pleural effusions: evaluation and management

854 CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 72 • NUMBER 10 OCTOBER 2005

JOSÉ C. YATACO, MDDepartment of Pulmonary and Critical CareMedicine, The Cleveland Clinic Foundation

RAED A. DWEIK, MDDirector, Pulmonary Vascular Program,Department of Pulmonary, Allergy andCritical Care Medicine, The Cleveland ClinicFoundation

Pleural effusions:Evaluation and management

REVIEW

■ ABSTRACT

Pleural effusions are very common, and physicians of allspecialties encounter them. A pleural effusion representsthe disruption of the normal mechanisms of formationand drainage of fluid from the pleural space. A rationaldiagnostic workup, emphasizing the most commoncauses, will reveal the etiology in most cases.

■ KEY POINTS

Symptoms depend on the amount of fluid accumulatedand the underlying cause of the effusion. Many patientshave no symptoms at the time a pleural effusion isdiscovered. Possible symptoms include pleuritic chestpain, dyspnea, and dry nonproductive cough.

A key question in evaluating an effusion is whether theexcess pleural fluid is transudative or exudative.

Treatment depends on the severity and the cause.Thoracentesis is done to relieve symptoms. Chest tubesprovide continuous drainage in cases of pneumothorax,hemothorax, penetrating chest trauma, complicatedparapneumonic effusion or empyema, or chylothorax.Pleural sclerosis (pleurodesis) is usually indicated forpatients with uncontrolled symptomatic malignanteffusions.

ANY CONDITIONS can cause pleural effu-sions, including diseases that are local

(in the lungs or pleura), extrapulmonic, or sys-temic. In many cases the cause is a chroniccondition for which the patient is alreadyreceiving treatment; therefore, a patient withpleural effusion may present to a pulmonolo-gist—or to a general internist, other medicalspecialist, or surgeon. In up to 20% of casesthe cause remains unknown despite a diagnos-tic workup.

■ AN IMBALANCE OF FLUIDFORMATION AND DRAINAGE

A pleural effusion—an excessive accumula-tion of fluid in the pleural space—indicates animbalance between pleural fluid formationand removal.

The normal pleural space contains a rela-tively small amount of fluid, 0.1 to 0.2 mL/kgof body weight on each side.1,2

Pleural fluid is formed and removed slow-ly, at an equivalent rate, and has a lower pro-tein concentration than lung and peripherallymph. It can accumulate by one or more ofthe following mechanisms1–3:• Increased hydrostatic pressure in the

microvascular circulation: clinical datasuggest that an elevation in capillarywedge pressure is the most importantdeterminant in the development of pleur-al effusion in congestive heart failure.

• Decreased oncotic pressure in themicrovascular circulation due to hypoal-buminemia, which increases the tendencyto form pleural interstitial fluid.

• Increased negative pressure in the pleuralspace, also increasing the tendency for

M

CLEVELAND CL IN IC JOURNAL OF MEDICINE VOLUME 72 • NUMBER 10 OCTOBER 2005 855

pleural fluid formation; this can happenwith a large atelectasis.

• Separation of the pleural surfaces, whichcould decrease the movement of fluid inthe pleural space and inhibit pleural lym-phatic drainage; this can happen with atrapped lung.

• Increased permeability of the microvascu-lar circulation due to inflammatory medi-ators, which would allow more fluid andprotein to leak across the lung and viscer-al surface into pleural space; this has beendocumented with infections such as pneu-monia.

• Impaired lymphatic drainage from thepleural surface due to blockage by tumoror fibrosis.

• Movement of ascitic fluid from the peri-toneal space through either diaphragmat-ic lymphatics or diaphragmatic defects.

■ SIGNS AND SYMPTOMS

Accumulation of pleural fluid produces arestrictive ventilatory defect and decreasestotal lung capacity, functional capacity, andforced vital capacity.4 It may cause ventila-tion-perfusion mismatches due to partiallyatelectatic lungs in dependent areas and, iflarge enough, may compromise cardiac out-put5 by causing ventricular diastolic col-lapse.

The symptoms depend on the amount offluid and the underlying cause. Many patientshave no symptoms at the time a pleural effu-sion is discovered. Possible symptoms includepleuritic chest pain, dyspnea, and dry nonpro-ductive cough.

Physical findings are reduced tactilefremitus, dullness on percussion, and dimin-ished or absent breath sounds. A pleural rubmay also be heard during late inspirationwhen the roughened pleural surfaces cometogether.

■ IMAGING STUDIES

The evaluation of a pleural effusion beginswith imaging studies to assess the amount ofpleural fluid, its distribution and accessibility,and possible associated intrathoracic abnor-malities.

Chest radiographyStandard posteroanterior and lateral chestradiography remains the most important tech-nique for the initial diagnosis of pleural effu-sion. Free pleural fluid flows to the mostdependent part of the pleural space. In theupright position, this is the subpulmonicregion, and accumulation of fluid causesapparent elevation of the hemithorax, lateraldisplacement of the dome of the diaphragm,and blunting of the costophrenic angle.6However, at least 250 mL of fluid must accu-mulate before it becomes visible in a pos-teroanterior radiograph.

Lateral decubitus radiography is extremelyvaluable in the evaluation of a subpulmoniceffusion. It is very sensitive, detecting effu-sions as small as 5 mL in experimental stud-ies,7,8 and should be a routine test.

On supine chest radiography, commonlyused in intensive care, moderate to largepleural effusions may escape detection becausethe pleural fluid settles to the back, and nochange in the diaphragm or lateral pleuraledges may be noted. In these cases, a pleuraleffusion must be suspected when there isincreased opacity of the hemithorax withoutobscuring of the vascular markings. If an effu-sion is suspected, lateral decubitus radiographyor ultrasonography should be ordered, sinceboth are more reliable for detecting small

Chestradiographsremain themost importanttechnique forthe initialdiagnosis ofpleural effusion

FIGURE 1. Computed tomographic scanshowing cavitating retrocardiac infiltrate(white arrow) with adjacent pleuraleffusion (black arrows).


pleural effusions in the intensive care setting.Loculated effusions, defined as effusions

that do not shift freely in the pleural space,occur when there are adhesions between thevisceral and parietal pleura. The lateral decu-bitus view helps in differentiating free fluidfrom loculated fluid. The patient should bepositioned with the affected side down on thex-ray table.

Chest radiographs can also provide impor-tant clues to the cause of an effusion. Bilateraleffusions accompanied by cardiomegaly areusually caused by congestive heart failure.Large unilateral effusions without contralateralmediastinal shift suggest a large atelectasis,infiltration of the lung with tumor, a mesothe-lioma, or a fixed mediastinum due to tumor orfibrosis.6

UltrasonographyThe major advantage of ultrasonography overradiography is its ability to differentiatebetween solid components (eg, tumor orthickened pleura) and liquid components of apleural process. It is useful in detecting abnor-malities that are subpulmonic (under thelung) or subphrenic (below the diaphragm)and in differentiating them.9–11

A major use of ultrasonography is to guidethoracentesis in small or loculated pleural

effusions, thereby increasing the yield andsafety of the procedure. However, it is notpractical to recommend ultrasonography forall effusions. Portable ultrasound units can bebrought to the bedside of extremely illpatients.11

Computed tomographyComputed tomography (CT), with its cross-sectional images, can be used to evaluate com-plex situations in which the anatomy cannotbe fully assessed by plain radiography or ultra-sonography (FIGURE 1). For instance, CT ishelpful in distinguishing empyema from lungabscess, in detecting pleural masses (eg,mesothelioma, plaques), in detecting lungparenchymal abnormalities “hidden” by aneffusion, and in outlining loculated fluid col-lections.10

■ THORACENTESIS AND LABORATORYSTUDIES

Transudate vs exudateAlthough the history, physical examination,and radiographic studies may provide impor-tant clues to the cause of a pleural effusion,almost all cases should be evaluated with diag-nostic thoracentesis.12,13

Possible situations in which thoracentesis

Even largeeffusions maybe missed onsupine chestradiographsbecause thepleural fluidsettles to theback

Causes of pleural effusions

FREQUENCY TRANSUDATES EXUDATES

Common Congestive heart failure Parapneumonic effusionNephrotic syndrome MalignancyCirrhosis with ascites Pulmonary embolism

Collagen vascular diseasePancreatitisTuberculosisPostcardiac injury syndrome

Less common Peritoneal dialysis ChylothoraxUrinothorax UremiaAtelectasis Esophageal perforationPulmonary embolism Asbestos-related diseaseMyxedema Drug-induced reactions

Viral infectionYellow nail syndromeSarcoidosis

T A B L E 1

PLEURAL EFFUSIONS YATACO AND DWEIK


should not be done are when the effusion istoo small to be safely aspirated (< 10 mm thickon ultrasonography or lateral decubitus radi-ography) or when it can be explained byunderlying congestive heart failure (especiallybilateral effusions that improve with diuresis),recent thoracic or abdominal surgery, or post-partum status. However, the procedure maystill be indicated in these situations if thepatient’s clinical condition deteriorates.

After obtaining a sample of pleural fluid,the clinician should determine whether theeffusion is transudative (ie, due to hydrostaticforces, and with a low protein content) orexudative (due to increased permeability ofthe pleural surfaces and blood vessels, with arelatively high protein content). If the fluid isa transudate, the possible causes are relativelyfew, and further diagnostic procedures are notnecessary. In contrast, if the fluid is an exu-date, there are many possible causes, and more

diagnostic tests are required (TABLE 1).Several tests of the pleural fluid have been

proposed to differentiate transudates fromexudates. Light’s criteria (TABLE 2), originallypublished in 1972 and still the gold standard,require simultaneous measurement of the lev-els of protein and lactate dehydrogenase in thepleural fluid and in the serum.2,12,13 Newerproposed criteria are not much more sensitiveor specific (TABLE 3).14–16

A particular use for some of the newer cri-teria is to differentiate between transudatesand exudates in some patients with congestiveheart failure who receive diuretics—whichcan cause a transient increase in protein con-centration in the pleural fluid due to move-ment of water from the pleural fluid into theblood—and are found to have an exudativeeffusion by Lights’s criteria. If the clinicalappearance suggests an uncomplicated tran-sudative effusion, the albumin levels in the

Almost alleffusionsshould beevaluated witha diagnosticthoracentesis

Light’s criteria for distinguishing transudativefrom exudative pleural fluid

PLEURAL/SERUM PLEURAL/SERUM LACTATE SERUM LACTATEPROTEIN RATIO DEHYDROGENASE RATIO DEHYDROGENASE

Transudate ≤ 0.5 ≤ 0.6 ≤ 200 U/L*

Exudate† > 0.5 > 0.6 > 200 U/L*

*2/3 upper limit of normal serum level†A single positive criterion is enough to classify the fluid as an exudate

T A B L E 2

Sensitivity and specificity of teststo distinguish exudative from transudative effusions

SENSITIVITY SPECIFICITYFOR EXUDATES FOR EXUDATES(%) (%)

Light’s criteria 98 83Pleural-fluid cholesterol level > 60 mg/dL 54 92Pleural-fluid cholesterol level > 43 mg/dL 75 80Ratio pleural-fluid cholesterol/serum cholesterol > 0.3 89 81Serum albumin level minus pleural fluid albumin level ≤ 1.2 g/dL 87 92

MODIFIED WITH PERMISSION FROM LIGHT RW. PLEURAL EFFUSION. N ENGL J MED 2002; 346:1971–1977.

T A B L E 3



serum and the pleural fluid should be mea-sured. A difference of 1.2 g/dL or less indicatesan exudate, while a difference greater than 1.2g/dL indicates a transudate.17 A low concen-tration of cholesterol in the pleural fluid mayalso be more accurate in classifying this fluidas a transudate.

If a pleural effusion is likely to be a transu-date, initial laboratory tests can be limited tolevels of protein, cholesterol, and lactate dehy-drogenase in the pleural fluid (TABLE 4).14,15

These tests could be an alternative to all themeasurements required by Light’s criteria.

If the effusion is exudative, further studiesshould be undertaken to establish a diagnosis(TABLE 5, TABLE 6). FIGURE 2 provides an initialdiagnostic algorithm for pleural effusions.

Specific tests of pleural fluidThe glucose level in transudates and most

exudates is similar to that of serum. Few con-ditions can cause very low pleural fluid glu-cose levels (< 25 mg/dL), eg, rheumatoidarthritis, tuberculosis, empyema, and malig-nancies with extensive pleural involvement.The clinical presentation usually is helpful inidentifying the most likely cause.

The pH of the normal pleural fluid isaround 7.64, owing to active transport ofHCO3 into the pleural space. Depending onthe clinical setting, a low pleural fluid pH canbe useful in establishing a diagnosis, guidingtherapy, and determining prognosis. In gener-al, a lower pH is seen in inflammatory andinfiltrative processes such as infected para-pneumonic effusions, empyema, malignan-cies, collagen vascular disease, tuberculosis,and esophageal rupture. Urinothorax is theonly transudative effusion that can presentwith a low pleural fluid pH.

Measurement of pleural fluid pH is espe-cially important if one suspects that the effu-sion is parapneumonic, ie, due to pneumonia.A pleural fluid pH below 7.2 in this situationindicates the patient is at increased risk forpoor outcome and indicates the need fordrainage (TABLE 7).18

In the case of malignancy, patients withextensive tumor burden of the pleura have apleural fluid with a low pH (< 7.28) and lowglucose. In general, these patients have a poorshort-term survival rate, but pleural pH alonehas insufficient accuracy for clinical use inidentifying patients who should not undergopleural sclerosis, in view of poor proceduresuccess (see Pleural sclerosis, below).19,20

Amylase. A high pleural amylase level (>200 U/dL) usually indicates pancreatitis,malignancy, or esophageal rupture. The clini-cal setting usually separates these entities, butif needed, additional assay of isoenzymes canbe ordered (salivary vs pancreatic source).21,22

In esophageal rupture and up to 10% of non-pancreatic malignancies, the amylase is of thesalivary type. Esophageal rupture presentswith an amylase level approximately fivetimes higher than the serum level, while inpancreatitis and pancreatic cancer the amy-lase level in the pleural fluid is much higher(10–30 times the serum level).22

Newer criteria for classificationof exudates and transudates

LACTATE CHOLESTEROL PROTEINDEHYDROGENASE

Transudate ≤ 45%* ≤ 45 mg/dL ≤ 2.9 g/dLExudate > 45%* > 45 mg/dL > 2.9 g/dL

*Of serum upper limit of normal

T A B L E 4

Definitive diagnosisbased on pleural fluid analysis

DIAGNOSIS CRITERIA

Urinothorax pH < 7, transudate,pleural fluid-to-serum creatinine ratio > 1

Empyema Pus, positive Gram stains or cultures

Malignancy Positive cytologic testing

Chylothorax Triglycerides > 110 mg/dL, chylomicrons

Tuberculosis, Positive stains or culturesfungal infection

Hemothorax Hematocrit > 50% of blood

Esophageal rupture pH < 7, high amylase (salivary)

Each of these tests should be ordered based on clinical suspicion

T A B L E 5



■ OTHER DIAGNOSTIC TESTS

Pleural biopsyThe main conditions that can be establishedwith needle biopsy of the pleura are tubercu-lous pleuritis and malignancy of the pleura.Needle biopsy is currently recommendedwhen tuberculous pleuritis is suspected andthe pleural fluid adenosine deaminase or inter-feron-gamma levels are not definitive (seeTuberculosis, below). A parietal pleural biop-sy specimen is positive for granulomas in up to80% of cases of tuberculous pleurisy, acid-faststaining is positive in 26%, and culture is pos-itive in 56%. At least one of these three testsis positive in 91% of cases.23

The incidence of granuloma on pleuralbiopsy is comparable in patients with andwithout human immunodeficiency virus(HIV) infection (CD4+ counts below200/mm3). The pleural fluid in HIV patients ismore likely to be smear-positive and culture-positive for acid-fast bacilli.

Pleural biopsy is also recommended whenmalignancy is suspected but cytologic study ofthe pleural fluid is negative and thoracoscopyis not readily available.

ThoracoscopyThoracoscopy (or pleuroscopy) involves pass-ing an endoscope through the chest wall todirectly view and collect samples from thepleura.

The goal of medical thoracoscopy (per-formed by a pulmonologist with the patientunder conscious sedation) is to visualize theentire lung and, when needed, to performbiopsies of the parietal or visceral pleural sur-faces. The main indications include pleuraleffusions of unknown cause, particularly ifmesothelioma, lung cancer, or tuberculosis issuspected. It can also be done to introducesclerosing agents.

Video-assisted thoracoscopic surgery takesplace in an operating room with the patientunder general anesthesia and with single lungventilation. Several procedures can be per-formed in this way: stapled lung biopsy, lobec-tomy, pneumonectomy, resection of pul-monary nodules, repair of a bronchopleural fis-tula, and evaluation of mediastinal tumors oradenopathy.

The major contraindication to medical orsurgical thoracoscopy is lack of a pleural spacedue to pleural adhesions. Relative contraindi-cations include uncontrolled cough, hypox-emia, coagulopathy, and severe cardiac disease.

Complications from medical thora-coscopy (eg, persistent air leak, subcutaneousemphysema) are minor and infrequent. Deathis extremely rare.24,25

■ PLEURAL EFFUSIONSIN SPECIFIC DISEASES

It is important to initially evaluate the patientfor cardiac, renal, intra-abdominal, systemic,and inflammatory conditions that could elicita pleural effusion.

Collagen vascular diseasesPleural effusions develop in up to 75% ofpatients with systemic lupus erythematosus(SLE) and 5% of patients with rheumatoid

Pleuraleffusions areseen in up to75% of patientswith lupus

Pleural fluid testsOn all effusions

ProteinLactate dehydrogenaseCholesterolCell count and differential

On exudates*Cytologic analysispH†

Gram stain and cultureFungal stain/cultureAcid-fast bacteria stain/culture

Other testsGlucoseAmylaseAdenosine deaminase or gamma-interferon levelAntinuclear antibody titerHematocritTriglyceridesCreatinineAlbumin

*Fluid can be saved for further analysis based oninitial results, although for practical reasons manyclinicians order all tests at the same time

†pH need be measured in transudates only whenurinothorax is suspected

T A B L E 6


arthritis during the course of the disease.SLE. The pleural fluid antinuclear anti-

body (ANA) titer may help in separating SLEeffusions from effusions due to other causes,even in patients with known SLE. A pleuralfluid ANA titer greater than 1:160 or a pleur-al fluid-to-serum ANA ratio greater than 1.0suggests lupus pleuritis.26 Although these cri-teria appear to be highly specific, they are nothighly sensitive.

Rheumatoid arthritis. Pleural effusions inrheumatoid arthritis are often asymptomatic.They may be quite large and often persist formany months without change. Rheumatoideffusions usually occur in patients with highserum rheumatoid factor titers and rheuma-toid nodules. The fluid typically has a verylow glucose level. Pleural rheumatoid factortiters are not helpful in diagnosis because theymay be elevated in pneumonia, tuberculosis,malignancy, and SLE.

In patients with rheumatoid arthritisbeing treated with anti-tumor-necrosis factortherapy, special concern is warranted toexclude tuberculosis.

TuberculosisIn many areas of the world, tuberculosis con-tinues to be the most common cause of pleur-

al effusions in the absence of demonstrablepulmonary disease. Rupture of a subpleuralcaseous focus into the pleural space allowstuberculous protein to enter the pleural spaceand to generate a hypersensitivity reactionresponsible for most of the clinical manifesta-tions.

Pleural effusion in tuberculous pleuritismanifests as an acute illness that can mimicacute bacterial pneumonia. It is usually uni-lateral and can be of any size. Coexistence ofparenchymal disease is visible on standardradiographs in 19% of patients.23

The pleural fluid in tuberculosis is invari-ably an exudate with more than 50% lympho-cytes in the white cell differential count. Itrarely contains more than 5% mesothelialcells, which is explained by the extensiveinvolvement of the pleural surface by theinflammatory process.2 A definitive diagnosismay be difficult and depends on the demon-stration of acid-fast bacilli in sputum, pleuralfluid, or pleural biopsy specimen, or thedemonstration of granulomas in the pleura.Pleural fluid analysis and cultures for acid-fastbacilli are positive in less than 25% of cases.Pleural biopsy culture can increase the yield to55%.2,23

Additional measurements that suggest the

In tuberculouspleuritis,pleural effusioncan mimicacute bacterialpneumonia

Pleural effusion confirmed radiographically

Symptomatic effusion with significant volume of fluid

No YesDiagnostic thoracentesis with basic tests:Lactate dehydrogenaseProteinCholesterolCell count and differential*

Determine if fluid is a transudate or exudate (TABLE 2,TABLE 4), consider common causes (TABLE 1)

No need to perform thoracentesis unless clinicaldeterioration occurs

Very small effusion (< 10 mm thick on lateral decubitusview) or asymptomatic with obvious cause (eg, congestiveheart failure, postoperative status)

FIGURE 2. Approach to pleural effusions

*Cytology may be ordered if malignant disease is suspected. If infection is considered in the differential diagnosis, then testing of the pHand glucose in pleural fluid must be ordered on initial evaluation.



diagnosis include pleural fluid adenosine deam-inase, interferon-gamma, and polymerase chainreaction for mycobacterial DNA. Elevations ofpleural adenosine deaminase levels have beenobserved in tuberculous pleurisy, rheumatoidarthritis, and empyema. Adenosine deaminaselevels above 40 U/L distinguish tuberculouseffusions from other lymphocytic pleural effu-sions (ie, malignancies, lymphoma, collagenvascular diseases),27,28 as do interferon-gammalevels above 140 pg/mL.29

UrinothoraxUrinothorax, a rare cause of pleural effusion, isbelieved to occur when urine moves retroperi-toneally into the pleural space owing to uri-nary obstruction, trauma, a retroperitonealinflammatory or malignant process, failednephrostomy, or kidney biopsy.2,30 The pleuralfluid is a transudate with the unique feature ofhaving a pleural fluid-to-serum creatinineratio greater than 1.0. It also can have a lowpH (< 7.3) or low glucose level, both of whichare uncommon in transudative effusions.31

After coronary artery bypass graftingPleural effusions are common immediately

after coronary artery bypass grafting(CABG).32 The reported prevalence 1week after surgery has ranged from 40% to75%.

Most of these effusions are small, unilater-al, left-sided, and asymptomatic. In general,they gradually resolve over several weeks.Large pleural effusions (> 25% of hemithorax)not explained by any other cause occur in asmall proportion of patients.

The fluid is invariably an exudate andcan be classified according to its grossdescription.32 Bloody effusions tend to occurearlier (< 4 weeks after surgery) and are easyto control with one to three therapeutic tho-racenteses. Nonbloody effusions tend tooccur later (> 4 weeks after surgery) andhave a relatively low lactate dehydrogenaselevel and a high percentage of lymphocytes.Nonbloody effusions are more difficult tocontrol despite repeat thoracentesis and mayrequire anti-inflammatory agents or chemi-cal pleurodesis.

Chylous effusionA true chylous pleural effusion develops whenchyle enters the pleural space owing to disrup-

Suggested approach to classificationand management of parapneumonic effusions*

RISK OF POOR PLEURAL SPACE ANATOMY pH BACTERIOLOGY DRAINAGEOUTCOME (GRAM STAIN OR CULTURE) INDICATED†

Very low Minimal free-flowing effusion AND > 7.2 AND Negative or unknown No(< 10 mm on lateral decubitus)

Low Small to moderate AND ≥ 7.2 AND Negative No‡

free-flowing effusion(> 10 mm and < 1/2 hemithorax)

Moderate Large free-flowing OR < 7.2 OR Positive Yesor loculated effusion(≥ 1/2 hemithorax)

High Pus Yes

*It is not necessary to have a proven bacterial pneumonia: clinical diagnosis is enough†pH and bacteriologic study results have priority over amount of fluid‡If clinical condition deteriorates, repeating thoracentesis and drainage should be considered

ADAPTED FROM COLICE GL, CURTIS A, DESLAURIERS J, ET AL; FOR THE AMERICAN COLLEGE OF CHEST PHYSICIANS PARAPNEUMONIC EFFUSIONS PANEL. ACCP CONSENSUSSTATEMENT. MEDICAL AND SURGICAL TREATMENT OF PARAPNEUMONIC EFFUSIONS: AN EVIDENCE-BASED GUIDELINE. CHEST 2000; 118:1158–1171.

T A B L E 7

Pleuraleffusions arecommonimmediatelyafter CABG


tion of the thoracic duct by trauma (surgicalor nonsurgical) or by malignancy. Continuousdrainage of a chylous effusion may result inmalnutrition and immunosuppression due tosignificant loss of protein, fats, electrolytes,and lymphocytes.

Initial conservative treatment consists oflimiting dietary fat to medium-chain triglyc-erides that are absorbed through the portalvenous system and not carried by lymph, in anattempt to decrease the lymph flow rate. Ifnecessary, lymph flow can be further reducedby using total parenteral nutrition and avoid-ing oral intake.33,34

Surgical therapy for chylothorax, withthoracic duct ligation or pleuroperitonealshunt implantation, may be necessary beforethe patient becomes too cachectic to toleratethe intervention.33,34

Pleural effusionsdue to pulmonary embolismPleural effusions occur in 30% to 50% ofpatients with pulmonary emboli. It is possiblethat a significant number of undiagnosed effu-sions are due to pulmonary embolism.

The fluid may be transudative (24%) orexudative, depending on the mechanism. Atransudate occurs when there is right-sidedheart failure and increased capillary pressurein the parietal pleura. An exudate occurs dueto increased permeability of the capillaries inthe lung (caused by ischemia or inflammatorymediators from the platelet-rich thrombi).

Standard anticoagulation is the treatmentof choice.

Pleural effusions in the intensive care unitThe incidence of pleural effusions in theintensive care unit (ICU) varies according tothe screening method. One study, using rou-tine ultrasonography, found pleural effusionsin 62% of medical ICU patients, with a pre-dominance of transudates.35 On the otherhand, pleural effusions were detected by phys-ical examination and opacification of at leastone third of the lung field on radiography inonly 8.4% of ICU patients. With the lattermethod, exudates related to infection weremore common.36

Thoracentesis is not contraindicated inICU patients receiving mechanical ventila-

tion. In one study of clinically documentedeffusions, routine thoracentesis was compli-cated by pneumothorax in only 7% ofpatients.36 The same study showed that thora-centesis altered the diagnosis in 45% ofpatients and changed the treatment in 33%.Use of ultrasound guidance has been shown toimprove the safety of thoracentesis inmechanically ventilated patients.37

■ UNEXPLAINED EFFUSIONS

The cause of 15% to 20% of all pleural effu-sions will remain unknown despite intensivediagnostic efforts.38 An unexplained pleuraleffusion has been defined as one without anapparent cause despite repeat thoracentesis.The clinician should ensure that all theunusual causes of pleural effusion are consid-ered and requisite studies are obtained.

Roughly 50% of these effusions resolvespontaneously, and no disease is apparent onlong-term follow-up. Many unresolved pleuraleffusions will turn out to be caused by malig-nant disease, which is obvious clinically or isincurable in any event. The most commontreatable cause of an unexplained effusion istuberculosis.39

Thus, invasive procedures such as video-assisted thoracoscopy or thoracotomy withdirect sampling of the pleura are frequentlyrecommended for these patients.

■ THERAPY

Therapeutic thoracentesisAny pleural effusion large enough to causesevere respiratory symptoms should be drainedregardless of the cause and regardless of con-comitant disease-specific treatment. Relief ofsymptoms is the main goal of therapeuticdrainage in these patients.

The only absolute contraindication tothoracentesis is active cutaneous infection atthe puncture site. Some relative contraindica-tions include severe bleeding diathesis, sys-temic anticoagulation, and a small volume offluid.

Possible complications of the procedureinclude bleeding (due to accidental punctureof a vessel or lung parenchyma), pneumotho-rax, infections (soft-tissue infection or empye-

The cause of15% to 20% ofall pleuraleffusions willremainunknown



ma), laceration of intra-abdominal organs,hypotension, and pulmonary edema.2

In general, no more than 1,000 to 1,500mL of fluid should be removed at one time.Rapid drainage of fluid may predisposepatients to the rare complication of re-expansion pulmonary edema in the underly-ing lung or rapid fluid shift from the intravas-cular space to the pleural space (post-thora-centesis shock). These complications appearto be related to the creation of excessive neg-ative pressure in the pleural cavity duringthoracentesis. Large-volume thoracentesiscan be undertaken with monitoring of theintrapleural pressure.40,41

Tube thoracostomy (chest tube)Tube thoracostomy allows continuous, large-volume drainage of air or liquid from thepleural space.

Specific indications for placement of achest tube include spontaneous or iatrogenicpneumothorax (especially if large and sympto-matic), hemothorax, penetrating chest trau-ma, complicated parapneumonic effusion orempyema, chylothorax, and pleurodesis ofsymptomatic pleural effusions.

In symptomatic or clinically unstablepatients, there is no absolute contraindicationto chest tube placement. In patients withcomplicated pleural spaces due to multipleloculations or previous pleurodesis, a contrastchest CT scan should be obtained to guide theplacement of the chest tube.

For drainage and pleurodesis of malignantpleural effusions, a silicone polymer (Silastic)

chest tube 20 to 24 F is usually adequate,although small-bore catheters (8–14 F) placedunder fluoroscopic, ultrasound, or CT guid-ance have also been successful.42,43

Complicated parapneumonic effusionsand frank empyema (FIGURE 3) require drainagewith a large-bore chest tube (28–36 F) to con-trol the local pleural inflammatory reaction,which may not otherwise respond to intra-venous antibiotics.

In multiloculated complicated effusions,image-guided placement of small-borecatheters (10–14 F) should be considered.44–46

If appropriate drainage is not obtaineddespite correct positioning of the tubes (veri-fied with chest CT), fibrinolytic therapy canbe used.47,48 Agents such as streptokinase,urokinase, and alteplase can lyse fibrin andimprove drainage.

Pleural sclerosisPleural sclerosis (pleurodesis) is considered forpatients with uncontrolled and recurrentsymptomatic malignant effusions, and rarely,in cases of benign effusions after failure ofmedical treatment. A sclerosing agent isinstilled into the pleural cavity via a tube tho-racostomy to produce a chemical serositis andsubsequent fibrosis of the pleura.

Pleural sclerosis should be attempted onlyif the lung expands fully after fluid removal.The visceral and parietal pleura need to beapproximated closely, obliterating the pleuralcavity so that fibrotic healing achieves pleuralsymphysis.

The overall success rate with fibrosingagents (ie, talc, doxycycline, and tetracycline)is 75%, compared with a complete success rateof only 44% for antineoplastic agents (ie,bleomycin).49 Talc is the most effective agent,with a complete success rate of 93%.49,50

Pleurodesis failure is usually the result of sub-optimal technique or inability to approximatethe pleural surfaces.

Surgical therapyVideo-assisted thoracoscopic surgery is veryuseful in managing incompletely drained para-pneumonic effusions. With thoracoscopy, theloculi in the pleura can be disrupted, thepleural space can be completely drained, andthe chest tube can be optimally placed.51

FIGURE 3. Chest computed tomographicscan with a “split pleural sign” (arrow),seen in empyema. This patient neededdrainage with tube thoracostomy.

Any pleuraleffusion largeenough tocause severerespiratorysymptomsshould bedrained


In cases of empyema with uncontrolledsepsis or progression to the fibroproliferativephase, a full thoracotomy with decorticationis performed with removal of all the fibrous

tissue and evacuation of all the pus from thepleural space. Decortication in this situationwill eliminate the septic source and allow thelung to expand.


■ REFERENCES1. Sahn SA. State of the art. The pleura. Am Rev Respir Dis

1988; 138:184–234.2. Light RW. Pleural diseases. Baltimore: Williams & Wilkins,

1995.3. Sy BAC, Dweik RA. Pleural disease. The Cleveland Clinic

Disease Management Project, 2002. http://www.cleve-landclinicmeded.com/diseasemanagement/pulmonary/pleuraldisease/pleuraldisease.htm. Accessed 5/9/05.

4. Gilmartin JJ, Wright AJ, Gibson GJ. Effects of pneumo-thorax or pleural effusion on pulmonary function.Thorax 1985; 40:60–65.

5. Agusti AG, Cardus J, Roca J, Grau JM, Xaubet A,Rodriguez-Roisin R. Ventilation-perfusion mismatch inpatients with pleural effusion: effects of thoracentesis.Am J Respir Crit Care Med 1997; 156:1205–1209.

6. Pugatch RD, Spirn PW. Radiology of the pleura. ClinChest Med 1985; 6:17–32.

7. Gallardo X, Castaner E, Mata JM. Benign pleural dis-eases. Eur J Radiol 2000; 34:87–97.

8. Moskowitz H, Platt RT, Schachar R, Mellins H. Roentgenvisualization of minute pleural effusion. An experimen-tal study to determine the minimum amount of pleuralfluid visible on a radiograph. Radiology 1973;109:33–35.

9. O’Moore PV, Mueller PR, Simeone JF, et al. Sonographicguidance in diagnostic and therapeutic interventions inthe pleural space. AJR Am J Roentgenol 1987; 149:1–5.

10. McLoud TC, Flower CD. Imaging the pleura: sonogra-phy, CT, and MR imaging. AJR Am J Roentgenol 1991;156:1145–1153.

11. McGahan JP. Aspiration and drainage procedures in theintensive care unit: percutaneous sonographic guid-ance. Radiology 1985; 154:531–532.

12. Light RW, Macgregor MI, Luchsinger PC, Ball WC, Jr.Pleural effusions: the diagnostic separation of transu-dates and exudates. Ann Intern Med 1972; 77:507–513.

13. Light RW. Clinical practice. Pleural effusion. N Engl JMed 2002; 346:1971–1977.

14. Peterman TA, Speicher CE. Evaluating pleural effusions.A two-stage laboratory approach. JAMA 1984;252:1051–1053.

15. Costa M, Quiroga T, Cruz E. Measurement of pleuralfluid cholesterol and lactate dehydrogenase. A simpleand accurate set of indicators for separating exudatesfrom transudates. Chest 1995; 108:1260–1263.

16. Burgess LJ, Maritz FJ, Taljaard JJ. Comparative analysisof the biochemical parameters used to distinguishbetween pleural transudates and exudates. Chest 1995;107:1604–1609.

17. Roth BJ, O’Meara TF, Cragun WH. The serum-effusionalbumin gradient in the evaluation of pleural effusions.Chest 1990; 98:546–549.

18. Colice GL, Curtis A, Deslauriers J, et al; for theAmerican College of Chest Physicians ParapneumonicEffusions Panel. AACP consensus statement. Medicaland surgical treatment of parapneumonic effusions. Anevidence-based guideline. Chest 2000; 118:1158–1171.

19. Heffner JE, Nietert PJ, Barbieri C. Pleural fluid pH as apredictor of survival for patients with malignant pleuraleffusions. Chest 2000; 117:79–86.

20. Sahn SA, Good JT Jr. Pleural fluid pH in malignant effu-

sions. Diagnostic, prognostic, and therapeutic implica-tions. Ann Intern Med 1988; 108:345–349.

21. Kramer MR, Saldana MJ, Cepero RJ, Pitchenik AE. Highamylase levels in neoplasm-related pleural effusion.Ann Intern Med 1989; 110:567–569.

22. Sherr HP, Light RW, Merson MH, Wolf RO, Taylor LL,Hendrix TR. Origin of pleural fluid amylase inesophageal rupture. Ann Intern Med 1972; 76:985–986.

23. Valdes L, Alvarez D, San Jose E, et al. Tuberculouspleurisy: a study of 254 patients. Arch Intern Med 1998;158:2017–2021.

24. Boutin C, Viallat JR, Cargnino P, Farisse P. Thoracoscopyin malignant pleural effusions. Am Rev Respir Dis 1981;124:588–592.

25. Menzies R, Charbonneau M. Thoracoscopy for the diag-nosis of pleural disease. Ann Intern Med 1991;114:271–276.

26. Good JT Jr, King TE, Antony VB, Sahn SA. Lupus pleuri-tis. Clinical features and pleural fluid characteristicswith special reference to pleural fluid antinuclear anti-bodies. Chest 1983; 84:714–818.

27. Lee YC, Rogers JT, Rodriguez RM, Miller KD, Light RW.Adenosine deaminase levels in nontuberculous lympho-cytic pleural effusions. Chest 2001; 120:356–361.

28. Ocana I, Martinez-Vazquez JM, Segura RM, Fernandez-De-Sevilla T, Capdevila JA. Adenosine deaminase inpleural fluids. Test for diagnosis of tuberculous pleuraleffusion. Chest 1983; 84:51–53.

29. Villena V, Lopez-Encuentra A, Echave-Sustaeta J,Martin-Escribano P, Ortuno-de-Solo B, Estenoz-Alfaro J.Interferon-gamma in 388 immunocompromised andimmunocompetent patients for diagnosing pleuraltuberculosis. Eur Respir J 1996; 9:2635–2639.

30. Miller KS, Wooten S, Sahn SA. Urinothorax: a cause oflow pH transudative pleural effusions. Am J Med 1988;85:448–449.

31. Stark DD, Shanes JG, Baron RL, Koch DD. Biochemicalfeatures of urinothorax. Arch Intern Med 1982;142:1509–1511.

32. Light RW, Rogers JT, Cheng D, Rodriguez RM. Largepleural effusions occurring after coronary artery bypassgrafting. Cardiovascular Surgery Associates, PC. AnnIntern Med 1999; 130:891–896.

33. Sassoon CS, Light RW. Chylothorax and pseudochylo-thorax. Clin Chest Med 1985; 6:163–171.

34. Teba L, Dedhia HV, Bowen R, Alexander JC. Chylothoraxreview. Crit Care Med 1985; 13:49–52.

35. Mattison LE, Coppage L, Alderman DF, Herlong JO,Sahn SA. Pleural effusions in the medical ICU: preva-lence, causes, and clinical implications. Chest 1997;111:1018–1023.

36. Fartoukh M, Azoulay E, Galliot R, et al. Clinically docu-mented pleural effusions in medical ICU patients: howuseful is routine thoracentesis? Chest 2002;121:178–184.

37. Lichtenstein D, Hulot JS, Rabiller A, Tostivint I, MeziereG. Feasibility and safety of ultrasound-aided thoracen-tesis in mechanically ventilated patients. Intensive CareMed 1999; 25:955–958.

38. Ansari T, Idell S. Management of undiagnosed persis-tent pleural effusions. Clin Chest Med 1998; 19:407–417.

39. Ferrer JS, Munoz XG, Orriols RM, Light RW, Morell FB.Evolution of idiopathic pleural effusion: a prospective,long-term follow-up study. Chest 1996; 109:1508–1513.

40. Light RW, Jenkinson SG, Minh VD, George RB.Observations on pleural fluid pressures as fluid is with-drawn during thoracentesis. Am Rev Respir Dis 1980;121:799–804.

41. Villena V, Lopez-Encuentra A, Pozo F, De-Pablo A,Martin-Escribano P. Measurement of pleural pressureduring therapeutic thoracentesis. Am J Respir Crit CareMed 2000; 162:1534–1538.

42. Morrison MC, Mueller PR, Lee MJ, et al. Sclerotherapyof malignant pleural effusion through sonographicallyplaced small-bore catheters. AJR Am J Roentgenol1992; 158:41–43.

43. Parker LA, Charnock GC, Delany DJ. Small bore catheterdrainage and sclerotherapy for malignant pleural effu-sions. Cancer 1989; 64:1218–1221.

44. Moulton JS. Image-guided drainage techniques. SeminRespir Infect 1999; 14:59–72.

45. Reinhold C, Illescas FF, Atri M, Bret PM. Treatment ofpleural effusions and pneumothorax with cathetersplaced percutaneously under imaging guidance. AJRAm J Roentgenol 1989; 152:1189–1191.

46. Silverman SG, Mueller PR, Saini S, et al. Thoracicempyema: management with image-guided catheterdrainage. Radiology 1988; 169:5–9.

47. Bouros D, Schiza S, Siafakas N. Utility of fibrinolyticagents for draining intrapleural infections. SeminRespir Infect 1999; 14:39–47.

48. Sahn SA. Use of fibrinolytic agents in the managementof complicated parapneumonic effusions and empye-mas. Thorax 1998; 53(suppl 2):S65–S72.

49. Walker-Renard PB, Vaughan LM, Sahn SA. Chemicalpleurodesis for malignant pleural effusions. Ann InternMed 1994; 120:56–64.

50. Management of malignant pleural effusions. Am JRespir Crit Care Med 2000; 162:1987–2001.

51. Cassina PC, Hauser M, Hillejan L, Greschuchna D,Stamatis G. Video-assisted thoracoscopy in the treat-ment of pleural empyema: stage-based managementand outcome. J Thorac Cardiovasc Surg 1999;117:234–238.

ADDRESS: Raed A. Dweik, MD, Department of Pulmonary,Allergy and Critical Care Medicine, A90, The Cleveland ClinicFoundation, 9500 Euclid Avenue, Cleveland, OH 44195-5038;e-mail [email protected].


YATACO AND DWEIK

The long-awaited fourth editionof To Act As A Unit: The Story of TheCleveland Clinic is now available.

This 405-page book will rivet yourattention as it describes the originand maturation of one of the nation’sleading academic medical centers. Learnhow the founders incorporated the bestfeatures of military medicine into TheCleveland Clinic’s unique and innovativemodel of medicine. See how majoradversity helped to shape theorganization’s character.

Spanning most of the 20th century,this inspirational story recounts TheCleveland Clinic’s humble origins, itssurvival of the ravages of war, disaster,and ostracism by the medicalestablishment, and its emergence,stronger than ever, into the 21st centuryas a thriving and respected health careinstitution.

The fourth edition contains new sectionsdescribing the formation of TheCleveland Clinic Health System, thecreation of The Cleveland Clinic LernerCollege of Medicine of Case WesternReserve University, the new-foundsuccess of philanthropy, and phenomenalgrowth in all areas of activity.

Don’t miss the opportunity to getyour copy now. Get it fromAmazon.com or send us an e-mailrequest at [email protected]’t wait another day!

NOW AVAILABLE!

CME ANSWERSAnswers to the credit test on page 951

of this issue

1 C 2 B 3 D 4 C 5 E 6 B 7 D8 B 9 C 10 E 11 C 12 D

pleural effusions: evaluation and management

Documents