ANNOTATION (Neeta Shenai)

The Finding: Data from this large population-based cohort study, including sibling-controlled analysis, provide evidence against a clinically significant increase in overall major birth defects with in utero use of SSRIs or venlafaxine. Though the prevalence of major birth defects (13%) and cardiac birth defects (15%) was increased in exposed infants, the adjusted odds ratio decreased in the sibling-controlled analysis to 1.06 and 0.92 respectively. Recent studies [1,2] found no association between SSRI’s and right ventricular outflow tract obstruction defects, which corresponds to data from this study when controlled for confounders in the sibling analysis. While sertraline was associated with a 2.5-fold increased prevalence of anal atresia, the estimated absolute risk was low.

Strength and Weaknesses: This is the first study with large sample sizes of venlafaxine. Additionally, it is also the first to control for familial confounding through a sibling design. Although this reduces the bias from possible lifestyle and genetic factors that are difficult to control, the sibling design has lower statistical power given the sample size. The full cohort included a large multi-country population, and the prospective methods permitted analysis of individual SSRIs and birth defects without the possibility of recall bias. However, non-adherence of filled anti-depressant prescriptions and lack of information on the severity of depression could have resulted in misclassification. Though the study accounted for major confounders including maternal age, diabetes, use of teratogenic drugs, and smoking, other cofounders such as maternal use of alcohol, other psychiatric comorbidities, or lifestyle factors were not included. Further, women prescribed SSRIs or venlafaxine were found to be prescribed other drugs more frequently than women in the control group (14.6% vs 1.0%) leading to potential confounding.

Relevance: Though depression is common in pregnancy, data on the teratogenicity of antidepressants has yielded conflicting results. SSRIs are the most commonly prescribed antidepressants and venlafaxine is emerging as a suitable alternative. Previous studies have many limitations including recall bias and low statistical power. As this study analyzed a large data set, accounted for many confounding factors through the sibling design, and did not find a substantial teratogenic effect, this provides reassuring data for its use in pregnancy.

ANNOTATION (Meredith Leigh Spada)

The Finding: The results from this study of a very large cohort of publicly insured pregnant women indicate a potential increased risk of PPHN associated with maternal SSRI use in late pregnancy. This finding held true even after performing sensitivity analyses and controlling for depression and other potential confounding factors including diabetes and NSAID use. Despite this finding, the absolute risk of PPHN was small, with an odds ratio of 1.10 after restricting to women with depression and adjusting for several other potential confounders. As such, the risk of PPHN associated with SSRI use in late pregnancy appears to be less pronounced than suggested in previous studies.

Strength and Weaknesses: This is the largest cohort studied to-date seeking to discern the association between SSRI use during late pregnancy and the risk of PPHN, suggesting broad applicability of its results. Access to medical records for this cohort provided information regarding use of SSRI and non-SSRI antidepressants and several potential confounders. The authors utilized advanced epidemiologic methods to account for maternal depression to mitigate potential confounding by the underlying psychiatric illness and its associated conditions and behaviors. Other studies, by contrast, including a recent large population-based cohort study of 1.6 million pregnancies, do not restrict outcomes to women with a diagnosis of depression.[1,2] Furthermore, the authors included an objective assessment of drug exposure, as they defined exposure status as having filled one prescription during the 90 days prior to delivery and then redefined exposure status in the sensitivity analysis as having filled at least two prescriptions during the 90 days before delivery.
Despite its many strengths, the study also had some limitations. It made several efforts to mitigate confounding factors by including risk factors for PPHN such as diabetes and NSAID use; however, information on some potential confounders was incomplete (e.g., smoking) or absent (e.g., maternal BMI, diabetes severity). Another worry is that defining exposure as having filled one prescription may result in a misclassification, as filling one prescription does not ensure medication compliance. As such, this classification may bias results towards the null, with results appearing as though there is no relationship between SSRI use in late pregnancy and PPHN. The authors conducted a sensitivity analysis to address this, redefining exposure as having filled two prescriptions. Still, while it is likely that women who go through the effort to refill a prescription are more likely to be taking the medication, refilling a prescription does not ensure that the medication was actually taken. Though this was a very large cohort, as Medicaid covers close to half of all births in the US, it did not include women in commercially insured populations, who are older and less racially diverse than the Medicaid population, thus potentially decreasing applicability to the overall general population.

Relevance: Based on a single epidemiologic study that found a 6-fold increase in risk of PPHN associated with SSRI use after the 20th week of pregnancy, the FDA issued a public health advisory in 2006 on a potential increased risk of PPHN associated with late pregnancy exposure to SSRIs.[3] In 2011, the FDA then updated the advisory based on review of additional studies whose findings conflicted with one another, and subsequently concluded that it was premature to reach any conclusion about a possible link between SSRI use in pregnancy and PPHN.[4] This study provides further insight regarding the controversy and thus provides doctors with new and relevant information to consider when counselling pregnant patients on the usage of SSRIs. While there may be an elevated risk of PPHN with SSRI use in late pregnancy, this study has demonstrated that the absolute risk appears to be very low.