Perinatal psychopharmacology risk–benefit and class hierarchy (MEQ)

(i) Untreated illness. Do not accept “stop everything today.” Bipolar recurrence after mood-stabiliser discontinuation in pregnancy is high (Viguera prospective data); postpartum is a peak window for mania/psychosis (Bergink). Depression-spectrum untreated illness also links to preterm birth and growth restriction (Grote). Frame suicide, care capacity, and obstetric harm alongside fetal drug risk.[1][7][8]

(ii) Valproate. EURAP shows high, dose-related major congenital malformation risk with valproate; NEAD shows adverse cognitive outcomes after fetal valproate. Plan urgent specialist switch off valproate, teratogen counselling, obstetric dating and anomaly-scan pathway, and an alternative stabiliser plan (continue/adjust quetiapine and/or introduce lithium or lamotrigine based on polarity history). Document shared decision with obstetrics.[2][3]

(iii) Lithium alternative. Patorno NEJM: small absolute increase in cardiac malformations after first-trimester lithium versus non-exposed — shared decision, not superstition. If chosen: more frequent levels as GFR changes, anomaly scan/fetal echo per local protocol, peripartum level/neonatal observation plan, and postpartum prophylaxis mindset.[4][6]

(iv) If depression/SSRI later. Late-pregnancy SSRI exposure associates with poor neonatal adaptation (Grigoriadis); rare PPHN signal needs absolute-risk language. Observation plan beats panic third-trimester cessation without relapse planning.[5][6]

(v) Lactation. Prefer lowest effective monotherapy; use relative infant dose and infant maturity; sertraline often favoured among antidepressants when efficacy allows; lithium/clozapine need specialist monitoring frameworks; do not ban breastfeeding by psychotropic class label alone.[6]