Maternal Use of Antidepressants in Pregnancy and the Risk of Postpartum Hemorrhage: An Electronic Health Record-Based Cohort Study.

OBJECTIVES

Prior studies have raised concerns about postpartum hemorrhage (PPH) risk associated with selective serotonin and serotonin-norepinephrine reuptake inhibitors (SSRIs and SNRIs), although limitations include possible confounding by maternal indications. The current study examined the risk of PPH associated with prenatal S/NRIs accounting for diagnosed maternal depression during pregnancy.

METHODS

The current study is a secondary analysis of data from a cohort of 2213 pregnancies (1997-2010) identified from a large electronic medical record-linkage system. Exposures were identified from electronic prescriptions records and resulted in three main groups: S/NRI-exposed pregnancies (S/NRI users), antidepressant-unexposed pregnancies (non-users), and S/NRI exposure in the year prior to (but not during) pregnancy (former users). Data used to identify PPH occurrences included obstetrician chart diagnoses, diagnosis codes, and recorded estimated blood loss (EBL). The risk of PPH by exposure group was estimated using unadjusted, minimally adjusted (maternal age at delivery, delivery year, parity, and maternal smoking), and depression-adjusted models (previous covariates and maternal depression). Supplemental analyses included an examination of PPH risk with prenatal bupropion monotherapy.

RESULTS

The study cohort consisted of 837 S/NRI users, 401 former users, and 786 non-users, as well as 114 bupropion users identified for additional analyses. The frequency of PPH was 11.6% in the overall cohort, and was 14.7% for S/NRI users, 10.2% for former users, and 8.7% for non-users. The risk of PPH was significantly higher for S/NRI users than non-users in unadjusted (OR 1.82 [95% CI 1.33, 2.49]), minimally adjusted (aOR 1.66 [1.19, 2.31]), and depression-adjusted models (aOR 1.46 [1.02, 2.10]). The risk of PPH was also higher for S/NRI former users than non-users in unadjusted models (OR 1.54 [1.05, 2.26]) but not minimally adjusted (aOR 1.42 [0.95, 2.11]) or depression-adjusted models (aOR 1.41 [0.95, 2.10]). No significant differences in PPH risk were observed between S/NRI users and former users, between bupropion users and non-users, or between S/NRI users and bupropion users.

CONCLUSIONS

Prenatal S/NRI exposure was associated with an increased risk of PPH after adjusting for maternal depression, although residual confounding and confounding by other factors (including unmeasured factors) may be possible, as the risk of PPH appeared to be similar between S/NRI users and former users. Best-practice PPH risk management strategies should be adhered to for depressed patients, especially those who require treatment with antidepressants during pregnancy.