First-trimester benzodiazepine fills showed no consistent rise in malformations

Design

Population-based cohort using two US claims databases with propensity score weighting

Population

About 4.2 million pregnancies; 51,339 Medicaid and 27,779 MarketScan pregnancies with first-trimester benzodiazepine fills

Primary outcome

Any major congenital malformation

Effect

Pooled adjusted RR 0.99 (95% CI 0.96 to 1.03); 4.0% vs 4.1% (Medicaid), 4.9% vs 4.8% (MarketScan)

This US population-based cohort used Medicaid (2000 to 2018) and MarketScan (2003 to 2020) claims. It included 2,478,434 publicly insured and 1,750,721 commercially insured pregnancies, of which 2.1% and 1.6% had a benzodiazepine fill in the first trimester. Exposed and unexposed pregnancies were compared with propensity score weighting, and a relative risk of 1.3 or more was prespecified as clinically significant.

The adjusted absolute risk of any major malformation was 4.0% in exposed vs 4.1% in unexposed pregnancies (Medicaid) and 4.9% vs 4.8% (MarketScan), a pooled adjusted relative risk of 0.99 (95% CI 0.96 to 1.03). No consistent organ-specific associations or dose-response relationship appeared, and results were similar across individual benzodiazepines. Some analyses of co-exposure with other psychoactive drugs showed increases, but without a consistent pattern.

This helps counselling after an unplanned early exposure: it does not suggest a large malformation risk from the benzodiazepine alone. It does not address later-pregnancy or neonatal effects, or the case for using a benzodiazepine in the first place. Pharmacy fills are a proxy for taking the drug.

• Pooled adjusted relative risk for any major malformation was 0.99 (95% CI 0.96 to 1.03).
• Absolute risks were about 4% to 5% in both exposed and unexposed pregnancies.
• No dose-response relationship and no consistent organ-specific signal emerged.
• Use this to reassure a woman who took a benzodiazepine before she knew she was pregnant, without advising abrupt stopping.
• Co-exposure with other psychoactive drugs needs separate review, because those results were inconsistent.

Why it matters

Earlier data on this counselling question were conflicting, and this adds a very large analysis that found no consistent excess.

Don't overread it

Claims data show a fill, not that the tablets were taken, and cover only malformations present at birth in live-born infants.

The statistics, in plain English

A relative risk of 0.99 with an interval of 0.96 to 1.03 rules out a rise of 1.3 or more (the prespecified threshold) with good confidence. It does not rule out a very small effect or effects on outcomes not captured, such as stillbirths and later neurodevelopment.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.