring may be needed.

4 – Should not use the method. Condition represents an unacceptable health risk if method is used.

This classification provides a quick and easy reference when providing contraception for our patients with chronic illnesses (Figures 1, 2, and 3).


Should your hypertensive patient be taking hormonal contraceptives?

ConditionCombined methodsDepo-ProveraOther progestin methods
Well-controlled hypertension321
Systolic 140-159/Diastolic 90-99321
Systolic >160/Diastolic >100432
Vascular disease432


What kind of migraine is it?

Visual disturbance in both eyes (i.e., homonymous hemianopsia)Migraine w/aura
Flashing or moving scotomaMigraine w/aura
Unilateral numbnessMigraine w/aura
"Pins & needles" in extremitiesMigraine w/aura
Unilateral weaknessMigraine w/aura
Aphasia or other speech difficultiesMigraine w/aura
Nausea + vomitingMigraine
Watery eyesMigraine
Taste or smell sensationsMigraine


When should a patient who suffers from headaches avoid hormonal contraceptives?

ConditionCombined methodsDepo-Provera and MirenaProgestin-only pills
Nonmigraine headaches1/211
Migraine w/o aura, age <352/321/2
Migraine w/o aura, age >353/421/2
Migraine with aura, any age42/32/3


Cardiovascular disease

Risk factors for cardiovascular disease in women include cigarette smoking, obesity, diabetes, hypertension, elevated lipids, and older age. Some of these conditions, like obesity, are not contraindications for estrogen-containing contraceptives on their own. But according to the WHO guidelines, the combination of two or more of these risk factors may be reason to avoid using either estrogen-containing contraceptives or depot medroxyprogesterone acetate (DMPA, Depo-Provera).1

Advise women with ischemic or complicated valvular heart disease to avoid combination oral contraceptives (COC) (WHO category 4). Complicated valvular or congenital heart disease that carries excess risk includes conditions with a higher likelihood of embolism (seen with rhythm disturbances, pulmonary hypertension, Eisenmenger's syndrome, or cyanosis) or is associated with fluid retention, hypertension, or endocarditis.2 On the other hand, if a woman is on long-term anticoagulant therapy and the drug is at therapeutic levels—for instance, if she has a prosthetic heart valve—she can use estrogen because the excess risk of embolism is negated by the medication.3

Women at increased risk for thromboembolic events should also avoid estrogen (category 4). That would include anyone immobilized for a long time due to surgery, injury, or chronic disability, as well as women with a personal history of stroke, deep vein thrombosis (DVT), or pulmonary embolism (PE).1 The risk of venous events is also increased in women with factor V Leiden mutation, antithrombin III deficiency, or other clotting abnormalities. Factor V Leiden increases the risk of DVT fivefold in COC users, resulting in a DVT risk that is still lower than that seen during pregnancy. The mutation is present in only 5% of Caucasian females. Therefore, if there is a clinical suspicion that testing will be needed, it should be reserved for only those with a strong family history of clots.4 A family history of DVT or PE in itself is not a contraindication to estrogen use.

Progestin-only contraceptives offer an alternative for these high-risk patients. For some of the conditions already mentioned, DMPA was also given a cautionary classification (category 3) in the WHO guidelines. That's because DMPA injections contain a higher dose of hormone than other progestin-only methods like the IUD, pill, and implant, and thus there is a theoretical concern that DMPA may increase the risk of thromboembolic events.

Cross-sectional studies have shown that DMPA adversely affects a patient's lipid profile over time, increasing LDL and decreasing HDL.5 Lower-dose progestin pills do not have this effect.6 DMPA has not been associated with any increase in cardiovascular events in normotensive women, however. This may be because progestin does not affect the clotting cascade, fibrinogen, or platelet aggregation, as estrogen does.7 Progestin-only methods offer a safer alternative for patients with a personal history of or risk for cardiovascular disease. In some cases, more follow-up may be warranted to monitor lipid profiles or blood pressure.


Hypertension increases the risk of complications associated with COCs. In the WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception, the odds of hemorrhagic and ischemic stroke were three to 10 times higher for hypertensive COC users than for normotensive non-COC users. The odds of myocardial infarction were 68 times higher for European hypertensive COC users versus normotensive non-COC users. There was no significant increase in MIs among US women, however, although those studies included few hypertensive COC users.8 For women with hypertension that is severe or associated with vascular changes, COCs and other formulas containing estrogen are contraindicated. Widespread use of the combined injectable, the ring, and the patch is relatively recent so data on any complications has been scant.

Because these methods contain estrogen, often generating plasma levels similar to those seen with COCs, the WHO criteria for COC use is currently the same for these other methods until further data are obtained. WHO has concluded that DMPA should also be used with caution, possibly because of the theoretical concerns already mentioned. A multicenter, hospital-based, case–control study did find an increased odds of hemorrhagic stroke—but not MI or thromboembolism—in hypertensive women who used progestin-only injectables and pills compared to normotensive women not using any hormonal contraceptive methods.9 Women who experienced hypertension during pregnancy but are otherwise normotensive need not restrict their use of any contraceptive.1

WHO criteria assign a "3" for combination contraceptives among women with mild (< 59 systolic or <99 diastolic) or well-controlled hypertension, indicating that all combination contraceptives should be used with caution because the risks may outweigh the benefits. Treatment of hypertension in non-COC users has been proven to decrease risk for cardiovascular disease. Similar studies in COC users are limited, yet to date, three studies of hypertensive women using sub-50 µg pills and receiving regu-lar follow-up failed to show any increased risk of stroke.4 For patients with mildly elevated or well-controlled blood pressure, there are highly effective progestin-only and non-hormonal (i.e., the copper IUD) methods that may work well. If these methods are not suitable, either because of undesirable side effects or other concerns, clinicians must use their best judgment to individualize care. For instance, the use of COCs in an otherwise low-risk woman who receives regular follow-up may be safer than an unplanned pregnancy with an increased risk for hypertensive morbidity and mortality (Figure 1).


If your patient is suffering from headaches, determine if they are migraines, and if so, if they are associated with aura (previously called "classic migraine"). "Aura" refers to fully reversible symptoms that indicate focal neurologic dysfunction, and either accompany a migraine or precede it by less than 60 minutes. Single symptoms may develop gradually, or several symptoms may occur in succession (Figure 2).10 Women who have migraines with aura are at higher risk of stroke, and so combination contraceptives are contraindicated at any age. In three case–control studies, the odds of ischemic stroke were 2.1 to 16.9 times higher for COC users (most on sub-50 µg pills) with migraines, compared to non-COC users without migraines, although some results were not statistically significant.8 Women over age 35 with common migraines (without aura) are at increased risk for stroke as well, so you'll want to start COCs cautiously in this age group (category 3); continuing COC is not advised (category 4) (Figure 3).1

In one of these case–control studies, data were collected on headache frequency and severity; over 80% of COC users noted no change in either respect. This percentage was similar in the non-COC users.8 If migraines or other headaches develop or worsen on COCs, consider lowering the dose of either the estrogen or the progestin component, reviewing alternate methods, or both. If headaches or common migraines are associated or exacerbated by menstruation or the placebo-week of pills, continuous use with a COC (monophasic only), a ring, or a patch may improve symptoms.

Diabetes mellitus

Women with diabetes may want contraception either to avoid pregnancy altogether, or to postpone it until their disease is under good control, thereby decreasing the risk of birth defects and miscarriage associated with high glycemic levels in early pregnancy. COCs slightly increase glucose and insulin levels, although both still remain within normal limits.11 Several case–control studies of insulin-dependent diabetic women have shown that OCs do not affect glycemic control, HbA1c, or progression of the disease over time.12 A retrospective cohort of diabetic women showed no increased mortality associated with COCs, even in women with early onset disease.13 Although some studies have shown that DMPA does increase glucose and decrease insulin levels more than OCs, blood glucose levels were still within normal limits.7

Don't prevent women with diabetes mellitus from using combination contraceptives unless they also have vascular disease, other cardiovascular risk factors, or long-standing disease (over 20 years duration). In these situations, estrogen-containing contraceptives (or DMPA) may be initiated under careful follow-up if alternative methods are not possible (category 3), but should not be continued in the face of worsening disease (category 4).1 IUDs do not increase the risk of pelvic inflammatory disease in type 1 or 2 diabetics, when compared to nondiabetics. And even in this population, antibiotic prophylaxis is not required for IUD insertion or removal.12

Hormonal contraceptives are not contraindicated in women with a history of gestational diabetes (GD) (category 1). A prospective, randomized trial of women with previous GD found that long-term use of low-dose OCs did not increase the risk of subsequent diabetes.12 A prospective cohort of Latina women with a history of GD found a two- to threefold increased risk of developing diabetes among those who breastfed and used progestin-only pills, but there were only 78 women in this category, 36% of whom switched methods during the follow-up period and who differed significantly in several ways from the comparison group.14 More study is needed in this particular population to determine if a risk truly exists. Currently, WHO criteria assign a "1" for progestin-only contraceptives for both women with a history of GD and breastfeeding women after 6 weeks postpartum.1


Women with epilepsy are not more likely to have seizures when they go on hormonal contraceptives. In fact, DMPA may actually reduce the number of seizures they experience because progestin raises the seizure threshold. And since DMPA contains a higher dose of progestin than other hormonal methods, its efficacy is not compromised by anticonvulsants or other medications that induce P450 enzyme metabolism.15 However, anticonvulsants will decrease the efficacy of low-dose hormonal contraceptives, which include the low-dose combined and progestin-only pills as well as the patch and ring. With that in mind, advise women on combined contraceptives to use a pill with 50 µg of ethinyl estradiol if they are taking phenobarbital, phenytoin, carbamazepine, oxcarbazepine, felbamate, or topiramate.

Conversely, anticonvulsants that do not induce the P450 system (including valproic acid, gabapentin, ethosuximide, benzodiazepines) do not interact with OCs.16 There is no information to date on the effect of these antiepileptics on the efficacy of the combined injectable Lunelle, but the hormone dosages in this formulation are still not as high as in DMPA, and so an interaction may occur. A pilot study of epileptic women using the Mirena intrauterine system showed no significant decrease in contraceptive efficacy with simultaneous use of enzyme-inducing epileptics.17

Mood disorders

A small number of women will experience more mood swings, depression, or anxiety while on hormonal contraceptives, though women are more likely to note an improvement in these symptoms.18 The risk of worsening mood disorders is more likely in women with a previous history of psychiatric illness and with progestin-only methods. DMPA is of particular concern because of its higher dose and long-acting effects, but the data are not consistent on this issue: A prospective cohort of inner-city DMPA users found no significant increase in depression scores over time, while another prospective study did find a significant increase in depressive symptoms; neither study looked at psychiatric populations.5,19 And a multicenter, prospective trial saw no increase in depression scores among DMPA users after 1 year, although ironically, patients with the lower-dose implant Norplant had a significant increase in scores after 2 years.20 Although a pre-existing mood disorder is not a contraindication to initiating hormonal contraceptives, these patients may need more frequent follow-up to make sure their symptoms don't worsen.

If a patient does experience mood swings, depression, or anxiety, first rule out other causes. In the absence of other causes, consider switching from a progestin-only to a combination method, decreasing the dosage, or switching to a nonhormonal method. If your patient experiences these symptoms cyclically, or only during the pill-free week, you can advise her to use the monophasic pill, patch, or ring continuously in order to avoid the drop in hormone levels associated with the placebo week. Finally, be sure to ask about over-the-counter medications as well: St. John's wort, for instance, can induce P450 metabolism and decrease contraceptive hormone levels as well.21


There is no evidence to indicate that hormonal contraceptives cause HIV infection or AIDS to progress in women. Some have asked whether hormonal methods increase viral shedding or susceptibility to infection, but the current evidence is either weak or biased.8 Keep in mind, though, that the anti-retrovirals nevirapine and nelfinavir have been shown to decrease OC efficacy. Data on the interaction of other anti-retrovirals with hormonal contraceptives is scarce. However, many of these medications either affect P450 metabolism or induce GI side effects like diarrhea and vomiting and therefore have the potential to affect contraceptive hormone levels.22

Encourage patients on anti-retrovirals to use barrier methods, not only as a back-up for their hormonal contraception, but to prevent further transmission of HIV and other sexually transmitted infections (STIs). Other medications that have been shown to decrease the hormonal levels associated with OCs include the anti-tuberculosis drugs rifampin and rifabutin, and the antifungal griseofulvin.18,23 Most antibiotics and antifungals do not affect the P450 system and studies of doxycycline, tetracycline, ampicillin, metronidazole, and the quinolones failed to show any significant interaction with contraceptives.24 Yet again, if GI side effects are likely due to the antibiotic itself, concomitant use of another medication, or the underlying illness, we recommend a back-up method.

The WHO criteria assigns a classification of "3" for IUD use in women who are HIV-positive or at high risk of contracting HIV. That's because these high-risk groups are very likely to be at high risk for other STIs and therefore advised against IUD use.1 However, one prospective study has found that HIV-positive women, regardless of CD4 count, do not have a higher risk of infection, complications, or viral shedding with IUD use.8 Nonspermicidal latex male condoms are the only form of contraception that have been shown to protect against HIV. Discourage patients from using spermicides, alone or with barrier methods, if they are at high risk for HIV, or are HIV-positive with a negative partner; recent studies have found these agents do irritate the vaginal mucosa and are associated with an increase in viral transmission.1


As we mentioned previously, each condition in the WHO Medical Eligibility Criteria is assigned a numerical ranking for each of 10 contraceptive methods. (Guidelines for the use of both the combined patch and ring should currently follow those for COCs). These guidelines thus offer over 1,200 recommendations for contraceptive use in chronic conditions, well beyond the scope of this article—and well beyond the capacity of the average clinician's memory. They also contain information on a number of chronic conditions in which contraceptives neither exacerbate the disease nor lose their effectiveness (for example, sickle cell anemia and thyroid disease). We urge clinicians to download the summary tables from the Web sites so they can quickly and easily refer to the recommendations in daily practice.


Dr. Hatcher is on the Speaker's Bureaus for Berlex Laboratories, Organon, Wyeth-Ayerst.

Dr. Cwiak is on the Speaker's Bureau for Berlex Laboratories.



1. Improving Access to Quality Care in Family Planning: Medical Eligibility Criteria for Contraceptive Use. 2nd ed. Geneva: Reproductive Health and Research: World Health Organization; 2000.

2. Seifert-Klauss V, Kaemmerer H, Brunner B, et al. Contraception in patients with congenital heart defects. Z Kardiol. 2000;89:606-611.

3. Burkman RT, Darney PD, Kaunitz AM, et al. Practice guidelines for OC selection: update 2002. Dialogues in Contraception. 2002;7:1-10.

4. Burkman RT. Cardiovascular issues with oral contraceptives: evidenced-based medicine. Int J Fertil Womens Med. 2000;45:166-174.

5. Westhoff C. Depot medroxyprogesterone acetate contraception. Metabolic parameters and mood changes. J Reprod Med. 1996;41:401-406.

6. Basdevant A. [Metabolic effects of oral contraceptives: influence of the dosage and nature of the substance used.] Contracept Fertil Sex (Paris). 1979;4:781-785.

7. Frederiksen MC. Depot medroxyprogesterone acetate contraception in women with medical problems. J Reprod Med. 1996;41(5 suppl):414-418.

8. Curtis KM, Chrisman CE, Peterson HB. Contraception for women in selected circumstances. Obstet Gynecol. 2002;99:1100-1112.

9. Cardiovascular disease and use of oral and injectable progestogen-only contraceptives and combined injectable contraceptives: results of an international, multicenter, case-control study. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Contraception. 1998;57: 315-324.

10. ICD-10 Guide for Headaches. International Headache Classification Committee. Cephalalgia. 1997;17(suppl 19):1-82.

11. Godsland IF, Winkler U, Lidegaard O, et al. Occlusive vascular diseases in oral contraceptive users. Drugs. 2000;60:721-869.

12. Kjos SL. Postpartum care of the woman with diabetes. Clin Obstet Gynecol. 2000;43:75-86.

13. Klein BE, Klein R, Moss SE. Mortality and hormone-related exposures in women with diabetes. Diabetes Care. 1999;22:248-252.

14. Kjos SL, Peters RK, Xiang A, et al. Contraception and the risk of type 2 diabetes mellitus in Latina women with prior gestational diabetes mellitus. JAMA. 1998;280: 533-538.

15. Injectable contraception: Depo-provera. In: Speroff L, Darney PD. A Clinical Guide for Contraception. 2nd ed. Baltimore, MD: Williams and Wilkin; 1996;175-189.

16. Crawford P. Interactions between antiepileptic drugs and hormonal contraception. CNS Drugs. 2002;16: 263-272.

17. Bounds W, Guillebaud J. Observational series on women using the contraceptive Mirena concurrently with anti-epileptic and other enzyme-inducing drugs. J Fam Plann Reprod Health Care. 2002;28:78-80.

18. Hatcher RA, Guillebaud J. The pill: combined oral contraceptives. In: Hatcher RA, Trussel J, Stewart F, et. al. Contraceptive Technology. 17th ed. New York, NY: Ardent Media, Inc; 1998:405-466.

19. Civic D, Scholes D, Ichikawa L, et al. Depressive symptoms in users and nonusers of depot medroxypro-gesterone acetate. Contraception. 2000;61:385-390.

20. Kaunitz AM. Long-acting hormonal contraception: assessing impact on bone density, weight, and mood. Int J Fertil Womens Med. 1999;44:110-117.

21. Henderson L, Yue QY, Bergquist C, et al. St. John's wort (Hypericum perforatum): drug interactions and clinical outcomes. Br J Clin Pharmacol. 2002;54: 349-356.

22. Schutz M. Quick reference guide to antiretrovirals. Eur J Med Res. 2001;6:219-227.

23. Finch CK, Chrisman CR, Baciewicz AM, et al. Rifampin and rifabutin drug interactions: an update. Arch Intern Med. 2002;162:985-992.

24. The Use of Hormonal Contraception in Women with Coexistent Medical Conditions. In: 2003 Compendium of Selected Publications. Washington, DC; 2003:608-617.

Dr. Hatcher is Professor of Gynecology and Obstetrics, Emory University School of Medicine, Atlanta, Ga. Dr. Cwiak is Assistant Professor of Gynecology and Obstetrics, Emory University School of Medicine.

Key points

  • Some risk factors for cardiovascular disease, like obesity, are not contraindications for estrogen-containing contraceptives on their own. But the combination of two or more of these risk factors may be reason to avoid using either estrogen-containing contraceptives or DMPA.

  • Advise women with ischemic or complicated valvular heart disease to avoid combination oral contraceptives (COC), but if a patient is on long-term anticoagulant therapy and the drug is at therapeutic levels—for instance, if she has a prosthetic heart valve—she can use estrogen because the excess risk of embolism is negated by the medication.

  • The odds of hemorrhagic and ischemic stroke are three to 10 times higher for hypertensive COC users than for normotensive non-COC users. The odds of MI were 68 times higher for European hypertensive COC users vs. normotensive non-COC users.

  • Women who have migraines with aura are at higher risk of stroke, and so combination contraceptives are contraindicated at any age.

  • Several case–control studies of insulin-dependent diabetic women have shown that OCs do not affect glycemic control, HbA1c, or progression of the disease over time.


Important Web sites

1. To access the WHO Medical Eligibility Criteria:

2. For more information on all US contraceptive methods, including the WHO Criteria:


Robert Hatcher, Carie Cwiak. Cover Story: When a chronically ill patient needs contraceptives. Contemporary Ob/Gyn Oct. 1, 2003;48:70-82.