Clinical Commander

All dossiers
pulm.pleural-effusion.core.v1

Pleural Effusion

pulmonologyacutesubacutechronicadultoutpatientacuteinpatient
Start encounter →Print handoutPRODUCTION

§5.5.2 Bayesian engine: Light's criteria encoded as the transudate↔exudate decision fork (exudate any 1 of 3, sens ~98% / spec ~83%, Light 1972 PMID 4642731); per-test LRs wired with PMIDs — pleural ADA >40 U/L LR+ ~9 (PMID 30913213), cytology pooled sensitivity 58% single / ~80% repeat (PMID 35110369), pH <7.20 → complicated parapneumonic drainage decision, NT-proBNP LR+ 15.2 for cardiac (PMID 20511623), serum-pleural albumin gradient >1.2 g/dL reclassifies diuretic-treated HF (PMID 17875051). Cross-dossier routing via workups[].branches_to: pulm.pe.core.v1, pulm.cap.core.v1, pulm.tuberculosis.v1, cardio.acute-hf.core.v1, gi.cirrhosis.core.v1 (all confirmed on disk). PE caveat encoded — PE effusion can be exudate OR transudate; trapped lung post-empyema noted. Drainage + substrate ladder stepwise (transudate-substrate / diagnostic-tap / parapneumonic-empyema / MPE / IPC / hemothorax / TB) with doses; intrapleural tPA 10 mg + DNase 5 mg BID × 3 d (MIST2 — combination only, monotherapy ineffective; pleural-bleeding contraindication encoded). Special populations: HF-on-diuretics (albumin-gradient/NT-proBNP correction), hepatic hydrothorax (TIPS/IPC, AVOID conventional chest tube — high-mortality severity_trigger), renal/urinothorax, chylothorax. Pregnancy/immunocompromised handled via TB + cytology pretest pathways. Effect-size numbers with PMIDs: RAPID 3-mo mortality bands (2.3/9.2/29.3%, PILOT PMID 32675200), AMPLE IPC vs talc hospital days (10 vs 12, PMID 29164255), TIME2 initial LOS (0 vs 4 d, PMID 22610520), ADA test characteristics (PMID 30913213), NT-proBNP LR (PMID 20511623). RxCUIs RxNav-verified 2026-05-16 — CORRECTED prior hand-authored errors: furosemide 4337→4603 (4337 is fentanyl), piperacillin-tazobactam 18631→1659131 (18631 is azithromycin), silver nitrate 2606543→9789 (2606543 is elemental silver). PRODUCTION blockers remaining: (1) Light's-criteria interpreter, RAPID, LENT scores not yet in clinical-tools-registry.ts (recorded as schema gap in .depth.md); (2) no engine-specific test file beyond shared dossier-contract; (3) ampicillin RxCUI 733 is ingredient-level (ampicillin/sulbactam combination CUI 1009148 used where available).

Entry points (4)

  • symptom
    Dyspnea, pleuritic chest pain, dull cough (BTS 2023 Roberts)
    dyspnea_pleuritic_pain
  • imaging
    Pleural effusion on CXR / lung ultrasound / CT (BTS 2023)
    effusion_on_imaging
  • symptom
    Diminished breath sounds, dullness on percussion, reduced tactile fremitus (BTS 2023)
    diminished_breath_sounds
  • problem_list
    Known HF / cirrhosis / malignancy with new effusion (BTS 2023)
    hf_cirrhosis_malignancy

Required inputs (21)

  • agerequired
    demographic • used at CONTEXT
    Etiology pretest probability (cancer, parapneumonic) shifts with age (BTS 2023)
  • spo2required
    vital • used at CONTEXT
    Hypoxia for thoracentesis urgency (BTS 2023)
  • rrrequired
    vital • used at CONTEXT
    Tachypnea + work-of-breathing severity (BTS 2023)
  • temprequired
    vital • used at CONTEXT
    Pleural infection screen (BTS 2023)
  • sbprequired
    vital • used at CONTEXT
    Septic shock screen in pleural infection (BTS 2023)
  • hf_cirrhosis_malignancy_renalrequired
    history • used at CONTEXT
    Transudative substrate identification + drives Light's albumin-gradient correction
  • diuretic_use
    history • used at INITIAL_WORKUP
    Diuretics concentrate transudate protein/LDH → pseudo-exudate; triggers albumin-gradient / NT-proBNP correction (Bielsa 2007 PMID 17875051)
  • recent_pneumonia_or_chest_trauma
    history • used at CONTEXT
    Parapneumonic vs hemothorax differential (BTS 2023)
  • anticoagulationrequired
    history • used at CONTEXT
    Pre-procedural bleeding risk for thoracentesis (BTS 2023)
  • serum_protein
    lab • used at INITIAL_WORKUP
    Light's criterion denominator — pleural/serum protein >0.5 (Light 1972 PMID 4642731)
  • serum_ldh
    lab • used at INITIAL_WORKUP
    Light's criterion denominator — pleural/serum LDH >0.6 or pleural LDH >2/3 ULN (Light 1972 PMID 4642731)
  • serum_albumin
    lab • used at INITIAL_WORKUP
    Serum-pleural albumin gradient >1.2 g/dL reclassifies diuretic-treated HF pseudo-exudate (Bielsa 2007 PMID 17875051)
  • pleural_fluid_ph
    lab • used at INITIAL_WORKUP
    pH <7.20 → complicated parapneumonic requiring drainage (BTS 2023; collect anaerobically in heparinised ABG syringe)
  • pleural_fluid_ada
    lab • used at BRANCHING_WORKUP
    ADA >40 U/L → LR+ ~9 for TB pleuritis (Aggarwal 2019 PMID 30913213)
  • pleural_fluid_cytology
    lab • used at BRANCHING_WORKUP
    Cytology pooled sensitivity 58% single sample; repeat → ~80% (PMID 35110369)
  • pleural_fluid_ntprobnp
    lab • used at INITIAL_WORKUP
    NT-proBNP for cardiac effusion misclassified by Light (LR+ 15.2; PMID 20511623)
  • platelets
    lab • used at INITIAL_WORKUP
    Platelets >=50 generally accepted for thoracentesis (BTS 2023)
  • inr
    lab • used at INITIAL_WORKUP
    INR <2 generally accepted for thoracentesis (BTS 2023)
  • cxrrequired
    imaging • used at INITIAL_WORKUP
    Initial confirmation + lateral decubitus for layering (BTS 2023)
  • lung_ultrasound
    imaging • used at INITIAL_WORKUP
    Bedside characterization (loculation, septations) + always US-guided thoracentesis (BTS 2023)
  • ct_chest
    imaging • used at BRANCHING_WORKUP
    Loculation, malignant features (pleural thickening, nodularity, mediastinal pleural involvement) (BTS 2023; ATS/STS 2018)

12-phase flow (12)

  1. 1FRAME
    Confirm pleural effusion (vs consolidation, atelectasis, elevated hemidiaphragm, pleural thickening) on CXR/US and determine acuity + bilaterality (bilateral → transudate-leaning)
    inputs: cxr
    advance: Effusion confirmed by imaging
  2. 2ENTRY
    Trigger from dyspnea + pleuritic pain, incidental imaging finding, or known substrate disease (HF/cirrhosis/malignancy)
    inputs: age
    advance: Entry trigger present
  3. 3CONTEXT
    Substrate (HF, cirrhosis, malignancy, renal, lupus, RA, post-CABG, asbestos, recent pneumonia/trauma, pancreatitis), diuretic use, allergies, anticoag, platelets/INR
    inputs: spo2, rr, temp, sbp, hf_cirrhosis_malignancy_renal, recent_pneumonia_or_chest_trauma, anticoagulation
    advance: Substrate captured + thoracentesis safety screen complete
  4. 4RED_FLAGS
    Tension hydrothorax (rare), empyema with sepsis/septic shock, hemothorax with hemodynamic instability, massive effusion with hypoxic respiratory failure or mediastinal shift
    inputs: spo2, sbp
    advance: Emergent drainage performed or escalated
  5. 5INITIAL_WORKUP
    CXR ± lateral decubitus, lung ultrasound (always before tap), BMP, CBC, LFTs, BNP if HF question; US-guided diagnostic thoracentesis if effusion >1 cm decubitus/US or unclear etiology — send pleural protein, LDH, albumin, glucose, pH (anaerobic), cell count + diff, Gram stain + aerobic/anaerobic culture, cytology, plus ADA/AFB/TB-PCR, amylase, triglyceride, NT-proBNP per pretest
    inputs: serum_protein, serum_ldh, serum_albumin, pleural_fluid_ph, pleural_fluid_ntprobnp, platelets, inr, cxr, lung_ultrasound
    advance: Thoracentesis performed (if indicated) and Light's rule applied
  6. 6BRANCHING_WORKUP
    Apply per-test LRs (ADA→TB, cytology→malignancy, pH<7.2→drainage, NT-proBNP/albumin-gradient→cardiac). CT chest with contrast if exudate / suspected malignancy / loculation; medical thoracoscopy + parietal pleural biopsy if undiagnosed exudate (cytology-negative); RAPID score for pleural infection
    inputs: ct_chest, pleural_fluid_ada, pleural_fluid_cytology
    actions: pe_full
    advance: Etiology identified or thoracoscopy referral made
  7. 7DIFFERENTIAL
    Bayesian fork: Light's exudate (any 1 of 3 criteria) → parapneumonic/empyema, malignant, TB, PE, autoimmune, chylothorax, post-CABG; Light's transudate → HF, cirrhosis (hepatic hydrothorax), nephrotic, hypoalbuminemia, urinothorax. Apply albumin-gradient/NT-proBNP correction for diuretic-treated HF; remember PE can be exudate OR transudate and post-empyema can leave trapped lung
    inputs: serum_protein, serum_ldh, serum_albumin, pleural_fluid_ntprobnp, pleural_fluid_ada
    advance: Transudate vs exudate assigned with corrected interpretation; etiology posterior set
  8. 8RISK_STRATIFICATION
    RAPID score for pleural infection 3-month mortality (low 2.3%, medium 9.2%, high 29.3%; PILOT Corcoran 2020 PMID 32675200); LENT score for MPE survival; MELD/MELD-Na for hepatic hydrothorax
    actions: calc.meld_na, calc.meld3, calc.curb65
    advance: Risk score documented
  9. 9TREATMENT
    Transudate → treat underlying disease (diuresis for HF; spironolactone+furosemide ± TIPS for hepatic hydrothorax — AVOID chest tube; RRT for renal). Complicated parapneumonic/empyema → chest tube + abx; intrapleural tPA 10 mg + DNase 5 mg BID × 3 d (MIST2 PMID 21830966; monotherapy ineffective). MPE → IPC vs talc pleurodesis (AMPLE PMID 29164255 / TIME2 PMID 22610520); IPC preferred for trapped lung. Hemothorax → large-bore tube ± surgery. TB pleurisy → RIPE ATT.
    advance: Drainage strategy + targeted therapy initiated
  10. 10DISPOSITION
    Admit for chest tube/empyema; outpatient drainage for chronic MPE via IPC; surgical (VATS decortication) referral for trapped lung / failed pleurodesis / day-5–7 drainage failure
    advance: Disposition set
  11. 11MONITORING
    Drain output, daily post-tube CXR, repeat US, fever/CRP curve in pleural infection, IPC site care, reaccumulation rate
    advance: Drainage trend documented
  12. 12FOLLOWUP
    Outpatient IPC management (3×/week drainage, taper for spontaneous pleurodesis ~46% by 70 d on AMPLE), oncology re-imaging, HF/cirrhosis substrate optimisation, repeat thoracentesis if recurrent, asbestos counselling + smoking cessation
    advance: Drainage and substrate plan in place; clinic follow-up scheduled