Skip to main content
2024 ISPE Annual Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Back
Evaluation

Oral Presentation

Session: The Treatment and Prevention of Pregnancy-related Infections

Risk of Pregnancy Loss after Prenatal Exposure to Doxycycline: A Retrospective Cohort Study

Monday, August 26, 2024
1:45 PM – 2:00 PM CEST
Location: Room XV
Background: Doxycycline is approved for prophylaxis and treatment of various infections, although it is generally not recommended during pregnancy. Limited safety data are available about the use of doxycycline during pregnancy.

Objectives: To evaluate the risk of pregnancy loss following exposure to doxycycline.

Methods: This was a retrospective cohort study in MarketScan Commercial Claims data 2005−2021. Four separate cohorts consisting of pregnant persons aged 12-55 years with doxycycline or comparator antibiotic prescription fills between conception and first trimester end (index date) and concomitant diagnoses of respiratory tract infections (RTI) (oral doxycycline vs oral amoxicillin or macrolides) or acne (oral doxycycline vs oral macrolides or topical clindamycin) were evaluated for pregnancy loss using diagnosis and procedure codes for spontaneous abortion or stillbirth. Pregnancies with antibiotic use for indications other than RTI or acne were excluded. We fitted separate logistic regression models using high-dimensional propensity score (HdPS) with additional expert informed covariates (obstetric characteristics, maternal comorbidities, and pre-pregnancy healthcare utilization) for each indication and comparator group. Doxycycline exposed pregnancies were matched to active comparator pregnancies within each indication in a ratio of 1:4 for RTI and 1:1 for acne on HdPS and gestational age (GA). We estimated the adjusted odds ratio (aOR) via conditional logistic regression. Sensitivity analyses included a 15-day delay between index date and pregnancy loss to assess reverse causation, exclusion of COVID-19 period, requirement for pre-conception exposure for acne comparisons, and stratification by index GA (2-6 weeks versus 7-12 weeks).

Results: We included 112,589 pregnancies with RTI and 8,204 pregnancies with acne. Doxycycline exposed patients were slightly older and had earlier index GA than comparator cohorts. In patients with RTI, the incidence proportion of pregnancy loss was 32.8% among those with doxycycline (n=1252) versus 19.8% with amoxicillin (n=5008), aOR 1.98 [95% CI 1.73, 2.28]; and 32.8% with doxycycline (n=1269) versus 22.6% with macrolide (n=5076), aOR 1.71 [1.49, 1.97]. Among acne patients, the incidence proportion was 39.7% with doxycycline (n=700) compared to 20.0% with macrolide (n=700), aOR 2.75 [2.12, 3.57]; and 35.3% with doxycycline (n=1228) versus 16.8% with clindamycin (n=1228); aOR 2.77 [2.27, 3.40]. The results were consistent in sensitivity analyses.

Conclusions: Exposure to doxycycline during early pregnancy is associated with an increased risk of pregnancy loss, which appears not attributable to confounding by indication.