Mental health disorders, substance use, and pregnancy-associated mortality

BACKGROUND: Mental health disorders (MHDs) and substance-use disorders (SUDs) are implicated in preventable pregnancy-associated mortality (PAM) in the USA. However, the independent and joint associations of perinatal MHDs/SUDs on PAM are not well understood. METHODS: We conducted a retrospective cohort study using the Nationwide Readmissions Database from 2010 to 2020. We examined the risks of PAM within the calendar year of delivery among women with and without MHDs/SUDs at delivery and further contrasted MHDs (depression, anxiety, bipolar, or psychotic disorders) with and without SUDs, and SUDs alone. We estimated the confounder-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs). We conducted a probabilistic bias analysis to address exposure misclassification and unmeasured confounding. RESULTS: Of 19 703 242 women with a delivery hospitalization, 9.4% (n = 1 865 087) had an MHD/SUD diagnosed at delivery. Of 1989 pregnancy-associated in-hospital deaths, 21.7% (n = 432) occurred among women with an MHD/SUD diagnosis. Compared with those without MHDs/SUDs, the HR for mortality associated with MHDs/SUDs was 2.30 (95% CI 2.04, 2.59). Women with MHDs alone had an HR of 1.49 (95% CI 1.24, 1.79) for mortality, whereas those with both MHDs and SUDs had an even higher HR (2.90, 95% CI 2.28, 3.68). The mortality HR associated with SUDs alone was 3.55 (95% CI 3.00, 4.21). These associations increased slightly following correction for exposure misclassification and unmeasured confounding. CONCLUSION: This population-based study highlights the disproportionately high rates of PAM associated with MHDs/SUDs. Increased efforts to address MHDs/SUDs in pregnancy are essential to reduce PAM related to MHDs/SUDs.