Late pregnancy use of selective serotonin reuptake inhibitors and serotonin and norepinephrine reuptake inhibitors is associated with increased risk of postpartum haemorrhage
- Department of Obstetrics and Gynecology, College of Physicians and Surgeons, Columbia University, New York, New York, USA
- Correspondence to: Professor Cande V Ananth, Department of Obstetrics and Gynecology, College of Physicians and Surgeons, Columbia University, 622 West 168th Street, New York, NY 10032, USA;
Commentary on: 
Implications for research and practice
Use of antidepressant medication is associated with increased risk for postpartum haemorrhage (PPH).
Further research is needed to establish a causal role between antidepressant medication use and PPH.
Clinicians should be aware of possible increased risk of postpartum haemorrhage when treating depression during pregnancy.
Antidepressant medications are commonly used to manage psychiatric conditions in pregnancy. While extensive research related to teratogenesis and neonatal outcomes has been undertaken,1 there is relatively few data related to maternal obstetrical outcomes. Research from other specialties suggests that serotonin reuptake inhibitors (SSRIs) may be associated with increased risk of hospitalisation for bleeding and acute upper gastrointestinal bleeding.2 ,3 Prior research on antidepressants and obstetric haemorrhage is limited.4 ,5 Palmsten and colleagues set out to determine whether antidepressant medication use is associated with increased risk for PPH.
This retrospective cohort evaluated 106 000 pregnant women from 2000 to 2007 with a diagnosis of mood or anxiety disorder. Using pharmacy dispensing data, women were classified based on whether they were prescribed antidepressant medications currently or had used them recently or in the past. Antidepressants were classified based on serotonin reuptake inhibition. Covariates included risk factors for PPH (the primary outcome, classified based on International Classification of Diseases (ICD)-9 coding), demographic factors and psychiatric diagnoses, among other factors. A number of sensitivity analyses were performed to assess the robustness of their conclusions.
In this cohort, the use of SSRIs or non-SSRIs was associated with an approximate 40% increased risk of PPH in an analysis adjusted for risk factors. Risk for PPH was 2.8% for women with mood and/or anxiety diagnoses on no medications compared to 4% for women currently using SSRIs and 3.8% for women currently using non-SSRIs. The risk for recent users (up to 30 days before delivery date) was 3.2% for SSRIs and 3.1% for non-SSRIs.
Obstetric haemorrhage is a leading cause of maternal mortality and severe morbidity in the USA.6 Identification of further clinical risk factors for PPH could lead to novel clinical interventions and research initiatives to reduce adverse outcomes from this complication. This study demonstrates an association between PPH and medication that is relatively commonly prescribed in pregnancy. Several aspects of the methodology in this study enhance the validity of the findings: first, the control population is composed of women with similar psychiatric diagnoses not on antidepressant medications, minimising the potential for selection bias. Second, a temporal relationship between the exposure (pharmacy dispensing the medication) and the outcome (PPH status) was established. Third, the authors incorporated adjustment for obesity (data unavailable in the nationwide Medicaid Analytic eXtract database) as an ecological construct in their regression models. This counteracted the role unobserved confounding and selection bias can play in the exposure–outcome relationship of observational epidemiology. Fourth, the authors confirm their associations through a high-dimensional propensity score method. This latter analysis is particularly well-suited when both (unobserved) confounders and selection bias are suspected to distort the associations.
The findings from this study add considerably to limited prior research on this subject, which has found similar associations despite methodological shortcomings.4 ,5 The magnitude of increased haemorrhage risk in relation to serotonin exposure demonstrated in this study is clinically relevant. Further research both from an epidemiological and basic science standpoint is needed to more fully support a causal relationship between SSRIs, SNRIs and haemorrhage. Similarly rigorous methodology in other populations would validate the associations found in this well-designed study. While the benefits of antidepressants may outweigh the relatively small attributable maternal and neonatal risks for many women, clinicians should be aware of a modestly increased risk for this serious adverse obstetric outcome.