For providers · Clinical resource

Psychiatric Medication in Pregnancy: A Risk-Benefit Framework

By Josephine W. Hazeley, MSN, PMHNP-BC
Board-certified psychiatric mental health nurse practitioner with a perinatal & postpartum focus — nearly two decades in psychiatric nursing, including research nursing on the trial behind the first FDA-approved postpartum-depression therapy. About your clinician

Clinically reviewed by Josephine W. Hazeley, PMHNP-BC on · Last updated

A perinatal prescriber weighs the risk of a specific medication at a specific dose against the risk of the untreated illness that medication treats — not the medication against no exposure at all. ACOG’s 2023 clinical practice guideline frames the decision as risk against risk, and current FDA drug labeling supports that by describing each medication’s known pregnancy and lactation risks in a narrative summary rather than a single letter grade, because untreated depression, anxiety, and bipolar disorder in pregnancy carry documented risks of their own. What follows is how that reasoning works: a framework, not a set of directives. Every medication decision is individualized with the patient by her prescriber.

If you or someone you know is in crisis, call or text 988 (Suicide & Crisis Lifeline) or call 911. The Postpartum Support International HelpLine is 1-800-944-4773.

Why isn’t the question “medication or no medication”?

The comparison that feels intuitive — a medicated pregnancy versus an unmedicated one — is the wrong one, because the unmedicated arm is not risk-free. A pregnancy affected by untreated psychiatric illness is itself an exposure.

ACOG’s guideline is direct that undertreated perinatal mood and anxiety conditions are associated with poorer outcomes for the pregnant patient and the infant, which is part of why the guideline recommends prescribing and titrating pharmacotherapy rather than defaulting to avoidance (AHRQ summary of the ACOG evidence review). Some of the birth outcomes historically attributed to antidepressants may trace in part to the depression being treated rather than the drug. So the honest framing is risk versus risk: the risk profile of a chosen medication set against the risk profile of the condition left undertreated.

How does a prescriber actually weigh it?

The reasoning moves through a consistent set of questions rather than a lookup table.

  • How severe is the illness, and what is the history? Recurrent major depression, prior postpartum psychosis, or bipolar disorder carries a different untreated-risk than a first, mild episode.
  • What has this patient responded to before? A medication that already achieved remission for her is often preferable to switching to an unfamiliar agent for a theoretical safety margin.
  • Where is she in the reproductive timeline? Preconception, first trimester, later pregnancy, and lactation each shift the calculus.
  • Can the regimen be simplified? The general preference is monotherapy at the lowest effective dose over polypharmacy, because fewer agents means a narrower exposure to characterize.

For a patient who has not previously taken medication for a mood condition, ACOG’s guideline identifies the SSRIs sertraline and escitalopram among first-line options for perinatal depression. That is a starting point for a conversation, not an instruction — dose, timing, and monitoring are set by the treating prescriber with the individual patient.

What happened to the old A/B/C/D/X pregnancy categories?

They were removed. In 2015 the FDA’s Pregnancy and Lactation Labeling Rule (PLLR) retired the letter categories for prescription drugs and replaced them with narrative risk summaries covering pregnancy, lactation, and reproductive potential.

The letters were retired because they were read as a grade — as if a “C” drug were uniformly riskier than a “B” — when they never encoded that. The PLLR narrative instead describes what is actually known: human data if it exists, animal data, and the drug’s pharmacology. For a prescriber, this means the label supports a specific conversation about a specific medication rather than a letter that flattens the evidence. When you see a referral note or a patient asking about a drug’s “pregnancy category,” the accurate answer is that categories no longer exist and the current label reads as a risk summary.

What about breastfeeding decisions?

Lactation is weighed separately from pregnancy, and it has its own primary reference: the NIH Drugs and Lactation Database (LactMed), which compiles the measured drug levels in milk and infant effects for individual medications. Sertraline, for example, is documented in LactMed as producing low levels in breastmilk, and the database notes that most authoritative reviewers consider it a preferred antidepressant during breastfeeding.

That is background education, not a recommendation for any individual. Whether to breastfeed while taking a given medication is a decision the patient makes with her prescriber, weighing her illness, the specific drug, and her own feeding goals. For the patient-facing version of this question, see psychiatric medication in pregnancy & breastfeeding.

Why is stopping at the positive test its own clinical event?

Because abrupt discontinuation of effective treatment is an intervention, with consequences the risk ledger has to count. Many patients stop psychiatric medication on their own the day the test turns positive, before anyone has weighed what that choice trades away. The stakes show up in the guideline’s own evidence: in the bipolar disorder data ACOG CPG No. 5 reviews, postpartum relapse occurred in 66% of patients not treated with pharmacotherapy versus 23% of those in whom pharmacotherapy was continued.

The clinical move is not to tell the patient to keep taking anything — it is to treat the positive test as the trigger for the risk-benefit conversation rather than the end of it. “Do not change anything yet; let’s get you in front of a perinatal prescriber this week” is a plan. Silence until the next scheduled visit is also a plan, just an unexamined one.

When should you refer instead of managing it yourself?

A risk-benefit conversation this individualized belongs with a prescriber who does it routinely. If a screen is positive, if symptoms are moderate to severe, or if a patient is asking whether to start, stop, or change a psychiatric medication around pregnancy, that is the point to bring in a perinatal prescriber rather than adjusting medication at the edge of your scope.

For how these pieces fit together across screening and hand-off, see the perinatal mental health referral guide for NC providers. To distinguish the presentations that route here, the baby blues vs. PPD vs. postpartum anxiety guide and the EPDS referral pathway set the thresholds, and the signals that warrant a medication evaluation cover the rest.

Mindful Counseling & Wellness is a North Carolina telehealth psychiatric practice led by a board-certified PMHNP with a perinatal specialty. If you have a pregnant or postpartum patient who needs a medication risk-benefit conversation, send a referral and we will take it from there.

Sources