Pleural Effusion

Meet the patient

A 68-year-old former smoker arrives with two weeks of progressive breathlessness and a sharp, pleuritic pain over the right base that has eased as he became more breathless. He is tachypnoeic; the right base is stony dull with absent breath sounds and reduced vocal resonance, and the trachea is pushed slightly to the left.[2]

The two questions that decide his next 24 hours are the two that decide every pleural effusion: is the lung pushed away by fluid, or pulled in by collapse? (the trachea answers at the bedside — pushed away means effusion), and once you have confirmed fluid, is it a transudate or an exudate? (the tap answers that). Hold those two questions and everything below slots into place.[1]

Light's criteria — the one fork that runs everything

A pleural effusion is not a diagnosis; it is a sign of an underlying disease. The single most powerful diagnostic step, introduced by Richard Light in 1972 and unchanged in half a century, is to classify the fluid as a transudate or an exudate. That one distinction collapses a differential of dozens of causes into two manageable lists, and it dictates which further fluid tests are worth sending.[1][2]

Light's criteria compare simultaneous serum protein and LDH against the pleural fluid. The effusion is an exudate if any ONE of the three holds:[1]

• Pleural fluid protein divided by serum protein over 0.5
• Pleural fluid LDH divided by serum LDH over 0.6
• Pleural fluid LDH over two-thirds the upper limit of normal for serum LDH[1]

The criteria are highly sensitive (about 99 per cent — they will not miss an exudate) but less specific (about 75 per cent), and they misclassify about a quarter of cardiac effusions in patients on diuretics as exudates, because diuresis concentrates the protein and LDH in the residual fluid. The two rescue tests for that trap are the serum-to-pleural-fluid albumin gradient (over 1.2 g/dL supports a transudate) and pleural fluid NT-proBNP (over 1500 ng/L strongly supports a cardiac cause).[2][10]

The four mechanisms — and which fork they land on

A transudate forms when systemic Starling forces are disturbed across a normal pleura; an exudate forms when the pleura itself is diseased. Recognising the mechanism predicts the transudate-exudate split before the tap.[1][10]

1. Increased hydrostatic pressure (transudate). Pulmonary venous pressure rises in left heart failure, mitral stenosis, fluid overload and constrictive pericarditis. Filtration exceeds lymphatic reabsorption; the fluid is protein-poor because the intact pleura still excludes large molecules.
2. Decreased oncotic pressure (transudate). Hypoalbuminaemia from cirrhosis, nephrotic syndrome or severe malnutrition lowers plasma oncotic pressure. Hepatic hydrothorax is the paradigm — ascites tracks through small diaphragmatic defects, more common on the right where the tendinous fibres are thinner, into the negative-pressure pleural space.
3. Increased capillary permeability (exudate). Inflammation from pneumonia, autoimmune disease, pulmonary embolism with infarction, pancreatitis or drugs opens intercellular junctions and lets protein-rich fluid leak in. Malignancy increases permeability directly through tumour cytokines (vascular endothelial growth factor).
4. Impaired lymphatic drainage (exudate). Obstruction by tumour or nodes raises back-pressure and traps protein, so even normal-constituency fluid accumulates as an exudate.[1]

Etymology for viva gold: the pleura is from the Greek pleura, "rib" or "side" — the membrane lining the side of the chest. The normal space holds only 5 to 15 mL of lubricant; the lymphatic pump can raise its clearance roughly twenty-eight-fold before it is overwhelmed, which is why early effusions are silent.[2]

How common, who, and which side

Pleural effusion is one of the commonest problems in hospital medicine, with an estimated annual incidence in the United States of over 1.5 million. Heart failure alone accounts for roughly 40 per cent of all effusions, parapneumonic effusion about 25 per cent, malignancy 15 to 25 per cent, with pulmonary embolism, tuberculosis and autoimmune disease making up much of the rest.[2][10]

The epidemiology is the epidemiology of its causes, and the side matters at the bedside. Roughly two-thirds of cardiac effusions are bilateral, and when unilateral they are right-sided in about two-thirds — because the larger right lung and pleural space and the gravitational distribution of pulmonary venous congestion favour the right. A unilateral right-sided effusion in heart failure is the common explanation, not a reason to chase an alternative cause first.[10]

In tuberculosis-endemic regions (India, sub-Saharan Africa, South-East Asia) tuberculous pleuritis is a leading cause of a unilateral exudate, whereas in western series malignancy dominates that slot. A parapneumonic effusion complicates roughly 20 to 40 per cent of bacterial pneumonias that need hospital admission; a minority progress to empyema.[5][10]

At the bedside — stony dull and shifted away

The classical symptom triad is progressive exertional dyspnoea, pleuritic chest pain and a dry cough. Dyspnoea is the commonest and most reliable symptom; pleuritic pain arises from the parietal pleura (the visceral pleura is insensate) and therefore points to an exudative cause — pneumonia, pulmonary embolism, autoimmune — rather than a transudate. The pain often eases as fluid accumulates and separates the inflamed pleural surfaces.[2]

Examination yields one of the most distinctive bedside findings in medicine. The affected hemithorax is expanded, moves less on respiration, and is dull to percussion. The percussion note is "stony dull" — a dense, thud-like quality distinct from the woody dullness of consolidation, because fluid transmits vibration poorly. Breath sounds and vocal resonance are reduced or absent over the fluid, and a pleural rub may sit at the upper margin where the two pleural surfaces still rub. The trachea and apex beat are deviated away in a massive effusion.[2]

Everyone forgets the trachea. Effusion pushes the mediastinum away; collapse pulls it toward. That single sign separates the two at the bedside before any imaging, and it is the discriminator examiners listen for.[4]

The differential of dullness with reduced breath sounds

The bedside finding of dullness with reduced breath sounds has a focused differential, and the examiner expects you to distinguish each by the features that separate them.[2][4]

[2]

The discriminator line: trachea away equals effusion; trachea toward equals collapse; bronchial breathing with increased resonance equals consolidation.[2]

Diagnostic thoracentesis — what to send, what to skip

The investigation proceeds in two phases: confirm the effusion (imaging), then identify the cause (pleural fluid analysis). The cornerstone is the diagnostic thoracentesis with pleural fluid interpreted through Light's criteria.[1][4]

Tap every unexplained unilateral effusion, every bilateral effusion with discordant appearances, any effusion that does not resolve with heart failure treatment, and whenever infection or malignancy is suspected. Perform it under ultrasound guidance — ultrasound confirms fluid, detects loculation and septation a CXR cannot see, finds the safe entry point and depth, and cuts the iatrogenic pneumothorax rate from blind taps to under 1 per cent.[4]

The first and decisive step is to apply Light's criteria to the fluid, comparing simultaneous serum protein and LDH. Once classified as a transudate, no further fluid tests are usually needed — the workup shifts to the systemic cause. Once it is an exudate, a panel of additional tests refines the diagnosis:[1][2]

• Glucose — under 3.3 mmol/L (60 mg/dL): empyema, malignancy, rheumatoid, TB, oesophageal rupture.
• pH — under 7.2: empyema, complicated parapneumonic, oesophageal rupture; 7.2 to 7.3 borderline.
• Cell count — neutrophils point to bacterial, PE or pancreatitis; lymphocytes to TB, malignancy or chronic disease; eosinophils to drug, air, blood or parasitic causes.
• Gram stain and culture, AFB — bacterial and mycobacterial.
• Cytology — malignant cells; sensitivity 60 to 90 per cent, higher with repeat samples.
• ADA — over 40 U/L in a lymphocytic effusion supports TB.
• Amylase — high in pancreatitis, pancreaticopleural fistula, oesophageal rupture.
• Triglycerides — over 1.24 mmol/L (110 mg/dL): chylothorax.[2][4]

The macroscopic appearance gives the first clue: serous and straw-coloured (transudate or early exudate), bloody (malignancy, trauma, PE, iatrogenic), turbid or purulent (empyema), milky white (chylothorax or pseudochylothorax), greenish (rheumatoid pleuritis), black (Aspergillus, rheumatoid), food particles (oesophageal rupture).[2]

[1] [4]

The drainage thresholds — pH, glucose, pus

The decision to drain a parapneumonic effusion rests on three fluid values: a pH under 7.20, a glucose under 3.3 mmol/L (60 mg/dL), or frankly purulent fluid (or a positive Gram stain) — any one mandates a chest drain. The pH must be measured anaerobically in a heparinised blood-gas syringe, on fluid free of lidocaine, with no air introduced. A pH between 7.20 and 7.30 is a grey zone; drain if there is also loculation or a large effusion.[5]

Parapneumonic effusions stratify into three stages that map directly onto management:[5]

[5] [6]

Resuscitation — the four time-critical effusions

Most effusions are managed electively, but a small number present as time-critical emergencies in which the resuscitation reflex precedes the diagnostic workup. The dangerous scenarios are tension pleural effusion, massive haemothorax, empyema with septic shock, and severe hypoxaemia from a large malignant effusion.[2][5]

Tension pleural effusion is rare but lethal — a large effusion under pressure that shifts the mediastinum away, compresses the great veins, and produces hypotension, tachycardia, marked dyspnoea and hypoxaemia. Immediate management is emergency needle aspiration (a 14 to 16 gauge cannula in the safe triangle) followed promptly by an intercostal chest drain.[2]

Massive haemothorax (over 1500 mL drained immediately, or over 200 mL per hour for two to four hours) is a surgical emergency. Insert a large-bore (28 to 36 French) intercostal drain, resuscitate with intravenous crystalloid and blood, correct coagulopathy, and refer immediately to thoracic surgery.[2]

Empyema with septic shock requires simultaneous resuscitation and source control: intravenous fluids, vasopressors for refractory hypotension, broad-spectrum antibiotics (a beta-lactam-beta-lactamase inhibitor or carbapenem plus anaerobic and MRSA cover as appropriate), and urgent image-guided chest drain, with intrapleural tPA plus DNase added for loculated collections.[5][6]

The one procedural limit everyone must know: a single thoracentesis must not drain more than 1.5 L (or stop if the patient develops chest discomfort, persistent cough, vasovagal symptoms or a drop in oxygen saturation) because of re-expansion pulmonary oedema — too-rapid lung re-expansion injures the pulmonary capillaries and floods the alveoli, within 24 hours of drainage.[2]

Parapneumonic effusion and empyema — MIST-2 changed the rule

Antibiotics for pleural infection must cover typical community pathogens plus anaerobes — a beta-lactam-beta-lactamase inhibitor such as amoxicillin-clavulanate 1.2 g IV every 8 hours, or ceftriaxone plus metronidazole; hospital-acquired empyema needs MRSA cover (vancomycin or teicoplanin) and gram-negatives including Pseudomonas (piperacillin-tazobactam or meropenem). Duration is 2 to 6 weeks intravenously then orally.[5]

Intrapleural tPA plus DNase was established by the MIST-2 trial (Rahman 2011): alteplase 10 mg plus dornase alfa 5 mg, each daily for three days, significantly increased drainage and reduced surgical referral and length of stay compared with either agent alone or placebo. The trap examiners love: tPA alone or DNase alone was no better than placebo, and DNase alone was actually harmful. The combination is now first-line for a loculated, poorly draining empyema.[6]

Surgical decortication (VATS or open) is indicated for failed medical therapy and multiloculated empyema in a fit patient. The RAPID score (Rahman 2014) — Renal (urea), Age, Purulence, Infection source (hospital-acquired scores worse), Dietary (albumin) — stratifies three-month mortality (low 1 to 5 per cent, medium 10 to 15 per cent, high 25 to 45 per cent) and guides the threshold for early surgery.[5][9]

Malignant effusion — IPC or talc, take your pick

A malignant effusion is a symptom-control problem layered on an oncology problem. The commonest primaries are lung (about a third), breast (about a fifth), lymphoma and ovarian, with mesothelioma a less common but important asbestos-related cause. The fluid is usually a bloody exudate, with malignant cells on cytology in 60 to 90 per cent.[3][11]

The TIME2 trial (Davies 2012) compared an indwelling pleural catheter with chest tube and talc slurry pleurodesis and found equivalent dyspnoea relief at 42 days, with the IPC group reporting better dyspnoea at six months and a shorter initial hospital stay. The IPC-Plus trial (Bhatnagar 2018) then showed that 4 g of sterile talc slurry instilled through an existing IPC produced successful pleurodesis at 35 days in 43 per cent versus 23 per cent with placebo, without excess harm — combining the convenience of an IPC with the pleurodesis effect.[3][7][8]

The BTS 2010 guideline and current ATS, STS and STR guideline therefore present IPC and talc pleurodesis as equivalent first-line options, with the choice guided by lung re-expansion (a trapped lung precludes pleurodesis and mandates an IPC) and expected prognosis. Mesothelioma-related effusions deserve thoracoscopy with biopsy for diagnosis and talc poudrage pleurodesis in the same sitting where possible.[3][11]

The specific-cause one-liners that earn marks

Tuberculous pleuritis is treated with standard antituberculous therapy — 2 months of HRZE plus 4 months of HR. The effusion is usually small, unilateral, lymphocyte-predominant, with ADA over 40 U/L. Drainage is rarely needed; it is a delayed hypersensitivity reaction to mycobacterial antigens, which is why pleural fluid cultures are positive in only 20 to 30 per cent.[10]

Hepatic hydrothorax is a transudative effusion in cirrhotic portal hypertension, typically right-sided (about 70 per cent), tracking from the peritoneal cavity through tiny diaphragmatic defects. Never put a chest tube — it causes protein and fluid depletion, infection, and a persistent bronchopleural fistula. Treat the ascites (spironolactone 100 to 400 mg daily plus furosemide 40 to 160 mg daily), offer serial therapeutic thoracentesis, and refer for TIPS or liver transplantation. The dangerous complication is spontaneous bacterial empyema.[12]

Chylothorax is milky, alkaline, sterile, lymphocyte-predominant, with triglycerides over 1.24 mmol/L (110 mg/dL). Causes are trauma or surgery, malignancy (lymphoma is the classic non-traumatic cause), congenital and idiopathic. Conservative therapy — a low-fat, medium-chain triglyceride diet (medium-chain triglycerides bypass the thoracic duct, absorbed directly into the portal vein) or total parenteral nutrition, plus octreotide 50 to 200 micrograms subcutaneously three times daily — allows spontaneous closure in many cases. Persistent high-output chylothorax needs surgical thoracic duct ligation.[2][10]

Rheumatoid pleuritis is distinctive — a very low glucose (often under 1.6 mmol/L, sometimes under 0.3), very low pH (under 7.2), very high LDH (over 1000), high rheumatoid factor, low complement, and a lymphocytic predominance. Managed with NSAIDs and corticosteroids.[2]

A consultant confession: distinguish chylothorax from pseudochylothorax. A true chylothorax is acute, has triglycerides over 1.24 mmol/L, and is from a thoracic duct leak; a pseudochylothorax is chronic (rheumatoid, old TB), looks milky but has high cholesterol and low triglycerides with cholesterol crystals, and is not managed with a low-fat diet.[2]

Chest tubes — the smallest effective drain

The modern principle is the smallest effective drain. A small-bore (8 to 14 French) Seldinger drain, inserted under ultrasound guidance, is preferred for most free-flowing effusions, complicated parapneumonic effusion and many empyemas, because it is more comfortable and equally effective when correctly placed. A large-bore (24 to 32 French) drain is reserved for frank pus, haemothorax and large-volume air leaks. The BTS recommends ultrasound guidance for every pleural drain.[4]

Drain care matters: assess output and swing regularly, flush (20 mL normal saline every 6 to 12 hours to maintain patency), avoid clamping (a chest tube is never clamped in a ventilated patient or a pneumothorax), and remove when the effusion has resolved and the lung re-expanded.[4]

When it goes wrong — the preventable list

• Missing empyema — a parapneumonic effusion with pH under 7.2, glucose under 3.3 mmol/L or frank pus needs a chest drain immediately, not more antibiotics. Send the pH in a heparinised blood-gas syringe, anaerobically.[5]
• Over-diuresing hepatic hydrothorax or placing a chest tube — both cause protein loss, infection and hepatorenal syndrome; use serial thoracentesis and refer for TIPS or transplant.[12]
• Draining more than 1.5 L at one sitting or draining too fast — causes re-expansion pulmonary oedema.[2]
• Pleurodesis in a trapped lung — pleurodesis cannot succeed if the lung does not re-expand; choose an indwelling pleural catheter instead.[3]
• Mislabelling a diuretic-treated cardiac effusion as an exudate — rescue with the serum-pleural albumin gradient over 1.2 g/dL or pleural NT-proBNP over 1500 ng/L.[10]
• Missing a tuberculous effusion in a high-burden setting — a lymphocytic exudate with ADA over 40 U/L is TB until proven otherwise.[10]
• DNase alone in empyema — MIST-2 showed it is harmful; only tPA plus DNase together works.[6]

Prognosis, disposition, and the score that sets both

The prognosis of a pleural effusion is the prognosis of its cause. Transudative effusions generally improve with treatment of the underlying disease. Parapneumonic effusion and empyema have an excellent prognosis when drained and treated early, but mortality rises steeply with delay, hospital-acquired infection, age and immunocompromise; the RAPID score quantifies this at three months.[5][9]

Malignant pleural effusion is a marker of advanced disease with a median survival of 3 to 12 months (lung and mesothelioma worse; breast, ovarian and lymphoma better, because they respond to systemic therapy). The LENT score (LDH, ECOG, neutrophil-to-lymphocyte ratio, tumour type) refines the prognosis and informs the choice between pleurodesis and an IPC.[3]

Disposition follows the cause and the patient's physiology. Outpatient management suits a stable transudate with a treatable cause and a small exudate under investigation. Inpatient management is required for drainage, sepsis, massive or tension effusion, haemothorax, diagnostic uncertainty, and significant comorbidity. Every patient needs a follow-up chest X-ray to document resolution, because a non-resolving effusion mandates reinvestigation for malignancy.[4]

Special populations

In the elderly, heart failure, pneumonia and malignancy all rise with age, reserve is lower, and re-expansion oedema and iatrogenic pneumothorax are less well tolerated — lower the threshold for inpatient management. In pregnancy, a small asymptomatic effusion may be physiological (third trimester, pre-eclampsia), but a large or symptomatic effusion needs workup — peripartum cardiomyopathy, pre-eclampsia, pulmonary embolism, ovarian hyperstimulation and tuberculosis are the principal causes.[10]

The immunocompromised are at risk for a broader spectrum of infection — bacterial, mycobacterial (including non-tuberculous), fungal and parasitic — and a wider differential for an exudative effusion. The anticoagulated patient needs the INR checked and corrected before thoracentesis if over 1.5; a bloody effusion in an anticoagulated patient still needs investigation — iatrogenic haemothorax is excluded by a pleural fluid haematocrit under 50 per cent of the serum haematocrit.[4]

[10] [11]

The evidence — BTS, MIST-2, TIME2

The British Thoracic Society (BTS) 2010 Pleural Disease Guidelines remain the international reference standard — Hooper on investigation of a unilateral effusion, Davies on pleural infection, Roberts on malignant effusion — formalising the chest X-ray, ultrasound, thoracentesis, Light's criteria algorithm, the drainage thresholds (pH under 7.2, glucose under 3.3 mmol/L, frank pus), and the role of CT, biopsy and thoracoscopy.[3][4][5]

The 2011 MIST-2 trial transformed loculated empyema by establishing intrapleural alteplase 10 mg plus dornase alfa 5 mg daily for three days — neither alone works, DNase alone is harmful. The 2014 RAPID score then provided a validated three-month mortality risk score for pleural infection.[6][9]

For malignant effusion, TIME2 (2012) showed an IPC non-inferior to talc pleurodesis for dyspnoea at 42 days, and IPC-Plus (2018) showed talc instilled through an IPC increased pleurodesis rates (43 per cent versus 23 per cent). Together they established IPC and talc pleurodesis as equivalent first-line options.[3][7][8]

The ERS 2024 statement on benign pleural effusions (Sundaralingam) synthesises the contemporary approach to heart failure, hepatic hydrothorax, nephrotic and parapneumonic effusions, including the role of NT-proBNP and the serum-to-pleural albumin gradient in cardiac effusions mislabelled as exudate by diuretic-confounded Light's criteria.[10]

The mantra, and the mnemonic

[1] [4]

The mantra: transudate or exudate decides everything — Light's criteria first, tap under ultrasound, drain when pH, glucose or pus demand it.[1]

[1] [3] [5]

Ward-round test — three stems, thirty seconds each

Stem 1 — the patient from the top of the topic (answer)

The 68-year-old with two weeks of progressive dyspnoea, stony dull right base, trachea pushed left. CXR confirms a large right effusion. What is your next step, and what single result changes management most? Model: Confirm with ultrasound, then diagnostic thoracentesis under ultrasound guidance and apply Light's criteria. The single result that changes management most is the pH (in a heparinised blood-gas syringe, anaerobically) alongside glucose — if this is a parapneumonic effusion with pH under 7.2 or glucose under 3.3 mmol/L, it goes straight to a chest drain; otherwise the transudate-exudate split and the fluid panel (cell count, Gram stain and culture, cytology, ADA, amylase, triglycerides) direct the cause hunt. Never drain more than 1.5 L at the first sitting.[1][5]

Stem 2 — the loculated empyema that will not drain (answer)

A 55-year-old with a parapneumonic effusion has a chest drain in place but remains febrile, and the drain output is poor. CT shows a multiloculated collection. What do you add, and what must you NOT add alone? Model: This is a loculated complicated parapneumonic effusion or empyema failing simple drainage. Add intrapleural tPA 10 mg plus DNase 5 mg, each daily for three days (MIST-2), which increases drainage and cuts surgical referral. Do not give DNase alone — MIST-2 showed it was harmful; do not give tPA alone either, as it was no better than placebo. Continue IV antibiotics covering community pathogens plus anaerobes, and call thoracic surgery early if the RAPID score is high or drainage still fails.[6][9]

Stem 3 — the cirrhotic with a right effusion (answer)

A 60-year-old with decompensated cirrhosis has a large right pleural effusion and tense ascites. The registrar wants to put a chest drain. What is the right call? Model: This is hepatic hydrothorax — a transudate tracking through diaphragmatic defects. Do NOT put a chest drain: it causes protein and fluid depletion, infection, and a persistent bronchopleural fistula. Confirm with Light's criteria (transudate) plus the serum-to-pleural albumin gradient over 1.2 g/dL, then treat the ascites (spironolactone plus furosemide, large-volume paracentesis), offer serial therapeutic thoracentesis for symptoms, and refer for TIPS or liver transplantation. Watch for spontaneous bacterial empyema — the pleural equivalent of spontaneous bacterial peritonitis.[12]

References

1. [1]Light RW, Macgregor MI, Luchsinger PC, Ball WC Jr Pleural effusions: the diagnostic separation of transudates and exudates Ann Intern Med, 1972.PMID 4642731
2. [2]Feller-Kopman D, Light R Pleural Disease N Engl J Med, 2018.PMID 29719174
3. [3]Roberts ME, Neville E, Berrisford RG, Antunes G, Ali NJ Management of a malignant pleural effusion: British Thoracic Society Pleural Disease Guideline 2010 Thorax, 2010.PMID 20696691
4. [4]Hooper C, Lee YCG, Maskell N Investigation of a unilateral pleural effusion in adults: British Thoracic Society Pleural Disease Guideline 2010 Thorax, 2010.PMID 20696692
5. [5]Davies HE, Davies RJO, Davies CWH Management of pleural infection in adults: British Thoracic Society Pleural Disease Guideline 2010 Thorax, 2010.PMID 20696693
6. [6]Rahman NM, Maskell NA, West A, et al. Intrapleural use of tissue plasminogen activator and DNase in pleural infection N Engl J Med, 2011.PMID 21830966
7. [7]Davies HE, Mishra EK, Kahan BC, et al. Effect of an indwelling pleural catheter vs chest tube and talc pleurodesis for relieving dyspnea in patients with malignant pleural effusion: the TIME2 randomized controlled trial JAMA, 2012.PMID 22610520
8. [8]Bhatnagar R, Keenan EK, Morley AJ, et al. Outpatient Talc Administration by Indwelling Pleural Catheter for Malignant Effusion N Engl J Med, 2018.PMID 29617585
9. [9]Rahman NM, Kahan BC, Miller RF, Gleeson FV, Nunn AJ, Maskell NA A clinical score (RAPID) to identify those at risk for poor outcome at presentation in patients with pleural infection Chest, 2014.PMID 24264558
10. [10]Sundaralingam A, Grabczak EM, Burra P, et al. ERS statement on benign pleural effusions in adults Eur Respir J, 2024.PMID 39060018
11. [11]Porcel JM Pleural mesothelioma Med Clin (Barc), 2022.PMID 35636988
12. [12]Cilia BJ, Haridy J, Raj A, et al. Hepatic hydrothorax as a manifestation of decompensated cirrhosis: An update on current management and future directions World J Hepatol, 2025.PMID 41179726