A contraceptive that "triples the risk of a blood clot" and one that "raises the yearly chance of a clot from about 3 to about 9 per 10,000 users" can be the exact same method described from the exact same study. Neither statement is a lie. The first is a relative number, which reports how many times a baseline is multiplied. The second is an absolute number, which reports how many people in a defined group are actually affected. A relative figure printed without its baseline is a multiplier with nothing to multiply, and it is the version most likely to alarm.
This piece walks through two of the best-documented examples behind current contraception guidance, venous thromboembolism (VTE) and breast cancer, to show why the same evidence can feel frightening or reassuring depending on which number reaches the reader.
Key points#
- Relative risk (a multiplier such as "triples") and absolute risk (extra cases per 10,000 or 100,000) describe the same data from different angles.
- A large multiple of a rare event is still a rare event; the absolute change is usually a few extra cases per 10,000 users each year.
- Background risk changes with age, so an identical relative increase produces a small absolute excess at 19 and a larger one at 38.
- Pregnancy itself, the outcome contraception prevents, carries a clot risk several times higher than any combined pill.
- Three questions turn a scary percentage into something you can weigh: what baseline, how large the absolute change, and compared with which alternative.
Two numbers, one dataset#
Start with the vocabulary, because the confusion is almost always vocabulary. Relative risk answers "how many times more likely?" Absolute risk answers "how many people out of this many?" A method can honestly be reported both ways, and the two figures can feel wildly different even though they came from one line in one table.
An analogy helps. If a lottery doubles your odds of winning, the multiplier sounds thrilling, yet two-in-a-hundred-million is still practically zero. Doubling a tiny number leaves a tiny number. The multiplier is real and the near-zero is also real. Risk works the same way, and a responsible reader keeps both in view at once.
The clot example, worked through#
Combined hormonal contraceptives do raise the risk of VTE. That is not in dispute. Meta-analyses summarized in the American Society for Reproductive Medicine guideline place the increase at roughly a threefold to fourfold multiple of baseline, with relative risks near 2.8 for levonorgestrel formulations and 3.2 for norethindrone formulations compared with non-use. On those numbers, "triples the risk" is a defensible headline.
Now supply the baseline the headline dropped. Among healthy nonpregnant women of reproductive age who do not use these methods, VTE occurs at roughly 1 to 5 cases per 10,000 woman-years. Users sit at roughly 3 to 15 per 10,000 woman-years. The tripling therefore moves a small number to a still-small number. The absolute change is a handful of extra events per 10,000 women each year, not anything close to a coin flip.
The comparison that resets the whole conversation is pregnancy. The same ASRM guideline puts VTE risk at 5 to 20 per 10,000 woman-years during pregnancy and 40 to 65 per 10,000 in the weeks after delivery. Pregnancy, the very condition contraception prevents, carries a clot risk several times higher than any combined pill. That is why the CDC advises against combined hormonal methods in the first 21 days after delivery, when clotting risk peaks, and eases the restriction between 21 and 42 days for women without added risk factors. The guidance is not treating the pill as uniquely dangerous. It is ranking one risk against another.
Why age rewrites the breast cancer story#
Breast cancer follows the same logic with an added twist. A 2023 UK nested case-control study and meta-analysis by Fitzpatrick and colleagues in PLOS Medicine found that current or recent use of hormonal contraceptives, combined or progestogen-only, was associated with an odds ratio around 1.2 to 1.3. In relative terms that is a 20 to 30 percent increase, and a headline can accurately say the method "raises breast cancer risk."
The twist lives in the absolute numbers. The same authors estimated that five years of use adds roughly 8 extra breast cancers per 100,000 users among those starting between ages 16 and 20, rising to about 265 per 100,000 among those using it between ages 35 and 39, calculated across a 15-year window. Notice that the relative increase stays nearly constant across ages while the absolute excess does not. Background breast cancer risk climbs steeply with age, so the same 20 percent lands on a tiny baseline for a teenager and a much larger baseline for someone in her late thirties. A relative risk quoted alone erases that distinction, which is exactly why one honest percentage can be trivial for a 19-year-old and worth a real conversation for a 38-year-old.
The framework thinks in absolute terms#
The CDC medical eligibility framework sorts conditions into four categories, from no restriction to unacceptable risk. It is built on precisely this habit of weighing absolute risks against one another and against the risk of the outcome being avoided. It never asks merely whether a method changes risk, because almost everything changes risk. It asks by how much, from what starting point, and against what alternative. That is the same discipline that careful epidemiology applies to any risk factor: a striking ratio means little until the base rate and the competing risk are on the table.
A three-question habit#
When a striking relative figure appears, three questions convert it into something you can judge:
- Relative to what baseline? A percentage increase is meaningless until you know the starting rate.
- How large is the absolute change? Extra cases per 10,000 or per 100,000 is the honest unit, not a bare multiplier.
- Compared with which alternative? The risk of the option being declined, here pregnancy, belongs in the same frame.
None of this decides whether a given method suits a given person. That depends on individual history and belongs in a conversation with a clinician. The narrower point is about literacy. A number that multiplies a rare event is still describing a rare event, and the evidence behind contraception risk tends to reassure precisely because it survives that translation.
Sources and further reading
- CDC US Selected Practice Recommendations: Combined Hormonal Contraceptives
- Fitzpatrick et al., Hormonal Contraceptives and Breast Cancer, PLOS Medicine 2023
- ASRM: Combined Hormonal Contraception and the Risk of Venous Thromboembolism, a Guideline (2016)
- CDC US Medical Eligibility Criteria for Contraceptive Use, 2024
Questions and answers
Does "triples the risk" mean my risk is now high?
Not by itself. Tripling a baseline of about 3 per 10,000 users lands near 9 per 10,000, which is still low. A multiplier only tells you how much a starting number grows, so you have to know the starting number before you can tell whether the result is large or small.
Why do guidelines still restrict combined pills right after childbirth?
Because the postpartum weeks carry one of the highest clot risks in reproductive life, on the order of 40 to 65 per 10,000. Adding a method that also nudges clot risk during that specific window is what the CDC guidance avoids, then relaxes as the postpartum risk falls.
Is the breast cancer signal a reason to avoid hormonal contraception?
It is one factor among many, and its absolute size depends heavily on age. For younger users the extra cases per 100,000 are very few; the number grows with background risk later in life. Weighing it belongs with a clinician who knows your history, not with a headline.