(242) Levothyroxine Use and The Risk of Pregnancy Loss Among Pregnant Women with Subclinical Hypothyroidism: An Observational Study Emulating a Target Trial

Publication Number: 573

Background: Hypothyroidism is the most common endocrine disorder in pregnancy, with subclinical hypothyroidism (SCH) affecting ~2% of pregnancies. Levothyroxine is clearly indicated for overt hypothyroidism but its benefit-risk ratio for SCH is unclear.

Objectives: To determine if the use of levothyroxine is associated with a decreased risk of spontaneous abortions and stillbirths in women with subclinical hypothyroidism.

Methods: We conducted a population-based cohort study (1998-2017) of pregnant women aged 15-45 years using the Clinical Practice Research Datalink linked to the Pregnancy Registry. Women with a SCH diagnosis or abnormal TSH value one year prior or during pregnancy were included. Levothyroxine use was defined by at least 1 prescription during pregnancy. Pregnancy loss was defined as a spontaneous abortion or stillbirth. We emulated multiple nested trials with a new trial starting at every gestational week. Each trial consisted of women who initiated levothyroxine in the previous 7 days matched to non-initiators (1:4) on time of diagnosis, gestational age, and high-dimensional propensity score. We used an ITT approach for exposure and followed women from cohort entry (pregnancy start or date of diagnosis/abnormal lab value, whichever occurred latest) until pregnancy loss or censoring event for each trial. Unexposed women could be included in multiple trials and women could contribute ≥1 pregnancy during follow-up. We used a pooled logistic regression model with bootstrap resampling to estimate hazard ratios (HRs) and 95% confidence intervals (CIs).

Results: Among 5,781 women and 6,259 pregnancies, 2.2% of pregnancies were exposed to levothyroxine. At cohort entry, the mean age was 32.2 years (SD 5.4), 25% were obese, and 8% had diabetes. Women using levothyroxine were younger and more likely to be diagnosed during pregnancy and to have no prior pregnancies. Over 266.84 years, 57 pregnancy losses occurred. The HR for pregnancy loss among women with SCH with use of levothyroxine versus non-use was 0.87 (95% CI 0.43, 1.79).

Conclusions: Although findings were inconclusive, this study highlights the feasibility of using the target trial framework to assess the safety of medications used in pregnancy.