Scientists map use of epilepsy drug in England and Wales linked to birth defects

Scientists have created a unique map of England and Wales that reveals detailed variations in the use of a single medicine. The drug is sodium valproate, a treatment for severe cases of epilepsy that is also known to cause health problems for pregnant women.

The findings show drops in the use of the drug, in response to warnings about its effects, but the map reveals clear differences in dispensing patterns to women of child-bearing age, with the highest rates being found in local authorities in the north-west and across coastal regions in the east. These variations highlight potential inequalities that need further investigation, the researchers warn.

Valproate is an important drug for treating epilepsy and bipolar disorder, but it has also been linked to miscarriages, birth defects and cases of conditions such as ADHD and autism.

The medicines regulator, MHRA, said in 2016 that babies exposed to sodium valproate in the womb had a high risk of serious developmental disorders. About 20,000 children in the UK are estimated to have life-changing injuries caused by valproate. In addition, studies have suggested that children of men who take the drug have an increased risk of neurological disorders.

“This is the first time we have been able to combine different sets of health data and understand how a particular medicine is used in the population,” said the research’s team leader, Reecha Sofat, professor of clinical pharmacology and therapeutics at the University of Liverpool. “The fact that we succeeded suggests we could provide benefits to the health service by understanding, at a granular level, how medicines are being used by people in different areas.”

Valproate was chosen for the research because of recent concerns about its use. “It becomes a problem of balancing neurological issues with the paediatric concerns,” said Sofat. “If a person with epilepsy who is pregnant is resistant to all other medicines apart from valproate, there is a clear problem for doctors. What is needed, we realised, was a comprehensive picture of who was taking valproate and where.”

The project is also important because it demonstrates the power of data-driven research, argues the team, which was backed by Health Data Research UK (HDR UK). By combining separate national datasets, it will be possible to outline the way different medicines are used around the country. The potential for this use of data science is striking.

One future use would be to link anonymised health data of mothers to their children’s health records to improve doctors’ knowledge of how medicines given during pregnancy affect the health of the unborn child, said another team member, Dr Caroline Dale, a research fellow at the University of Liverpool.

“Pregnant women are often excluded from traditional research methods, such as clinical trials, given understandable concerns about unknown risks,” Dale said. “This means we have huge gaps in our knowledge of the outcomes of safe drug use in this group. Our data-led approach is non-invasive and could be expanded to understand maternal health in pregnancy better, linking this to child health records that would help us to understand the real-world effects of medicines.”

It should also be possible to monitor changes in the prescribing of other drugs and reveal variations in the numbers of different drugs that people may be taking in different areas, say the researchers, whose work is outlined in a paper published in the journal BMJ Medicine this week.

To carry out their research, scientists from the British Heart Foundation Data Science Centre – backed by HDR UK – were allowed to use anonymised NHS health records, including primary care and hospital records, which they then linked to dispense records for their target medicine: sodium valproate. This was the first time such data had been combined in this way.

The research showed that between 2019 and 2023, new use of sodium valproate fell from 7 to 5 per 100,000 women in those aged 15-19, from 11 to 7 in those aged 20-29, and 14 to 7 in those aged 30-39.

Pregnancy rates fell from 6.0 to 5.2 per 1,000 women dispensed sodium valproate over the same period. There was no evidence that epilepsy-related deaths increased in women aged 15-49 during 2015-2022, but there was some evidence for a slight increase in men during the period April 2018 to December 2022.

“These, on their own, are striking findings,” said Sofat. “They show that there has been an encouraging response, in general, to warnings about the risks involved in taking valproate, though these do not seem to be shared equally across the country.

“But what is also important is the demonstration that by providing access to different anonymised health databases, we can make telling insights into the ways that medicines are being used, allowing doctors to make all sorts of improvements to the way these are dispensed – and that could have enormous benefits for the health service.”