(Chest. 2001;119:385S-390S.)
© 2001
American College of Chest Physicians
Are Quantitative Cultures Useful in the Diagnosis of Hospital-Acquired Pneumonia?*
Gerry San Pedro, MD
*
From the Department of Internal Medicine, Section of Pulmonary and Critical Care Medicine, Louisiana State University Health Sciences Center at Shreveport, Shreveport, LA.
Correspondence to: Gerry San Pedro, MD, Section of Pulmonary and Critical Care Medicine, Louisiana State University Health Sciences Center at Shreveport, 1501 Kings Highway, PO Box 33932, Shreveport, LA 71130-3932; e-mail: gerrsp{at}prysm.net
 |
Abstract
|
|---|
Noninvasive and invasive tests have been developed and studied for
their utility in diagnosing and guiding the treatment of
hospital-acquired pneumonia, a condition with an inherently high
mortality. Early empiric antibiotic treatment has been shown to reduce
mortality, so delaying this treatment until test results are available
is not justifiable. Furthermore, tailoring therapy based on results of
either noninvasive or invasive tests has not been clearly shown to
affect morbidity and mortality. This may be related to quantitative
limitations of these tests or possibly to a high false-negative rate in
patients who receive early antibiotic treatment and may therefore have
suppressed bacterial counts. Results of these tests, however, do
influence treatment. It is therefore hoped that they may ultimately
provide a rational basis for making therapeutic decisions. In the
future, outcomes research should be a part of large-scale clinical
trials, and noninvasive and invasive tests should be incorporated into
the design in an attempt to provide a better understanding of the value
of such tests.
Key Words: bacteriology • BAL • bronchoscopy • diagnosis • endotracheal aspiration • hospital-acquired pneumonia
|
Introduction
|
|---|
This
article examines the utility of quantitative and qualitative
(nonquantitative) cultures and of invasive and noninvasive techniques
in the diagnosis of hospital-acquired pneumonia (HAP). I shall seek to
address whether quantitative cultures are truly necessary and how
important it is to identify a specific agent as a cause of the
pneumonia. Ventilator-associated pneumonia as well as hospital-acquired
infections, in general, are considered.
How good are we at making a diagnosis of HAP? Our clinical skills at
making this diagnosis seem to be suboptimal at times. Consider the
patient who is in the hospital and receiving mechanical ventilation and
develops fever, leukocytosis, pulmonary infiltration on chest
radiography, and purulent tracheobronchial secretions. In this clinical
setting, we can certainly entertain a diagnosis of HAP. Andrews and
colleagues1
discussed this issue in patients with
preexisting abnormal chest radiographic findings and symptomatology
compatible with pneumonia. Their study1
utilized necropsy
material to validate the presence or absence of pneumonia. It was
demonstrated, however, that 29% of these cases were misdiagnosed
clinically.1
Other conditions, such as "pulmonary
fibroproliferation," sinusitis, cholecystitis, and various
noninfectious conditions, can also produce findings that might suggest
the presence of pneumonia.2
Similarly, the chest
radiograph may be misleading at times.3
In the report by
Andrews and colleagues,1
36% of patients had pneumonia at
necropsy, even though their chest radiographs may not have changed.
Work from our own group4
5
has also clearly shown that
there are many other potentially lethal conditions that can create a
picture resembling pneumonia. Access to pulmonary tissue for making a
histologic diagnosis would clearly be a desirable objective. The
histopathologic findings that characterize nosocomial pneumonia have
been described in detail.6
|
Diagnosis of HAP
|
|---|
Clinical findings and chest radiography alone may not always
suffice for making a definitive diagnosis of pneumonia. However, we can
attempt to make a diagnosis by examining respiratory secretions. The
number of organisms present in an infection, which may in turn be
dependent on the virulence of the organism and the host’s response, is
of prime importance. How many organisms must be detected to assume the
presence of infection vs mere colonization? Johanson and
colleagues7
have suggested that about
103 cfu/g of tissue indicate the presence of
infection, although this could depend on the type of bacteria present.
Tissue is often not available, however, and counts may have to be
obtained in secretions. In this case, about
105 bacteria per milliliter of exudate may
have similar import.8
Sources of sampling for making a
diagnosis include blood and other body fluids as well as several
techniques available for obtaining samples. These techniques include
needle aspiration, endotracheal aspiration, and both blind and invasive
procedures to obtain bronchial secretions (Table 1
).
|
Percutaneous Needle Aspiration
|
|---|
How useful in general is percutaneous needle aspiration? As noted
above, much depends on the bacterial counts in the area being
aspirated.9
The sensitivity of the technique for patients
who are not receiving mechanical ventilation is about
60%,10
with the figure somewhat lower (approximately
40%) for patients receiving mechanical ventilation.11
Thus, the procedure may prove to be less useful in the population of
patients in which it is most needed. From the standpoint of predictive
value, a positive test result is helpful, but a negative test result
may not be helpful since about a third of patients with a negative test
result may actually have the disease.10
The main objection
to percutaneous needle aspiration is the high incidence of
complications, including hemorrhage, hemoptysis, and pneumothorax. This
is true even though the procedure itself may be fairly easy to carry
out. The incidence of pneumothorax is about 20%, but may be up to 60%
in patients receiving mechanical ventilation.12
About half
of these patients will require a chest tube.
|
Endotracheal Aspiration
|
|---|
Results of endotracheal aspiration may be compared with those of
other invasive and noninvasive techniques (Table 2
), including open-lung biopsy, postmortem lung cultures, blood and
pleural fluid analysis, and protected specimen brushing (PSB) and
BAL.13
Sensitivity and specificity of routine endotracheal
aspiration have a broad scatter in relation to these other
measurements, with sensitivity ranging from 57 to 88%, but with
specificity in a relatively low range, from 0 to 33%. Thus,
nonquantitative endotracheal cultures may have limited utility. One of
the chief values of endotracheal cultures is that they exclude certain
types of infection when the organism is absent. For example, absence of
Pseudomonas in an endotracheal aspirate makes it unlikely that this
organism is the cause of an infection. Attempts have also been made to
perform quantitative endotracheal aspiration (QEA). Jourdain and
colleagues14
have attempted to establish a cutoff for
assigning significance to colony counts in endotracheal aspirates by
examining sensitivity, specificity, and overall accuracy (Table 3
). They concluded that a level of 105 to
106 bacteria per milliliter of exudate was
optimal for diagnostic purposes. Although counts higher than this
figure improved specificity, there was an unacceptable decline in
sensitivity.
|
Noninvasive Techniques
|
|---|
Can noninvasive techniques that can be conducted at the bedside be
used to obtain bronchial samples? We can employ blind bronchial
sampling,15
blind BAL fluid sampling,16
or
blind performance of PSB.17
We know somewhat more about
the sensitivity and specificity of these techniques. For blind
bronchial sampling, the sensitivity is about 80%; for blind BAL, the
sensitivity is about 73% with a specificity of 96%. The figures are
also good for blind PSB, at 66% and 91%, respectively, and with a
high diagnostic yield (66%) and good concordance with
bronchofibroscopic PSB (85%).17
Then why not use these
techniques routinely? Unfortunately, they are relatively new, and
experience with their use is limited. These procedures, however, may
have considerable utility in the future since they are easy to perform;
in some centers, the specimens are even obtained by respiratory
technicians.
|
PSB
|
|---|
PSB is a standard invasive method for diagnosing HAP. The brush is
small and obtains a sample in the range of 0.001 mL. After the sample
is obtained, it is diluted in about 1 mL of transport medium, an
approximately 1,000-fold dilution.18
Consequently, the
standards for colony counts must be appropriately adjusted, and counts
of 103 cfu/mL of transport medium would be
considered diagnostic. Overall, this technique has about 82%
sensitivity and 89% specificity.19
However, the
repeatability of this procedure in the same patient is not perfect, and
quantitative discordance may be considerable.20
21
Another
problem with this technique and others already discussed is that colony
counts fall in patients previously treated with
antibiotics.22
23
The incidence of such treatment is high,
and most patients entering the intensive-care environment are already
being treated with one or more antibiotics. However, PSB could play a
role in identifying patients with recurrent or resistant infection. In
a study by Montravers and colleagues,24
76 patients with
positive specimen findings who were treated with antibiotics were
restudied 3 days later. Of 173 organisms originally identified, 11
organisms (6%) were shown to be still present, 3 of which were in
significant numbers. However, 32 new organisms were also identified, 9
of which were in significant numbers. Resistance was observed in 26 of
these 32 organisms (81%). Thus, a total of 12 old or new
organisms were noted to be present in significant numbers, and
resistance did appear to represent a problem. Of the 9 patients with
significant numbers of new organisms, only 44% subsequently responded
to treatment; however, in the other 67 patients, the response was 93%,
a highly significant difference. Thus, perhaps there may be a role for
repeating PSB in the setting of poorly responding patients.
However, further studies need to document that repeating PSB in these
patients results in changes in management that lead to improved
outcomes.
|
BAL
|
|---|
How effective is BAL as a diagnostic tool? This technique samples
a large number (approximately 106) of alveoli and
can recover a larger number of organisms. The threshold is in the range
of 104 organisms per milliliter of lavage fluid,
which is equivalent to 106 organisms per
milliliter of alveolar fluid from the periphery.25
Sensitivity and specificity are high, in the range of 72 to 100% for
sensitivity and 69 to 100% for specificity.13
Qualitatively, the same organisms are recovered repeatedly in the range
of 95%, but as in the case with PSB, there is considerable
quantitative discordance.26
Sensitivity and specificity of
BAL have also been compared with those of PSB (Table 4
), and the results are comparable. The question also arises about the
comparability of results of PSB vs BAL in patients already receiving
antibiotics. As can be seen in Table 5
,13
the range was broad, but the two
studies27
28
in patients receiving antibiotics showed
values similar to the two studies29
30
in patients not
receiving antibiotics. In general, however, the yield of organisms is
believed to be lower in patients receiving antibiotics than in those
not receiving antibiotics.
|
Outcomes Research
|
|---|
Treatment outcome is an important consideration in determining
whether invasive or noninvasive methods are best for obtaining
specimens for bacterial culture. Does making a specific etiologic
diagnosis make a difference in outcome? Do results of culture
techniques influence antibiotic choices and, ultimately, outcome? Does
the use of a quantitative technique result in improved outcomes
compared with nonquantitative techniques?31
Many efforts have been made to compare the bacteriology of quantitative
vs nonquantitative techniques; some research is beginning to address
patient outcomes as well. Luna and colleagues32
from
Argentina investigated outcomes in 132 patients with
ventilator-associated pneumonia who underwent BAL within 24 h of
clinical diagnosis. Of this number, 107 patients (81%) were previously
receiving antibiotics. Using the criterion of 104
cfu/mL as a cutoff, 65 of the 132 patients (49%) had positive
findings; of this number, 46 patients (71%) died. However, mortality
did not differ significantly in patients who either had or had not
previously received antibiotics, in those with positive vs negative
findings by BAL, or in those whose antibiotic regimen was changed after
the lavage procedure. A significantly lower mortality was found in
patients who were deemed to have received adequate vs inadequate
therapy before their procedure (38% vs 91%). In patients receiving no
antibiotic therapy before the procedure, mortality was 60%. It is also
noteworthy that half of the mortality in this study in patients with
positive BAL findings was experienced within 48 h of the
procedure, before culture results were available, so that this
information could not always be used to direct therapy. In this study,
early initiation of adequate therapy for ventilator-associated
pneumonia reduced mortality, although overall mortality was high and
not influenced by culture results or by switching antibiotics based on
these results. Also, if therapy is delayed until bronchoscopy is
performed or while awaiting BAL results, mortality is
higher.32
In a second study from Spain, Sanchez-Nieto and
colleagues33
studied 51 patients who were randomly
classified into two groups. One group received extensive investigation,
including QEA, BAL, and PSB; the other group received only QEA. Of the
24 patients in the first group, 16 patients (67%) had positive
findings by both QEA and BAL and 14 patients by PSB (58%), although
there were substantial within-patient inconsistencies. In this group,
10 of the patients (42%) had their regimens modified as the result of
testing, and overall mortality was 11 of 24 patients (46%). In the
second group of 27 patients, QEA findings were positive in 20 patients
(74%), 4 of 27 patients had their regimens modified, and the mortality
was 7 of 27 patients (26%). Although more patients in the extensively
tested group had their regimens modified after testing, the difference
in mortality between the two groups was not significant. A potentially
confounding factor in this study, however, was that a much higher
percentage of the first group had culture findings that were
positive for Pseudomonas, an organism known to be associated with a
high mortality. Despite the more frequent change in antibiotic regimen
in the former group, the authors concluded that there was no
substantial difference in mortality between patients undergoing
intensive, compared with less intensive, investigation, although the
numbers in the study were too small to warrant drawing strong
conclusions.
Despite the fact that neither of these two studies provides strong
evidence favoring an aggressive diagnostic approach in HAP, they are
indicative of the type of investigations needed to improve our
understanding and treatment of this condition. The database regarding
sensitivity and specificity of invasive and noninvasive tests is now
well characterized. We should begin to direct our energies toward
determining the outcomes of applying these therapies.
|
Appendix 1
|
|---|
Dr. George Eliopoulos:
You have convinced me of the difficulty
of using this kind of information to make clinical decisions. However,
let me ask this question: If you were to design a study to show
superiority of one regimen over another for the treatment of pneumonia,
what would you demand of the investigators for making as ironclad as
possible a definition of nosocomial pneumonia or ventilator-associated
pneumonia?
Dr. Gerry San Pedro:
You can define nosocomial pneumonia in
various ways, but what most investigators are seeking is to know that
they are studying comparable groups of patients in the different arms
of their studies.
Dr. Eliopoulos:
Not entirely true. For example, if you have
1,000 patients in both arms, and 999 of them have viral pneumonia, the
groups are comparable. But what I want to know is how many patients
have Pseudomonas infection and how many have pneumococcal infection,
etc. These definitions, from my point of view, are crucial.
Dr. David Bowton:
The key question is how to deal with the
disquieting information concerning histology, BAL, PSB, correlates, and
reclassification. There is no correlation between any of our diagnostic
studies whether you use pathology or quantitative culture, and if you
use different criteria for histologic analysis, you reclassify 30% of
the patients. Coming up with a clinically relevant diagnostic schema at
the moment seems almost impossible, and we must accept that we often
simply cannot make the diagnosis. One approach is to simply pick
criteria that are reproducible, obtain whatever microbiological
information is available, accept that the method might be flawed, and
then consistently apply it to all patients. There must also be a large
enough patient population to achieve adequate power for subgroup
analysis to detect outcome differences. This will, to some extent,
avoid the question of whether the microbiology is sufficiently
accurate. The microbiological classification can initially be
considered a matter that is primarily of academic interest but may, in
addition, provide some scientific underpinning if patients have
comparable microbiology. The key feature is a large sample size.
Patients can also be matched in other ways, such as having comparable
APACHE (acute physiology and chronic health evaluation) scores.
However, the mortality level in a large group is the key.
Dr. Eliopoulos:
What criteria would you use?
Dr. San Pedro:
I would employ a mix. There would have to be
both clinical criteria and invasive techniques. Personally, I would use
BAL.
Dr. David Weber:
One must be careful about taking
risk profiles that have been developed in one group of patients
and assuming they will work in another group. We evaluated the risk
factors for ventilator-associated pneumonia in elderly patients
admitted to either a medical ICU or surgical ICU. APACHE II scores,
which were developed and validated in surgical patients, were
predictive for the development of pneumonia in elderly persons in our
surgical ICU but were not predictive for elderly persons in our medical
ICU. Another point regarding resistance, which will have a major impact
on the development of resistance, is knowing which antibiotic a patient
might have been taking before pneumonia developed.
Dr. Louis Rice:
I have a question about the adjustments that
were made in the antibiotic regimen. Was the therapy narrowed to target
specific organisms or was it expanded to deal with resistant organisms?
Dr. San Pedro:
That was not dealt with in the article by
Sanchez-Nieto et al.33
Dr. Bowton:
In a study40
41
that was
similar to the study by Luna et al,32
when antibiotic
therapy was discontinued based on BAL, there was a low mortality, only
14%.
Dr. San Pedro:
The goal there was to narrow down antibiotic
coverage, but in most hospitals, I suspect the practice would be to
expand coverage by adding an antibiotic.
Dr. Rice:
None of these would be techniques that would
prevent physicians from narrowing down therapy.
Dr. Steve Nelson:
In a patient doing poorly in the ICU
receiving broad-spectrum therapy, I will often send an endotracheal
aspirate, and if it returns in 2 to 3 days without showing a suspected
pathogen, I will often trim back my therapy. This has not been studied
very carefully, however.
Dr. Bowton:
That should be a part of the study design. A study
should have a standardized initial empiric approach and then perhaps be
modified subsequently based on findings from testing such as QEA.
Although some have argued that this may not lead to a difference in
outcome, if you employ an empiric approach and narrow it back, there
may be new information.
|
Footnotes
|
|---|
Abbreviations: APACHE = acute
physiology and chronic health evaluation; HAP = hospital-acquired
pneumonia; PSB = protected specimen brushing; QEA = quantitative
endotracheal aspiration
|
References
|
|---|
-
Andrews, CP, Coalson, JJ, Smith, JD, et al (1981) Diagnosis of nosocomial bacterial pneumonia in acute, diffuse lung injury. Chest 80,254-258[Abstract/Free Full Text]
-
Meduri, GU, Mauldin, GL, Wunderink, RG, et al (1994) Causes of fever and pulmonary densities in patients with clinical manifestations of ventilator-associated pneumonia. Chest 106,221-235[Abstract/Free Full Text]
-
Wunderink, RG, Woldenberg, LS, Zeiss, J, et al (1992) The radiologic diagnosis of autopsy-proven ventilator-associated pneumonia. Chest 101,458-463[Abstract/Free Full Text]
-
San Pedro, GS, McCracken, G, Campbell, GD (1989) 61% of deaths in pneumonia are potentially preventable [abstract]. Am Rev Respir Dis 139,A428
-
San Pedro, GS, Briery, C, Whitson, B, et al (1998) The clinical diagnosis of pneumonia is a poor predictor for an attributable cause of death [abstract]. Am J Respir Crit Care Med 157,A292
-
Coalson, JJ (1995) The pathology of nosocomial pneumonia. Clin Chest Med 16,13-28[ISI][Medline]
-
Johanson, WG, Jr, Seidenfeld, JJ, Gomez, P, et al (1988) Bacteriologic diagnosis of nosocomial pneumonia following prolonged mechanical ventilation. Am Rev Respir Dis 137,259-264[ISI][Medline]
-
Bartlett, JG, Finegold, SM (1978) Bacteriology of expectorated sputum with quantitative culture and wash technique compared to transtracheal aspirates. Am Rev Respir Dis 117,1019-1027[ISI][Medline]
-
San Pedro, GS, Campbell, GD, Tryka, AF (1989) Factors affecting the sensitivity of needle aspiration in a rodent model of Gram-negative bacillary pneumonia. Chest 95,198S-199S[Free Full Text]
-
Dorca, J, Manresa, F, Esteban, L, et al (1995) Efficacy, safety, and therapeutic relevance of transthoracic aspiration with ultrathin needle in nonventilated nosocomial pneumonia. Am J Respir Crit Care Med 151,1491-1496[Abstract]
-
Torres, A, Jiménez, P, Puig de la Bellacasa, J, et al (1990) Diagnostic value of nonfluoroscopic percutaneous lung needle aspiration in patients with pneumonia. Chest 98,840-844[Abstract/Free Full Text]
-
Mayhall, CG (1997) Nosocomial pneumonia: diagnosis and prevention. Infect Dis Clin North Am 11,427-457[CrossRef][ISI][Medline]
-
Meduri, GU (1995) Diagnosis and differential diagnosis of ventilator-associated pneumonia. Clin Chest Med 16,61-93[ISI][Medline]
-
Jourdain, B, Novara, A, Joly-Guillou, M-L, et al (1995) Role of quantitative cultures of endotracheal aspirates in the diagnosis of nosocomial pneumonia. Am J Respir Crit Care Med 152,241-246[Abstract]
-
Papazian, L, Thomas, P, Garbe, L, et al (1995) Bronchoscopic or blind sampling techniques for the diagnosis of ventilator-associated pneumonia. Am J Respir Crit Care Med 152,1982-1991[Abstract]
-
Pugin, J, Auckenthaler, R, Mili, N, et al (1991) Diagnosis of ventilator-associated pneumonia by bacteriologic analysis of bronchoscopic and nonbronchoscopic "blind" bronchoalveolar lavage fluid. Am Rev Respir Dis 143,1121-1129[ISI][Medline]
-
Bello, S, Tejada, A, Chacón, E, et al (1996) "Blind" protected specimen brushing versus bronchoscopic techniques in the aetiological diagnosis of ventilator-associated pneumonia. Eur Respir J 9,1494-1499[Abstract]
-
Marquette, C-H, Ramon, P, Courcol, R, et al (1988) Bronchoscopic protected catheter brush for the diagnosis of pulmonary infections. Chest 93,746-750[Abstract]
-
Chastre, J, Fagon, J-Y, Bornet-Lesco, M, et al (1995) Evaluation of bronchoscopic techniques for the diagnosis of nosocomial pneumonia. Am J Respir Crit Care Med 152,231-240[Abstract]
-
Marquette, CH, Herengt, F, Mathieu, D, et al (1993) Diagnosis of pneumonia in mechanically ventilated patients: repeatability of the protected specimen brush. Am Rev Respir Dis 147,211-214[ISI][Medline]
-
Timsit, J-F, Misset, B, Francoual, S, et al (1993) Is protected specimen brush a reproducible method to diagnose ICU-acquired pneumonia? Chest 104,104-108[Abstract/Free Full Text]
-
Marrie, TJ (1994) Community-acquired pneumonia. Clin Infect Dis 18,501-515[ISI][Medline]
-
. et alÖrtqvist, Å, Kalin, M, Lejdeborn, L (1990) Diagnostic fiberoptic bronchoscopy and protected brush culture in patients with community-acquired pneumonia. Chest 97,576-582[Abstract/Free Full Text]
-
Montravers, P, Fagon, J-Y, Chastre, J, et al (1993) Follow-up protected specimen brushes to assess treatment in nosocomial pneumonia. Am Rev Respir Dis 147,38-44[ISI][Medline]
-
Marquette, CH, Copin, M-C, Wallet, F, et al (1995) Diagnostic tests for pneumonia in ventilated patients: prospective evaluation of diagnostic accuracy using histology as a diagnostic gold standard. Am J Respir Crit Care Med 151,1878-1888[Abstract]
-
Gerbeaux, P, Ledoray, V, Boussuges, A, et al (1998) Diagnosis of nosocomial pneumonia in mechanically ventilated patients: repeatability of the bronchoalveolar lavage. Am J Respir Crit Care Med 157,76-80[Abstract/Free Full Text]
-
Torres, A, Puig de la Bellacasa, J, Xaubet, A, et al (1989) Diagnostic value of quantitative cultures of bronchoalveolar lavage and telescoping plugged catheters in mechanically ventilated patients with bacterial pneumonia. Am Rev Respir Dis 140,306-310[ISI][Medline]
-
Violán, JS, de Castro, FR, Luna, JC, et al (1993) Comparative efficacy of bronchoalveolar lavage and telescoping plugged catheter in the diagnosis of pneumonia in mechanically ventilated patients. Chest 103,386-390[Abstract/Free Full Text]
-
Chastre, J, Fagon, J-Y, Soler, P, et al (1988) Diagnosis of nosocomial bacterial pneumonia in intubated patients undergoing ventilation: comparison of the usefulness of bronchoalveolar lavage and the protected specimen brush [published erratum appears in Am J Med 1989; 86:258] Am J Med 85,499-506[ISI][Medline]
-
Jiménez, P, Saldías, F, Meneses, M, et al (1993) Diagnostic fiberoptic bronchoscopy in patients with community-acquired pneumonia. Chest 103,1023-1027[Abstract/Free Full Text]
-
Wunderink, RG (1998) Mortality and the diagnosis of ventilator-associated pneumonia: a new direction. Am J Respir Crit Care Med 157,349-350[Free Full Text]
-
Luna, CM, Vujacich, P, Niederman, MS, et al (1997) Impact of BAL data on the therapy and outcome of ventilator-associated pneumonia. Chest 111,676-685[Abstract/Free Full Text]
-
Sanchez-Nieto, J-M, Torres, A, Garcia-Cordoba, F, et al (1998) Impact of invasive and noninvasive quantitative culture sampling on outcome of ventilator-associated pneumonia: a pilot study. Am J Respir Crit Care Med 157,371-376[Abstract/Free Full Text]
-
Hill, JD, Ratliff, JL, Parrott, JCW, et al (1976) Pulmonary pathology in acute respiratory insufficiency: lung biopsy as a diagnostic tool. J Thorac Cardiovasc Surg 71,64-71[Abstract]
-
Berger, R, Arango, L (1985) Etiologic diagnosis of bacterial nosocomial pneumonia in seriously ill patients. Crit Care Med 13,833-836[ISI][Medline]
-
Villers, D, Derriennic, M, Raffi, F, et al (1985) Reliability of the bronchoscopic protected catheter brush in intubated and ventilated patients. Chest 88,527-530[Abstract/Free Full Text]
-
Lambert, RS, Vereen, LE, George, RB (1989) Comparison of tracheal aspirates and protected brush catheter specimens for identifying pathogenic bacteria in mechanically ventilated patients. Am J Med Sci 297,377-382[ISI][Medline]
-
Chastre, J, Viau, F, Brun, P, et al (1984) Prospective evaluation of the protected specimen brush for the diagnosis of pulmonary infections in ventilated patients. Am Rev Respir Dis 130,924-929[ISI][Medline]
-
Vallés, J, Rello, J, Fernández, R, et al (1994) Role of bronchoalveolar lavage in mechanically ventilated patients with suspected pneumonia. Eur J Clin Microbiol Infect Dis 13,549-558[CrossRef][ISI][Medline]
-
Guerra, LF, Baughman, RP (1990) Use of bronchioalveolar lavage to diagnose bacterial pneumonia in mechanically ventilated patients. Crit Care Med 18,169-173[ISI][Medline]
-
Fagon, JY, Chastre, J, Hance, AJ, et al (1993) Nosocomial pneumonia in ventilated patients: a cohort study evaluating attributable mortality and hospital stay. Am J Med 94,281-299[CrossRef][ISI][Medline]
TOP
Abstract
Introduction
