Clinical Commander

All dossiers
cardio.hfmref.core.v1

Heart failure with mildly reduced EF (HFmrEF, LVEF 41–49)

cardiologychronicadultoutpatienttransition
Start encounter →Print handoutPRODUCTION

Phenotype-specific MECE-on-EF split from cardio.hf.core.v1 — HFmrEF (LVEF 41–49) only. Dynamic phenotype: serial echo at 3–6 mo re-routes to cardio.hfref.core.v1 (EF ≤40) or cardio.hfpef.core.v1 (EF ≥50); prior LVEF ≤40 → HFimpEF (continue full HFrEF GDMT — do NOT relabel/de-escalate, TRED-HF). Chronic parent cardio.hf.core.v1 (DEPTH-PASS-2 complete) cross-referenced not edited; siblings cardio.hfref/hfpef.core.v1 deepened in parallel. Manifest points at existing sibling cardio.acute-hf.core.v1.ts per nearest-ID precedent so the audit broken_pointers check passes; decision surface (2 regimen_axes + workups + panels + calculators), test_files, full evidence object, and chronic phases all present. DEPTH-PASS-2 2026-05-16 (shard-07-cardio-chronic, golden-template-mirrored on cardio.htn.core.v1) added: (1) co-located _design-brief.md + _research-bundle.md per §5.5 items 1+2 (18 verified PMIDs live PubMed-MCP-checked, named trials + effect sizes + 95% CI + retrieval-dated, subgroup-vs-dedicated evidence weighting explicit, Consensus→WebSearch fallback logged, bidirectional EF-band reclassification routing); design_brief: repointed to src/lib/dossiers/cardio.hfmref.core.v1._design-brief.md (was non-existent _briefs/ path). (2) cardio.hfmref.core.v1 differentials+ros+finding-lrs seed files NEW (10 differentials w/ cohort-anchored priors on the mid-range partition: stable-HFmrEF/HFimpEF/HFpEF-borderline/HFrEF-declining + ischaemic/non-ischaemic + tachy/infiltrative/valvular/look-alike; 15 ROS; 33 LR rows = 26 LR+/24 LR− incl. serial-EF-trend, prior-EF≤40→HFimpEF, NT-proBNP bands, ischaemic substrate; 3 conditional-dependency rules incl. EF-band|serial-trend and BB-prognostic|rhythm; T_test≈5%/T_treat≈60%, prior-≤40 a hard gate). (3) 2nd regimen axis hfmref_phenotype_trajectory_matrix (drug × EF-trajectory/phenotype gating as data: HFimpEF-continue-GDMT / ischaemic / AF / CKD / DM2 / lower-mid-range-41-44 vs upper-45-49). (4) RxCUI bugs fixed vs RxNav-validated DrugEffectProfile registry: sac/val 1656340→1656339, metoprolol-succ 866427→221124, lisinopril 18867→29046, ARB anchor losartan→candesartan 214354 (CHARM-Preserved trial drug). (5) Guideline content refresh (not just tag): SGLT2i Class I (ESC 2023)/2a (AHA 2022) first-line ALL; ARNi/MRA/BB demoted to explicit Class 2b "may be considered" subgroup-derived w/ CI-crosses-1 honesty; finerenone (FINEARTS-HF) preferred MRA; HFimpEF=continue-GDMT hard gate (TRED-HF); evidence.pmids fixed (Heidenreich 35379504→35363499/35379503, CHARM 13678868→13678871, FAIR-HF dropped for verified AFFIRM-AHF, EAST-AFNET4 32673028→32865375, TRANSFORM-HF 36342178→36648467, SOLVD/STRONG-HF added) → 19 verified. INTEGRATED (not PRODUCTION): all RxCUIs now RxNav-validated (research:rxnav:validate, IN/PIN/MIN:OK; sac/val reverse-MISMATCH is the accepted documented registry annotation); non-pharm entries (IV iron, revascularization) marked non_pharm; 96-fail rxnav baseline + carvedilol_cr SCD:REJECT out-of-scope per prompt. PEP-CHF (16693536) dropped from evidence.pmids — no perindopril profile, identifier-pending, flagged next-pass. 10 special-population / migration severity triggers (EF migration ↓/↑, CKD, DM2, AF, pregnancy, geriatric-frailty, hepatic, ischemic etiology, hyperkalemia). DEPTH-PASS-3 2026-05-26 (lane-E): +NMA (Kotecha Lancet 2014 BB IPD HF+sinus HR 0.73 vs HF+AF HR 0.97 — directly anchors the HFmrEF Class 2b BB-sinus-only framing; Song Postgrad Med J 2024 SGLT2i/GLP-1RA/DPP4i hypoglycemic NMA — SGLT2i ranked first across EF) +Cochrane (Long 2019 CD003331 exercise-based CR for HF) +USPSTF (HF screening NOT a USPSTF topic — explicitly flagged; HFmrEF is a dynamic-phenotype diagnosis emerging from serial echo follow-up, not amenable to population screening; HTN A-2021 + statin B-2022 + smoking A-2021 hooks for ischemic-HFmrEF aetiology prevention) +ICER (Bhatt JAMA Cardiol 2023 sac/val HFmrEF subset $67,331/QALY at EF ≤55 — meaningfully better CE than HFpEF subset; Davis/McEwan EJHF 2024 dapa £6,470/QALY UK NHS HFmrEF subset retains CE) +Pauker-Kassirer decision thresholds explicit (RAS/MRA/BB "may be considered" Class 2b T_treat≈50–70% — subgroup CIs cross 1, raised T_treat = transparent decision-arithmetic basis for the AHA Class 2b language; HFimpEF prior-≤40 is a hard gate not a threshold per TRED-HF); side-car at cardio.hfmref.core.v1._depth-pass-3.md. Zero schema churn; 6 new PMIDs live-verified via PubMed MCP 2026-05-26. HFmrEF-specific finerenone/spironolactone/BB CE flagged for W2 ICER cache backfill rather than fabricated.

Entry points (6)

  • imaging
    Echo LVEF 41–49 — HFmrEF phenotype
    echo_lvef_41_49
  • symptom
    Dyspnea on exertion / orthopnea / PND
    dyspnea
  • symptom
    Lower-extremity edema
    lower_extremity_edema
  • lab_abnormality
    NT-proBNP elevated with mildly reduced LVEF
    nt_probnp_elevated
  • problem_list
    Known HFmrEF — surveillance / titration visit
    hfmref_existing
  • history
    Recent HF hospitalisation (transition of care)
    recent_hf_admission

Required inputs (19)

  • agerequired
    demographic • used at CONTEXT
    Age-adjusted NT-proBNP thresholds + drug tolerability
  • sbprequired
    vital • used at CONTEXT
    RAS/ARNi/MRA initiation requires SBP ≥100; monitor for hypotension
  • hrrequired
    vital • used at CONTEXT
    BB titration target; AF vs sinus changes BB indication (Cleland IPD 2018)
  • weightrequired
    vital • used at CONTEXT
    Weight diary for congestion; loop diuretic titration
  • creatininerequired
    lab • used at CONTEXT
    eGFR gates SGLT2i (≥20), finerenone (≥25), ACEi/ARNi titration
  • potassiumrequired
    lab • used at CONTEXT
    Finerenone/spironolactone contraindicated K >5.0; baseline before RAS/MRA
  • lvefrequired
    imaging • used at DIFFERENTIAL
    HFmrEF gate (LVEF 41–49); migration in/out of band drives re-routing
  • prior_lvef_le_40required
    history • used at FRAME
    If LVEF was ever ≤40 → HF-improved (continue HFrEF GDMT), NOT HFmrEF
  • nt_probnp
    lab • used at INITIAL_WORKUP
    Diagnostic confirmation + monitoring; age-adjusted thresholds
  • iron_panel
    lab • used at INITIAL_WORKUP
    IV iron in symptomatic HF + iron deficiency (AFFIRM-AHF)
  • tsh
    lab • used at INITIAL_WORKUP
    Thyroid dysfunction as reversible HF cause/aggravator
  • a1c
    lab • used at INITIAL_WORKUP
    DM co-prevalence drives SGLT2i + finerenone selection
  • ecg
    imaging • used at INITIAL_WORKUP
    AF vs sinus (BB indication), ischemic Q-waves, low-voltage (amyloid)
  • cad
    history • used at CONTEXT
    Ischemic etiology more common in HFmrEF — drives revascularization workup
  • atrial_fibrillation
    history • used at CONTEXT
    AF abolishes BB mortality benefit; drives rate/rhythm + AC
  • ckd
    history • used at CONTEXT
    CKD stage gates SGLT2i, finerenone, RAS titration
  • dm2
    history • used at CONTEXT
    Dual cardiorenal indication — favors SGLT2i + finerenone
  • current_meds
    medication • used at CONTEXT
    Detect existing HF therapy; flag NSAIDs / non-DHP CCB / glitazones
  • nyha_classrequired
    symptom • used at RISK_STRATIFICATION
    NYHA II–IV drives visit cadence + advanced-HF thresholds

12-phase flow (12)

  1. 1FRAME
    Confirm chronic HFmrEF; if prior LVEF ≤40 → route cardio.hf-improved.core.v1; if decompensated → cardio.acute-hf.core.v1
    inputs: prior_lvef_le_40, nyha_class, sbp
    advance: EF in 41–49 band, no prior ≤40, hemodynamically stable
  2. 2ENTRY
    Recognise triggering echo/lab/symptom and capture identity
    inputs: age
    advance: one entry trigger present
  3. 3CONTEXT
    Vitals, K/Cr, comorbidities (CAD, HTN, AF, DM, CKD), current meds
    inputs: sbp, hr, weight, creatinine, potassium, cad, atrial_fibrillation, ckd, dm2, current_meds
    advance: vitals/labs and comorbidity screen complete
  4. 4RED_FLAGS
    Screen for decompensation, cardiogenic shock, hyperkalemia, AKI on RAS/MRA
    inputs: sbp, creatinine, potassium
    actions: cardiogenic_shock, acute_pulm_edema
    advance: no red flags or routed to acute pathway
  5. 5INITIAL_WORKUP
    NT-proBNP, BMP, TSH, iron panel, A1c, ECG, echo with global longitudinal strain
    inputs: nt_probnp, tsh, iron_panel, a1c, ecg
    actions: panel.cardiac, panel.renal, panel.thyroid, panel.iron
    advance: baseline labs returned; LVEF + strain documented
  6. 6BRANCHING_WORKUP
    Ischemic (stress/cath) vs non-ischemic (CMR) etiology; infiltrative screen (PYP, serum/urine free light chains) when red-flag features
    inputs: lvef, cad
    actions: le_edema, afib_new_onset
    advance: etiology established or referred
  7. 7DIFFERENTIAL
    HFmrEF vs HFpEF (EF ±5% band) vs HFrEF vs HF-improved vs high-output vs valvular vs infiltrative vs tachycardia-mediated
    inputs: lvef
    advance: phenotype confirmed as de-novo HFmrEF
  8. 8RISK_STRATIFICATION
    NYHA, MAGGIC, NT-proBNP trajectory; ICD only with an independent indication (no routine primary prevention at EF >35)
    inputs: nyha_class
    advance: NYHA documented; high-risk flag if applicable
  9. 9TREATMENT
    SGLT2i (Class I 2023 ESC) → finerenone/MRA → ARNi/ARB/ACEi + BB (2b; BB sinus-only) → loop diuretic; comorbidity-directed therapy + revascularization where indicated
    inputs: lvef, sbp, hr, creatinine, potassium, dm2, ckd
    advance: SGLT2i started (unless contraindicated) + comorbidity-directed plan documented
  10. 10DISPOSITION
    Visit cadence by NYHA; advanced-HF referral if migrating to HFrEF Stage D
    inputs: nyha_class
    actions: preop_cardiac
    advance: next visit interval set; referral made if criteria met
  11. 11MONITORING
    BMP within 1–2 wk of RAS/MRA change; weight diary; serial echo 3–6 mo for phenotype migration; NT-proBNP q3–6 mo
    inputs: creatinine, potassium
    actions: panel.renal
    advance: monitoring plan documented incl. echo re-phenotype interval
  12. 12FOLLOWUP
    Cardiac rehab, vaccinations, re-phenotype: re-route to HFrEF if EF ≤40 or HFpEF if EF ≥50
    inputs: lvef
    advance: follow-up scheduled with re-phenotype trigger documented