The new england journal of medicine

clinical practice

Combination Estrogen–Progestin
Oral Contraceptives
Diana B. Petitti, M.D., M.P.H.
This Journal feature begins with a case vignette highlighting a common clinical problem.
Evidence supporting various strategies is then presented, followed by a review of formal guidelines,
when they exist. The article ends with the author’s clinical recommendations.

A healthy, sexually active, 35-year-old woman presents for advice about the use of oral
contraceptives. She does not smoke cigarettes and has no personal or family history of
venous thromboembolism, myocardial infarction, or stroke. Her blood pressure is
120/80 mm Hg. Should an oral contraceptive be prescribed, and if so, how should a
formulation be chosen?

the clinical problem

Most oral contraceptives now in widespread use are combinations of an estrogen and a From the Research and Evaluation Depart-
progestin. At least 10 million women in the United States and 100 million women world- ment, Kaiser Permanente Southern Califor-
nia, Pasadena. Address reprint requests to
wide use combination oral contraceptives. They are highly effective in preventing preg- Dr. Petitti at Research and Evaluation, Kaiser
nancy: about 5 women per 100 typical users1 and fewer than 1 per 100 women with per- Permanente Southern California, 100 S. Los
fect use become pregnant per year. Robles, 2nd Fl., Pasadena, CA 91188, or at
[Link]@[Link].
The first oral-contraceptive formulations marketed in the United States, in 1960 and
1961, contained 2 to 5 times as much estrogen and 5 to 10 times as much progestin as N Engl J Med 2003;349:1443-50.
the oral contraceptives now in use. The use of these high-dose formulations was linked Copyright © 2003 Massachusetts Medical Society.

to increased risks of ischemic stroke, myocardial infarction, and pulmonary embolism

in healthy young women. The doses of estrogen and progestin were reduced rapidly
during the 1960s and 1970s because of concern about safety and because the reduction
of the doses did not reduce the contraceptive effectiveness.
The reductions in the dose of estrogen are believed to have decreased the risk of
venous thrombosis. Because the effects of combination oral contraceptives on cardio-
vascular risk factors such as lipid levels and glucose tolerance vary with the type of
progestin used in conjunction with the same dose of estrogen, oral-contraceptive for-
mulations with fewer adverse effects on these metabolic variables were developed.
The combination estrogen–progestin oral contraceptives that are now on the market
contain estrogen at doses ranging from 20 to 50 µg of ethinyl estradiol or, uncommonly,
mestranol. These estrogens are combined with any of several different progestins,
which may be given at the same dose every day (“monophasic”) or at varying doses ac-
cording to the phase of the cycle (“biphasic” or “triphasic”) to mimic more closely the
production of progesterone during the normal menstrual cycle. Details of formulations
and doses are provided in Supplementary Appendix 1 (available with the full text of this
article at [Link]
Recently, oral contraceptives have been classified by some according to “generation”
(first, second, third, and most recently, fourth generation). These terms sometimes re-
fer to the timing of the introduction of a product (given both the dose of estrogen and
the type of progestin), sometimes refer to the timing of the market introduction of the
progestin, sometimes refer to the structure of the carbon ring from which the progestin
is derived (estrane or gonane), and sometimes lack a clear definition. Thus, the same

n engl j med 349;15 [Link] october 9, 2003 1443

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The new england journal of medicine

formulation may be classified as being of different and arterial vascular disease, which may vary ac-
generations in different studies (Table 1). Because cording to the formulation. The challenge for the
of this confusion, I avoid the use of such terms in clinician is to identify women in whom the risks
this article. associated with oral-contraceptive use outweigh
The use of combined oral contraceptives has the benefits and to select formulations for other
many noncontraceptive benefits, but there are also women that minimize the risks and maximize the
risks, including risks of venous thromboembolism benefits.

Table 1. Types of Progestin in Combination Estrogen–Progestin Oral Contraceptives Marketed in the United States or Mentioned in Studies
of Types of Progestin and Cardiovascular Disease.

Type of Progestin Generation Brand Names of Selected Products*

According to According to
Time of Market Published Studies
Introduction† of Vascular Disease
Ever marketed in the
United States
Chlormadinone acetateठFirst Other C-Quens
Desogestrel Third Third Desogen, Ortho-Cept, Apri (monophasic)
Cyclessa (triphasic)
DimethisteroneठFirst Not mentioned Oracon
Drospirenone Fourth Not studied Yasmin (monophasic)
Ethynodiol diacetate First First or second Demulen 1/35, Zovia 1/35E (monophasic)
Levonorgestrel Second Second Levlen, Levora, Nordette, Portia (monophasic)
Alesse, Aviane, Lessina, Levlite (monophasic)
Enpresse, Tri-Levlen, Triphasil, Trivora (triphasic)
Medroxyprogesterone First Not mentioned Provest
acetate‡
Norethindrone First First or second Necon 1/35, Norinyl 1+35, Nortrel 1/35, Ortho-Novum 1/35 (monophasic)
Brevicon, Modicon, Neocon 0.5/35, Nortrel 0.5/35 (monophasic)
Ovcon-35 (monophasic)
Necon 7/7/7, Ortho-Novum 7/7/7, Nortrel 7/7/7, Tri-Norinyl (triphasic)
Norethindrone acetate‡ First First or second Loestrin 21 1.5/30, Loestrin Fe 1.5/30, Microgestin Fe 1.5/30 (monophasic)
Loestrin 21 1/20, Loestrin Fe 1/20, Microgestin Fe 1/20 (monophasic)
Estrostep 21, Estrostep Fe (triphasic)
Norethynodrel‡ First First Enovid
Norgestimate‡ Third Second, third, Mononessa, Ortho-Cyclen, Sprintec (monophasic)
or other Ortho Tri-Cyclen, Ortho Tri-Cyclen Lo (triphasic)
Norgestrel First First or second Cryselle, Lo/Ovral, Low-Ogestrel (monophasic)
Never marketed in the
United States
Gestodene Third Third —
Lynestrenol First First or second —

* Among products currently marketed in the United States, the lists include only monophasic and triphasic formulations containing less than
50 µg of ethinyl estradiol; each list of monophasic formulations containing a given type of progestin includes products containing the same
amount of estrogen and progestin; separate lists of monophasic preparations containing a given type of progestin appear in order of decreas-
ing dose of ethinyl estradiol or progestin. The listed products are provided as examples and were not selected on the basis of cost or market
share.
† According to this classification, the first generation includes contraceptives approved for marketing in the United States before 1973, the sec-
ond generation those approved for marketing in the United States between 1973 and 1989, the third generation those approved for marketing
in the United States or Europe between 1990 and 2000, and the fourth generation those approved for marketing in the United States after 2000.
‡ This type of progestin is not contained in any combination estrogen–progestin oral contraceptives currently marketed in the United States.
§ In the United States, this type of progestin has been marketed only in combination oral contraceptives involving the sequential administration
of estrogen and progestin.

1444 n engl j med 349;15 [Link] october 9 , 2003

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 clinical practice

risks associated with oral-contraceptive use

strategies and evidence
Myocardial Infarction and Ischemic Stroke
benefits of oral-contraceptive use Whereas older oral-contraceptive preparations in-
Ovarian Cancer creased the risk of myocardial infarction and ische-
The risk of ovarian cancer is reduced by at least half mic stroke, studies reported in the past seven years
among women who use oral contraceptives, includ- have yielded conflicting results.15-23 The relative
ing those who use low-estrogen formulations.2,3 risks of myocardial infarction and ischemic stroke
The reduction in risk occurs after relatively short- among current users of oral contraceptives, as com-
term use (5 years) and persists for 10 to 20 years af- pared with nonusers, were 1.0 and 1.1, respectively,
ter use has been discontinued. This benefit extends according to a pooled analysis of data from two case–
to women with a family history of ovarian cancer4 control studies in the United States.18-22 The Myo-
and women with a mutation in the BRCA1 or BRCA2 cardial Infarction and Oral Contraceptives (MICA)
gene.5,6 The suggested mechanism for this effect is study15 reported a relative risk of myocardial infarc-
the suppression of ovulation. tion among current users of oral contraceptives of
1.4, which was not statistically significant. Other
Endometrial Cancer studies reported statistically significant increases
Older formulations of oral contraceptives, which by a factor of two to five in the relative risks of myo-
contained higher doses of estrogen, reduce the risk cardial infarction17,19 and ischemic stroke20,21,23
of endometrial cancer, presumably by progestin- among current users of oral contraceptives. The dif-
mediated suppression of estrogen-induced prolif- ferences are probably due to the higher prevalence
eration of endometrial cells. The reduction in risk of smoking and of untreated and undiagnosed hy-
occurs after a relatively short period of use (five years) pertension in the populations of the latter studies.
and persists long after discontinuation. Newer for- In keeping with these findings, large increas-
mulations have not been studied, but the presumed es (by a factor ranging from 7 to more than 100)
mechanism suggests that these formulations would have been observed in the relative risks of myocar-
also reduce the risk of endometrial cancer.7 dial infarction and ischemic stroke among users
of oral contraceptives who smoke or have hyperten-
Acne sion.16,19,20,23 Table 2 shows the estimated num-
Randomized, double-blind, placebo-controlled tri- ber of excess cases of ischemic stroke or myocar-
als show substantial reductions in the severity of dial infarction attributable to oral-contraceptive use
acne among both patients given oral contraceptives among nonsmokers, smokers, and women with hy-
and patients given placebo, but the patients who pertension according to age.24 For comparison, the
received oral contraceptives had greater improve- pregnancy-related rate of death per 100,000 live
ment.8 Randomized trials comparing low-estrogen births in 199925 is also shown.
oral contraceptives including different progestins It has been proposed that the type of progestin
do not show consistent differences among formu- used may modify the arterial vascular risk associ-
lations.9 Some formulations have been approved for ated with the use of oral contraceptives. However,
a marketing claim regarding their beneficial effect when the dose of estrogen was less than 50 µg, the
on acne, but all low-dose combination oral contra- risk of ischemic stroke did not differ between wom-
ceptives cause a similar decrease in the concentra- en taking oral-contraceptive formulations contain-
tion of free testosterone, the presumed mechanism ing desogestrel or gestodene and women taking
for the improvement of acne.10 formulations containing levonorgestrel or noreth-
indrone.20
Menstrual Disorders, Loss of Blood, and Anemia
One randomized, double-blind, placebo-controlled Venous Thromboembolism and Cerebral Venous
trial of low-dose oral contraceptives showed that Thrombosis
their use reduces the severity of dysfunctional uter- The risk of venous thromboembolism (pulmonary
ine bleeding.11 Oral-contraceptive use decreases embolism and deep venous thrombosis) is increased
menstrual blood flow and is associated with a re- by a factor of three to four among current users
duced prevalence of anemia and increased hemo- of low-estrogen oral contraceptives26-32 containing
globin concentrations in anemic women.12-14 norethindrone, norethindrone acetate, lynestrenol,

n engl j med 349;15 [Link] october 9, 2003 1445

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The new england journal of medicine

gestrel than among users of other formulations is

Table 2. Age-Specific Estimates of the Excess Rates of Myocardial Infarction,
uncertain.40,41
Ischemic Stroke, and Venous Thromboembolism Attributable to the Use
of Low-Estrogen Oral Contraceptives and Pregnancy-Related Mortality.* The risk of venous thromboembolism among
users of oral contraceptives who have thrombophil-
Variable Age ia — defined as a deficiency in protein C or protein
20–24 30–34 40–44 S or the presence of factor V Leiden or the prothrom-
Yr Yr Yr bin G20210A mutation — is much higher (by a fac-
No. of excess cases of myocardial infarction tor of 6 to 40) than the risk among nonusers who
and ischemic stroke attributable to oral- do not have thrombophilia.27,29
contraceptive use (per 100,000 woman-yr
of use)†
evidence of no or minimal effect
Among nonsmokers 0.4 0.6 2
Low-estrogen oral contraceptives do not appear to
Among smokers 1 2 20 increase the risk of hemorrhagic stroke among
Among women with hypertension 4 7 29 women who do not have hypertension. There is no
No. of pregnancy-related deaths (per 100,000 10 12 45 increase in the risk of hepatocellular carcinoma42
live births) and only a small increase in the risk of hepatocellu-
No. of excess cases of venous thromboembolism lar adenoma among users of lower-dose oral con-
attributable to oral-contraceptive use traceptives.43
(per 100,000 woman-yr of use)
With norethindrone, norethindrone acetate, 6 9 12
levonorgestrel, or ethynodiol diacetate areas of uncertainty
With desogestrel or gestodene 16 23 30
particular progestins and vascular disease
* Low estrogen was defined as less than 50 µg.
Definitive data are lacking regarding the vascular
† Data are from Farley et al.24 risks associated with the newer progestins, cypro-
terone and drospirenone. Drospirenone is derived
from spirolactone and lacks androgenic activity in
ethynodiol diacetate, levonorgestrel, or norges- cell systems. Although there are theoretical advan-
trel. Two recent meta-analyses33,34 both concluded tages to progestins that have less androgenicity, they
that the use of low-estrogen oral contraceptives have not been proved to have practical implications.
containing one of the so-called third-generation The possibility that low-estrogen oral contracep-
progestins, desogestrel or gestodene, increases the tives containing desogestrel or gestodene might be
risk of venous thromboembolism more than low- associated with a lower risk of myocardial infarc-
estrogen formulations containing levonorgestrel tion than those containing levonorgestrel cannot
— by a factor of 1.5 to 1.8. These differences in risk be established or ruled out on the basis of current
appear to be real and cannot be accounted for by evidence.15,16,19,44 The risk of venous thromboem-
methodologic problems in the studies or the analy- bolism in formulations containing norgestimate
ses.35 Table 2 shows the number of excess venous remains uncertain, because in some studies it is con-
thromboembolic events among users of low-estro- sidered to be a second-generation progestin, and in
gen formulations containing desogestrel or gesto- others a third-generation progestin.
dene as compared with users of low-estrogen for-
mulations containing levonorgestrel.24 screening for thrombophilia
The elevation in the risk of venous thromboem- Tests for the factor V Leiden and prothrombin-gene
bolism is greatest in the first year after use is initiat- mutations, which increase the risk of venous throm-
ed, but an elevated risk persists beyond the first boembolism, are not routinely available and are
year.36 Women are not at increased risk for venous costly where they are available. The role and cost
thromboembolism, myocardial infarction, or ische- effectiveness of screening for these and other pro-
mic stroke after they cease taking oral contracep- thrombotic polymorphisms remain uncertain. A
tives.35,37 family history of venous thromboembolism has
Currently used oral contraceptives also increase a poor positive predictive value for thrombophilic
the risk of central retinal venous thrombosis.38,39 gene defects.45 Screening is not routinely recom-
Whether this risk is increased more among users mended before the initiation of oral-contracep-
of formulations containing gestodene or deso- tive use.

1446 n engl j med 349;15 [Link] october 9 , 2003

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 clinical practice

hypertension and other cardiovascular cervical cancer

risk factors Human papillomavirus (HPV) is a causative agent
The risks of myocardial infarction and stroke among in both squamous-cell and adenomatous cervical
users of oral contraceptives who have hypertension cancer,55 yet until recently, HPV infection was not
and take medications for blood-pressure control are taken into account in studies of oral-contraceptive
not known. The risks of myocardial infarction and use and cervical neoplasia. Because of this and oth-
ischemic stroke are 7 to 15 times as high among us- er methodologic problems, there are conflicting
ers of oral contraceptives who have diabetes or hy- data on the association between oral-contraceptive
percholesterolemia as among those who do not use and the risk of cervical cancer.56
have these conditions.16,20 As with hypertension, Recent studies suggest that oral-contraceptive
whether treatment of these conditions or associat- use, especially long-term use, may increase the risk
ed vascular risk factors modifies the risks associat- of cervical cancer in women who are positive for
ed with oral-contraceptive use remains uncertain. HPV DNA but not in women who are negative for
HPV DNA.57 The data are consistent with a promo-
migraine tional effect of oral contraceptives in women with
Some studies have reported an increased risk of HPV. Oral-contraceptive use may cause persistence
ischemic stroke among oral-contraceptive users of infection with the particular types of HPV that
who have a history of migraine headache, as com- confer a high risk of cervical cancer58 and may also
pared with women who do not have such a histo- affect the acquisition of HPV infection, although
ry.46-48 However, the available studies have had these conclusions are still not firm. Any effect of
methodologic limitations, including the possibili- oral-contraceptive use on the risk of cervical neopla-
ties that hemiplegic migraines were misdiagnosed sia will be small if neoplasia is detected early and
as ischemic stroke or that women with ischemic treated.
stroke were more likely to report headaches, in ret-
rospect, as having been migraines. Headaches may other neoplasia
be classified as migraine without meeting the estab- The risk of colorectal cancer may be decreased with
lished criteria for the diagnosis, and self-reported the use of oral contraceptives,59 although studies
migraine without associated neurologic deficits is have had conflicting results. A lower risk of benign
not considered to be a contraindication to the use breast disease among long-term users of oral con-
of oral contraceptives. traceptives was observed in studies of older formu-
lations, but it is uncertain whether such an associ-
breast cancer ation holds true for newer formulations. Whether
Despite at least 60 epidemiologic studies of breast oral-contraceptive use affects the risk of uterine my-
cancer and oral-contraceptive use, the effect of oral- oma is also uncertain.60,61
contraceptive use on the risk of breast cancer re-
mains controversial. A pooled analysis of data from other formulation-specific effects
54 studies, reported in 1996, showed a small in- Numerous studies have evaluated contraceptive ef-
crease in the risk of breast cancer with current use fectiveness, cycle control, bleeding patterns, or mi-
(relative risk, 1.24; 95 percent confidence interval, nor side effects such as weight gain with specific
1.15 to 1.33); the relative risk decreased to 1.0 with- formulations; have compared these effects with
in 10 years after use had been discontinued.49 In those of placebo; or have compared the effects of
contrast, a recent study in the United States that in- two or three of the dozens of different formulations.
volved 4575 women with breast cancer50 showed There is no evidence of differences in contraceptive
no increase in the risk of breast cancer among wom- effectiveness among different formulations, but the
en who were using oral contraceptives at the time equivalence of all currently marketed products has
of the study, women who had ever used oral contra- not been established through rigorous head-to-head
ceptives, or women with long-term use. The possi- comparisons. Studies of cycle control with various
bility that women with a strong family history of formulations have used different measures of bleed-
breast cancer or BRCA1 or BRCA2 gene mutations ing, rendering it impossible to make valid compar-
may be at higher risk for breast cancer than other isons.62 The superiority of any specific formulation
women when using oral contraceptives has not been in reducing the risk of any minor side effect of oral-
ruled out.51-54 contraceptive use has not been established.

n engl j med 349;15 [Link] october 9, 2003 1447

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 The new england journal of medicine

The cost to the patient of a cycle of oral contra-

conclusions
ceptives depends on the formulation, insurance cov- and recommendations
erage, and local pricing practices. It ranges from no
cost to the patient to more than $50 per cycle. For most women who are, like the woman in the
vignette, healthy and free of cardiovascular disease
guidelines and major cardiovascular risk factors, the use of
combination estrogen–progestin oral contracep-
In 2000, the American College of Obstetricians and tives is associated with low relative and absolute
Gynecologists (ACOG) issued practice guidelines risks of cardiovascular disease. Even when the health
regarding hormonal contraception.63 Also in 2000, risks are taken into account, the net health benefit
the World Health Organization (WHO) published of oral-contraceptive use in these women is great,
guidelines on medical eligibility for contraceptive especially given the effect on the risk of ovarian can-
use.64 Table 3 summarizes the ACOG and WHO rec- cer and its effectiveness in preventing pregnancy.
ommendations for the use of combination oral con- The favorable risk–benefit ratio for healthy women
traceptives in women with characteristics that might applies to all oral contraceptives containing a low
put them at particularly high risk for adverse vas- dose of estrogen (less than 50 µg). However, among
cular events and according to personal and family formulations containing less than 50 µg of ethi-
history. The two sets of guidelines are similar. nyl estradiol, those containing desogestrel or gesto-

Table 3. Summary of Guidelines for the Use of Combination Estrogen–Progestin Oral Contraceptives in Women with Characteristics
That Might Increase the Risk of Adverse Effects.*

Variable ACOG Guidelines WHO Guidelines

Smoker, >35 yr of age

<15 cigarettes/day Risk unacceptable Risk usually outweighs benefit
≥15 cigarettes/day Risk unacceptable Risk unacceptable
Hypertension
Blood pressure controlled Risk acceptable; no definition Risk usually outweighs benefit if systolic blood pressure
of blood-pressure control is 140–159 mm Hg and diastolic blood pressure is
90–99 mm Hg
Blood pressure uncontrolled Risk unacceptable; no definition Risk unacceptable if systolic blood pressure is ≥160
of uncontrolled blood pressure mm Hg or diastolic blood pressure is ≥100 mm Hg
History of stroke, ischemic heart disease, Risk unacceptable Risk unacceptable
or venous thromboembolism
Diabetes Risk acceptable if no other cardio- Benefit outweighs risk if no end-organ damage and
vascular risk factors and no diabetes is of ≤20 yr duration
end-organ damage
Hypercholesterolemia Risk acceptable if LDL cholesterol Benefit–risk ratio is dependent on the presence or
<160 mg/dl and no other absence of other cardiovascular risk factors
cardiovascular risk factors
Multiple cardiovascular risk factors Not addressed Risk usually outweighs benefit or risk unacceptable,
depending on risk factors
Migraine headache
Age ≥35 yr Risk usually outweighs benefit Risk usually outweighs benefit
Focal symptoms Risk unacceptable Risk unacceptable
Breast cancer
Current disease Risk unacceptable Risk unacceptable
Past disease, no active disease for 5 yr Risk unacceptable Risk usually outweighs benefit
Family history of breast or ovarian cancer Risk acceptable Risk acceptable

* The American College of Obstetricians and Gynecologists (ACOG) guidelines recommend the use of formulations containing less than 50 µg
of ethinyl estradiol with the “lowest progestin dose,” without mention of the type of progestin. The World Health Organization (WHO) guide-
lines pertain explicitly to formulations containing 35 µg or less of ethinyl estradiol and do not mention the dose or type of progestin. To con-
vert values for low-density lipoprotein (LDL) cholesterol to millimoles per liter, multiply by 0.02586.

1448 n engl j med 349;15 [Link] october 9 , 2003

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 clinical practice

dene are associated with an increase by a factor of when using oral contraceptives. Smoking cessation
1.5 to 1.8 in the risk of venous thromboembolism as should be encouraged routinely, and the clinician
compared with formulations containing norethin- should document that older women who choose to
drone, norethindrone acetate, ethynodiol diacetate, use oral contraceptives while continuing to smoke
or levonorgestrel. Initiating oral-contraceptive use have been counseled about the vascular risks as-
with the low-estrogen oral contraceptive that is the sociated with oral-contraceptive use. All oral-con-
least costly for the patient is a reasonable clinical traceptive users should have their blood pressure
practice. checked before initiating use and periodically
The risks associated with oral-contraceptive use thereafter. Women with well-controlled hyperten-
outweigh the benefits for women with a history of sion who elect to use oral contraceptives should be
stroke, ischemic heart disease, or venous thrombo- counseled that it is uncertain whether blood-pres-
embolism. Oral contraceptives are contraindicated sure control eliminates the associated increases in
in women who are known to carry the factor V Leiden the risks of stroke and myocardial infarction. Oral-
gene mutation, although screening for this condi- contraceptive use should be discontinued immedi-
tion is not recommended. Oral-contraceptive use ately in any woman with symptoms suggestive of
should be discouraged among women older than stroke, myocardial infarction, or venous thrombo-
35 years of age who smoke (especially if they smoke sis. Oral-contraceptive users should be screened
more than 15 cigarettes per day) because they clearly regularly for cervical neoplasia but do not require
have an increased risk of arterial vascular disease more frequent screening than nonusers.

references
1. Fu H, Darroch JE, Haas T, Ranjit N. Con- oral contraceptives: criteria for pill selec- cular Disease and Steroid Hormone Contra-
traceptive failure rates: new estimates from tion. Cutis 2000;66:281-6. ception. Acute myocardial infarction and
the 1995 National Survey of Family Growth. 11. Davis A, Godwin A, Lippman J, Olson combined oral contraceptives: results of an
Fam Plann Perspect 1999;31:56-63. W, Kafrissen M. Triphasic norgestimate- international multicentre case-control study.
2. Ness RB, Grisso JA, Klapper J, et al. Risk ethinyl estradiol for treating dysfunctional Lancet 1997;349:1202-9.
of ovarian cancer in relation to estrogen and uterine bleeding. Obstet Gynecol 2000;96: 20. Kemmeren JM, Tanis BC, Van Den Bosch
progestin dose and use characteristics of oral 913-20. MA, et al. Risk of Arterial Thrombosis in Re-
contraceptives. Am J Epidemiol 2000;152: 12. Milman N, Clausen J, Byg KE. Iron status lation to Oral Contraceptives (RATIO) study:
233-41. in 268 Danish women aged 18-30 years: in- oral contraceptives and the risk of ischemic
3. Royar J, Becher H, Chang-Claude J. Low- fluence of menstruation, contraceptive meth- stroke. Stroke 2002;33:1202-8.
dose oral contraceptives: protective effect on od, and iron supplementation. Ann Hematol 21. Heinemann LA, Lewis MA, Spitzer WO,
ovarian cancer risk. Int J Cancer 2001;95: 1998;77:13-9. Thorogood M, Guggenmoos-Holzmann I,
370-4. 13. Rivera R, Almonte H, Arreola M, et al. Bruppacher R. Thromboembolic stroke in
4. Walker GR, Schlesselman JJ, Ness RB. The effects of three different regimens of young women: a European case-control study
Family history of cancer, oral contraceptive oral contraceptives and three different intrau- on oral contraceptives. Contraception 1998;
use, and ovarian cancer risk. Am J Obstet Gy- terine devices on the levels of hemoglobin, 57:29-37.
necol 2002;186:8-14. serum iron and iron binding capacity in ane- 22. Schwartz SM, Petitti DB, Siscovick DS,
5. Modan B, Hartge P, Hirsh-Yechezkel G, mic women. Contraception 1983;27:311-27. et al. Stroke and use of low-dose oral contra-
et al. Parity, oral contraceptives, and the risk 14. Effects of contraceptives on hemoglobin ceptives in young women: a pooled analysis
of ovarian cancer among carriers and non- and ferritin. Contraception 1998;58:262-73. of two US studies. Stroke 1998;29:2277-84.
carriers of a BRCA1 or BRCA2 mutation. 15. Dunn N, Thorogood M, Faragher B, et al. 23. WHO Collaborative Study of Cardiovas-
N Engl J Med 2001;345:235-40. Oral contraceptives and myocardial infarc- cular Disease and Steroid Hormone Contra-
6. Narod SA, Risch H, Moslehi R, et al. tion: results of the MICA case-control study. ception. Ischaemic stroke and combined oral
Oral contraceptives and the risk of heredi- BMJ 1999;318:1579-83. contraceptives: results of an international,
tary ovarian cancer. N Engl J Med 1998;339: 16. Tanis BC, van den Bosch MAAJ, Kemme- multicentre, case-control study. Lancet 1996;
424-8. ren JM, et al. Oral contraceptives and the risk 348:498-505.
7. Burkman RT. Oral contraceptives: cur- of myocardial infarction. N Engl J Med 2001; 24. Farley TM, Collins J, Schlesselman JJ.
rent status. Clin Obstet Gynecol 2001;44: 345:1787-93. Hormonal contraception and risk of cardio-
62-72. 17. Lewis MA, Heinemann LA, Spitzer WO, vascular disease: an international perspec-
8. Olson WH, Lippman JS, Robisch DM. MacRae KD, Bruppacher R. The use of oral tive. Contraception 1998;57:211-30.
The duration of response to norgestimate contraceptives and the occurrence of acute 25. Chang J, Elam-Evans LD, Berg CJ, et al.
and ethinyl estradiol in the treatment of acne myocardial infarction in young women: re- Pregnancy-related mortality surveillance —
vulgaris. Int J Fertil Womens Med 1998;43: sults from the Transnational Study on Oral United States, 1991–1999. MMWR Surveill
286-90. Contraceptives and the Health of Young Summ 2003;52:1-8.
9. Thorneycroft IH, Stanczyk FZ, Bradshaw Women. Contraception 1997;56:129-40. 26. Farmer RD, Lawrenson RA, Thompson
KD, Ballagh SA, Nichols M, Weber ME. Ef- 18. Sidney S, Siscovick DS, Petitti DB, et al. CR, Kennedy JG, Hambleton IR. Population-
fect of low-dose oral contraceptives on an- Myocardial infarction and use of low-dose based study of risk of venous thromboem-
drogenic markers and acne. Contraception oral contraceptives: a pooled analysis of 2 US bolism associated with various oral contra-
1999;60:255-62. studies. Circulation 1998;98:1058-63. ceptives. Lancet 1997;349:83-8.
10. Koulianos GT. Treatment of acne with 19. WHO Collaborative Study of Cardiovas- 27. Bloemenkamp KW, Rosendaal FR, Buller

n engl j med 349;15 [Link] october 9, 2003 1449

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.
 clinical practice

HR, Helmerhorst FM, Colly LP, Vanden- vein thrombosis in carriers of a prothrom- 51. Grabrick DM, Hartmann LC, Cerhan JR,
broucke JP. Risk of venous thrombosis with bin-gene mutation and in users of oral con- et al. Risk of breast cancer with oral contra-
use of current low-dose oral contraceptives traceptives. N Engl J Med 1998;338:1793-7. ceptive use in women with a family history of
is not explained by diagnostic suspicion and 39. de Bruijn SF, Stam J, Koopman MM, Van- breast cancer. JAMA 2000;284:1791-8.
referral bias. Arch Intern Med 1999;159:65- denbroucke JP. Case-control study of risk of 52. Heimdal K, Skovlund E, Moller P. Oral
70. cerebral sinus thrombosis in oral contracep- contraceptives and risk of familial breast
28. Lidegaard O, Edstrom B, Kreiner S. Oral tive users and in carriers of hereditary pro- cancer. Cancer Detect Prev 2002;26:23-7.
contraceptives and venous thromboembo- thrombotic conditions. BMJ 1998;316:589- 53. Ursin G, Henderson BE, Haile RW, et al.
lism: a case-control study. Contraception 92. [Erratum, BMJ 1998;316:822.] Does oral contraceptive use increase the risk
1998;57:291-301. 40. de Bruijn SF, Stam J, Vandenbroucke of breast cancer in women with BRCA1/
29. Vandenbroucke JP, Koster T, Briet E, JP. Increased risk of cerebral venous sinus BRCA2 mutations more than in other wom-
Reitsma PH, Bertina RM, Rosendaal FR. In- thrombosis with third-generation oral con- en? Cancer Res 1997;57:3678-81.
creased risk of venous thrombosis in oral-con- traceptives. Lancet 1998;351:1404. 54. Narod SA, Dube MP, Klijn J, et al. Oral
traceptive users who are carriers of factor V 41. Martinelli I, Taioli E, Palli D, Mannucci contraceptives and the risk of breast cancer
Leiden mutation. Lancet 1994;344:1453-7. PM. Risk of cerebral vein thrombosis and in BRCA1 and BRCA2 mutation carriers.
30. Parkin L, Skegg DC, Wilson M, Herbi- oral contraceptives. Lancet 1998;352:326. J Natl Cancer Inst 2002;94:1773-9.
son GP, Paul C. Oral contraceptives and fa- 42. Oral contraceptives and liver cancer: re- 55. Ngelangel C, Munoz N, Bosch FX, et al.
tal pulmonary embolism. Lancet 2000;355: sults of the Multicentre International Liver Causes of cervical cancer in the Philippines:
2133-4. Tumor Study (MILTS). Contraception 1997; a case-control study. J Natl Cancer Inst 1998;
31. Spitzer WO, Lewis MA, Heinemann LA, 56:275-84. 90:43-9.
Thorogood M, MacRae KD. Third genera- 43. Heinemann LA, Weimann A, Gerken G, 56. Delgado-Rodriguez M, Sillero-Arenas M,
tion oral contraceptives and risk of venous Thiel C, Schlaud M, DoMinh T. Modern oral Martin-Moreno JM, Galvez-Vargas R. Oral
thromboembolic disorders: an internation- contraceptive use and benign liver tumors: contraceptives and cancer of the cervix uteri:
al case-control study. BMJ 1996;312:83- the German Benign Liver Tumor Case-Con- a meta-analysis. Acta Obstet Gynecol Scand
8. trol Study. Eur J Contracept Reprod Health 1992;71:368-76.
32. World Health Organization Collaborative Care 1998;3:194-200. 57. Moreno V, Bosch FX, Munoz N, et al. Ef-
Study of Cardiovascular Disease and Steroid 44. Lewis MA, Spitzer WO, Heinemann LA, fect of oral contraceptives on risk of cervical
Hormone Contraception. Venous throm- MacRae KD, Bruppacher R, Thorogood M. cancer in women with human papillomavi-
boembolic disease and combined oral con- Third generation oral contraceptives and risk rus infection: the IARC multicentric case-
traceptives: results of international multi- of myocardial infarction: an international control study. Lancet 2002;359:1085-92.
centre case-control study. Lancet 1995;346: case-control study. BMJ 1996;312:88-90. 58. Thomas DB, Ray RM, Koetsawang A, et
1575-82. 45. Cosmi B, Legnani C, Bernardi F, Coccheri al. Human papillomaviruses and cervical can-
33. Hennessy S, Berlin JA, Kinman JL, Marg- S, Palareti G. Role of family history in identi- cer in Bangkok. I. Risk factors for invasive
olis DJ, Marcus SM, Strom BL. Risk of venous fying women with thrombophilia and high- cervical carcinomas with human papilloma-
thromboembolism from oral contraceptives er risk of venous thromboembolism during virus types 16 and 18 DNA. Am J Epidemiol
containing gestodene and desogestrel ver- oral contraception. Arch Intern Med 2003; 2001;153:723-31.
sus levonorgestrel: a meta-analysis and for- 163:1105-9. 59. Fernandez E, La Vecchia C, Franceschi S,
mal sensitivity analysis. Contraception 2001; 46. Chang CL, Donaghy M, Poulter N. Mi- et al. Oral contraceptive use and risk of colo-
64:125-33. graine and stroke in young women: case- rectal cancer. Epidemiology 1998;9:295-300.
34. Kemmeren JM, Algra A, Grobbee DE. control study: the World Health Organisa- 60. Chiaffarino F, Parazzini F, La Vecchia C,
Third generation oral contraceptives and risk tion Collaborative Study of Cardiovascular Marsico S, Surace M, Ricci E. Use of oral con-
of venous thrombosis: meta-analysis. BMJ Disease and Steroid Hormone Contracep- traceptives and uterine fibroids: results from
2001;323:131-4. tion. BMJ 1999;318:13-8. a case-control study. Br J Obstet Gynaecol
35. Hannaford P. Cardiovascular events as- 47. Tzourio C, Kittner SJ, Bousser MG, Alp- 1999;106:857-60.
sociated with different combined oral con- erovitch A. Migraine and stroke in young 61. Samadi AR, Lee NC, Flanders WD, Bor-
traceptives: a review of current data. Drug women. Cephalalgia 2000;20:190-9. ing JR III, Parris EB. Risk factors for self re-
Saf 2000;22:361-71. 48. Tzourio C, Tehindrazanarivelo A, Igle- ported uterine fibroids: a case-control study.
36. Bloemenkamp KW, Rosendaal FR, sias S, et al. Case-control study of migraine Am J Public Health 1996;86:858-62.
Helmerhorst FM, Vandenbroucke JP. High- and risk of ischaemic stroke in young wom- 62. Thorneycroft IH. Cycle control with oral
er risk of venous thrombosis during early use en. BMJ 1995;310:830-3. contraceptives: a review of the literature. Am
of oral contraceptives in women with inher- 49. Collaborative Group on Hormonal Fac- J Obstet Gynecol 1999;180:280-7.
ited clotting defects. Arch Intern Med 2000; tors in Breast Cancer. Breast cancer and hor- 63. ACOG Committee on Practice Bulletins-
160:49-52. monal contraceptives: collaborative reanaly- Gynecology. The use of hormonal contra-
37. Stampfer MJ, Willett WC, Colditz GA, sis of individual data on 53 297 women with ception in women with coexisting medical
Speizer FE, Hennekens CH. Past use of oral breast cancer and 100 239 women without conditions: number 18, July 2000. Int J Gy-
contraceptives and cardiovascular disease: a breast cancer from 54 epidemiological stud- naecol Obstet 2001;75:93-106.
meta-analysis in the context of the Nurses’ ies. Lancet 1996;347:1713-27. 64. Improving access to quality care in fam-
Health Study. Am J Obstet Gynecol 1990;163: 50. Marchbanks PA, McDonald JA, Wilson ily planning: medical eligibility criteria for
285-91. HG, et al. Oral contraceptives and the risk of contraceptive use. 2nd ed. Geneva: World
38. Martinelli I, Sacchi E, Landi G, Taioli E, breast cancer. N Engl J Med 2002;346:2025- Health Organization, 2000:1-11.
Duca F, Mannucci PM. High risk of cerebral- 32. Copyright © 2003 Massachusetts Medical Society.

1450 n engl j med 349;15 [Link] october 9 , 2003

Downloaded from [Link] on February 23, 2004. This article is being provided free of charge for use in Thailand:NEJM Sponsored.
Copyright © 2003 Massachusetts Medical Society. All rights reserved.