Pleural Effusions in the Medical ICU

Pleural Effusions in the Medical ICU* Prevalence, Causes, and Clinical Implications Lalaine E. Mattison, MD; Lynn Coppage, MD; Daniel F. Alderman, MD; John 0. Herlong , MD; and Steven A. Sahn, MD, FCCP

Objective: To determine the prevalence and causes of pleural effusions in patients admitted to a medical ICU (MICU). Design: Prospective. Setting: MICU in a tertiary care hospital. Patients: One hundred consecutive patients admitted to the MICU at the Medical University of South Carolina whose length of stay exceeded 24 h had chest radiographs reviewed daily and chest sonograms performed within 10 h of their latest chest radiograph. Results: The prevalence of pleural effusions in 100 consecutive MICU patients was 62%, with 41% of effusions detected at admission. Fifty-seven of 62 (92%) pleural effusions were small. Causes of pleural effusions were as follows: heart failure, 22 of 62 (35%); atelectasis, 14 of 62 (23%); uncomplicated parapneumonic effusions, seven of 62 (11 %); hepatic hydrothorax, five of 62 (8%); hypoalbuminemia, five of 62 (8%); malignancy, two of 62 (3%); and unknown, three of 62 (5%). Pancreatitis, extravascular catheter migration, uremic pleurisy, and empyema caused an effusion in one instance each. Heart failure was the most frequent cause of bilateral effusions (13/34 [38%]). When compared with patients who never had effusions during their MICU stay, patients with pleural effusions were older (54±2 years, mean±SEM, vs 47±2 years [p=0.04]), had lower serum albumin concentration (2.4±0.1 vs 3.0±0.01 gldL [p=0.002]), higher acute physiology and chronic health evaluation II scores during the initial 24 h of MICU stay (17.2± 1.1 vs 12± 1.2 [p=O.OlO]), longer MICU stays (9.8±1.0 vs 4.6±0.7 days [p=0.0002]), and longer mechanical ventilation (7.0± 1.3 vs 1.9±0.7 days [p=0.004]). No patient died as a direct result of his or her pleural effusion. Chest radiograph readings had good correlation with chest sonograms (p
Although most patients are admitted to the medical ICU (MICU ) for conditions other than pleural disease, the pleura is often s econdarily affected by pulmonary parenchymal disorders and dysfunction of other organ systems. Most patients *From the Division of Pulmonary and Critical Care Medicine (Drs. M attison and Sahn ) and the Department of Radiology (Drs. Coppage, Alde rman, and Herlong), Medical University of South Carolina, Charleston. Supported b y the Division of Pulmonary and Critical Care Medicine and th e Department of R daiology, Medical University of South Carolina, Charleston. Manuscript received June 20, 1996; revision accepted O ctober 8. Reprint requests: Dr. Sahn, Prof essor of Medicine, DirectorPulnwnary and Critical Care, Medical University of South Carolina, 171 Ashley Avenue, Charleston, SC 29425 1018

will have several chest radiographs during their MICU stay to detect pneumothorax and new infiltrates and to assess endotracheal tubes, chest tubes, and catheters. Detection of small pleural effusions is problematic in these patients, because chest radiographs are usually obtained with the patient in the supine or semirecumbent position and a pleural fluid volume of < 500 mL may produce only a subtle increased haziness over the lower lung zone while the bronchovascular markings remain unobliterated.1 Lower lobe parenchymal collapse or consolidation without air bronchogram visualization can be difficult to distinguish from pleural fluid on chest radiographs taken in the supine position.2 Clinical Investigations in Critical Care

The prevalence, causes, and clinical significance of pleural effusions in MICU patients are unknown. MICU patients appear to be at risk for developing pleural effusions because of the clinical setting and the severity of their primary disease. Many MICU patients present with hypotension and hemodynamic instability and are treated with aggressive hydration leading to fluid overload. Often, MICU patients are immobile because of pain, sedation, or paralytic drugs , thus putting them at risk for developing atelectasis. We hypothesized that if actively sought, pleural effusions in critically ill patients would be common. Therefore, we prospectively studied patients admitted to the MICU and reviewed their serial chest radiographs and clinical presentations to assess prevalence, causes, and clinical significance of pleural effusions.

MATERIALS AND METHODS \Ve prospectively determined the prevalence and causes of pleural effusions in critically ill patients in a tertiary care center. All patients admitted over a 4.5-month period to the MICU of the Medical University of South Carolina, whose length of stay exceeded 24 h, were enrolled. A patient readmitted >30 days from their last MICU discharge date was considered a separate admission. One investigator (L.E.M. ) collected clinical and laboratory data on all enrolled patients during their entire stay in the MICU. The data collected included the following: age, sex, primary reason for admission to the MICU , vital signs, CBC count, serum electrolytes, serum albumin concentration, microbiologic results, hemodynamic parameters, and pleural fluid characteristics. Acute physiology and chronic health evaluation (APACHE ) II score was recorded on all patients during the first 24 h of MICU stay. An unbiased attending radiologist (L.C. ) and attending pulmonologist (S.A.S.) independently reviewed all portable chest radiographs to determine the existence of pleural effusions. All chest radiographs were performed with the patient in the upright or semi-upright position. In instances in which the two radiographic readings differed, the serial chest radiographs were presented to another unbiased attending radiologist for a third and deciding interpretation. When the chest radiographs were analyzed for existence of pleural effusion , no historic, laboratory, or chest sonographic data were available to the readers. The volume of the effusion was estimated and the laterality recorded. A pleural effusion was defined as small if the fluid obliterated the costaphrenic angle or obscured the lower lung zone; moderate if it opacified the lower and middle lung zones; and large if all three lung zones were opacified. In each patient in whom pleural fluid was detected, an expert in pleural diseases (S.A.S. ) reviewed all available clinical data and determined the cause of the effusion. Chest radiographs up to the time of hospital discharge were also reviewed to aid in the confirmation of the cause of the pleural effusion. The presence or absence of pleural effusions was recorded based only on chest radiographic readings. A bedside chest ultrasonogram was performed on each patient after consent was obtained within 72 h of admission, except in patients who were in hemodynamically unstable conditions. The ultrasonographer did not have knowledge of any patient information or access to the chest radiograph interpretation. Although

difficult to achieve in every patient because of the severity of the underlying disease and the presence of indwelling catheters and support tubes, the ultrasonographer attempted to optimize patient position. The criteria for diagnosing the causes of the pleural effusions were as follows: heart failure-S 3 gallop, basilar crackles and jugular venous distention, chest radiograph showing cardiomegaly, excess extravascular lung water and usually bilateral effusions, transudative fluid, and elevated pulmonaty capillary wedge pressure (PCWP) (when available); atelectasis-compatible chest radiograph findings (plate-like changes, volume loss, and small ipsilateral effusion) and rapid disappearance of the fluid with resolution of the atelectasis; no other cause of the effusion evident; hepatic hydrothorax-an effusion in the setting of liver failure and clinically or ultrasound documented ascites; hypoalbuminemia-serum albumin level :Sl.8 gldL and small to moderate bilateral effusions without pulmonary parenchymal disease; parapneumonic effusion-fever, new localized pulmonary infiltrate, ipsilateral free-flowing or loculated polymorphonuclear predominant exudate >vith or without pleural fluid acidosis, and bacterial isolation from sputum or BAL specimens; empyema-pus or compatible pleural fluid chemistries or positive pleural fluid Gram stain or culture; uremic effusion-a patient receiving long-term hemodialysis \vith a unilateral effusion and compatible clinical course that resolved with continued dialysis and without another cause identified; acute pancreatitis- historic, physical, and laboratory evidence of acute pancreatitis and a left -sided effusion that resolved as the pancreatitis resolved; extravascular catheter migration (EVCM)-when aspirated fluid was similar to the infused fluid and dye was noted to accumulate in the pleural space shortly after injection through the distal port of the catheter; and malignancy-cytologic specimen positive for malignant cells.

Statistics

Age, serum albumin concentration, length of MICU stay, APACHE II scores, and duration of intubation in patients with and without pleural effusions detected by serial chest radiographs were compared using the unpaired, two-tailed Student's t test. Mortality rates between the two groups were assessed using x2 analysis (Statworks; Cricket Software; Philadelphia). All values are expressed as mean±SEM. To determine the degree of interobserver agreement on the chest radiograph and chest sonogram readings, we used Cohen's kappa statistical formula, which is a nonparametric measure of association.

RESULTS

One hundred forty-nine patients were admitted to the MICU at the Medical University of South Carolina over the 4.5-month study period. Forty-nine patients were excluded because their MICU stay was :::::24 h. One hundred consecutive patients who were enrolled completed the study protocol. As designed, the investigators made no attempt to influence the diagnostic or therapeutic management of enrolled subjects. The study patients had diverse primary diseases that led to their admission to the MICU (Table 1) and no patient was admitted to the MICU because of primary pleural disease. Three patients CHEST I 111 I 4 I APRIL, 1997

1019

Table !-Causes of Pleural Effusions and Primary Admitting Diagnosis to the MICU (n=62) Disease Category Primary Admitting Diagnosis to the MICU Hypercapneic respiratory failure (n=7) Myasthenia gravis Status asthmaticus Bronchial stent migration Rhabdomyolysis Idiopathic pulmonary fibrosis Hypoxic respiratory failure (n = 16) Malignancy

No. of Patients

2 2 1 1 1

6

Pneumonia

3

Lung biopsy Alveolar hemorrh age

2

Aspiration Bronchiolitis obliterans with organizing pneumonia ARDS

2

Sepsis (n=6)

6

Cardiovascular disorder (n = 16) Pulmonary edema Cardiac arrest

7 4

Arrhythmia Acute myocardial infarction Hypertensive ctisis Cardiac tamponade Neurologic disorder (n= 4) Encephalopathy Status epilepticus Delirium tremens Cerebrovascular accident GI disorder (n= 10) GI Bleed

Liver failure Metabolic disorder (n=3) Diabetic ketoacidosis Acute renal failure Drug overdose

2 1 1

1

6

4

Cause of Pleural Effusion (No. of Cases)

Atelectasis Unknown (1) Heart failure (l) Atelectasis Atelectasis Atelectasis

Atelectasis (1) Heart failure (2) Malignant effusion (2) Hypoalbuminemia (1) Parapneumonic (2) Empyema (l) Unknown Parapneumonic (1) Unknown (1) Parapneumonic Parapneumonic Pancreatitis (1) Hypoalbuminemia (1) Atelectasis (2) Heart failure (1) EVCM (1) Parapneumonic effusion (l) Hypoalbuminemia (1) Heart failure (7) Heart failure (3) Atelectasis ( 1) Hypoalbuminemia (1) Heart failure (1) Heart failure Heart failure Heart failure Atelectasis Heart failure Atelectasis Heart failure Heart failure (1) Hepatic hydrothorax (1) Uremia (1) Atelectasis (2) Hypoalbuminemia (1) Hepatic hydrothorax Parapneumonic Atelectasis Heart failure

were enrolled twice because their readmission date was :2::30 days from their MICU discharge date. There were 41 male and 59 female patients. Only 1020

two cases required a third review of the chest radiographs, since the interpretations by the radiologist and pulmonologist concurred in 98 of the 100 cases. Forty-one of 100 patients had effusions on admission, and an additional 21 patients developed effusions dming their MICU stay. The mean age of patients with pleural effusion was 54:±:2 years (mean±SEM) and the age of those without effusion was 47:±:2 years (p=0.04). The albumin concentration in the group with pleural effusions was lower (2.4:±:0.1 gldL) than in the patients without effusions (3.0:±:0.1 gldL; p=0.002). Additionally, those with effusions had longer MICU stays (9.8:±: 1.0 vs 4.6:±:0.7 days; p=0.0002) and longer mechanical ventilation (7.0:±:1.3 vs 1.9:±:0.7 days; p=0.004) (Table 2). Using the APACHE II classification system, patients with pleural effusions had higher scores (17.2:±:1.1 vs 12:±:1.2; p=O.OlO) during the initial 24 h in the MICU. Small pleural effusions were seen in 57 of 62 (92%) patients, moderate effusions were detected in four patients with heart failure, and a large effusion was caused by adenocarcinoma of the lung with pleural metastasis in one patient. Pleural effusions resulted from noninfectious causes in 51 of 62 (82%) patients . Heart failure was the leading cause (22 of 62, 35%) of all effusions; atelectasis was next most frequent occurring in 14 of 62 (23%) patients (Table 3). The most common cause of bilateral effusions in this cohort was heart failure, seen in 13 of 34 (38%) patients, while atelectasis, noted in 10 of 28 (36%) patients, was the most frequent cause of a unilateral effusion. Twentyfive of the 100 patients were known to have a history of heart failure at the time of MICU admission. Fourteen of the 25 (56%) patients had effusions caused by heart failure during their MICU stay. Ten of the remaining 25 (40%) patients never developed an effusion, while one patient eventually developed an atelectatic effusion during her MICU stay. An additional eight patients were discovered to have

Table 2-Differences Between MICU Patients With and Without Pleural Effusions With Without Pleural Pleural Effusions Effusion

p Value

52:<::2* 47:!:2 0.04 Age, yr 2.4±0.1 3.0±0.1 0.002 Serum albumin, gldL 9.8:!:1.0 4.6±0.7 0.0002 Length of MICU stay, d 17.2± l.l 12.0:<::1.2 0.010 APACHE II points during the first 24 h of MICU stay Duration of mechanical ventilation, d 7.0±1.3 . 1.9±0.7 0.004 *Mean±SEM. Clinical Investigations in Critical Care

Table 3-Causes of Pleural Effusions in Patients in the MICU (n=62) No. of Patients

Laterality of Effusions (n=62)

Causes of Pleural Effusion

(%)

Bilateral

Left

Right

Heart failure Atelectasis Uncomplicated parapneumonic effusion Hepatic hydrothorax Hypoalbuminemia Malignancy Pancreatitis EVCM Uremic pleurisy Empyema Unknown

22 (35) 14 (23) 7 (11)

18 4 4

3 9

1 1 2

5 (8) 5 (8) 2 (3) 1 (2) 1 (2) 1 (2) 1 (2) 3 (5)

0 5 0 0 0 0

2 0 1 0 0 0 0 0

3 0 1 1 1 1 0 2

underlying heart failure during their MICU stay. Therefore, 67% (22133) of patients with heart failure developed pleural effusions during their MICU admission. Effusions due to hypoalbuminemia, seen in five of 62 (8%) patients, were caused by severe malnutrition. The single iatrogenic pleural effusion was caused by indwelling catheter erosion of the wall of the superior vena cava 4 days after placement for chemotherapy access. The causes of three patients' pleural effusions were undetermined despite careful review of all relevant historic and clinical data; no diagnostic thoracenteses were performed on these three patients. There were eight infectious causes of pleural effusions. Seven patients had uncomplicated parapneumonic effusions, while one patient had bilateral empyemas. Only three of the eight patients had successful thoracenteses despite optimization of patient positioning and ultrasound guidance at the bedside. When thoracentesis was unsuccessful (three of eight patients) in patients with very small pleural effusions, or pleural fluid volume was deemed too small to safely pursue (two of eight patients), these patients were observed carefully. In all five patients, fever resolved within 48 h of starting IV antibiotic therapy and their effusions improved over the course of their MICU stay. The clinically documented pneumonia in five of the eight (63%) patients was community acquired. The parapneumonic effusions were caused by pneumococcal pneumonia in two patients, Legionella pneumonia in 1:\vo, and staphylococcal and enterococcal pneumonia, in one each. No bacterial, fungal, or viral isolate could be identified in one patient who was neutropenic and febrile after a bone marrow transplant despite repeated sputum cultures and BAL. Inability to grow the

offending organism in this neutropenic patient with a new localized pulmonary parenchymal infiltrate was likely related to suppression by antibiotics. The one case of bilateral empyemas was caused by Streptococcus pneumoniae. This patient had sequential bilateral thoracenteses, and fluid from both pleural spaces showed Gram-positive cocci on Gram's stain and the pleural fluid cultures grew S pneumonia. The bilateral thoracenteses were performed immediately on admission despite the small size of the effusions because of a strong suspicion for infected pleural spaces. A total of 14 diagnostic thoracenteses were attempted on 13 patients during the study period, and 11 of 14 (79%) thoracenteses were successful in obtaining fluid. Three of 22 (14%) patients with effusions secondary to heart failure had pleural fluid analyses because of increasing effusions despite response to diuretic therapy. All three effusions were transudative and the effusions decreased with the addition of inotropic agents and more aggressive diuretic therapy. Only one of 14 (7%) patients with acute atelectasis as the cause of pleural effusion underwent a diagnostic thoracentesis. The patient was admitted to the MICU for heart failure, but developed a new right-sided pleural effusion after her PCWP was normal for 3 days. The fluid was transudative and resolved following resolution of the atelectasis. The remainder of the diagnostic thoracenteses revealed parapneumonic effusions, malignant effusions, and extravasation of hyperalimentation fluid. In 49 of 62 (79%) patients with pleural effusions who did not undergo thoracentesis, the effusions improved with therapy of the underlying disease. At the time of admission to the MICU, 41 of 100 (41 %) patients had pleural effusions detected by chest radiograph. Seventeen of the remaining 21 (76%) patients developed an effusion by the sixth MICU day. Only 10 of 62 (16%) pleural effusions completely resolved by the time of MICU discharge.

Table 4-Chest Radiograph and Sonogram* Detection of Pleural Effusion in MICU Patients 1 (n=74)

Positive by chest sonogram Negative by chest sonogram Total

Positive by Chest Radiograph

Negative by Chest Radiograph

Total

25 3 28

10 36 46

35 39 74

*Chest sonograms were performed within 10 h of the latest chest radiographs. 1 Cohen's kappa statistic coefficient=0.64 and a p value <0.0001, which indicate that the good degree of agreement between the chest radiograph and chest sonogram readings occurred beyond chance. CHEST I 111 I 4 I APRIL, 1997

1021

None of the patients died or were readmitted to the MICU because of their pleural disease. There were 74 patients who had a chest sonogram within 10 h of their latest chest radiograph (Table 4). In 61 of the 74 (82%) chest ultrasounds, the results correlated with chest radiograph interpretation. In six of the remaining 13 patients, the chest sonogram recognized small pleural effusions 24 h prior to detection by chest radiograph. An additional four patients had small effusions never detected by serial chest radiographs. These four patients were not included in the group of patients with pleural effusions. The chest sonogram failed to confirm the radiographic readings in three patients with small effusions. This may have been explained by the inability to optimize patient position. To demonstrate the degree of agreement between the chest radiograph and chest sonogram readings by independent reviewers, Cohen's kappa coefficient was calculated. The method revealed a kappa coefficient of 0.64, and standard error of 0.11. These values yield a normal variate (z) of 5.65, and its corresponding p value is <0.0001. These results indicate that the good degree of agreement between the chest radiograph and chest sonogram readings occurred beyond chance alone. Ten (16%) of the 62 patients with pleural effusions died and four ( ll%) of 38 patients without effusions died (p=not significant) in the MICU. No patient died as a direct result of his or her pleural disease. DISCUSSION

Patients in ICUs are rarely admitted for primary pleural disease, such as hemothorax, spontaneous pneumothorax, or large pleural effusions causing respiratory failure. In this study, the causes of pleural effusions were as varied as the spectrum of primary diseases that prompted the MICU admission. Heart failure, the most common cause of pleural effusions, was bilateral (18/22 cases, 82%) and transudative. These characteristics are consistent vvith previous reports. 3 -5 A total of 33 patients were diagnosed as having heart failure, with 25 of 33 (76%) patients known to have left ventricular dysfunction at the time of MICU admission. The diagnosis of heart failure was established in eight of the remaining 33 (24%) patients because pleural effusions developed after IV hydration. Subsequent echocardiograms showed left ventricular dysfunction. In contrast to some studies that reported a 39 to 51% prevalence of pleural effusions in heart failure patients,4 our study showed a higher prevalence (22133 patients, 67%). These patients had positive fluid balance especially during the first few days of their MICU stay, worsening alveolar-arterial oxygen gradient, and elevated PCWP (22.2±4.7 mm Hg; n=lO). Our higher prevalence may be explained by our aggressive search for 1022

pleural effusions coupled with a clinical setting wherein patients with underlying, albeit unsuspected, heart disease were aggressively hydrated. Alveolar collapse causes increased negative pleural pressure that theoretically would result in an increased volume of interstitial ultrafiltrate moving into the pleural space. In atelectasis, fluid will continue to accumulate in the pleural space until the balance of Starling forces returns to steady state. This presumably happens quickly as the presence of pleural fluid increases pleural pressure and decreases pleural fluid formation. Atelectatic effusions, the most common cause of a unilateral pleural effusion in the MICU, resulted from severe muscle weakness, airway obstruction, or patient immobility. Patient movement is often impeded by pain, sedation, restraints, or alteration of mental status as a consequence of the primary disease. Follow-up chest radiographs both in and after discharge from the MICU showed disappearance of pleural effusions with resolution of atelectasis. In fact, four of seven patients with hypercapneic respiratory failure, and two of four patients with neurologic disorders had atelectatic effusions, supporting the proposed risk factors stated above. Suspicion for a complicated parapneumonic effusion was raised when the patient remained in a clinically toxic condition despite appropriate antibiotic therapy. These patients should have an immediate diagnostic thoracentesis. One of eight (13%) patients with parapneumonic effusion had bilateral empyemas necessitating bilateral chest thoracostomy tubes. The conditions of the other seven patients with uncomplicated parapneumonic effusions improved with antibiotic therapy alone. Patients with pleural effusions were more ill compared with patients who never developed pleural fluid. The former had extended MICU stays, required longer ventilatOiy support, and higher APACHE II scores during the first 24 h of their MICU stay. The APACHE II classification system attempts to quantify the severity of the disease in critically ill patients, the higher the score being indicative of worse outcome. 6 The lower serum albumin level in the patients with effusion was also an index of their overall poorer condition7 •8 and probably was contributory to the formation and volume of pleural fluid in some instances. Thirteen of 62 (21%) patients with documented pleural effusions had fluid analysis. According to the study design, the physicians caring for the patients were not informed of our data, and they made independent decisions whether to obtain pleural fluid . Diagnostic thoracenteses were done to confirm malignant involvement of the pleura or to exclude an infected pleural space in the setting of a new or increasing effusion. In four of 62 (6%) patients with effusions, the thoracentesis made a difference in Clinical Investigations in Critical Care

their management; two had malignant effusions, one had bilateral empyema, and one had extravasation of fluid infused through a central catheter. We have shown previously that thoracentesis in the ICU in mechanically ventilated patients is safe and results in no more complications than in non-ICU patients. 9 The value of daily portable chest radiographs in ICUs is controversial. The image obtained from pmtable chest radiographs is technically inferior to the standard posteroanterior chest radiograph because the shorter distance between the patient and the source of the X-ray beam results in magnification of the cardiac silhouette and blurring of thoracic shadows.5 •10 However, the portable chest radiograph is the standard radiologic imaging technique in critically ill patients and its proponents cite studies that show that routine chest radiographs alter management in mechanically ventilated patients. 10-12 There are contrasting opinions on the sensitivity of chest radiographs taken with the patient in the supine position in the detection of pleural effusions. 2 ·13 There is evidence that chest ultrasound increases the detection rate of pleural fluid compared with portable chest radiographs.10 To maximize assessment of the pleural space, we included chest ultrasound in the evaluation. All subjects did not have chest ultrasound because of hemodynamic instability or refusal to undergo the procedure. There was a good correlation between the radiographic and sonographic interpretations, as illustrated by the Cohen's kappa coefficient value of 0.64. Three patients with pleural effusions not seen on ultrasound had very small atelectatic effusions. Sonograms in these patients had to be pe1formed while the patients were in supine position due to their severe underlying disease. Small effusions (<100 mL) may not be detectable on supine chest sonograms 14 but can be documented by CT. 5 .I5 Our study has limitations. Our findings and recommendations may not be applicable to patients in the surgical or trauma ICUs, wherein patients' underlying proble ms are different from MICU patients. In addition, the true incidence of pleural effusions may be underestimated by the study design. The use of serial chest ultrasound coupled with serial chest radiographs could have increased detection of small pleural effusions.

cially with cardiomegaly and without very low albumin levels, the most likely cause is heart failure. Atelectasis accounts for most unilateral effusions. This prospective study showed that pleural effusions in critically ill patients are common and that the portable chest radiograph is sensitive in detecting these effusions. In the present era of cost containment, observation without thoracentesis in most MICU patients \Nith small pleural effusions may be warranted initially. This is particularly true in the afebrile patient with bilateral effusions, such as heart failure, atelectasis, and hypoalbuminemia. However, if infection is a consideration, a thoracentesis should be done without delay, especially if the effusion increases. ACKNOWLEDGMENT: We acknowledge Yuko Y. Palesch, PhD, Department of Biometry and Epidemiology at the Medical University of South Carolina, Charleston, for her help with statistical analysis. REFERENCES

2 3 4 5 6

7

8 9 lO 11 12

CONCLUSIONS

In 100 consecutive admissions to the MICU of a tertiary care center with stays greater than 24 h, the prevalence of pleural effusions was 62%, with 41% being present on admission. Most pleural effusions were small, and two thirds were due to noninfectious causes, namely heart failure and atelectasis. When bilateral effusions are seen in MICU patients, espe-

13 14 15

Salm SA. Pleural disease in the critically ill patient. In: Rippe JM , In..in RS, Fink MP, et al, eds. Intensive care medicine. 3rd ed. Boston: Little Brown, 1995; 720-37 Woodring JH. Recognition of pleural effusion on supine radiographs: how much fluid is required? AJR 1984; 142: 59-64 Sahn SA. The pleura. Am Rev Respir Dis 1988; 138:184-234 Wiener-Kronish JP, Matthay MA. Pleural effusions associated with hydrostatic and increased permeability pulmonary edema. Chest 1988; 93:852-58 Wiener MD, Garay SM, Leitman BS, et al. Imaging of the intensive care unit patient. Clin Chest Med 1991; 12:169-97 Knaus WA, Draper EA, Wagner DP, et al. APACHE II: a severity of disease classification system. Ctit Care Med 1985; 13:818-29 Hermann FR, Safran C, Levkoff SE, et al. Serum albumin level on admission as a predictor of death, length of stay, and readmission. Arch Intern Med 1992; 152:125-30 Corti MC, Guralnik JM, Salive ME, et al. Serum albumin level and physical disability as predictors of mortality in older persons. JAMA 1994; 272:1036-42 Godwin JE, Sahn SA. Thoracentesis: a safe procedure in mechanically ventilated patients. Ann Intern Med 1990; 113:800-02 Yu CJ, Yang PC, Chang DB, et al. Diagnostic and the rapeutic use of chest sonography: value in critically ill patients. AJR 1992; 159:695-701 Bekemeyer WB , Crapo RO , Calhoon S, et al. Efficacy of chest radiography in a respiratory intensive care unit: a prospective study. Chest 1985; 88:691-96 Hall JB , White SR, Kanison T. Efficacy of daily routine chest radiographs in intubated, mechanically ventilated patients. Crit Care Med 1991; 19:689-93 Ruskin JA, Gurney JW, Thorsen MK, et al. Detection of pleural effusions on supine chest radiographs. AJR 1987; 148:681-83 Eibenberger KL, Dock WI, Ammann ME, et al. Quantification of pleural effusions: sonography versus radiography. Radiology 1994; 191:681-84 Mirvis SE, Tobin KD, Kostrubiak I, et al. Thoracic CT in detecting occult disease in critically ill patients. AJR 1987; 148:685-89

CHEST I 111 I 4 I APRIL, 1997

1023