Combined contraceptive pills 'increase risk of blood clots'

"Women who take the latest generations of contraceptive pills are at a greater risk of potentially lethal blood clots," The Times reports. While the increase in risk is statistically significant, it is very small in terms of individual risk

The combined oral contraceptive pill, commonly referred to as "the pill", is already well known to be linked to increased risk of blood clots in the veins, such as deep vein thrombosis (DVT), as we discussed back in 2014.

A new study, using two large GP databases, set out to refine the assessment of the risk. It identified women who had had a venous blood clot, matched them by age to unaffected women, and examined use of the pill in the previous year.

Overall it found that use of any contraceptive pill almost tripled risk of blood clot; though the baseline risk is small. And risk was generally higher with the newer third generation pills, compared with older pills. Encouragingly, risk was lowest for pills containing levonorgestrel, which is by far the most common prescribed. This pill carried risk of around six extra cases of blood clot for every 10,000 women prescribed.

Risk was more than double this for pills containing desogestrel, gestodene, drospirenone and cyproterone, though these aren't normally pills of first choice in practice, and are normally used when there are reasons to treat other symptoms such as acne.

The combined oral contraceptive pill remains a safe and effective form of contraception for most women, but it is not suitable for all – such as women with a history of heart disease or high blood pressure. Read more about who can, and who shouldn't, use the combined oral contraceptive pill.

Where did the story come from?

The study was carried out by researchers from the Division of Primary Care, University Park in Nottingham. It received no external sources of funding. The study was published in the peer-reviewed British Medical Journal as an open-access article. This means it can be read online or downloaded by anyone for free.

The reporting of the study by the UK media was accurate and refreshingly took steps to put the small increase in risk in context.

What kind of research was this?

This was a case-control study of women identified through two general practice databases in the UK. The researchers were aiming to look at the link between use of the combined oral contraceptive pill ("the pill") and risk of blood clots in the veins (e.g. deep vein thrombosis, or DVT), specifically taking into account the type of progestogen in the pill.

Use of the pill is already well known to be associated with increased risk of blood clots in the veins (venous thromboembolism). Different types of pill combine different types of the hormone progestogen with another hormone called oestrogen. It is recognised that the different progestogens have differing influence on the risk of blood clots, though previous study has not been able to quantify the risks of the different pills, particularly newer ones.

This case-control study investigated this by looking at women diagnosed with a blood clot, matching them to unaffected women and then looking at type of pill used.

What did the research involve?

The study used two large GP databases, QResearch and Clinical Practice Research Datalink (CPRD), both of which have previously been used to look into links between different drugs and blood clot risk. QResearch covers 618 general practices in the UK, and CPRD covers 722.

Researchers identified women aged 15-49 years registered between 2001 and 2013 who had a first instance of a venous blood clot. They matched these "cases" with up to five unaffected age-matched "controls" from the same database. They excluded women who were pregnant around the time, or those who had a hysterectomy or sterilisation. They excluded women who were pregnant around the time, who had hysterectomy or sterilisation, or who had a history of use of blood thinning medicine – suggesting history of or susceptibility to blood clots.

Use of the combined oral contraceptive pill was examined in the year prior to the blood clot record. They included all the most commonly used preparations in the UK, containing the different types of progestogen. They also included the combination of oestrogen with cyproterone acetate (brand name Dianette), which acts as a contraceptive pill, but its main indication is for the treatment of acne. They looked at when the pill had been used in relation to the time of the clot (e.g. current or past use) and how long it had been used for. 

They took into account potential confounding factors that may influence clot risk, including:

  • chronic medical conditions (e.g. cancer, heart or lung disease, arthritis or inflammatory conditions)
  • recent immobilisation such as through trauma, surgery or hospital admission
  • smoking and alcohol
  • obesity
  • polycystic ovary syndrome (associated with pill use and increased risk of clots)
  • social deprivation

What were the basic results?

After exclusions they had a sample of 5,500 cases and 22,396 controls in the QResearch database, and 5,062 cases and 19,638 matched controls in the CPRD database. The incidence of venous blood clots in the two databases was around six per 10,000 women per year. Just over half (58%) the blood clots in the two databases were DVTs.

In the two databases 28-30% of cases had used the pill in the past year, compared with 16-18% of controls. Overall, any pill use in the past year was associated with an almost tripled risk of venous blood clot compared to no use (adjusted odds ratio (OR) 2.97, 95% confidence interval (CI) 2.78 to 3.17).

The most common pill was one containing the progestogen levonorgestrel, which accounted for roughly half of prescriptions in cases and controls.

By type of progestogen the researchers found the following to be associated with lower risk:

  • levonorgestrel (OR 2.38, 95% CI 2.18 to 2.59)
  • norethisterone (OR 2.56, 95% CI 2.15 to 3.06)
  • norgestimate (OR 2.53, 95% CI 2.17 to 2.96)

The following were associated with higher risks:

  • desogestrel (OR 4.28, 95% CI 3.66 to 5.01)
  • gestodene (OR 3.64, 95% CI 3.00 to 4.43)
  • drospirenone (OR 4.12, 95% CI 3.43 to 4.96)
  • cyproterone acetate (OR 4.27, 95% CI 3.57 to 5.11)

Pills are sometimes termed according to generations of when they were manufactured. The bottom list are newer "third generation" pills, while the former group includes mostly earlier generations. The exception is norgestimate in the former list, which is also third generation.

The number of extra cases of venous blood clot per year was lowest for levonorgestrel and norgestimate (both six extra per 10,000 women) and highest for desogestrel and cyproterone (both 14 extra per 10,000 women).

How did the researchers interpret the results?

The researchers conclude: "In these population-based, case-control studies using two large primary care databases, risks of venous thromboembolism associated with combined oral contraceptives were, with the exception of norgestimate, higher for newer drug preparations than for second generation drugs."


It is already well known that the combined oral contraceptive pill ("the pill") is associated with increased risk of venous blood clots. It is also already known that risk can differ according to the type of progestogen in the pill. This study adds further evidence helping to quantify these risks.

The study has numerous strengths. It has used two large GP databases covering large samples of the UK population, and containing reliable information on medical diagnoses and prescriptions made. The analyses were also adjusted for various confounders known to be associated with risk of blood clots.

It demonstrates pill use in the previous year almost tripled risk of venous blood clot, with risk generally higher with the newer pills than the older ones – though there were some exceptions.

Encouragingly, preparations containing levonorgestrel – which is by far the most common pill prescribed – had the lowest associated risk; around six extra cases of blood clot for every 10,000 women prescribed.

The preparations associated with the highest risks in this study – desogestrel, gestodene, drospirenone and cyproterone – were already recognised to be linked to higher risk, though this study has helped to better quantify these risks. These are not usually pill preparations of first choice in practice and are normally used when there are specific indications (e.g. women who have acne, particularly those taking cyproterone), or who have had side effects with other preparations.

The organisation in charge of regulating medicines in England quantified the risks of the combined pill last year and came up with very similar results.

That review said the benefits of the pill far outweigh the risks, but added: "Prescribers and women should be aware of the major risk factors for thromboembolism, and of the key signs and symptoms."

This study again highlights the need for careful prescribing of the combined oral contraceptive pill, taking into account the individual woman’s risk factors such as lifestyle and medical history. Women should also be aware of the signs and symptoms of venous blood clots, such as DVT. If a woman taking the pill experiences unexplained swelling or pain in the leg, or sudden breathlessness and/or chest pain, they should seek medical help immediately.

The combined pill may not be suitable for you if you have a history of certain chronic diseases, such as heart disease, diabetes or high blood pressure. Other alternative methods of contraception, such as a contraceptive implant may be a more suitable option.

Your GP should be able to advise you on the safest method for your individual circumstances.

NHS Attribution