The contraceptive methods most commonly used in the US (in order of popularity) are oral contraceptives, female sterilization, condoms, male sterilization, progestin injections, withdrawal (coitus interruptus), intrauterine devices (IUDs), periodic abstinence, spermicides, and diaphragms (see Table 1: Family Planning: Comparison of Common Contraceptive Methods). Over several years, pregnancy rates are < 1%/yr with methods unrelated to coitus (IUDs, progestin injections, subdermal progestin implants, and oral contraceptives when taken consistently) and about 5%/yr with coitus-related methods (eg, condoms, diaphragms, spermicides, withdrawal); pregnancy rates tend to be higher during the first year, then decrease in later years as users become more facile and women's fertility decreases. In contrast, the pregnancy rate is 90%/ yr with frequent, unprotected intercourse. Condoms are often preferred despite their relatively high failure rate because they protect against sexually transmitted diseases, especially HIV. Emergency contraception, done postcoitus, should not be used as a regular method of contraception.
Oral contraceptives (OCs) simulate ovarian hormones: They provide negative feedback to the hypothalamus; its release of gonadotropin-releasing hormone is inhibited, thus inhibiting pituitary release of gonadotropins that stimulate ovulation. The endometrium becomes thin, and cervical mucus becomes thick and impervious to sperm.
An OC may be a combination of an estrogen and a progestin or a progestin alone. Most combination formulations are taken daily for 3 wk, then no pill or a placebo is taken during the 4th wk to allow for withdrawal bleeding. However, one formulation is taken daily for 12 wk, then not taken during the 13th wk, so that withdrawal bleeding occurs only 4 times/yr. Two new formulations provide active hormones for 24 days and placebo for 4 days so that a tablet is taken every day.
Progestin alone is taken daily; it often results in irregular bleeding and may be less effective than combination OCs. Progestin alone is not used unless estrogen is contraindicated—eg, during breastfeeding.
Combination formulations have similar efficacy; the pregnancy rate after 1 yr is < 0.3% with perfect use and about 8% with typical (ie, inconsistent) use. Low-dose formulations (20 to 35 μg of estrogen) are usually preferred to higher-dose formulations (50 μg of estrogen) because low-dose formulations appear equally effective and, except for a higher incidence of bleeding during the first few months of use, have fewer adverse effects.
Intermittently stopping OCs appears to have no benefit, so OCs can be taken continuously until menopause, which is indicated by an elevated follicle-stimulating hormone level. Combination OCs are not prescribed for women > 35 who smoke cigarettes or have other contraindications (see Table 2: Family Planning: Contraindications to Combination Oral Contraceptive Use).
|Contraindications to Combination Oral Contraceptive Use
Abnormal uterine bleeding with an undiagnosed cause
Diabetes mellitus (or other disorders) with vascular complications
History of estrogen-dependent cancer, coronary artery disease, hepatic adenoma, idiopathic recurrent jaundice of pregnancy, stroke, thromboembolism, deep venous thrombosis, MI, or active SLE (because risk of cardiovascular sequelae is increased)
Hypertension if uncontrolled
Immobilization of a lower extremity if prolonged
Liver disorders if active
Major surgery for 1 mo after (and if elective, for 1 mo before)
Smoking after age 35
Triglyceride levels > 250 mg/dL
Amenorrhea with an undiagnosed cause
Cigarette smoking if heavy in women < 35
Family history of blood clots
History of liver disorders with recovery
Migraines with neurologic symptoms
Precancerous cervical lesions (treated)
Seizure disorder (drug-treated)
OCs may cause breakthrough bleeding (which may resolve with time or when the estrogen dose is increased) or amenorrhea (which may resolve when the progestin dose is decreased). In a few women, ovulation remains inhibited for a few months after they stop taking OCs. OCs do not adversely affect the outcome of pregnancy when conception occurs during or after their use.
Estrogens increase aldosterone production and cause Na retention, which can produce dose-related, reversible increases in BP and weight, up to about 2 kg; weight gain may be accompanied by bloating, edema, and breast tenderness. Most progestins used in OCs are related to 19-nortestosterone and are androgenic; norgestimate and desogestrel are less androgenic than levonorgestrel, norethindrone, norethindrone acetate, and ethynodiol diacetate. Androgenic effects may include acne, nervousness, and an anabolic effect resulting in weight gain. If a woman gains > 4.5 kg/yr, a less androgenic OC should be used. Drospirenone, a new progestin, is related to spironolactone, not 19-nortestosterone; it is antiandrogenic and diuretic.
OCs increase the risk of certain disorders and decrease the risk of others (see Table 3: Family Planning: Some Risks and Benefits of Combination Oral Contraceptives).
|Some Risks and Benefits of Combination Oral Contraceptives
Benign hepatic adenomas (rare; dose- and duration-related)
Cervical cancer (when OCs are taken > 5 yr)
Deep vein thrombophlebitis (dose-related)
Insulin resistance (related to the progestin dose)
Stroke (possibly) in women who take high-dose OCs and have other stroke risk factors
Abnormal uterine bleeding
Benign breast disorders
Benign ovarian tumors
Functional ovarian cysts
Iron deficiency anemia
OCs = oral contraceptives.
Incidence of deep venous thrombosis and thromboembolism increases in relation to estrogen dose; formulations with 20 to 35 μg increase risk to about 3 to 4 times normal (but risk is still only about half of that during pregnancy). Formulations containing the less androgenic progestin desogestrel may have a slightly higher risk than formulations containing levonorgestrel, but this difference has not been established. Varicose veins do not appear to further increase risk. Hypercoagulability is probably caused by increases in clotting factors, particularly VII and X, and possibly increased platelet adhesion. If deep vein thrombophlebitis or pulmonary embolism is suspected, OCs should be stopped, pending results of diagnostic tests. OCs should be stopped as soon as possible before any major surgery and 1 mo before elective major surgery; they should not be restarted until 1 mo after the surgery.
Current use of OCs does not increase overall risk of breast cancer, nor does former use in women aged 35 to 65. Also, use does not increase risk in high-risk groups (eg, women with certain benign breast disorders or a family history of breast cancer). Risk of cervical cancer is increased in women who have used OCs for > 5 yr; the reason is unknown. OC users should have annual cervical cytology screening (eg, Papanicolaou [Pap] test).
Although increased stroke risk has been attributed to OCs, low-dose combination OCs do not appear to increase risk in healthy, normotensive, nonsmoking women. Nonetheless, if focal neurologic symptoms, aphasia, or other symptoms that may herald stroke develop, OCs should be stopped. CNS effects of OCs include nausea, vomiting, headache, depression, and sleep disturbances.
Although progestins may cause reversible, dose-related insulin resistance, use of current OCs, which have a low progestin dose, rarely results in hyperglycemia. Serum high-density lipoprotein (HDL) cholesterol levels may decrease when OCs with a high progestin dose are used but usually increase when OCs with low progestin and estrogen doses are used. Most alterations in serum levels of other metabolites are not clinically significant. Thyroxine-binding globulin capacity may increase; free thyroxine levels, thyroid-stimulating hormone levels, and thyroid function are not affected. Levels of pyridoxine, folate, most other B vitamins, ascorbic acid, Ca, manganese, and zinc decrease; vitamin A levels increase.
OCs accelerate growth of existing gallstones but do not cause new stones to form. Thus, incidence of cholelithiasis increases during the first few years of OC use, then decreases. Women who develop idiopathic recurrent jaundice of pregnancy (cholestasis of pregnancy) may become jaundiced if they take OCs; these women should not take OCs.
Rarely, benign hepatic adenomas that can spontaneously rupture develop. Incidence increases as duration of use and OC dose increase; adenomas usually regress spontaneously after the OC is stopped.
Melasma occurs in some women; it is accentuated by sunlight and disappears slowly after OCs are stopped. Because treatment is difficult (see Pigmentation Disorders: Melasma (chloasma)), OCs are stopped when melasma first appears. OCs do not increase risk of malignant melanoma.
OCs decrease risk of endometrial and ovarian cancers by about 50% for at least 20 yr after OCs are stopped. They also decrease risk of benign ovarian tumors, abnormal uterine bleeding, dysmenorrhea, premenstrual syndrome, iron deficiency anemia, benign breast disorders, and functional ovarian cysts. Ectopic pregnancy and salpingitis, which can impair fertility, are also less likely.
Although OCs can slow the metabolism of certain drugs (eg, meperidine), the effects are not clinically important.
Some drugs (eg, cyclophosphamide, rifampin) can induce liver enzymes that accelerate transformation of OCs to less biologically active metabolites; women who take these drugs should not be given OCs concurrently. Whether certain antibiotics (eg, penicillin, ampicillin, sulfonamides) and anticonvulsants (eg, carbamazepine, phenytoin, phenobarbital, primidone, topiramate) reduce the effectiveness of OCs is less clear. If therapeutic doses of antibiotics are prescribed, using a barrier method in addition to OCs may be prudent. Women who take anticonvulsants should use a 50-μg estrogen formulation because a lower dose often causes breakthrough bleeding.
Before OCs are started and annually thereafter, breast and pelvic examinations, liver palpation, and measurement of BP and weight are necessary. BP is also measured 3 mo after starting OCs. OC users should have annual cervical cytology screening (eg, Pap test), particularly if they have taken OCs > 5 yr or if they are young, sexually active women. For women with a history of a liver disorder, normal liver function must be documented before OCs are prescribed. Women at risk of diabetes (eg, those who have a family history or who have had high-birth-weight infants or unexplained fetal deaths in previous pregnancies) require annual plasma glucose screening and a complete serum lipid profile (repeated annually if results are abnormal). Use of low-dose OCs is not contraindicated by abnormal glucose or lipid test results, except for triglycerides > 250 mg/dL.
When to start OCs after pregnancy is guided by when ovulation is expected. After an abortion, ovulation usually occurs after 2 to 4 wk and before the first menses; older gestational age predicts later ovulation. In women who have a term delivery and are not breastfeeding, ovulation occasionally occurs as early as 4 wk after delivery but usually not until after menses. In women who are breastfeeding, ovulation usually occurs ≥ 10 wk after delivery and after menses.
After spontaneous or induced abortion of a fetus < 12 wk gestation, OCs are started immediately. If the fetus is 12 to 28 wk gestation, OCs are delayed 1 wk.
After a term delivery, risk of thromboembolism, which normally increases postpartum, also influences when OCs are started. If women deliver after 28 wk and are not breastfeeding, combination OCs are delayed 2 wk because combination OCs may enhance risk of thromboembolism. For women who are breastfeeding, progestin-only OCs are started a few days after delivery because progestin-only OCs are not thrombogenic. Progestin-only OCs are used because OCs that contain estrogen reduce the amount of milk produced as well as the protein and fat concentration in milk.
Barrier contraceptives include condoms, diaphragms, cervical caps, vaginal spermicides (foams, creams, suppositories), and the contraceptive sponge.
Condom use is the only reversible male method other than withdrawal, which is probably much less effective. Condoms decrease risk of sexually transmitted diseases (only latex condoms can fully protect against HIV) and may prevent precancerous changes in the cervix.
The condom is applied before penetration; the tip should extend about 1 cm beyond the penis to collect the ejaculate. During removal, care is taken to avoid spilling condom contents. A new condom is used for each act of coitus. Pregnancy rates in the 1st yr are 2% with perfect use but 15% with typical (ie, inconsistent) use. Adding a spermicide, which may be included in the condom's lubricant or inserted into the vagina, may lower these rates.
The diaphragm, a dome-shaped rubber cup with a flexible rim that fits over the cervix, is a barrier to sperm. Diaphragms are made in various sizes. A health care practitioner fits a diaphragm to a woman so that it is comfortable for her and her partner. After childbirth or a significant weight change, the woman is refitted. The diaphragm should remain in place for > 8 h after the last coitus. Spermicides should be used before each coitus in case the diaphragm is displaced. Pregnancy rates in the 1st yr are about 6% with perfect use but about 16% with typical use.
The cervical cap resembles the diaphragm but is smaller and more rigid. It comes in several sizes and is fitted by a health care practitioner. It can be left in place for 48 h. For nulliparous women, pregnancy rates in the 1st yr are 9% with perfect use and about 18% with typical use. Pregnancy rates for parous women are about twice as high because obtaining a secure fit is difficult.
Vaginal foams, creams, and suppositories provide a physical barrier to sperm and contain a spermicide, usually nonoxynol-9. They are combined with other barrier methods and are placed in the vagina before each coitus. These products appear to have similar efficacy.
The contraceptive sponge, previously off the market, is available again. It contains nonoxynol-9, does not need to be fitted by a health care practitioner, and can be inserted before coitus. Its efficacy is less than that of the diaphragm.
(Natural Family Planning)
Although the ovum can be fertilized for only about 12 h after ovulation, sperm can fertilize an ovum for up to 5 days after coitus. Thus, periodic abstinence requires abstinence from coitus during the 5 days before ovulation. Several methods can be used to identify the time of ovulation; they include the calendar method, basal body temperature, and the characteristics of cervical mucus.
The calendar rhythm method aims to predict ovulation solely by menstrual dates. Ovulation occurs about 14 days before onset of menses; but even when menstrual cycles are regular, this method is often inaccurate. The interval of abstinence during the menstrual cycle is determined by subtracting 18 days from the shortest of the previous 12 cycles and 11 days from the longest (see Fig. 1: Family Planning: Natural family planning methods.). For example, if cycles vary between 26 and 29 days, abstinence is required from days 8 through 18 of each cycle. The greater the variance in cycle length, the longer abstinence lasts.
Methods that incorporate better predictors of ovulation are more effective but require training and effort:
The 1-yr pregnancy rates with of any of these methods is 25% with typical use. With perfect use, rates are 9% for the calendar method, 2% for the temperature method, 3% with the mucus method, and 2% with the symptothermal method. The symptothermal method is considered the most effective method of periodic abstinence because achieving perfect use is easier.
Depot medroxyprogesterone acetate (DMPA) is a long-acting injectable formulation of medroxyprogesterone acetate. Pregnancy rates in the 1st yr are 0.3% with perfect use and are slightly higher with typical use (ie, delays between injections).
The injection must be given during the first 5 days of the menstrual cycle to prevent ovulation. The dose is 150 mg q 3 mo by deep gluteal or deltoid IM injection. The injection site is not massaged, ensuring slow absorption. Alternatively, a formulation can be injected sc; the dose is 104 mg sc q 3 mo. Levels are usually effective starting 24 h after IM or sc injection and are maintained up to 4 mo or more. If the interval between injections is > 13 wk, a negative pregnancy test is required before giving the next injection.
The most common adverse effect is disruption of the menstrual cycle. In the 3 mo after the first DMPA injection, about 30% of women have amenorrhea. Another 30% have spotting or irregular bleeding (usually light) > 11 days/mo; anemia does not usually result. As use continues, bleeding and menses tend to decrease; after 2 yr, about 70% of women receiving DMPA have amenorrhea. Because DMPA has a long duration of action, ovulation may be delayed for up to 1 yr after the last injection; after ovulation occurs, fertility is usually rapidly restored.
Women typically gain 1.5 to 4 kg during the 1st yr of DMPA use and continue to gain weight thereafter. Usually, women who want to take DMPA are advised to decrease caloric intake and increase energy expenditure. Headache is a common reason for stopping DMPA, but severity tends to decrease over time. Most women using DMPA do not have headaches, and preexisting tension headaches and migraines usually do not worsen. Mild, reversible, clinically insignificant deterioration of glucose tolerance and lipid profile may occur. There is no evidence of increased fracture risk. Adolescents and young women using DMPA should consume 1500 mg of Ca and 400 units of vitamin D daily; supplements should be taken if necessary. Unlike OCs, DMPA does not contribute to hypertension or thromboembolism.
DMPA does not appear to increase risk of breast, ovarian, or invasive cervical cancer. DMPA reduces risk of endometrial cancer and pelvic inflammatory disease. By stimulating erythropoiesis, DMPA reduces risk of iron deficiency anemia and, for women with sickle cell disease, risks of anemia and crises.
Transdermal and Intravaginal Steroid Contraceptives
A 20-cm2 transdermal patch can deliver 150 μg of the progestin norelgestromin (the active metabolite of norgestimate) and 20 μg of ethinyl estradiol daily into the systemic circulation for 7 days; then, the patch is removed, and a new patch is applied to a different area of the skin. Steroid blood levels are much more constant than with OCs. After 3 patches, no patch is used for the 4th wk to allow withdrawal bleeding. Overall contraceptive efficacy, incidence of bleeding, and adverse effects are similar to those of OCs, but adherence may be better. The patch may be less effective in overweight women.
Use of a flexible vaginal ring containing ethinyl estradiol and the progestin etonogestrel, which are absorbed through the vaginal epithelium, provides relatively constant blood levels. The ring is 58 mm in diameter and 4 mm thick. It is inserted and removed by the woman; it is left in place for 3 wk, then removed for 1 wk to allow for withdrawal bleeding. Bleeding with the ring in place is uncommon. Contraceptive efficacy and adverse effects are similar to those of OCs, and adherence may be better.
Progestin subdermal implants are available as a new, single-rod implant that can be inserted through a trocar without a skin incision. The implant releases etonogestrel at a rate of 50 μg/day. The implant provides effective contraception for 3 yr.
The most common adverse effects are similar to those of other progestins (uterine bleeding, amenorrhea, headache). Whether weight gain results is unclear. Removing the implant requires a skin incision. After implant removal, ovarian activity normalizes immediately.
Only about 1 million women in the US use intrauterine devices (IUDs), despite their advantages over OCs:
IUDs induce endometrial inflammation; this inflammation attracts neutrophils, which are toxic to sperm and prevent fertilization of the ovum.
In the US, 2 types are used: levonorgestrel-releasing IUD (effective for 5 yr; cumulative pregnancy rate of 0.5%) and copper-bearing T380A (effective for ≥ 10 yr; cumulative pregnancy rate of < 2%). IUDs are inserted high in the fundus at any time during the cycle. Contraindications include pregnancy and untreated cervicitis or vaginitis. After IUD removal, fertility rate returns to normal after about 1 yr.
During the 1st yr, 10 to 15% of women stop use; fewer stop thereafter. Of those who stop, > 50% do so because of bleeding and pain, which occur in about 15% of users during the 1st yr and 7% during the 2nd yr. Bleeding stops completely within 1 yr in 20% of women using the levonorgestrel-releasing IUD.
Spontaneous expulsion occurs in about 5% during the 1st yr (usually within the first few weeks) and in fewer women thereafter. Expulsion is more common among younger women and among nulligravidas. If another IUD is inserted, it is usually retained. Because about 20% of expulsions are unnoticed, a plastic string is attached to the IUD so that the user can check for its presence periodically.
Uterine perforation occurs in about 1/1000. Perforation occurs only during IUD insertion. Sometimes only the distal portion of the IUD penetrates; then over the next few months, uterine contractions force the IUD into the peritoneal cavity. Perforation is considered if a woman cannot feel the string but did not notice expulsion. If the string is not visible during pelvic examination, the uterine cavity is probed with a sound or biopsy instrument unless pregnancy is suspected. If the IUD cannot be felt, ultrasonography is done. If the IUD is not seen, an abdominal x-ray is taken to exclude an intraperitoneal location. Intraperitoneal IUDs may cause intestinal adhesions. IUDs that have perforated the uterus are removed, often via laparoscopy.
Occasionally, salpingitis develops during the 1st mo because of contamination during insertion; risk is not high enough to warrant routine antibiotic prophylaxis. IUD strings do not provide access for bacteria. Pelvic infections that occur after 1 mo are sexually transmitted; unless severe, they can be treated with the IUD in place.
Pregnancies that occur with an IUD in place are usually intrauterine (95%). About 5% are an ectopic pregnancy, which must be ruled out (see Abnormalities of Pregnancy: Ectopic Pregnancy) in any woman who becomes pregnant with an IUD in place. With intrauterine pregnancies, there is a high risk of spontaneous abortion (about 55%) and preterm delivery but not of congenital defects, fetal death, or pelvic infection during pregnancy. Risk of abortion decreases to 20% if the IUD is removed.
IUDs do not increase and may decrease risk of endometrial adenocarcinoma and cervical cancer.
Emergency contraception is use of contraceptive hormones within 72 h of unprotected coitus. Emergency contraception can decrease the pregnancy rate for a single act of unprotected coitus at midcycle, which is typically about 8%. The most commonly used regimen is 2 doses of levonorgestrel 0.75 mg; one is taken within 72 h of unprotected intercourse, followed by one 12 h later. If taken within 72 h of one act of unprotected intercourse, levonorgestrel reduces risk of pregnancy by 89%, but if taken within 24 h, it reduces risk by 95%. Taking levonorgestrel 1.5 mg once seems to be equally effective. Another regimen—2 tablets, each containing ethinyl estradiol 50 μg and levonorgestrel 0.25 mg, followed by 2 more tablets 12 h later—has been used, but it is slightly less effective. The high estrogen dose often causes nausea and may cause vomiting.
Inserting a copper IUD within 10 days of coitus is more expensive but more effective than hormone tablets; pregnancy rate is 0.1%.
Last full review/revision August 2007 by Daniel R. Mishell, Jr., MD; Megan A. Economidis, MD