Bronchoscopy with bronchoalveolar lavage in tuberculosis and fungal infections.

R P Baughman, M N Dohn, R G Loudon and P T Frame Chest 1991;99;92-97 DOI 10.1378/chest.99.1.92 The online version of this article, along with updated information and services can be found online on the World Wide Web at: http://chestjournal.chestpubs.org/content/99/1/92

Chest is the official journal of the American College of Chest Physicians. It has been published monthly since 1935. Copyright1991by the American College of Chest Physicians, 3300 Dundee Road, Northbrook, IL 60062. All rights reserved. No part of this article or PDF may be reproduced or distributed without the prior written permission of the copyright holder. (http://chestjournal.chestpubs.org/site/misc/reprints.xhtml) ISSN:0012-3692

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians

Bronchpscopy with Bronchoalveolar Lavage in Tuberculosis and Fungal Infections*

Robert P. Baughman, M.D., F.C.C.P.; Michael N. Dohn, M.D.; Robert G. London, M.B. Ch.B., F.C.C.P.; and Peter T. Frame, M.D.

Study objective: To determine the utility of bronchoscopy with bronchoalveolar lavage for diagnosing M tuberculosis and fungal infections. Design: Retrospective review of patients over a six-year period. Setting: In- and outpatients of one University hospital and affiliated Veterans Administration Medical Center. I'atii'ntx: Those who were subsequently found to have either M tuberculosis or fungal infections. Interventions: Bronchoscopy with bronchoalveolar lavage specimens were compared to prebronchoscopy sputum, when available. Specimens were sent for smear and culture for both acid-fast bacilli and fungi. In the case of lavage, an aliquot also was studied for cellular differential. Measurements and Results: For TB, sputum was smearpositive in 6/47 (34 percent) and culture positive in 24/47

(51 percent), while bronchoscopy was smear positive in 34/50 (68 percent) and culture positive in 46/50 (92 percent). For fungal infections, no sputum was smear-positive and only 1/22 (5 percent) was sputum culture-positive, while bronchoscopy was smear-positive in 14/41 (34 percent) and culture positive in 35/41 (85 percent). Bronchoscopy wash ings and BAL provided complementary specimens. Eightythree patients had adequate lavages and the cellularity was significantly different from controls (lymphocytes: TB 1811.2 percent [meanSD]; fungal: 1311.1 percent; controls 63.1 percent; p<0.001; neutrophils: TB 911.5 percent; fungal: 6 9.1 percent controls: 2 1.5 percent, p<0.01); however, there was overlap and no pattern was characteristic for TB or fungal infections. Conclusion: Bronchoscopy with BAL is useful in diagnosing tuberculosis and fungal infections. (Chest 1991; 99:92-97)

"D ronchoscopy

with BAL has proved to be a valuable

-^-* tool in evaluating the inflammatory reaction of the lung in several diseases.1 In particular, it is useful in assessing the nature of the alveolitis in pulmonary sarcoidosis and hypersensitivity pneumonitis. Both diseases are characterized by a granulomatous reaction in the lung as well as interstitial inflammation. More over, both diseases are characterized by an increased percentage of lymphocytes retrieved by BAL, espe cially in patients with evidence of active, progressive disease.2-3 The study the lung by concentrated to diagnose of infectious granulomatous diseases of BAL is limited. Studies to date have on bronchoscopy with lavage as a method TB.J"6 The results to date have been

The current study reports the use of bronchoscopy with BAL in infectious granulomatosis. The report deals with the use of bronchoscopy to diagnose infec tious agents in patients with fever and obscure diag nosis. In addition, we report the BAL findings in tuberculous and fungal infections.

METHODS
All bronchoseopies of patients with a final diagnosis of Mycobiictcrium tuberculosis or fungal infections who underwent bron choscopy at the University of Cincinnati Medical Center or the Veterans Administration Hospital, Cincinnati were reviewed. Mycobacterial and fungal culture results of all specimens, regardless of source, were reviewed. Patients found to have documented M tuberculosis or fungul infections who underwent bronchoscopy were enrolled in this study. There were two groups of patients, those undergoing bronchoscopy for diagnosis of their lung disease (30 patients with M tuberculosis and 41 with fungal infections) and those with known M tuberculosis infection who underwent bron choscopy for study purposes (20 M tuberculosis patients). Patients with sputum positive for AFB by smear were encouraged to participate in this study and gave informed consent of a protocol approved by our Institutional Review Board. A control group of 35 smoking adults undergoing bronchoscopy for localized lung lesions also was studied by BAL. Patients underwent bronchoscopy and attempted lavage in the usual manner of our institution." During the entire time of the bronchoseopy, the suctioned specimen was collected and this was considered the bronchial wash. In addition, the bronchoscope was advanced and wedged in the area of the infiltrate in those patients with locali/ed lesions. In patients with diffuse infiltrates or normal chest roentgenograms, the bronchoscope was wedged in the right middle lobe or lingula. In the control group, the lavage was

conflicting, and in some cases, recommendations have been made not to use bronchoscopy for diagnosing TB.7 Fungal infection represents another infectious gran ulomatous disease. To date, most reports have cen tered upon the utility of bronchoscopy and BAL isolating various fungi," especially in immunocompromised patients.9'"

From University of Cincinnati Medical Outer, Cincinnati. the Supported in part by National Institutes of Health grants RR 00068 and AI 25897. Manuscript received April 2; revision accepted June 26. Reprint requests: Dr. Baughman, 231 Bethesda Avenue, Rm 7511, Cincinnati 45267-0564

92

Bronchoscopy and BAL in Tuberculosis and Fungal Infections (Baughman etal)

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians

performed in the lung opposite the lesion. Aliquots of 60 ml of normal saline solution were instilled and immediately aspirated using a hand-held syringe. A total of 120 to 240 ml was instilled in each patient. The aspirated fluid was pooled. In those patients in whom less than 10 ml could be aspirated, the lavage was believed to be unsuccessful and no further analysis of lavage fluid was performed. An aliquot of the bronchial wash and BAL fluid was sent for AFB smear and culture as well as fungal culture. An additional aliquot was sent for cytologic examination of both the wash and lavage fluid using methenamine silver stain to identify fungal forms. Postbronchoscopy sputum specimens, when available, also were sent for AFB smear and culture as well as fungal culture. Another aliquot of the BAL fluid was analyzed for cellular content. Two to four hundred microliters of unconcentrated BAL fluid was spun onto glass slides using a cytocentrifuge (Cytospin II, Shandon Southern Instruments, Sewickley, PA). Slides were stained with a modified Wright-Giemsa stain (Diff-Quik, American Scientific Prod ucts, McGaw Park, IL). Cell differential counts were performed on 200 nucleated cells per slide using standard morphologic criteria established for our laboratory.12 Chest roentgenograms were performed in all patients 24 h prior to bronchoscopy. The chest roentgenograms were divided into six possible groups: (1) cavitary (one or more cavities of greater than 5 mm in diameter); (2) infiltrate (infiltrate in two or fewer lobes of the lung without cavities seen); (3) nodule (one or more discrete nodules without associated infiltrate); (4) miliary (diffuse reticulonodular infiltrate involving more than two lobes); (5) effusion (pleural effusion without associated parenchyma! disease except volume loss); and (6) normal (no abnormalities seen). In addition, the patients' charts were reviewed to identify any possible underlying disease which may have led to increased risk for infection. Among the possible causes identified were AIDS, leukemia, lymphoma, solid organ malignancies, chronic corticosteroid administration and solid organ transplantation.

patients

with

culture-proven

M tuberculosis

were

RESULTS Of 2,524 bronchoscopies performed over a six-year period at the University of Cincinnati Medical Center, 91 patients were identified and studied. Fifty patients were eventually found to have M tuberculosis and 41 patients were found to have fungal infections. Of the fungal infections, 25 had histoplasmosis, 11 had cryptococcosis, four had blastomycosis and one had coccidioidomycosis. During that same period of time, 99
50 N u M B E R O F P A T I E N T S

diagnosed. Of the 49 patients with M tuberculosis who did not undergo bronchoscopy, 16 had extrapulmonary tuberculosis without lung disease and 33 were diag nosed by sputum alone. Figure 1 shows the underlying diagnosis in the patients who were diagnosed as having infection. Eighty percent of the patients with M tuberculosis had no underlying disease, while 53 percent of the fungal infections occurred in immunocompromised patients, the major group (19 patients) having AIDS. Figure 2, top summarizes the diagnostic yield for direct examination of specimens obtained from expec torated sputum, bronchoscopy and other techniques. Of the 50 patients eventually proven to be infected with M tuberculosis, positive stains were identified in 16 of 47 (34 percent) sputum smears, 34 of 50 (68 percent) bronchoscopy specimens and eight of ten (80 percent) other specimens. The other tissue examined with transbronchial lung biopsy (three of four positive), open lung biopsy (three positive), bone marrow biopsy (one of two positive) and pleural biopsy (one positive). For patients eventually shown to have pulmonary fungal infections, specimens were positive in none of 22 expectorated sputum, 14 of 41 (66 percent) of bronchoscopy specimens, and 12 of 18 (67 percent) of other tissue specimens. Other tissues studied for fungus included bone marrow (two of seven positive), lymph nodes (two of three positive), open lung (three positive), cerebrospinal fluid (two positive) and one positive each for adrenal aspirate, vocal cord biopsy and transbronchial lung biopsy. The diagnostic yield for culture of specimens is shown in Figure 2 bottom. Of the 50 patients with M tuberculosis, 24 of 47 (51 percent) had positive sputum cultures, 46 of 50 (92 percent) had positive bronchos copy, 12 of 13 (92 percent) patients had other speci mens positive. Of the other specimens, two transbron chial biopsies were culture-positive, four of five urine

40

30

20

10 FIGURE 1. The underlying diagnosis of patients eventually diagnosed as having M tuberculosis (M Tb) or fungal infections. Transplant patients are those who received solid organ transplants. Cancer patients include those with leukemia, lymphoma or undergiong chemotherapy for solid organ malig nancy.

AIDS

CANCER

STEROIDS

TRANSPLANT

NONE

IM.

Tb

I FUNGAL

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians

CHEST / 99 / 1 / JANUARY 1991

93

100

p o s I
T I V

B Y S M E A R

SPUTUM

BRONCH
IM Tb. IFUNGAL

OTHER

120 P O S 100

T I V

E B Y C U
L

U R

SPUTUM

BRONCH iMTb.H! FUNGAL

OTHER

FIGURE Comparison of sputum to bronchoscopy or 2. other tissue sample for diagnosis of M tuberculosis (M Th) or fungal infections by eitber smear (top) or culture (bottom).

specimens were positive and three open lung, two bone marrow biopsies and one pleural biopsy were all positive for M tuberculosis. There were 30 patients whose prebronchoscopy expectorated sputum samples were smear-negative for AFB before bronchoscopy. The bronchoscopy speci men was smear-positive in 15 (50 percent) and culture20
N U M B
R 0 P A

positive in 26 (87 percent). Only nine (30 percent) of the patients had a prebronchoscopy sputum specimen which was subsequently culture-positive. Bronchos copy specimens were superior to sputum culture alone for diagnosing M tuberculosis (chi square =19.8, p<0.0001). However, one patient had only the pre bronchoscopy sputum culture as the positive specimen

T I E N T S

CAVITARY

INFILTRATE

NODULE

MILIARY

EFFUSION

NORMAL

FIGURE The chest roentgenogram pattern of the 3. patients with either A/tuherculosix (M Tb) or fungal pneumonia.
94

. Tb

FUNGAL

Bronchoscopy and BAL in Tuberculosis and Fungal Infections (Baugh,;ian et al)

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians

lymphocytes 3.08 percent; neutrophils 5.8 2 1.50 percent). As a group, the patients with fungal infections also had elevated lymphocytes and neutro SputumSmearM phils (normal: lymphocytes16.4 3.07 percent; neutrophils3.82.64 percent; immunocompro mised: lymphocytes12.5 1.78 percent; neutro tuberculosisCavitaryInfiltrateMiliaryEflusionNodulefungalCavitaryInfiltrateMiliaryNoduleNormal8/168/140/41/50/30/20/70/30/50/3(50)*(36)(0)(20)(0)(0)(0)(0)(0)(0)Cult10/168/141/41/50/30 phils7.0 1.79 percent). DISCUSSION The use of bronchoscopy with BAL allows one to better diagnose and understand the inflammatory nature of many diseases. '-9In patients with infectious granulomatous disease, bronchoscopy with lavage can be useful both as a diagnostic tool as well as a way to examine inflammatory response in the lung. In this report, we found that bronchoscopy supplemented with lavage was useful in diagnosing TB and fungal infections. We also found that most patients had abnormalities in their BAL fluid which have been previously described for noninfectious granulomatous diseases. There is some controversy regarding the use of bronchoscopy to diagnose M tuberculosis. Although rigid bronchoscopy had been reported as useful, it was seldom necessary.13 As TB became less common and its manifestations more nonspecific, flexible fiber optic bronchoscopy has been reported more fre quently for diagnosing TB. This is especially true in miliary TB, where sputum often is negative.14 The overall yield of bronchoscopy for diagnosing TB has been greater than 90 percent when cultures were included in the analysis.15"' In addition, patients who were sputum smear-negative for TB still had a greater than 90 percent yield for M tuberculosis.I7J* Since this study was not prospective in nature, we cannot be sure that all patients with TB were identified in our survey. However, our institution is the sole source of care for most of the patients studied. All 99 patients with documented M tuberculosis were iden tified and for those 50 patients who underwent bron choscopy, the bronchoscopy specimen yielded a diag nosis in more than 90 percent. Although empiric antituberculosis drugs were used in some patients who underwent bronchoscopy, antifungal agents were not employed for prolonged empiric trials. In order to provide a rapid diagnosis, several tech niques have been proposed. These include a radiometric method, which is sensitive but not specific for M tuberculosis.1' Measurements of antigens and anti bodies to M tuberculosis also have been used, but are not specific enough for general use.192" Recently, an assay for tuberculostearic acid has proved very sensi tive and specific.21 An argument has been made that bronchoscopy is done too frequently in patients who are smear-negative but likely to have TB. One expert recommends "most
CHEST / 99 / 1 / JANUARY. 1991 95

Table 1Comparison of Sputum and Bronchoscopy Results to Chest Roentgenogram

*Values given are positive/total tested with percentages in paren theses.

and three patients had only the postbronchoscopy sputum culture specimen-positive. Because adequate volumes of specimens were not always obtained from each bronchoscopy, only 47 (52 percent) patients with M tuberculosis or fungal infec tions had both a BAL and bronchial wash sent for fungal and AFB smear and culture. For the 29 patients with M tuberculosis, seven had positive wash speci mens only while one had a positive BAL only. Three of the patients in this group had negative bronchoscopy. Of the 18 patients with fungal infection, four had negative bronchoscopies and two each had either the wash or BAL specimen positive. The chest roentgenogram pattern for the two groups is shown in Figure 3. Almost all patterns were seen with the infections, although cavitary, infiltrate and nodules were the most common. Table 1 compares the sputum and bronchoscopy results with the more common roentgenogram patterns for both M tuber culosis and fungal infections. For patients with miliary patterns on roentgenogram, sputum smear and culture were low yield, but bronchoscopy specimens were high yield with cultures being positive in 100 percent of cases for both M tuberculosis and fungal infections. There were 83 patients who had adequate speci mens for cell differential counts. These included 44 pa tients with M tuberculosis and 39 patients with fungal infections. We have separated the patients into normal hosts and immunocompromised patients. There was significant overlap of both groups in terms of both the neutrophils and lymphocytes retrieved by lavage. For M tuberculosis, the percentage of lymphocytes (normal: 16.9 11.19 percent [meanSD]; immuno compromised: 22.9 13.25percent) and neutrophils (normal: 9.811.53 percent; immunocompromised: 9.012.68 percent) was significantly higher than that in control subjects who smoked (control subjects:

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians

patients culture

should not have bronchoscopy until sputum results are available."7 At our institution,

patients with smear-negative TB are still unlikely to have a positive culture. We found a significantly higher yield for bronchoscopy, both smear and culture, than for the sputum culture alone. For diagnosing fungal infections, sputum culture appears to be inadequate in most cases. In the current series, we had no sputum smear-positive for fungus, although some patients were eventually culture-posi tive. The yield of bronchoscopy for fungal infections was significantly higher than noted by others,22 al though several groups have reported on the use of bronchoscopy to diagnose fungal infections in smaller series.*'10 The addition of BAL to bronchoscopy alone appears not to have influenced the diagnostic yield for our patients with M tuberculosis. This may be misleading, since the lavage process itself will increase the volume of the bronchial wash and often will induce cough productive of sputum during the bronchoscopy itself. Our results of the diagnostic yield of bronchoscopy (90 percent) are similar to other series which supplement their bronchoscopies with transbronchial biopsy.16 The trausbronchial biopsy was the only positive specimen in only one patient in this series. The use of transbronchial biopsy increases the risk of bronchoscopy, and although it has a high diagnostic yield, it usually is not necessary even in patients with miliary tuber culosis. For patients with fungal infections, lavage also does not increase the yield of the procedure. However, for both M tuberculosis and fungal infections, our studies would suggest that both bronchial wash and lavage specimens, when available, should be sent for smear and culture. Bronchoalveolar lavage has been used to character ize inflammatory response in many diseases. In our study, patients with M tuberculosis and fungal infec tions had an increased percentage of lymphocytes found in the BAL fluid. This is not surprising, since these infections usually elicit a granulomatous re sponse. Others studying tuberculosis and fungal infec tions have reported increased lymphocytes in the BAL.4-23-2'1 This increased percentage also was seen in the imiminocompromised patients, including AIDS patients. Not all patients had increased lymphocytes, and there was a spectrum of response. The increased percentage of neutrophils usually seen in these patients represents a diflerent inflam matory response. In an animal model of acute histoplasmosis, we described an initial influx of neutrophils in the BAL, followed by lymphocytes which persisted for a much longer period of time.25 Our analysis of the clinical data does not allow us to determine whether patients with increased neutrophils had more acute
96

infections, while those with lymphocytes had more chronic disease. We conclude that bronchoscopy with BAL can be useful in diagnosing M tuberculosis and fungal infec tions. Patients with these diseases tend to have an inflammatory response characterized by increased lymphocytes and neutrophils.

REFERENCES
1 Reynolds HY. Bronchoalveolar 135:250-63 lavage. Am Rev Respir Dis 1987;

2 Hunninghake GW, Crystal RG. Pulmonary sarcoidosis: a disor der mediated by excess helper T-lymphocyte activity at sites of disease activity. N Engl J Med 1981; 305:429-34 3 Leatherman JW, Michael AW, Schwartz BA, Hoidal JR. Lung T cells in hypersensitivity pneumonitis. Ann Intern Med 1984; 100:390-92 4 Dhand R, De A, Gupta N, Jaswal S, Kohli KK, Malik SK, et al. Cellular composition of bronchoalveolar lavage fluid in patients with tuberculosis. Indian J Med Res 1988; 87:555-60 5 Norman E. Keistinin T, Uddenfeldt M, Rydstrom R, Lundgren R. Bronchoalveolar lavage is better than gastric lavage in the diagnosis of pulmonary tuberculosis. Scand J Infect Dis 1988; 20:77-80 6 Russell MD, Torrington KG, Tenholder MF. A ten-year experi ence fiberoptic bronchoscopy for mycobacterial isolation: impact of the Bactec system. Am Rev Respir Dis 1986; 133:1069-71 7 NeffTA. Bronchoscopy and Bactec for the diagnosis of tuber culosis: state of the art, or a brief dissertation on the efficient search for the tubercle bacillus? Am Rev Respir Dis 1986; 133: 962 8 George RB, Jenkinson SG, Light RG. Fiberoptic bronchoscopy in the diagnosis of pulmonary fungal and nocardial infections. Chest 1978; 73:33-36 9 Stover DE, Zaman MB, Hajdu SI, Lange M, Gold J, Armstrong D. Bronchoalveolar lavage in the diagnosis of diffuse pulmonary infiltrates in the immunosuppressed host. Ann Intern Med 1984; 101:1-7 10 Gal AA, Koss MN, Hawkins J, Evans S, Einstein H. The pathology of pulmonary cryptococcal infections in the acquired immunodeficiency syndromes. Arch Pathol Lab Med 1986; 110: 502-07 11 Dohn MN, Baughman RP. Effect of changing instilled volume for bronchoalveolar lavage in patients with interstitial lung disease. Am Rev Respir Dis 1985; 131:390-92 12 Baughman RP, Strohofer S, Kim CK. Variation of differential cell counts of bronchoalveolar lavage fluid. Arch Pathol Lab Med 1986; 110:341-43 13 Shipman SJ. Diagnostic bronchoscopy Am Rev Tuberculosis 1939; 39:629-32 in occult tuberculosis.

14 Burk JR, Viroslav J, Bynum LJ. Miliary tuberculosis diagnosed by fiberoptic bronchoscopy and transbronchial biopsy. Tubercle 1978:59:107-09 15 Danek SJ, Bower JS. Diagnosis of pulmonary tuberculosis by flexible fiberoptic bronchoscopy. Am Rev Respir Dis 1979; 119: 677-79 16 Wallace JM, Deutsch AL, HarrellJH, MoserKM. Bronchoscopy and transbronchial biopsy in evaluation of patients with sus pected active tuberculosis. Am J Med 1981; 70:1189-94 17 Willcox PA, Benatar SR, Potgieter PD. Use of the flexible fiberoptic bronchoscope in diagnosis of sputum-negative pul monary tuberculosis. Thorax 1982; 37:598-601 18 Chawla R, Pant K, Jaggi OP, Chandra-Shekhar S, Thukral SS. Fiberoptic bronchoscopy in smear-negative pulmonary tuberculosis. Eur Respir J 1988; 1:804-06

Bronchoscopy and BAL in Tuberculosis and Fungal Infections (Baughman et al)

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians

19 Raja A, Baughman

RP, Daniel TM. The detection

by immnno-

assay of antibody to mycobacterial antigens and mycobacterial antigens in bronchoalveolar lavage Huid from patients with tuberculosis and control subjects. Chest 1988; 94:133-37 20 Levy H, Wadee RA, Feldman C, Rabson AR. En7,yme-linked imniunosorbent assay for the detection Mycobacterium tuberculosis in bronchial Chest 1988; 93:762-66 of antibodies against washings and serum.

22 Prechter GC, Prakash UBS. Bronchoscopy in the diagnosis of pulmonary histoplasmosis. Chest 1989; 95:1033-36 23 Leatherman JVV,Michael AF, Simpson M, Hoidal J. Characteri/.ation of cell-mediated immunity in the lung in histoplasmosis by monoclonal antilx>dies. Am Rev Respir Dis 1983; 127:S195 24 Jacobs RF, Marnier DJ, Balk RA, Bradsher HVV.Lymphocyte stibpopnlations of bhxxl and alveolar lavage in blastomycosis. Chest 1985; 88:579-85 25 Baughman RP, Kim CK, Vinegar A, llendricks DE, Schmidt DJ, Bullock WE. The pathogenesis of experimental pulmonary histoplasmosis: correlative studies of histopathology, bronchoal veolar lavage, and respiratory function. Am Rev Respir Dis 1986; 134:771-76

21 Pang JA, Chan US, Chan CY, Cheung SW, French GL. A tuberculostearic acid assay in the diagnosis of sputum smearnegative pulmonary tuberculosis: a prospective stud) ot bronchoscopic aspirate and lavage specimens. 111:650-54 Ann Intern Med 1989;

30th Annual Durrance-Waring Mid-Winter Chest Conference

The American Lung Association of Colorado will present this course February 27-March 2 in Aspen. Topics include acute respiratory failure, adult cystic fibrosis and lung transplantation. For information, contact ALAC, 1600 Race Street, Denver 80206 (303:388-4327).

CHEST / 99 / 1 / JANUARY 1991

97

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians

Bronchoscopy with bronchoalveolar lavage in tuberculosis and fungal infections. R P Baughman, M N Dohn, R G Loudon and P T Frame Chest 1991;99; 92-97 DOI 10.1378/chest.99.1.92 This information is current as of July 27, 2011
Updated Information & Services Updated Information and services can be found at: http://chestjournal.chestpubs.org/content/99/1/92 Cited Bys This article has been cited by 6 HighWire-hosted articles: http://chestjournal.chestpubs.org/content/99/1/92#related-urls Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.chestpubs.org/site/misc/reprints.xhtml Reprints Information about ordering reprints can be found online: http://www.chestpubs.org/site/misc/reprints.xhtml Citation Alerts Receive free e-mail alerts when new articles cite this article. To sign up, select the "Services" link to the right of the online article. Images in PowerPoint format Figures that appear in CHEST articles can be downloaded for teaching purposes in PowerPoint slide format. See any online figure for directions.

Downloaded from chestjournal.chestpubs.org by guest on July 27, 2011 1991 American College of Chest Physicians