How to Choose LAIA Versus Oral Antipsychotics in Pregnancy
How should clinicians choose between second-generation long-acting injectable and oral antipsychotics for a pregnant patient with serious mental illness?
Pregnant patients with schizophrenia or bipolar disorder may need ongoing antipsychotic treatment, and relapse from undertreatment can create substantial maternal and fetal risk. This guide applies when a clinician is deciding whether formulation choice itself changes obstetric risk and whether a long-acting injectable antipsychotic remains a reasonable option during pregnancy.
-
Confirm that ongoing antipsychotic treatment is clinically necessary
Start from the clinical reality that untreated or inadequately treated maternal psychiatric illness is associated with poor treatment adherence, reduced prenatal care, increased alcohol or tobacco use, and disruptions to family dynamics. In patients with severe illness, weigh the risk of peripartum relapse against the risk of in utero antipsychotic exposure before considering a formulation change.
-
Identify whether a long-acting injectable is appropriate for adherence needs
Consider a second-generation long-acting injectable antipsychotic when the patient has demonstrated response and tolerability to the corresponding oral formulation, particularly if the patient prefers this modality or has a history of poor adherence. The article notes that long-acting injectables are associated with reduced psychiatric relapse hospitalization rates, improved treatment retention, and longer time to medication discontinuation compared with oral antipsychotics.
-
Compare obstetric risk by formulation rather than assuming LAIAs are riskier
Use the study's matched comparison to frame formulation-specific counseling. After 1:1 propensity score matching, the composite obstetric outcome of gestational diabetes, preeclampsia, eclampsia, or newly diagnosed hypertensive disorder occurred in 8.7% of oral second-generation antipsychotic exposures and 8.3% of long-acting injectable exposures, with no significant difference between groups (OR 0.95; 95% CI, 0.76-1.18; P=.612).
-
Explain that individual obstetric complications also did not differ
Counsel the patient that no statistically or clinically significant differences were observed for the individual components of the composite outcome. Specifically, the study did not find formulation-based differences in newly diagnosed gestational diabetes, preeclampsia, eclampsia, or gestational hypertension.
-
Include delivery planning in the discussion without overattributing risk to formulation
If the patient asks whether a long-acting injectable changes the chance of cesarean delivery, explain that this study did not find a significant difference. Cesarean section odds were similar between long-acting injectable and oral groups (OR 1.09; 95% CI, 0.83-1.43; P=.537).
-
Monitor for gestational diabetes during pregnancy if antipsychotics are required
Do not interpret the absence of a between-group difference as meaning metabolic monitoring is unnecessary. The article states that the American College of Obstetricians and Gynecologists recommends monitoring for gestational diabetes in pregnant patients who require antipsychotic treatment.
-
State clearly what these data do and do not cover
Use these findings to compare obstetric complications between long-acting injectable and oral second-generation antipsychotics, not to infer safety for all pregnancy outcomes. The study did not assess spontaneous abortion, congenital malformations, or complete neonatal outcomes, so formulation counseling should remain limited to the obstetric outcomes actually studied.
-
Base final formulation choice on relapse prevention and adherence needs
When a pregnant patient needs a second-generation antipsychotic, use adherence history, prior response, tolerability, and relapse risk to guide whether to continue or initiate a long-acting injectable. In this study, long-acting injectable antipsychotics had similar observed rates of gestational diabetes and hypertensive obstetric complications as oral formulations, so formulation choice can be made without an observed obstetric penalty in these outcomes.
Clinical Considerations
- The study did not assess spontaneous abortion because those events are more likely to occur in the first trimester.
- Congenital malformations and full neonatal outcomes were not evaluated because of limitations in maternal-neonatal chart linkage within the database.
- Oral antipsychotic overlap was allowed in the long-acting injectable cohort, so the findings reflect real-world long-acting injectable use rather than purely oral-free exposure.
- Relevant confounders such as severity of maternal psychiatric illness and the duration or extent of individual antipsychotic exposure could not be fully characterized.
Bottom Line
Do not avoid a second-generation long-acting injectable antipsychotic in pregnancy solely because of concern for higher rates of gestational diabetes or hypertensive obstetric complications compared with the corresponding oral formulation.
