THE PHYSICAL EXAMINATION DAVID ROLSTON, MD, EDITORENRIQUE DIAZ-GUZMAN, MD MARIE M. BUDEV, DO, MPH CME CREDIT Department of Pulmonary, Allergy, and Critical Care Medicine, Pulmonary Institute, Cleveland Department of Pulmonary, Allergy, and Critical Care Medicine, Pulmonary Institute, Cleveland Clinic Clinic Accuracy of the physical examination in evaluating pleural effusion ■ A B S T R AC T pleural I N DETECTING AND EVALUATING effusion, technology has not replaced A careful physical examination is a valuable and clinical skills. Yet, despite centuries of lore, noninvasive means of assessing pleural effusion and data are limited on the role of the physical should be routinely performed in every patient in whom examination and on its accuracy compared this condition is suspected. Although physical with other noninvasive tests such as conven- examination is less accurate than ultrasonography or tional chest radiography or ultrasonography. computed tomography in detecting a pleural effusion, the The following is an overview of the value sensitivity and specificity of the different physical signs of of the clinical history and physical examination pleural effusion may be comparable to those of in detecting pleural effusion and a brief review of the available information regarding its accu- conventional chest radiography. racy compared with other diagnostic methods. ■ KEY POINTS ■ POTENTIAL CAUSES ARE MANY The potential causes of pleural effusion are many and include congestive heart failure, pneumonia, cancer, and The pleurae consist of two membranes that pro- tect the lungs, allow them to move, contribute pulmonary embolism. to their shape, and prevent the alveoli at the pleural surface from becoming overdistended. Cardinal symptoms of pleural effusion are cough, chest Between the visceral pleura (covering the lung) pain, and dyspnea, but these are not very sensitive or and the parietal pleura (covering the diaphragm specific. and the chest wall) is the pleural space. In healthy adults, the pleural space con- Common signs of pleural effusion are asymmetric chest tains an estimated 5 to 10 mL of pleural fluid expansion, asymmetric tactile fremitus, dullness to (0.1 mg/kg body weight).1 Pleural effusion is percussion, absent or diminished breath sounds, and rubs. an accumulation of an abnormal amount of The larger the effusion, the more sensitive these signs are. fluid in the pleural space. Although the potential causes are many, Some have advocated auscultatory percussion (tapping the most common are congestive heart failure, pneumonia (40% of patients hospitalized with on the manubrium while listening on the patient’s back) pneumonia have pleural effusion),2,3 cancer, as being more sensitive than conventional percussion for and pulmonary embolism.4 detecting the dullness to percussion of pleural effusion. Because many diseases affecting different organs can cause a pleural effusion, we cannot overemphasize the importance of a thorough history and physical examination to uncover clues that will help identify its cause and nar- row the diagnostic workup. For example, sig- nificant weight loss and cachexia could be due CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 75 • NUMBER 4 APRIL 2008 297 skin.9 298 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 75 • NUMBER 4 APRIL 2008 . and many patients have many cases referred to the ipsilateral shoulder. and nitrofurantoin lung function is poor. a con.75–0.5 manifestations of the underlying process but Pain may also be referred to the abdomen.5 0. nonproductive cough. Trexall].59 0.42–0. ed with dyspnea.74 0.57 0. if accompanied by purulent sputum it suggests pneumonia. parapneumonic effusion. FIGURES 3–5 to cancer. disease of the pleura such as mesothelioma.19–0. proportion to the size of the effusion.2 [Macrobid]) can be associated with exudative Chest pain accompanying a pleural effu- effusions.710.89 0. and joint. and cancer are ■ SYMPTOMS ARE NEITHER SENSITIVE Pain of pleural origin can remain localized common causes NOR SPECIFIC to the adjacent area of the chest.05 0. viral pleurisy. Although large pleural effusions case series.68 0.55 0. methotrexate also have shortness of breath if their baseline [Rheumatrex. FEW DATA monary embolism. Many patients with a pleural effu.0 0. A dull.32–0. and pleuritic chest pain.85–1.96 Pleural rub10 0. PLEURAL EFFUSION DIAZ-GUZMAN AND BUDEV TA B L E 1 Common physical signs of pleural effusion SIGN SENSITIVITY SPECIFICITY POSITIVE NEGATIVE PREDICTIVE PREDICTIVE VALUE VALUE Asymmetry of expansion10 0. Dyspnea is a consequence of a combina. Although aching chest pain may be due to an underlying this symptom is rarely helpful in diagnosing a pleural malignancy. It is common in patients with pul- Cough. ■ PHYSICAL EXAMINATION: cated by hemoptysis it suggests cancer or pul. monary embolism. asbestos can raise the suspicion of a malignant chest wall and ribs—or to a benign disease of heart failure. amiodarone [Cordarone]. but it can also occur in patients sequence of inflammation of the pleurae or with pneumothorax or pericarditis. not of the effusion itself.89 Absent breath sounds10.90 0. LONG TRADITION. or eye symptoms reduce lung volume and are generally associat- could be due to a connective tissue disorder.8.82 0.86 0.81 0. The most common Pleuritic chest pain is described as being symptoms directly related to effusion are worse with deep inspiration or when lying cough. If the of pleural The symptoms of pleural effusion are neither diaphragmatic pleura is involved. the thoracic wall such as rib fracture or costo- chondritis. a ventilation. since several medications patients with small to moderate effusions may (eg.7 pleural effusion. sion suggests inflammation of the parietal al history must be ascertained. Our knowledge of the role of physical exami- tion of a restrictive lung defect.83–0. and if compli. In addition.6 but could be due to cancer in the Pneumonia. but some- times it is referred to other areas. the patient’s occupation.15 0.91 0.5 down. and a decrease in cardiac based mostly on expert opinion and on small output. dyspnea.53–0.88 0. nation in detecting pleural effusion is still perfusion mismatch. the symptoms may be out of A thorough review of the patient’s medica.5 0.99 0.89 *See text.93 Asymmetry of tactile fremitus10 0. since exposure to pleura.97 Auscultatory technique10.11 0. or sion have a dry.12–14 0.95 Dullness to percussion* Comparative technique10. and tions is mandatory. the pain is in effusions sensitive nor specific.95 0. compression of the bronchial walls. Physical findings are generally normal if patients. clinical examination had a large effusions (> 1. sensitivities and specificities of the different percussion. the size of the COPYRIGHT ELSEVIER 1985. and bulging of intercostal spaces. JAY SJ. and tactile fremitus) with bedside signs in detecting pleural effusion have not ultrasonography and found that physical exam- been extensively studied. in cases of bilateral effusions. TA B L E 2 Anecdotal physical signs of pleural effusion SIGN DESCRIPTION Skodaic resonance14 Area of hyperresonance above a pleural effusion Succussion splash16 Splashing sound produced by violently shaking patients with hydropneumothorax Grocco triangle15 Right-angle triangle of dullness found over the posterior region of the chest opposite a large pleural effusion Garland triangle 11 Small area of resonance next to the spine found in patients with large unilateral pleural effusions TA B L E 3 Findings of pleural effusion according to size FINDING SIZE OF EFFUSION < 300 mL 300–1. typically detected on chest radiography proportion to cAt the upper part of the effusion ADAPTED FROM CLINICS IN CHEST MEDICINE 1985.500 mL > 1.500 mL Tachypnea No Present Significant Chest expansion Normal Decreaseda Significantly decreaseda Tactile fremitus Normal Decreased Absent Breath sounds Vesicular Decreased Absent or bronchial Contralateral tracheal Absent Absent Present or mediastinal shiftb Bulging intercostal No Sometimes Present spaces Dyspnea may Egophonyc No Yes Yes be out of aOn the affected side or.11. both hemithoraces bMediastinal shift opposite to the side of the effusion. signs of pleural effusion10–15. TABLE 2 lists some Patterson et al11 prospectively compared less common (anecdotal) signs. a specificity of 60%. DIAGNOSTIC PROCEDURES FOR PLEURAL DISEASE. 6(1): 34. Compared with ultrasonography as less than 300 mL of fluid is present. and selecting pleural puncture sites in 67 lungs. whereas the gold standard. suggest that clinical acumen is less accurate respectively) but a similar specificity (71%). and a neg. but cer- tain reports found it about as accurate as Bigger effusions are easier to detect standard chest radiography.14–16 The physical examination (including auscultation. CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 75 • NUMBER 4 APRIL 2008 299 . The physical findings are related to the vol- Diacon et al17 assessed the accuracy of ume of fluid in the pleural effusion and its clinical examination and ultrasonography for effects on the chest wall. diaphragm. pleural effusion TABLE 1 lists the most common physical ative predictive value of 45%.500 mL) can be associated sensitivity of 76%. than ultrasonography of the chest. The limited data ination had a lower sensitivity (53% vs 80%. a with significant asymmetry of chest expansion positive predictive value of 85%. 20 In the absence of emphy.10 its sensitivity was effusion than the amount of overall chest 82%. “ninety-nine”). Chest wall tumors or skin by palpation include asymmetric chest expan. Inspection sion and asymmetric tactile fremitus. Furthermore. sym- underlying lung In addition. Other signs. TABLE 3 shows some of the common physi- FIGURE 2. the difference should be at least 2 inch. In a Small effusions tion adopted by the patient (patients with a recent study by Kalantri et al10 in 278 patients can cause large pleural effusion may have orthopnea). a sign of pleural effusion. dyspnea if rior diameter (“barrel shape”) seen in patients when the pretest probability of disease based with chronic obstructive pulmonary disease. the clinician asks the patient to say spe- the chest at the level of the nipples to com. specific for pleural effusion. asking the patient to breathe (FIGURE 1). PLEURAL EFFUSION DIAZ-GUZMAN AND BUDEV “Ninety-Nine. air. predictive value was low at 59%. 74% and a specificity of 91%. abscesses may be related to underlying empye- 300 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 75 • NUMBER 4 APRIL 2008 . it (of whom 57% had pleural effusions). An expansion of 1. repeatedly (eg. The utility of very low probability (8%) of pleural effusion. is available about its accuracy. although in the sidered abnormal. or tumors. fluid.21 More relevant to pleural study by Kalantri et al. “ninety-nine”). which we can assess by palpation. Tactile fremitus—the examiner cal findings. by inspection we can assess metrical chest expansion was associated with a disease is the expansion of the thorax. present inspecting chest expansion to detect lung Tactile fremitus is defined as the vibra- restriction was first noted by Laennec19 in tion felt by the clinician’s hand resting on the 1821. Palpation of the chest can also help in detecting underlying disease of the Palpation thorax sometimes associated with pleurisy or Signs of pleural effusion that can be detected pleural effusions. its specificity was 86%. Little information es. A simple method of evaluating chest chest wall of a patient (FIGURE 2). Palpation to detect asymmetry of chest expan- sion. and thus this sign is not sema. pare the circumference at end-inspiration and Asymmetry of tactile fremitus can be due to at end-expiration. asym- can also reveal abnormalities in the shape of metric chest expansion had a sensitivity of significant the thorax such as the increased anteroposte.22 To elicit the expansion is to place a measuring tape around sign. it can placing your hands on the patient’s back with provide other relevant information such as your thumbs pointed towards the spine and the respiratory rate and the breathing posi. Although inspection of the chest is not very Chest expansion can be evaluated by helpful in detecting a pleural effusion. metrical.5 inches or less is con. cific words repeatedly (eg. and its positive expansion is whether the expansion is sym.18 on other clinical findings was applied. depending on the amount of asks the patient to say specific words pleural fluid present.” FIGURE 1. 2. The comparative percussion technique involves comparing the sounds (dullness or hyperresonance) on the right vs the left hemithorax. In the original description. ventricle) that would interfere with the trans- sion by physical examination. some authors advocate percussion in the lateral This method consists of tapping lightly supine position to detect a shift in the dullness the manubrium sterni with the distal phalanx that would indicate movement of fluid in the of the index or middle finger while listening chest.25 over the posterior chest wall with a stetho. The pulp of a finger) along three or more parallel original technique was subsequently modified lines from the apex of each hemithorax per- for the examination of the chest by Guarino. pleural effusion.ma.26 percussive sounds penetrate a maxi. Chest percussion—the examiner bilateral. pain is worse cussion and its accuracy related to the size of ting or standing position with the arms resting the effusion are unknown. suggesting that tates each posterior hemithorax from top to or when lying the absence of the sign is very helpful in rul. bottom.27 pendicularly downward toward the base to CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 75 • NUMBER 4 APRIL 2008 301 .8 Most of these mission of the sounds. effusion (FIGURE 5): with the patient sitting up The auscultatory percussion technique and his or her back facing the examiner. tions. Kalantri et al10 at the sides or on the thighs. and crepi- tus may be due to subcutaneous emphysema. a was first described by Laennec and used to stethoscope is placed approximately 3 cm delineate the size of several organs by placing below the last rib in the mid-scapular line. Percussion is with deep found that dullness to percussion had a posi.23 who proposed that dullness is always present in a pleural effusion. localized tenderness may be associated with rib fractures or costochondritis. comparing the sounds on the two down ing out an effusion. Pleuritic chest The sensitivity of comparative chest per. percussion was mum of 6 cm (2 cm of body wall thickness and limited to the manubrium in an attempt to 4 cm of lung). and at least 500 mL of fluid must avoid other solid structures (such as the left be present in order to be able to detect an effu. the stethoscope directly above the structure to The physician then proceeds to percuss with be outlined. scope (FIGURE 4). Dullness may indicate pleural effusion (FIGURE 3). sion. There are two main techniques used to detect pleural effusions: comparative percussion and auscultatory percussion. manubrium sterni while the physician auscul- tive predictive value of 97%. followed by percussion from the his or her free hand (by finger flicking or the periphery towards the organ of interest. tion of the lung and atelectasis can also be associated with dullness to percussion. The patient must be in the sit. sides and trying to identify dullness to percus- According to classic textbook descrip.27 The authors modified descriptions are based on original studies done this technique for the detection of pleural in cadavers more than 100 years ago.24 taps the patient’s chest on alternating sides to detect the characteristic dullness of Since other conditions such as consolida. This is the technique intro- duced in the 18th century by Auenbrugger and Forbes. Percussion The chest can be percussed directly with the tips of the fingers of one hand or indirectly by placing a third finger against the surface to be percussed. applied with equal intensity over the inspiration tive predictive value of only 55% but a nega. although it may be difficult to detect if the effusion is FIGURE 3. examiner taps on the patient’s manubrium was as McDermott et al16 compared convention- while listening with a stethoscope to the al and auscultatory percussion in detecting sensitive as patient’s back. lated better with the findings on ultrasonogra- masses and consolidations. patients with obesity. They found Lichtenstein et al15 performed a study of that auscultatory percussion was 100% sensi. prospective blinded study in 118 patients.19 auscultation is perhaps Bohadana et al13 compared auscultatory the physical examination technique most used and conventional percussion with chest radio. when Bourke et al14 compared conventional and auscultatory percussion in 21 patients with abnormal radiographs. Guarino’s second method of auscultatory percussion. respec- ill patients. tively). pleural effusion in 14 hospitalized patients. Guarino and phy than did those on conventional percus- Guarino12 found this technique to be highly sion.1%. Auscultatory percussion: the sion. auscultation in critically ill patients and found 302 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 75 • NUMBER 4 APRIL 2008 . Of note. However. In a sensitivity or specificity. graphic findings in 281 patients. PLEURAL EFFUSION DIAZ-GUZMAN AND BUDEV FIGURE 5. to detect pleural effusion. It is important to mention that in this series only a few patients had a pleural effu- auscultation FIGURE 4. Although auscultatory percussion was The findings on auscultatory percussion corre- used initially to try to detect lung lesions. this Kalantri et al10 found that auscultatory method was highly (95%) sensitive and 100% percussion had a sensitivity of 58% and a specific in detecting pleural effusion. The authors gave no information about effective in detecting pleural effusion.2%) In critically but high specificity (97. radiography using ultrasonography instead of chest radiog- identify dullness to percussion. their findings Auscultation suggested that auscultatory percussion can Originally described by Laennec (who invent- detect as little as 50 mL of pleural fluid. tive for detecting large pleural effusions. pneumonia.12 raphy as the gold standard for comparison. even in specificity of 85%.3% vs 85.4% vs 19. both methods had low sensitivity (15. ed the stethoscope). or other pleural abnormalities. tion of Auenbrugger’s original treatise entitled “Inventum novum ex ical signs in the diagnosis of pleural effusion. Guarino JC. Coimbra FT. De Waele M. This sound. Clinically Oriented Anatomy. bos detegendi” (Vienna. Saunders. 44:S112. Clinical practice. Auenbrugger L. Auscultatory percussion: a simple method to 25. Ir J Med Sci 1989. Krumbhaar EB. 15. 1966. CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 75 • NUMBER 4 APRIL 2008 303 . Santiago JR. and significant interob- had similar sensitivity but higher accuracy. Critical Care Medicine. 50:218–225. The reactive enthesopathies. 4. Philadelphia. Mourgeon E. from any disease that causes direct inflamma- phonism as a pathognomonic sign associated tion of the pleurae can be associated with a with a moderate degree of effusion. pleural effusions. Leopold SS. Auenbrugger L. McCarthy M. 1962. tional percussion. Evidence-based physical diagnosis. 9500 Euclid Avenue. drome. Weil A. 11. Pleural Disease. Chen H. A Treatise On the Disease[s] of the Chest. Little is known about the predictive with chest radiography. J Gen Intern Med 1994. Nunes D. 3d ed. 158:82–84. 23. Leipzig: Vogel. 100:9–15. About egophony. Bohadana AB. The word pleural rub. A90. it is used to describe the change monic of pleural disease but not of pleural in the pronounced sound of E to A.B. On percussion of the chest: being a transla- 10. Patterson LA. Pleural effusion. 9. Graham I.10 The However. Costantino TG. Accuracy and reliability of phys. excursion: reliability of the cloth tape measure technique. Landis HRM. OH 44195. if the effusion is small. Cluzel P. 2004. Satz WA. 107:1598–1603. mesothelioma. 107:191–196. rev. Svandova E. for Dawsons of Pall Mall. London: Reprinted sonography [abstract]. Kalantri S. chest radiog. Philadelphia: Lippincott Williams lung and obstructive pulmonary emphysema. the senile 5. The barrel deformity of the chest. Leopold’s Principles and Methods of Osteopath Assoc 2007. Of note. Bouros D. Melinova L. Shifting percussion dullness of the chest: a sign of pleural detect pleural effusion. 18. 13. 2007. einer Darstellung der Lehre vom Percussionsschall. New York: ed. 14. MPH. Dalley AF. tasize to the pleura. New York: Marcel Dekker. Experience with clinical studies. Diagnosis of Company. & Wilkins. Chest 1995. Percussion of the 27. W. DO. Moore KL. Joshi R. effusion. Fries JF. Cleveland. Light RW. Hurley G. the differential diagnosis includes pleuritis. Ebert RV. Guarino JR. Parapneumonic effusions and empyema. compared effusion. Norris GW. Light RW. Forbes J. of the Chest and the Principles of Physical Diagnosis. 1761). 75:193–194. Grenier P. Am J Med 1958. J Kansas Med Soc 1974. 6. et al. and conventional chest percussion. a prospective comparison of clinical examination with ultrasound. pneu- consolidation will not be large enough to monia. auscultatory findings value of this sign. ADDRESS: Marie M. Pierce JA. 26. 1880. 346:1971–1977. Saunders. Auscultatory percussion. Brutsche MH.19 Pleural rub. Philadelphia: W. and tumors that metas- produce this sign. Stastny B. Similarly. 4th ed. 3. Comparative diagnostic performances of auscultation. Montgomery CM. and lung ultrasonography in acute respiratory distress syn. Cleveland Clinic. Volume and cellular content of 16. 25:13–22. 3:75–80. Hopkins HU. Library of the New York Academy of Medicine and Hafner Publishing 7. a pleural rub will disappear mechanism responsible for finding this sign in once an effusion develops. in the study displacement and compression or consolida. tion of pleural effusions of hospitalized patients. Ann Emerg Med 2004. other lung con. e-mail budevm@ccf. 1965. Am J of conventional percussion and auscultation percussion in the detec- Respir Crit Care Med 2000. Julliard KN. Goldstein I. South Med J 1997. Saunders. Handbuch und Atlas der topographischen Percussion nebst ties by auscultatory percussion: a comparative study with conven. percussione thoracis humani. 5th ed. Accuracy of pleural puncture sites: Soc 2006. Dis Mon 1985. 21. Philadelphia: Lippincott Williams & Wilkins. 2001. Inventum Novum. In fact. 1986 to 1990. Neuburger M. Anesthesiology 2004. Diseases 22. 1929. Agur AMR. ■ ■ REFERENCES 1. Pleural Diseases. Pract 1997. Bockenhauer SE. 31(1):1–46. server variability needs to be taken into Absent or diminished breath sounds account.B. Physical Diagnosis. is pathogno- means goat. 9:483–486. classically described egophony comes from the Greek “ego. Department of Pulmonary and raphy. with an overall diag.28 strongly suggest an effusion. Baltimore: The Johns Hopkins Press. et al. Bourke S. 24. Gilbert VE. J Am 8. by Kalantri et al it had a very low sensitivity tion of the lung at the top of the effusion. Light RW. Detection of lung abnormali. Chest 2003. (5%) but a very high specificity (99%). Li R. Pleural effusions that result Egophonism.org. a new aid in the examination of chest re-visited: a comparison of the diagnostic value of auscultatory the chest. may produce egophonism without a pleural nostic accuracy of 60%. Noppen M.it to have a very low sensitivity (42%) but a ditions associated with large consolidations higher specificity (90%). McDermott TD. The effusion. Laennec RTH. J Am Acad Nurse 2. 20. Soler M. Philadelphia: W. Proc Am Thorac 17. Stafford F. 90:1255–1256. N Engl J Med 2002. A comparison normal pleural fluid in humans examined by pleural lavage. Little is known massive pleural effusions is probably upward about the accuracy of this finding. Measuring thoracic Chest 1995. Respiration 1986. McGee SR. 9:71–74. ut signo abstrusos interni pectoris mor- 101:431–438. 2006. Weedon J. a prospective comparison of physical examination with bedside ultra. 123:436–441. Sapira JD. 12. 28.” which as rubbing of unoiled leather.B. Lokhande T. 162:1023–1026. Rouby JJ. 5th 19. Chestnut T. Respir Med 2007. 108:865–867. Laennec also described ego. Light RW. Guarino JR. Diacon AH. Diagnosing pleural effusion: 1936. Lichtenstein D. Budev. Marel M. Schumann L.