(C-113) Use of sodium valproate and other antiseizure drug treatments in England and Wales: quantitative analysis of nationwide linked electronic health records

Background: Sodium valproate is an evidence-based treatment for idiopathic generalised epilepsy (IGE), and is also used for bipolar affective disorder. Sodium valproate is known to have teratogenic effects and has been subject to regulatory changes.

Objectives: To investigate the use of sodium valproate in England and Wales, including during pregnancy, compared with other antiseizure drug treatments, based on national level electronic health records.

Methods: Quantitative analysis of nationwide linked electronic health records.

Individual level, population scale data from NHS England's Secure Data Environment, from the British Heart Foundation Data Science Centre's CVD-COVID-UK/COVID-IMPACT Consortium (for England), and the Secure Anonymised Information Linkage Databank (for Wales), 1 January 2019 to 31 December 2023.
1 200 000 individuals dispensed any selected antiseizure drug treatment (ie, sodium valproate, lamotrigine, levetiracetam, carbamazepine, or topiramate); 304 000 women, aged 15-49 years, dispensed any selected antiseizure drug treatment and 28 400 women, aged 15-49 years, dispensed sodium valproate.

Prevalent (current) and incident (new) uses of sodium valproate and other antiseizure drug treatments before and during the covid-19 pandemic (1 January 2019 to 31 December 2023), grouped by age and sex. Pregnancy rates per 1000 women, aged 15-49 years, who used antiseizure drug treatments, and timing and dose of sodium valproate dispensed during pregnancy. Geographical variation in use of sodium valproate and disease indications (epilepsy and bipolar affective disorder). Trends in deaths related to epilepsy for 2015-22.

Results: Prevalent use of sodium valproate in women of childbearing potential decreased and use of most other antiseizure drug treatments increased between 2019 and 2023. Incident use of sodium valproate per 100 000 women decreased from seven to five in women aged 15-19 years, from 11 to seven in women aged 20-29 years, and from 14 to seven in women aged 30-39 years between 2019 and 2022. Incident use also decreased in men of the same age but remained at much higher levels (from 53 to 43 in men aged 15-19 years, 59 to 47 in men aged 20-29 years, and 57 to 42 in men aged 30-39 years, per 100 000 men). Pregnancy rates decreased from 6.0 to 5.2 per 1000 women of childbearing potential who were dispensed sodium valproate over the same period. No clear evidence was found that deaths related to epilepsy increased in women aged 15-49 during 2015-22, but a slight increase was found in men aged 15-49 during the later period between April 2018 and December 2022.

Conclusions: This approach, linking national dispensing data to health records at the individual level, could help monitor changes to medicines affected by regulatory changes.