Postpartum depression dossier — ACOG 757 2018/2024 + USPSTF 2019 (Curry JAMA PMID 30747971) + NICE NG192 2024 + APA 2010 PPD + PSI 2024 + Meltzer-Brody 2018 + Deligiannidis 2023 + Cox 1987 + Levis 2020 + Wisner 2013 + Stewart 2016 + Trost MMWR 2022 + Paulson 2010 + Hirschfeld 2000. Distinct from postpartum psychosis (psych.postpartum-psychosis.v1) — PPD incidence 10–15% births vs PP 1–2 per 1000 births; PPD predominantly outpatient (SSRI + psychotherapy + neurosteroid for severe) vs PP mandatory inpatient (antipsychotic + mood stabiliser). Emerging psychotic features in PPD route to psych.postpartum-psychosis.v1 per sibling_differentiation. Distinct from general MDD (psych.depression.core.v1) — perinatal-context drives screening cadence (EPDS / PHQ-9 at intake + 24–28 wk + delivery + 6-wk postpartum + 1/2/4/6-mo well-child visits), lactation-aware SSRI selection (sertraline / paroxetine first-line per LactMed 2024 + PSI 2024 + APA 2010 PPD), brexanolone / zuranolone pathway, partner / paternal PPD screening, and mother-infant attachment monitoring. Longitudinal MDD maintenance after 12-mo postpartum window transitions to psych.depression.core.v1. Axis 1 first-line pharmacotherapy + psychotherapy: CBT / IPT first-line for mild PPD (EPDS 10–12) per USPSTF 2019 B-grade. SSRI for moderate PPD (EPDS 13–19) — sertraline 50 mg PO daily titrate 100–200 mg first-line in lactation; paroxetine alternative first-line (avoid in next pregnancy — FDA 2015 category D); citalopram / escitalopram / venlafaxine / bupropion / nortriptyline acceptable; fluoxetine cautioned (long half-life + high M/P ratio) but not contraindicated. Axis 2 neurosteroid pathway for severe PPD: brexanolone (Zulresso) IV 60-h continuous infusion at FDA-Zulresso-REMS-accredited site for severe PPD (Meltzer-Brody Lancet 2018 PMID 30253856; FDA approval 2019); zuranolone (Zurzuvae) 50 mg PO daily × 14 days outpatient (Deligiannidis AJP 2023 PMID 37467049; FDA approval Aug 2023). Both gating on EPDS ≥ 20 + functional impact + (adequate SSRI trial failed at 4–6 wk OR rapid relief needed). REMS requirements: brexanolone continuous pulse-ox + 2-hourly cognitive assessment + no driving during + 48 h after; zuranolone CNS-depressant + driving advisory (no driving 12 h after each dose) + CYP3A4 caution + take with fatty food. Axis 3 lactation-aware selection: sertraline + paroxetine first-line per LactMed 2024; citalopram + escitalopram + venlafaxine + bupropion + nortriptyline acceptable; fluoxetine cautioned; brexanolone + zuranolone require lactation consult (pump-and-discard practices conservative; data limited); doxepin + MAOI avoided in breastfeeding. Severity triggers (8): ppd_with_si_or_infant_harm_ideation (life-threatening — ED + Stanley-Brown SPI + admit consideration; routes to psych.suicidality.ed.core.v1), ppd_with_psychotic_features (life-threatening — re-evaluate for postpartum psychosis; routes to psych.postpartum-psychosis.v1), severe_ppd_brexanolone_or_zuranolone_eligible (severe — brexanolone IV OR zuranolone PO at EPDS ≥ 20 + functional impact), bipolar_emergence_with_ppd_treatment (severe — STOP SSRI; routes to psych.bipolar-disorder.core.v1), breastfeeding_medication_selection_critical (moderate — sertraline / paroxetine first-line; do not separate mother-infant), postpartum_thyroiditis_overlap (mild — TSH + free T4 at 6 wk postpartum visit if symptomatic), infant_attachment_disturbance_emerging (moderate — mother-infant dyadic therapy + perinatal social work), paternal_partner_ppd_screening_positive (mild — partner referral; family-system approach). Four setting playbooks: outpatient (universal EPDS / PHQ-9 screening + perinatal psych Q2 wk × 12 wk then monthly × 12 mo + CBT/IPT + SSRI + lactation-aware selection + C-SSRS at every visit), ED (acute SI / crisis triage + 1:1 sitter + Stanley-Brown SPI + means restriction + cross-route), inpatient (severe PPD + SI admit + brexanolone REMS infusion + bridge ≤ 1 wk + Caring Contacts cascade), transition (bridge ≤ 1 wk after psychiatric admission + Caring Contacts + Stanley-Brown SPI follow-up + means re-confirmation). Sibling differentiation vs psych.depression.core.v1 (general MDD outside perinatal), psych.postpartum-psychosis.v1 (PP escalation if psychotic features), psych.suicidality.ed.core.v1 (SI carryover), psych.bipolar-disorder.core.v1 (bipolar emergence on SSRI). Cross-route via carryover state — EPDS, PHQ-9, C-SSRS, MDQ, breastfeeding status, current meds, prior trials. Phenotype matrix (severity EPDS × suicidality C-SSRS × bipolar screen MDQ × prior PPD/MDD × breastfeeding × anxiety comorbidity × infant attachment × paternal/partner PPD × access to brexanolone × access to zuranolone) encoded indirectly via severity_triggers + setting_playbooks + sibling_differentiation. First-class TS field for phenotype matrix is schema-blocked. Bayesian linkage: pre-test priors — PPD ~ 10–15% of births (ACOG 757 2018/2024; Wisner 2013); ~ 25–30% with prior MDD; ~ 35–50% with prior PPD (Wisner 2013; Stewart 2016). LR+ — EPDS ≥ 10 LR+ ~ 4 (Levis BMJ 2020); EPDS ≥ 13 LR+ ~ 7 (Levis 2020); PHQ-9 ≥ 10 LR+ ~ 5 (Kroenke 2001); MDQ ≥ 7 LR+ ~ 4 for bipolar spectrum (Hirschfeld 2000). T_treat = EPDS ≥ 13 OR functional impact; T_treat_brexanolone_zuranolone = EPDS ≥ 20 + functional impact + (SSRI failure OR rapid relief). T_test (supportive only) = EPDS < 10 + no functional impact + no SI. Cross-dossier routing edges (4): psych.postpartum-psychosis.v1, psych.suicidality.ed.core.v1, psych.bipolar-disorder.core.v1, psych.depression.core.v1. ROS/DDx LR seed data audited by npm run audit:ros-ddx-coverage (cross-cutting; not touched by this shard). CRITICAL anti-pattern warnings baked in: (1) DO NOT start SSRI in postpartum patient without bipolar screen (MDQ + manic history ask) — manic switch in unrecognised bipolar (STEP-BD Sachs NEJM 2007); (2) DO NOT default to discontinuing antidepressant during pregnancy or postpartum without psychiatric consult — relapse risk often exceeds in utero exposure risk (ACOG 757 2018/2024); (3) DO NOT discharge a postpartum patient with active SI plan / intent home from ED — admit consideration mandatory (Trost MMWR 2022 PMID 35587456); (4) DO NOT assume PPD without screening — universal EPDS / PHQ-9 at intake + 24–28 wk + delivery + 6-wk postpartum + well-child visits (ACOG 757 2018/2024; USPSTF 2019); (5) DO NOT prescribe fluoxetine first-line in actively-breastfeeding patient if alternative available — long half-life + high M/P ratio (LactMed 2024; PSI 2024); (6) DO NOT initiate brexanolone or zuranolone without REMS / informed-consent / driving / sedation precautions (FDA Zulresso REMS 2019; FDA Zurzuvae 2023); (7) DO NOT miss postpartum thyroiditis — symptom onset 3–6 mo postpartum + fatigue → TSH + free T4 (ATA 2017 PMID 28056690); (8) DO NOT overlook paternal / partner PPD — partner EPDS ≥ 10 → referral; paternal PPD ~ 10% (Paulson JAMA 2010 PMID 20483968); (9) DO NOT separate mother from infant for breastfeeding-compatible medications (LactMed 2024; PSI 2024); (10) DO NOT skip C-SSRS — maternal suicide leading cause of mortality through 1 y postpartum (Trost MMWR 2022); (11) DO NOT delay psychotic-feature re-evaluation — re-route to psych.postpartum-psychosis.v1 (ACOG 757 2018/2024); (12) DO NOT treat severe PPD (EPDS ≥ 20 + functional impact) as outpatient-monitor-only (Meltzer-Brody 2018; Deligiannidis 2023). PRODUCTION blockers: (1) prisma/seed/manifests/psych.postpartum-depression.v1.ts pointer out-of-shard-scope per psych.postpartum-psychosis.v1 / psych.bipolar-disorder.core.v1 / psych.first-episode-psychosis.core.v1 / psych.alcohol_withdrawal.core.v1 / psych.suicidality.ed.core.v1 precedent — manifest repointed to psych.depression.core.v1.ts; (2) calc.epds + calc.cssrs + calc.mdq not in clinical-tools-registry — flagged for future calc-registry batch (EPDS workflow embedded in setting-playbook required_assessments + non_drug_actions; C-SSRS via workup.suicide_risk umbrella; MDQ workflow embedded in CONTEXT phase required_inputs + outpatient required_assessments); (3) RxCUIs need RxNav validation via scripts/research/rxnav-validate.ts (brexanolone 2625139 + zuranolone 2570683 specifically flagged given recent FDA approvals); (4) targeted test file pending (relies on dossier-contract.test.ts); (5) brexanolone infusion-protocol atoms not yet registered — referenced inline in Axis 2 Step 1. Authored 2026-05-15 in shard-5-obped-id Phase C wave-2 dispatch — companion brief at src/lib/dossiers/_briefs/psych.postpartum-depression.v1.md + research bundle at src/lib/dossiers/_research-bundles/psych.postpartum-depression.v1.md.