CONTRACEPTION

Combined hormonal
contraception

Whats new ?
Transdermal, vaginal and injectable methods
containing both estrogen and progestogen are now
available

Anna Glasier

Amenorrhoea is becoming more acceptable among

women using hormonal contraception
WHO medical eligibility criteria are replacing the
concept of relative and absolute contraindications to
use of all contraceptive methods
The combined oral contraceptive pill (COCP) was introduced in
the early 1960s and is currently used by more than 60 million
women worldwide. In the UK, 25% of women using a reversible
method of contraception take an oral contraceptive pill. The COCP
is popular because of its efficacy, ease of use and additional health
benefits beyond those of contraception (see page 23). In recent
years, new delivery systems for combined hormonal contraception
have become available. The contraceptive patch Evra was marketed in the UK in 2004. The combined contraceptive vaginal ring
Nuvaring is available in much of Europe and in the USA, and may
be licensed in the UK in 2006. There are no plans to market the
combined injection Lunelle in the UK. The mechanism of action
and efficacy of these relatively new delivery systems are similar
to those of the COCP. Few data are available on the long-term
risks, but it is sensible to assume that they too are no different
from the COCP.

The risk of cervical cancer is increased in women using

the combined pill, particularly if they have persistent
human papillomavirus infection

first-generation progestogens (norethindrone)

second-generation norgestrel derivatives (e.g. levonorgestrel)
third-generation progestogens (e.g. gestodene, desogestrel,
norgestimate)
the newest progestogen, drospirenone (has anti-androgenic and
antimineralocorticoid activity; the term fourth generation is
avoided by the manufacturer).
Different progestogens differ in potency and thus contraceptive
effectiveness is achieved at different doses. At equivalent contraceptive efficacies, the progestogens are said to have slightly different side-effects, but the evidence is unconvincing, except possibly
for side-effects clearly associated with relative differences in androgenicity (e.g. acne). The anti-androgenic properties of drospirenone
may make it particularly beneficial for acne and hirsutism.

Available preparations
Combined hormonal contraceptives contain both estrogen (usually
ethinylestradiol) and a progestogen (a synthetic compound that
behaves like progesterone).

Formulations: classically, the COCP is taken for 21 days followed by

a 7-day break (the pill-free interval, PFI), when withdrawal bleeding usually occurs. Combined pills are available in monophasic
preparations (in which every pill in the packet contains the same
dose of steroids), and biphasic and triphasic preparations (in which
the dose of both estrogen and progestogen changes once or twice
over the 21-day period). Biphasic and triphasic pills were introduced to reduce the total dose of progestogens, and in the belief
that a regimen that mimicked the normal cycle would produce
better cycle control. However, there is no evidence for better cycle
control, and some women find such preparations confusing, particularly when they want to take two packets of pills consecutively
to postpone menstruation. In an attempt to improve compliance,
everyday preparations are used widely in the USA and Australia.
These regimens involve the taking of inactive tablets, rather than
7 days without pills.
In the USA, an 84-day packet of pills (followed by a 7-day PFI)
is marketed as Seasonale. Only four withdrawal bleeds per year
is a pattern increasingly popular with many European women
and can be achieved simply by taking 21-day packets of pills
consecutively.

Combined oral contraceptive pill

Estrogen: the dose of estrogen used in the COCP is 1550 g; most
women now use so-called low-dose pills containing 3035 g.
Low-dose pills are potentially safer, because the cardiovascular
risks of the COCP result mainly from estrogen. However, the lower
the dose of estrogen, the higher the risk of poor cycle control,
breakthrough bleeding and, at least theoretically, pregnancy if
compliance is poor. This is reflected in the new rules for missed
pills, which distinguish between pills containing less than or more
than 30 g of ethinylestradiol (Figure 1).
Progestogens used in currently available COCPs are divided into
four groups:

Anna Glasier is Honorary Professor in the Department of Obstetrics and

Gynaecology at the University of Edinburgh, UK and in the Department
of Public Health and Policy at the University of London School of Hygiene
and Tropical Medicine, London, and Director of Family Planning and
Well Woman Services for NHS Lothian. She qualified from the University
of Bristol, and trained in obstetrics and gynaecology, specializing
in reproductive medicine. Her research interests are contraception,
including emergency contraception, and reproductive health care.
Conflicts of interest: none declared.

MEDICINE 34:1

Combined contraceptive patch

Only one contraceptive patch is currently available. Evra is
20 cm2 in area and delivers 20 g of ethinylestradiol and 150 g
of norelgestromin (17-deacetylnorgestimate) per day. Each patch
1

2006 Elsevier Ltd

CONTRACEPTION

Guidance for women who have missed pills

If one or two 3035 g ethinylestradiol pills
have been missed at any time

If three or more 3035 g ethinylestradiol pills have been

missed at any time

or

or

One 20 g ethinylestradiol pill is missed

Two or more 20 g ethinylestradiol pills are missed

She should take the most recent missed pill as soon as

she remembers

She should take the most recent missed pill as soon as she remembers

She should continue taking the remaining pills daily at

her usual time1

She should be advised to use condoms or abstain from sex until she has
taken pills for 7 consecutive days

She does not require additional contraceptive

protection

In addition (because extending the pill-free interval is risky)

She should continue taking the remaining pills daily at her usual time 1

She does not require emergency contraception

If pills are missed in week 1

(days 17) (because the pill-free
interval has been extended)

If pills are missed in week 3

(days 1521) (to avoid extending
the pill-free interval)

Emergency contraception should

be considered if she had
unprotected sex in the pill-free
interval or in week 1

She should finish the pills in her

current pack and start a new
pack on the next day, thereby
omitting the pill-free interval

1Depending

on when she remembers her missed pill, she may take two pills on the same day (one at the moment of remembering and the other at the regular time)
or even at the same time

Bleeding episodes occur 1822 days after injection when estrogen

concentrations decline to 50 pg/ml or less. About 70% of women
experience one bleeding episode per month.

lasts 7 days; three patches are used consecutively, followed by a

placebo patch or patch-free interval in week four, when withdrawal
bleeding occurs. Contraceptive protection lasts for up to 10 days,
allowing for errors in changing the patch. In a randomized trial
comparing the patch with a COCP, there was no significant difference in effectiveness; overall pearl indices were 1.24/100 womenyears for the patch and 2.18/100 women-years for the COCP. The
efficacy may be reduced in heavier women. Bleeding patterns and
side-effects are similar to those associated with the COCP.

Mode of action
Combined hormonal contraception acts mainly by inhibiting
ovulation. The estrogen component inhibits secretion of pituitary
follicle-stimulating hormone, thereby suppressing the development of ovarian follicles; progestogen inhibits the development
of the luteinizing hormone surge. Other mechanisms of action of
combined hormonal contraception are shown in Figure 2.
In some women, the 7-day hormone-free interval is long enough
to allow follicle growth; on the last day of the PFI, 25% of women
using the COCP exhibit ultrasonographical evidence of follicles of
10 mm in diameter. If the PFI is extended beyond 7 days, these
follicles continue to grow and, despite restarting the pill, ovulation may occur. The same is likely to be true of the hormone-free
interval following patch or ring use. In women who appear to have
conceived as a result of a genuine COCP failure (rather than as a
result of an error in pill-taking) and who wish to continue using
the COCP after the pregnancy is over, the PFI can be shortened to
4 or 5 days to ensure suppression of follicular development. Some
brands of low-dose COCP have a shorter PFI as standard.

Combined contraceptive vaginal ring

Nuvaring releases 15 g of ethinylestradiol and 120 g of
etonorgestrel per day and is now licensed in much of Europe. The
ring is made of soft ethylene-vinyl-acetate co-polymer and measures 54 mm in diameter. It is designed to last for 3 weeks; a 7-day
ring-free interval is associated with bleeding patterns that appear
superior to those associated with the COCP. In all other respects,
including efficacy, the ring is no different from the COCP.
Combined injectable contraceptive
A once-monthly injectable contraceptive containing 25 mg of
medroxyprogesterone acetate and 5 mg of estradiol cypionate
(Lunelle) is available in some parts of the world, but not in the
UK. Injections are administered intramuscularly every 28 days.
MEDICINE 34:1

2006 Elsevier Ltd

CONTRACEPTION

Mechanisms of action of combined oral

contraceptives

WHO medical eligibility criteria category 3 and 4

conditions for combined oral contraception

Category 3 conditions
Breast-feeding 6 weeks to 6 months post-partum
Before 3 weeks after childbirth
Smoking < 15 cigarettes/day and > 35 years of age
Adequate controlled hypertension
Blood pressure > 140/90 mm Hg, < 160/100 mm Hg
Severe hyperlipidaemia
Non-focal migraine and > 35 years of age
Previous breast cancer 5 years without recurrence
Current or medically treated gallbladder disease
Past combined oral contraception-related cholestasis
Mild cirrhosis

Inhibition of ovulation
Changes in cervical mucus that interfere with sperm transport
Tubal motility may be altered
Atrophy of endometrium
Uterine receptivity essential for successful implantation may
be impaired

Because the risks of the COCP are mainly from estrogen, manufacturers have steadily reduced the dose. In recognition of the risk
of compromising contraceptive efficacy, one formulation (Mircette,
currently available only in the USA) comprises 21 days of 20 g
ethinylestradiol in combination with desogestrel, but includes five
daily doses of 10 g ethinylestradiol alone during the pill-free
week to maintain suppression of follicle growth.

Category 3/4 conditions

Multiple risk factors for cardiovascular disease
Diabetes with retinopathy, nephropathy, neuropathy or other
vascular disease, or of > 20 years duration
Category 4 conditions
Breast-feeding before 6 weeks post-partum
Smoking > 15 cigarettes/day and > 35 years of age
Blood pressure > 160/100 mm Hg
Hypertension with vascular disease
History of or current deep vein thrombosis
Major surgery with prolonged immobilization
History of or current myocardial infarction
History of stroke
Complicated valvular heart disease
Focal migraine
Current breast cancer
Severe cirrhosis
Liver tumour

Efficacy
When used correctly, combined hormonal contraceptives are
almost 100% effective. The failure rate during perfect use is
0.1%. In practice, methods of contraception that make demands
of users have a higher failure rate, because mistakes are made
during typical use. Because of errors in pill-taking, the failure
rate associated with typical use of the COCP is about 8%. There
is evidence that use of the patch or ring may be less demanding,
so typical-use failure rates (not yet available for either method)
may be lower. This should be a clear advantage for the combined
injectable contraceptive.

Contraindications
Absolute contraindications (WHOMEC category 4 conditions)
to the COCP are listed in Figure 3. Relative contraindications
(WHOMEC category 3 conditions) include serious or multiple
risk factors for arterial disease, including hypertension, family
history, diabetes mellitus, smoking, age over 35 years, obesity
and migraine (Figure 3).
Women with hyperprolactinaemia who want to avoid pregnancy
should be advised to use progestogen-only contraception (see
page 6), because estrogen stimulates lactotrophs (prolactin-secreting
cells in the pituitary gland), thereby increasing the prolactin
concentration.

system. The patch, for example, should be associated with less

nausea. In women who use the COCP, the dose of estrogen or type
of progestogen can be altered by changing to a different brand
of pill, and it is worth trying this if time alone does not solve
the problem. Chloasma is much less common and is definitely
related to steroid hormones. It can also occur with progestogenonly methods.
Serious side-effects involve mainly the cardiovascular system.
Combined hormonal contraception affects both the venous (venous
thromboembolism (VTE)) and the arterial (myocardial infarction
(MI), cerebrovascular accident) circulation. Although the aetiology
and epidemiology of venous and arterial disease differ, in both
cases the increased risk appears to be related to an increased
thrombotic tendency. Contraceptive steroids are metabolized by
the liver and affect the metabolism of carbohydrates, lipids, plasma
proteins, amino acids, vitamins and clotting factors. Changes in
clotting factors create a tendency to hypercoagulability, which is
partly balanced by an increase in fibrinolysis. The adverse effect
on clotting is related to the dose of estrogen; low-dose pills are

Risks and side-effects

Combined hormonal contraception has an effect on almost every
system in the body. Most side-effects are minor; mood change,
weight gain or fluid retention, nausea and vomiting, headache,
loss of libido, mastalgia, breast enlargement and greasy skin are
common complaints that also occur in the absence of contraceptive
use. Many improve or disappear within 36 months of starting
the method, but side-effects often lead to discontinuation. Some
side-effects may be alleviated by changing to a different delivery

MEDICINE 34:1

2006 Elsevier Ltd

CONTRACEPTION

associated with a reduced risk compared with pills containing

50 g of ethinylestradiol.
In a recent analysis of 25 years follow-up of 46,000 women
who took part in the Royal College of General Practitioners (RCGP)
Oral Contraceptive Study (comparing 517,519 years of COCP use
with 335,998 years of never-use), the risk of death from all causes
was similar in ever-users and never-users of oral contraception.
In current or recent users, however, there was an increase in the
relative risk of death from two conditions cervical cancer (relative risk 2.5) and haemorrhagic stroke (1.9).

The risk of MI was increased in women using second-generation

COCPs, and in smokers (by nine times) and women with
hypertension (tenfold).
The risk of stroke attributable to COCP use is small. The relative risk of haemorrhagic stroke is not increased in women under
35 years of age and is only slightly increased in older women. The
risk of ischaemic stroke is slightly increased (relative risk 1.5) and
is also slightly higher in the over-35s. Smoking and hypertension
increase the risk of stroke by tenfold and threefold, respectively.
Migraine
Women who have migraine with aura may be at increased risk
of stroke. Migraine with aura is a category 4 condition for use of
combined hormonal contraception.

Venous disease
The COCP is associated with a threefold increase in the relative
risk of VTE. Risk is unaffected by age, smoking and duration of
COCP use, but is higher in obese women (body mass index, BMI
> 30 kg/m2) and in women with a history of pregnancy-induced
hypertension. Opinion varies acrimoniously on the validity and
interpretation of several studies published in the mid-1990s showing differing VTE risk with different types of progestogen. On
balance, most experts agree that the difference in risk of VTE is
real but small. COCPs containing the third-generation progestogen
gestodene or desogestrel carry about a twofold increased risk of
VTE compared with pills containing levonorgestrel. Although the
reason for this difference is unclear, it is known that oral contraceptive use reduces the efficiency with which activated protein C
down-regulates in vitro thrombin formation, and this phenomenon
appears to be more pronounced in women using a COCP containing desogestrel rather than levonorgestrel.
The absolute risk of VTE in COCP users is very low
(15/10,000 woman-years) and considerably less than that during
pregnancy (60/10,000 woman-years). The risk returns to normal
within 3 months of stopping the COCP. The Committee on Safety of
Medicines warns that women should be informed of the increased
risk of VTE when third-generation pills are prescribed. It is possible
that COCPs containing anti-androgens may be associated with an
even higher risk of VTE. This certainly appears to be the case for
Dianette, a combination of ethinylestradiol and cyproterone acetate
licensed for treatment of hirsutism and severe acne that acts as
a combined oral contraceptive. There is no reason to believe that
the risk of VTE is any different with other routes of administration
of combined hormonal contraception (patch, ring or injectable).
Women with inherited thrombophilias (e.g. factor V Leiden)
are at increased risk of VTE, and a family history of VTE is an
indication for testing for various thrombophilias. Population-level
screening is considered neither practical nor economical.

Breast cancer
Overviews of the risks of COCPs are dominated by breast cancer.
Data are difficult to interpret because COCP formulations and
patterns of reproduction (particularly age at first pregnancy) have
changed with time. A meta-analysis of 54 studies involving more
than 53,000 women with breast cancer and 100,000 controls concluded that use of the COCP was associated with a small increased
risk of breast cancer that persisted for 10 years after stopping the
COCP. The relative risk was 1.24 in current users, 1.16 14 years
after stopping and 1.07 59 years after stopping. After 10 years,
the relative risk was the same as that in never-users. Although the
relative risk was higher in women who started taking the COCP
at a young age (because breast cancer is uncommon in this age
group), there was little added effect from the duration of use, or
the dose or type of hormone. Women who had ever used the COCP
were significantly less likely (relative risk 0.88) than never-users
to develop cancer that spread beyond the breast, even if they had
stopped the COCP more than 10 years earlier. In the 25-year RCGP
follow-up study, ever-users were not more likely to die from breast
cancer than never-users.
A large case-control study from the USA involving 8000 women
(published after the 1996 meta-analysis) suggested no increase in
breast cancer risk (relative risk 1.0, 95% CI 0.81.3); however, the
upper limit of the confidence intervals is in line with the findings of
the much larger meta-analysis. Again, the association with breast
cancer is probably independent of the route of administration.
The relationship between the COCP and breast cancer is difficult to explain, because the risk appears to increase soon after
exposure, does not increase with duration of use, and returns to
normal after 10 years of no exposure. It has been suggested that
starting the COCP may accelerate the appearance of breast cancer in
susceptible women. It is also possible that tumours are diagnosed
earlier in women who are using the COCP, though it is difficult
to explain why a tendency to earlier diagnosis would persist for
years after stopping. A biological effect of combined hormonal
contraception has not been excluded.

Arterial disease
Arterial disease in COCP users is much more serious than venous
disease, though the absolute risk of MI and stroke in young
women is tiny. There is widespread agreement that the risk of MI
is increased in women who take the COCP and smoke or have
hypertension, but the relationship between MI and COCP use is
controversial in women with no other risk factors.
A recent meta-analysis of 23 studies concluded that the risk
of MI was increased in current COCP users vs never-users (odds
ratio 2.48, 95% CI 1.913.22).
The risk in past-users was not increased, nor was the risk in
women using third-generation COCPs or COCPs containing 20 g
of ethinylestradiol.

MEDICINE 34:1

Cervical cancer
Data on the risk of cervical cancer in COCP users are difficult to
interpret, because barrier methods confer some protection and the
aetiology of cervical cancer is connected with sexual activity. In
the 25-year follow-up study, the relative risk of dying from cervical
cancer was 2.5 in ever-users. A meta-analysis of ten case-control
studies showed an increased risk of cervical cancer in women
4

2006 Elsevier Ltd

CONTRACEPTION

with persistent human papillomavirus infection using hormonal

contraception. The relative risk was 2.8 after 5 years of use of the
COCP and 4.0 after 10 years.

the COCP was started (e.g. polycystic ovary syndrome), but which
were masked by regular artificial withdrawal bleeds.
There is no evidence for any adverse effect on the fetus as a
result of previous COCP use. When conception occurs during COCP
use, the risk of teratogenesis is low or non-existent.

Liver cancer
Benign hepatic adenoma is an uncommon consequence of COCP
use. In countries where hepatocellular carcinoma is rare, this
disease is occasionally associated with the COCP. In populations
where liver cancer is common (e.g. the Far East), short-term use of
COCPs does not affect the incidence of hepatocellular carcinoma;
data on long-term use are scarce.

FURTHER READING
Beral V, Hermon C, Kay C et al. Mortality associated with oral
contraceptive use: 25 year follow up of a cohort of 46,000 women
from Royal College of General Practitioners Oral Contraception Study.
BMJ 1999; 318: 96100.
Etminan M, Takkouche B, Isorna F C et al. Risk of ischaemic stroke
in people with migraine: systematic review and meta-analysis of
observational studies. BMJ 2005; 330: 635.
Faculty of Family Planning and Reproductive Health Care Clinical
Effectiveness Unit. FFPRHC guidance (October 2003) first prescription
of combined oral contraception. J Fam Plann Reprod Health Care
2003; 29: 20923.
Khader Y S, Rice J, John L et al. Oral contraceptive use and risk of
myocardial infarction: a meta-analysis. Contraception 2003; 68:
1117.
Marchbanks P A, McDonald J A, Wilson H G et al. Oral contraceptives and
the risk of breast cancer. N Eng J Med 2002; 346: 202532.
OBrien P A. The third generation oral contraceptive controversy. BMJ
1999; 319: 7956.
Rosing J, Middeldorp S, Curvers J et al. Low-dose oral contraceptives and
acquired resistance to activated protein C: a randomised cross-over
study. Lancet 1999; 354: 203640.
Skegg D C G. Third generation oral contraceptives. BMJ 2000; 321:
1901.
WHO. Improving access to quality care in family planning. Selected
practice recommendations for contraceptive use. Geneva: WHO, 2005.
WHO. Improving access to quality care in family planning. Medical
eligibility criteria for contraceptive use. Geneva: WHO, 2004.
WHO. Cardiovascular disease and steroid hormone contraception. Report
of a scientific group. WHO Tech Rep Ser 1998; 877.
WHO. Oral contraceptives and neoplasia. WHO Tech Rep Ser 1992; 817.

Practical prescribing
History: a full history, including family history, should be taken to
exclude risk factors that might contraindicate combined hormonal
contraception use or indicate further investigations.
Examination: blood pressure should be measured, and it may
be helpful to record baseline weight. BMI of more than 30 kg/m2
is considered a relative contraindication to combined hormonal contraception, and BMI more than 35 kg/m2 is an absolute
contraindication because of the increased risk of VTE.
Pelvic examination is not routinely indicated at the first (or any)
visit unless gynaecological pathology is suspected. Women do not
like pelvic examinations and some, particularly the young, may be
deterred from starting or continuing with the COCP if examination
is seen as a prerequisite. Breast examination is also unnecessary
unless the woman has symptoms of breast disease. Cervical smears
should be taken in accordance with national policy.
Choice of preparation: new users should usually start with a
low-dose (3035 g) pill containing a second-generation progestogen. If breakthrough bleeding occurs and persists beyond
the first 3 months, and a gynaecological cause is excluded, a
COCP containing a higher dose of estrogen or a different type of
progestogen may be tried.
Women taking long-term enzyme-inducing drugs (e.g. some
anticonvulsants) should use a preparation containing 50 g of
estrogen, to ensure best efficacy (this may necessitate taking
two pills each day, one containing 30 g and the other 20 g of
ethinylestradiol, if a 50 g preparation is not available).
The patch should be considered if nausea or vomiting is a persistent side-effect of pill use or if there is a potential problem with
oral absorption. The patch may also be useful in women who want
to use combined hormonal contraception, but who have difficulty
remembering to take a pill every day.
Progestogen-only contraception may be considered in women
with contraindications to the COCP (see page 6).

Practice points
The COCP is extremely safe; mortality in ever-users is no
different from that in never-users 10 years after stopping it
The COCP is a highly effective method of contraception; the
failure rate is only 0.1% when it is used correctly
When used typically, failure rates are about 8%, because
compliance is often poor
A combined contraceptive patch is now available, with mode of
action, efficacy and side-effects similar to those of the COCP;
because each patch lasts for 7 days, compliance may be better
The absolute risk of all serious side-effects, particularly MI and
stroke, is extremely small
Although there is probably a slight increase in the risk of VTE
associated with COCPs containing desogestrel or gestodene,
choice of COCP should be a matter for individual women
guided by their doctor
Current and recent users are at increased risk of diagnosis of
breast cancer, but the risk returns to that seen in never-users
10 years after stopping the pill

Information: women should be carefully instructed how to use

the COCP and what to do when pills are forgotten (Figure 1).
Many women choose (or are advised) to take a break from using
the COCP for a few months. Although most cardiovascular risks
decrease when the pill is stopped, they recur as soon as it is started
again, and unplanned pregnancies commonly occur during such
breaks. Most women who stop the COCP regain normal fertility
within 3 months. Secondary, so-called post-pill amenorrhoea is
almost always the result of abnormalities that were present before

MEDICINE 34:1

2006 Elsevier Ltd