15.11 — Contraception

Two numbers are quoted for every contraceptive method, and confusing them is the single commonest source of poor decisions in this area.

Perfect use — the failure rate when the method is used exactly as intended, every time.

Typical use — the failure rate in real life, including forgotten pills, late injections and misjudged timing.

For some methods the two numbers are almost identical. For others they differ by a factor of ten, and that gap — not the perfect-use figure — is what should drive the choice.

The failure rates

Pregnancies per 100 women in the first year of use.

Method	Perfect use	Typical use
Implant	0.05	0.05
Vasectomy	0.1	0.15
Hormonal IUD	0.2	0.2
Copper IUD	0.6	0.8
Female sterilisation	0.5	0.5
Injection	0.2	4
Combined pill	0.3	7
Progestogen-only pill	0.3	7
Patch / ring	0.3	7
Male condom	2	13
Diaphragm	6	17
Fertility awareness	0.4–5	2–23
Withdrawal	4	20
No method	85	85

Read the first four rows. Implants, intrauterine devices and sterilisation have perfect and typical use rates that are the same, because there is nothing for the user to do. These are the long-acting reversible contraceptives (LARC), and they are 20 to 100 times more effective in practice than the pill.

The pill's typical failure rate of around 7 percent per year is far higher than most users believe. Over five years that is a substantial cumulative probability.

Which is why LARC methods are recommended first in most guidelines — not because they suit everyone, but because the gap between theory and practice is where unintended pregnancies actually come from.

Hormonal methods

Combined hormonal contraception — pill, patch, ring

Oestrogen plus progestogen.

Three mechanisms, in order of importance:

Suppression of ovulation — the main one. Steady hormone levels suppress GnRH pulsing, so there is no FSH rise, no follicle selection, and no LH surge (Chapter 15.4).

Thickened cervical mucus, preventing sperm passage.

Thinned endometrium, reducing implantation.

Benefits beyond contraception, and they are substantial:

Lighter, less painful, predictable periods.Reduced acne.Reduced ovarian cyst formation.And a large reduction in ovarian and endometrial cancer risk — around 30 to 50 percent, persisting for decades after stopping. This is one of the most significant and least-known benefits of any medication, and it follows directly from reducing the number of ovulations (Chapter 15.3).

Reduced colorectal cancer risk.

Risks:

Venous thromboembolism — the main one, and it needs stating with actual numbers rather than adjectives.

Baseline risk in a young woman: about 2 per 10,000 per year. On the combined pill: about 5 to 12 per 10,000 per year. In pregnancy: about 29 per 10,000. In the six weeks after delivery: about 300 to 400 per 10,000.

So the pill roughly triples a small risk, and pregnancy raises it far more. The comparison that matters for a woman choosing contraception is the pill against pregnancy, not the pill against nothing.

Risk is higher with older age, smoking, obesity, immobility, and thrombophilia.

Cardiovascular — a small increase in stroke and heart attack, concentrated almost entirely in women who smoke and are over 35, which is why the combined pill is not prescribed in that group.

Breast cancer — a small increase in current users, returning to baseline within about 10 years of stopping. In absolute terms this is a small number of extra cases, and it is set against the substantial reductions in ovarian and endometrial cancer.

Migraine with aura is an absolute contraindication, because the stroke risk is meaningfully increased. Migraine without aura is not. Distinguishing them matters, and aura means the visual or sensory disturbance preceding the headache.

And the traditional pill-free week is not physiologically necessary. The bleed on the pill is a withdrawal bleed, not a period — there is no endometrium built up in the usual way and no ovulation occurred.

It was included in the original design to make the method more acceptable, on the reasoning that a monthly bleed would seem more natural. Continuous or extended use is safe and is now explicitly offered, and it reduces the number of hormone-free intervals during which ovulation can escape — which is where a proportion of failures occur.

Progestogen-only pill

Two types, and they work differently.

Traditional (levonorgestrel, norethisterone) — works mainly by thickening cervical mucus. Must be taken within a 3-hour window each day, which is a demanding requirement.

Desogestrel — suppresses ovulation in most cycles, and has a 12-hour window. This type is now available without prescription in several countries.

Suitable when oestrogen is contraindicated — smokers over 35, migraine with aura, breastfeeding, high clot risk.

Irregular bleeding is the commonest reason for stopping, and it should be discussed in advance: around 20 percent have no bleeding, 40 percent have regular bleeding, and 40 percent have irregular bleeding.

Injection

Depot medroxyprogesterone acetate, every 12 to 13 weeks.

Highly effective in perfect use, and the typical-use rate of around 4 percent reflects late injections.

Advantages: private, nothing to remember daily, and most users stop having periods entirely after a year, which many regard as a benefit.

And two specific disadvantages that must be discussed.

Reduced bone mineral density during use, which recovers after stopping. It matters most in adolescents, who are still accruing peak bone mass (Chapter 5.1), and in women approaching menopause.

Delayed return of fertility — a median of around 9 months after the last injection, and occasionally up to 18. This is unique among reversible methods, and it is the single most important thing to tell someone who may want to conceive within a year or two.

Implant

A single flexible rod placed under the skin of the upper arm, lasting 3 years.

The most effective reversible method available — failure below 0.1 percent.

Works by ovulation suppression plus mucus thickening.

Fertility returns immediately on removal.

Irregular bleeding is the main drawback and the main reason for early removal, and counselling about it beforehand substantially improves continuation rates.

Intrauterine methods

Copper IUD

Effective for 5 to 10 years depending on the device.

No hormones at all.

Mechanism: copper ions are toxic to sperm and impair their motility, and they produce a sterile inflammatory reaction in the endometrium that is hostile to both sperm and fertilisation.

And it is worth being accurate here, because the mechanism is politically contested. The primary action is preventing fertilisation, not preventing implantation. Studies examining fertilised eggs in the tubes of IUD users find far fewer than expected, indicating that fertilisation is largely prevented. A post-fertilisation effect probably contributes when used as emergency contraception, which is the honest position.

Advantages: hormone-free, immediately reversible, long-lasting, and the most effective form of emergency contraception (below).

Disadvantage: periods are typically heavier and more painful, particularly in the first 3 to 6 months. In a woman who already has heavy periods, this is the wrong choice.

Hormonal IUD (intrauterine system)

Releases levonorgestrel locally, lasting 5 to 8 years depending on the device.

Mechanism is predominantly local — thickened mucus, thinned endometrium — with ovulation suppressed in only a minority of cycles.

Which is why systemic side effects are minimal: the blood level is a fraction of that from an oral progestogen.

And its effect on bleeding is the opposite of the copper device. It reduces menstrual blood loss by around 90 percent, and around 20 percent of users have no periods at all after a year.

Which makes it a treatment as well as a contraceptive. It is first-line for heavy menstrual bleeding (Chapter 15.4), it is used for endometrial protection with HRT, and it is used in endometriosis and adenomyosis.

Facts about IUDs worth correcting

They can be used by women who have never been pregnant. The old restriction was based on outdated devices and has been abandoned.

They do not cause infertility. The association with pelvic inflammatory disease traced almost entirely to one poorly designed device withdrawn in the 1970s. Modern devices carry a small increased infection risk confined to the first 3 weeks after insertion, from organisms introduced at the time.

Insertion is uncomfortable and the pain has been systematically underestimated. This is a genuine and recently acknowledged failing — women reporting severe pain were frequently dismissed, and guidance has changed to require that pain relief options are discussed in advance.

Expulsion occurs in around 3 to 5 percent, most in the first year, and users are taught to check the threads.

If pregnancy occurs with an IUD in place, ectopic pregnancy must be excluded. The device does not cause ectopic pregnancy — it reduces the absolute number — but it prevents uterine pregnancy far more effectively than tubal, so a higher proportion of the few pregnancies that occur are ectopic (Chapter 4.3).

Barrier methods

Male condoms — and their unique advantage is that they are the only method that also protects against sexually transmitted infection.

Typical failure of around 13 percent reflects inconsistent use, incorrect use, and breakage.

Correct use matters and is rarely taught: check the expiry date, open carefully without teeth or nails, squeeze the tip to expel air, apply before any genital contact, use water-based or silicone lubricant only — oil-based products destroy latex within minutes — and withdraw while still erect, holding the base.

Female condoms — user-controlled, can be inserted in advance, and also protect against infection. Less widely available and more expensive.

Diaphragm and cap — used with spermicide, and considerably less effective. Must be left in for 6 hours afterwards.

Spermicide alone — poor efficacy, and nonoxynol-9 with frequent use can irritate the vaginal and rectal lining and increase HIV transmission risk, which is why it is not recommended for people at risk.

Fertility awareness

Identifying the fertile window and avoiding intercourse or using a barrier during it (Chapter 15.4).

And the enormous range in the table — 0.4 to 23 percent — reflects the difference between methods and between users.

Symptothermal methods, combining cervical mucus with basal body temperature, taught properly and used consistently, achieve perfect-use rates around 0.4 to 1 percent. Calendar-only methods do not.

The requirements are real: daily observation and recording, regular cycles, a cooperative partner, and acceptance of a fertile window of about 8 to 10 days requiring abstinence or barriers.

And a caution about apps. Most predict from calendar data using an assumed 14-day pattern, which is exactly the assumption that fails in variable cycles. Only apps that incorporate temperature or mucus data and have been formally evaluated should be relied on for contraception, and one such app is regulated as a medical device.

Lactational amenorrhoea is a genuine method with three strict conditions, all of which must hold: the baby is under 6 months, is exclusively breastfed with no long gaps including at night, and periods have not returned. Under those conditions it is around 98 percent effective. Break any one and it is not.

Emergency contraception

Three options, and they are not equivalent.

Copper IUD — the most effective by a wide margin, over 99 percent, up to 5 days after intercourse or up to 5 days after the earliest estimated ovulation.

And it has the advantage of continuing as ongoing contraception. It is substantially underused because it requires a clinic appointment, while pills are available immediately.

Ulipristal acetate — up to 5 days. More effective than levonorgestrel, particularly closer to ovulation.

Levonorgestrel — up to 3 days, and most effective the sooner it is taken.

Both pills work primarily by delaying or preventing ovulation. Neither is effective once ovulation has occurred, and neither disrupts an established pregnancy — a point worth stating clearly because it is the source of persistent confusion. They are not abortifacients.

Two practical points that reduce failures:

Body weight affects efficacy. Levonorgestrel is less effective above around 70 kg and ulipristal above around 85 kg, and the copper IUD is unaffected — which is a strong argument for offering it.

And ulipristal and hormonal contraception interact. Starting or restarting hormonal contraception within 5 days reduces ulipristal's effect, so hormonal contraception is delayed 5 days and barriers used. With levonorgestrel, hormonal contraception can be started immediately.

Permanent methods

Vasectomy — Chapter 15.2. Simpler, safer, cheaper and more effective than female sterilisation, and considerably less used, which is a pattern with social rather than medical explanations.

Female sterilisation — occluding or removing the fallopian tubes.

And practice has changed: complete removal of the tubes (salpingectomy) is now preferred over clipping, because it is more effective and because a substantial proportion of ovarian cancers appear to originate in the fallopian tube, so removal reduces that risk.

Both are counselled as permanent. Regret is commonest in those sterilised young, and reversal is difficult and often unsuccessful.

Choosing

The questions that actually determine the answer:

Do you want children in the future, and when? — which rules out sterilisation and makes the injection a poor choice if within a year.

How reliably would you use a daily or weekly method? — the honest answer to this is the most important single input.

Do you need protection from infection? — condoms, alone or in addition.

Do you have heavy or painful periods? — hormonal IUD.

Do you want to avoid hormones? — copper IUD, barriers, fertility awareness.

Are there medical contraindications? — migraine with aura, smoking over 35, clot history, breast cancer, liver disease.

And the criteria for those are formalised. The WHO Medical Eligibility Criteria and its national equivalents categorise every method against every condition from 1 (no restriction) to 4 (unacceptable risk), and they are freely available and used routinely.

Contraception after childbirth: progestogen-only methods can be started immediately, including while breastfeeding. Combined methods are avoided in the first 6 weeks because of clot risk, and until 6 months if breastfeeding is being established. An IUD can be inserted within 48 hours of delivery or after 4 weeks.

And the timing matters: ovulation can return as early as 3 to 4 weeks in a woman who is not breastfeeding (Chapter 15.7).

What the next page fixes

Chapter 15.12 covers the opposite problem — infertility, what causes it, what is actually done about it, and how well assisted reproduction works, stated with real numbers rather than clinic marketing.