Category: Medical/Surgical/Diseases/Complications
Poster Session III
People with mental health conditions are at increased risk of severe maternal morbidity (SMM), yet little is known about how this differs by psychiatric diagnosis. We evaluated the relationship between psychiatric conditions and SMM in U.S. Veterans - a population with a high prevalence of psychiatric disease and universal screening.
Study Design: This observational cohort study used diagnosis codes from an administrative database of births to Veterans reimbursed by the Veterans Health Administration between 2010 and 2020. The exposure was a psychiatric condition - defined as depression, anxiety, post-traumatic stress disorder, or severe mental illness (schizophrenia or bipolar disorder) - classified in a mutually exclusive fashion into “active” (codes recorded in clinical encounters within 1 year of delivery) and “historical” (those who only had codes >1 year prior to delivery). The primary outcome was SMM, defined using a validated composite; non-transfusion SMM was also evaluated. A multivariable logistic regression model was conducted adjusting for demographic confounders. Mediation was assessed by separately adding an obstetric comorbidity score (a validated predictor of SMM) and substance use disorders to these models.
Results: Of 47,833 eligible births, 3.6% had historical, 17.5% had active, and 79% had no maternal psychiatric conditions (Table 1). Compared to people with no psychiatric conditions, SMM was significantly increased in people with active psychiatric conditions (3.1% active versus 1.8% without, aOR 1.58, 95% CI 1.35-1.84) but not historical psychiatric conditions (1.3% versus 1.8%, aOR 0.89, 95% CI 0.57-1.39, Table 2). Findings were similar for non-transfusion SMM. The risk was increased across all psychiatric diagnoses. After accounting for potential mediators, associations were reduced.
Conclusion: SMM was significantly increased in Veterans with active psychiatric conditions during pregnancy. Obstetric comorbidities may play a role in the causal pathway and warrant further investigation.
Danielle M. Panelli, MD
Stanford University
Palo Alto, California, United States
Aryan Esmaeili, MD, PhD
VA HSR&D Health Economics Resource Center
Menlo Park, California, United States
Vilija Joyce, MS
VA HSR&D Health Economics Resource Center
Menlo Park, California, United States
Caitlin Chan, MS
VA HSR&D Health Economics Resource Center
Menlo Park, California, United States
Kritee Gujral, PhD
VA HSR&D Health Economics Resource Center
Menlo Park, California, United States
Susan Schmitt, PhD
VA HSR&D Health Economics Resource Center
Menlo Park, California, United States
Nicole Murphy, BA
VA HSR&D Health Economics Resource Center
Menlo Park, California, United States
Rachel Kimerling, PhD
VA National Center for PTSD
Menlo Park, California, United States
Stephanie A. Leonard, PhD (she/her/hers)
Assistant Professor
Stanford University
Stanford, California, United States
Jonathan G. Shaw
Stanford University
Stanford, California, United States
Ciaran S. Phibbs, PhD
Professor
Stanford University
Palo Alto, California, United States