Key Takeaways
Extended Takeaways
- This comparison was strengthened by propensity score matching on 26 features, producing 2 cohorts of 2,025 patients with all covariates balanced at SMDs below 0.1 and a maximum standardized difference of 0.095.
- The matched population reflects a high-comorbidity perinatal psychiatry cohort: bipolar disorder was present in 47%, schizophrenia in 30%, anxiety disorders in 44%, major depressive disorder in 39%, nicotine use disorder in 42%, circulatory disorders in 34%, and preexisting diabetes mellitus in 11%.
- The absolute event rates for the composite obstetric outcome were similar between formulations, occurring in 8.7% of oral SGA exposures and 8.3% of LAIA exposures, which can help frame counseling beyond the nonsignificant OR 0.95; 95% CI, 0.76–1.18; P=.612.
- Cesarean delivery rates also did not differ significantly between treatment modalities in the matched analysis (OR 1.09; 95% CI, 0.83–1.43; P = .537), suggesting formulation choice alone should not be expected to alter operative delivery risk.
- The study only evaluated outcomes occurring from 1 day after the index event through 280 days and did not assess spontaneous abortion or congenital malformations, so these findings should not be used to infer first-trimester loss or teratogenic risk.
- Because oral antipsychotic overlap was allowed in the LAIA cohort, these results are most applicable to real-world long-acting injectable use patterns rather than to completely oral-free LAIA exposure.
