The desk's full answer, unedited.

DEMO MODE — fictional scenario, no real PHI. Short answer: there is no single right answer, but for a patient with bipolar I who has been well for 3 years on lithium, abrupt discontinuation is usually the worst option. The decision is a maternal-relapse-risk versus fetal-risk tradeoff. In many such cases, a reasonable option is to continue lithium with informed consent and tighter monitoring, especially if prior illness was severe, recurrent, psychotic, or postpartum-associated. If she strongly prefers to stop, taper rather than abruptly stop; relapse risk rises substantially after discontinuation. A switch is possible, but evidence that switching early in pregnancy improves overall outcome is limited and switching itself can destabilize mood. Evidence is strongest for teratogenic risk being increased above baseline but lower than older historical estimates; evidence is thinner on the exact best switching strategy in an already-stable patient. How I would frame the options: 1. Continue lithium - Pros: preserves the treatment that has kept her euthymic; avoids destabilization during pregnancy, when bipolar relapse carries meaningful risk. - Cons: first-trimester fetal exposure is associated with an increased risk of congenital malformations, especially cardiac malformations, compared with unexposed pregnancies, though the absolute risk appears modest in modern cohort data. - When this option becomes more attractive: bipolar I, history of severe mania, psychosis, suicidality, rapid relapse off lithium, multiple prior episodes, strong prior lithium response, or prior postpartum episodes. 2. Taper and discontinue lithium - Pros: reduces fetal lithium exposure in organogenesis. - Cons: substantially increases maternal relapse risk, particularly if done rapidly. For a patient stable on lithium for years, this is often the main clinical hazard. - If chosen: taper over weeks, not abruptly, and have a contingency plan for early relapse detection. 3. Switch to another maintenance option - Pros: may reduce lithium-specific fetal risk. - Cons: not risk-free; evidence is limited for benefit of switching an already-stable lithium responder after pregnancy is discovered, and cross-tapering can precipitate relapse. Lamotrigine is often considered when the polarity is predominantly depressive and prior mania burden is lower, but it is generally weaker for classic mania prevention than lithium. Atypical antipsychotics are alternatives in some cases, but comparative pregnancy safety and efficacy vary by agent. - Net: switching is individualized, not automatically preferred just because pregnancy is recognized. What the evidence supports: Fetal risk with lithium - FDA labeling for lithium notes fetal/neonatal toxicity concerns and an increased risk of cardiac malformations with first-trimester exposure; labeling does not mandate discontinuation but emphasizes individualized risk-benefit assessment and serum monitoring. Source: FDA Prescribing Information for lithium products. Strength: regulatory labeling summarizing human and pharmacologic data. - The best-known modern cohort is Patorno et al., New England Journal of Medicine, 2017, which found first-trimester lithium exposure associated with an increased risk of cardiac malformations versus unexposed infants, with absolute risk notably lower than older reports suggested and a dose-response signal. Strength: large observational cohort; stronger than case series, but still nonrandomized. - Older literature linked lithium to Ebstein anomaly at much higher relative risk, but contemporary data suggest the absolute risk is lower than previously believed. Strength: historical observational evidence, superseded in part by newer cohort data. Maternal relapse risk if lithium is stopped - Viguera et al. showed high recurrence risk in pregnant women with bipolar disorder after mood stabilizer discontinuation, particularly with rapid discontinuation. Strength: observational prospective data; clinically influential though not randomized. - This is one of the strongest practical reasons not to stop lithium abruptly in a stable bipolar I patient. Guideline direction - ACOG Clinical Practice Guideline on treatment and management of mental health conditions during pregnancy/postpartum supports individualized risk-benefit counseling and does not treat pregnancy as an automatic reason to stop effective treatment; lithium may be continued when benefits justify fetal risk, with maternal serum monitoring and fetal cardiac evaluation considered. Strength: guideline consensus informed by observational evidence. - NICE guideline on antenatal/postnatal mental health similarly emphasizes individualized decision-making; if lithium is continued, use the lowest effective dose, monitor levels frequently, and coordinate obstetric care. Strength: guideline consensus. - APA-related guidance and perinatal psychiatry reviews generally align: continuation is often reasonable in severe lithium-responsive bipolar disorder. Strength: expert guidance/consensus, because randomized pregnancy data are sparse. My clinical synthesis for this case - She has bipolar I and sustained stability for 3 years on lithium with a therapeutic level of 0.8. On these facts alone, continuation is a strong option and often the option with the best chance of maintaining maternal wellness. - I would generally not recommend abrupt cessation. - If her prior course included severe mania, psychosis, hospitalization, suicidality, or postpartum episodes, that pushes further toward continuation. - If she had a comparatively mild course, no prior severe mania, strong aversion to fetal exposure, and willingness for close follow-up, taper/discontinue or switch could be considered, but that is a less evidence-secure path for maintaining stability. Monitoring schedule if lithium is continued Maternal lithium/renal/thyroid monitoring - Check immediately now: lithium trough, creatinine/BUN or CMP with renal function, TSH, calcium; consider urinalysis if clinically indicated. Document current dose, adherence, hydration, caffeine changes, vomiting, and interacting medications. - Lithium level target: use the lowest effective therapeutic concentration based on her prior response; many clinicians aim to remain in her established effective range rather than lowering reflexively into subtherapeutic territory. Evidence is thin on an exact pregnancy-specific target. - Frequency during pregnancy: - At least every 3-4 weeks until about 34 weeks. - Then weekly from 34 weeks until delivery. - Recheck sooner with hyperemesis, dehydration, preeclampsia, renal change, medication interactions, or clinical change. - Renal function: with the same cadence as lithium levels or at minimum once each trimester, and more often if there are obstetric/medical complications. - Thyroid: at baseline and at least once each trimester, because both pregnancy and lithium can affect thyroid function. - Calcium: baseline and periodically if clinically indicated, given lithium-associated hyperparathyroid/hypercalcemia risk. Why frequent levels are needed - Pregnancy increases GFR and volume of distribution, often lowering lithium levels during early/mid pregnancy; the dose sometimes must increase to maintain the pre-pregnancy therapeutic level. After delivery, renal clearance falls back quickly, so the same pregnancy dose can become toxic postpartum. Source support: named perinatal psychiatry reviews and guideline consensus; strength: physiologic and observational evidence. Fetal monitoring - Detailed anatomic ultrasound. - Fetal echocardiography is commonly recommended when first-trimester lithium exposure has occurred, typically around 18-22 weeks. This is based on the association with cardiac malformations, though evidence for exact yield is observational rather than trial-based. - Coordinate with maternal-fetal medicine/obstetrics. Peripartum management - Around labor/delivery, lithium management is individualized. Older practice often held or reduced lithium at labor onset to reduce maternal/neonatal toxicity risk; newer reviews suggest individualized management based on volume status, renal function, and delivery plan rather than universal discontinuation. Evidence here is thinner and practice varies. - Avoid dehydration. - Avoid interacting drugs that can raise lithium, especially NSAIDs; also watch ACE inhibitors/ARBs and thiazide diuretics if relevant. - Check lithium level close to delivery and again about 24 hours postpartum, then adjust dose downward toward the pre-pregnancy requirement as renal clearance normalizes. Postpartum - Postpartum is a very high-risk relapse period for bipolar disorder. If lithium was continued, close monitoring in the first 2 postpartum weeks is important because serum levels can rise quickly. If lithium was discontinued during pregnancy, restarting promptly postpartum is often considered in very high-risk patients. Strength: observational evidence and guideline consensus. If she prefers tapering off now - Avoid abrupt stop. - A gradual taper over at least several weeks is preferred; faster discontinuation appears associated with higher recurrence risk in observational data. - Increase visit frequency and have a low threshold for symptom-triggered reassessment. - If considering substitution, choose based on her polarity pattern and prior responses; evidence is limited for a universally superior switch strategy in this exact scenario. Bottom line - For a 34-year-old with bipolar I, euthymic for 3 years on lithium 900 mg nightly with level 0.8, newly 6 weeks pregnant and planning to continue the pregnancy, continuation of lithium with informed consent and intensified monitoring is often a reasonable, evidence-supported choice, especially if past illness was severe or clearly lithium-responsive. - If stopping is chosen, taper rather than abruptly discontinue. - Monitoring if continuing: lithium trough plus renal function every 3-4 weeks until 34 weeks, then weekly until delivery; thyroid each trimester; detailed anatomy scan and fetal echo around 18-22 weeks after first-trimester exposure; dose/readjust closely postpartum with a level within about 24 hours after delivery. Sources named - FDA Prescribing Information: lithium labeling. - Patorno E, et al. Lithium Use in Pregnancy and the Risk of Cardiac Malformations. New England Journal of Medicine. 2017. - Viguera AC, et al. Risk of recurrence in women with bipolar disorder during pregnancy after discontinuation of lithium/mood stabilizer treatment. Observational prospective study. - ACOG Clinical Practice Guideline: Treatment and Management of Mental Health Conditions During Pregnancy and Postpartum. - NICE Guideline: Antenatal and postnatal mental health. Evidence caveats - No randomized trials answer continue-versus-stop lithium in pregnancy. - Comparative evidence for switching an already-stable lithium responder after conception is limited. - Peridelivery lithium holding strategies vary; evidence is thinner than for first-trimester malformation risk and relapse risk after discontinuation.