| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Medicine (Dr. Heyland), Queen's University, Kingston, Ontario; Department of Medicine (Drs. Cook and Jaeschke), McMaster University, Hamilton, Ontario; Department of Surgery (Dr. Marshall), University of Toronto, Toronto, Ontario; Department of Medicine (Dr. Heule), University of Alberta, Edmonton, Alberta; Departments of Anesthesia and Pulmonary Medicine (Dr. Guslits), Henry Ford Hospital, Detroit, MI; and the Department of Medicine (Dr. Lang), Joseph Brant Memorial Hospital, Burlington, Ontario.
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Design: Prospective cohort study.
Setting: Ten tertiary care ICUs in Canada.
Patients: Ninety-two mechanically ventilated patients with a clinical suspicion of VAP who underwent bronchoscopy were compared with 49 patients with a clinical suspicion of pneumonia who did not.
Interventions: None.
Measurements and results: We compared antibiotic use, duration of mechanical ventilation, ICU stay, and mortality. In addition, for patients who received bronchoscopy, we administered a questionnaire (before and after bronchoscopy) to evaluate the effect of PBC or BAL on (1) physician perception of the probability of VAP, (2) physician confidence in the diagnosis of VAP, and (3) changes to antibiotic management. After bronchoscopy results became available, from the physician's perspective, the diagnosis of VAP was deemed much less likely (p < 0.001), confidence in the diagnosis increased (p = 0.03), and level of comfort with the management plan increased (p = 0.02). Following the results of invasive diagnostic tests, in the group that underwent bronchoscopy, patients were receiving fewer antibiotics (31/92 vs 9/49, p = 0.05) and more patients had treatment with all their antibiotics discontinued (18/92 vs 3/49, p = 0.04) compared with the group that did not undergo bronchoscopy. Duration of mechanical ventilation and ICU stay were similar between the two groups, but mortality was lower in the group that underwent bronchoscopy with PBC or BAL (18.5% vs 34.7%, p = 0.03).
Conclusions: Invasive diagnostic testing may increase physician confidence in the diagnosis and management of VAP and allows for greater ability to limit or discontinue antibiotic treatment. Whether performing PBC or BAL affects clinically important outcomes requires further study.
Key Words: bronchoalveolar lavage • bronchoscopy • cohort study • diagnosis • outcomes assessment • protected brush catheter • technology evaluation • ventilator-associated pneumonia
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Bronchoscopy with use of protected brush catheters (PBCs) or BAL may be a more accurate method of diagnosing VAP. Initial reports suggested sensitivity and specificity were high (pooled estimates, 89.9% and 94.5%, respectively, for PBC and 53.3 to 100% and 98.6% for BAL),6 while recent studies question the accuracy of invasive techniques in patients receiving antibiotics prior to bronchoscopy.7 ,8 ,9 While the utilization of invasive techniques has both advocates and antagonists, prior to dissemination and integration of PBC and BAL into day-to-day practice, we believe that this diagnostic technology requires more rigorous assessment.
Guyatt et al10 have established a framework for evaluating diagnostic technologies. This framework describes different levels of assessment from how the diagnostic tool performs to what impact it has on health-care workers, patient management, and patient outcomes (Table 1 ). Although numerous studies have assessed the diagnostic accuracy of PBC/BAL, there are few controlled studies that have assessed the effect of PBC/BAL on health-care providers, on therapeutic decision-making, or on patient outcomes. In this study, we asked the following questions. (1) Does PBC/BAL change physician estimates of the likelihood of VAP? (2) Does PBC/BAL change physician confidence in the diagnosis of VAP? (3) Are physicians more confident and/or comfortable with their management plan knowing the results of PBC/BAL? (4) Does antibiotic management change as a result of PBC/BAL? (5) Given a negative result of PBC/BAL, how comfortable are physicians withholding antibiotic therapy?
|
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An adjudication committee, blinded to treatment allocation and study outcomes, reviewed pertinent clinical, microbiological, laboratory, radiographic, and pathologic reports to determine whether patients with clinical suspicion of pneumonia did have pneumonia. A new and/or persistent radiographic infiltrate was required to make the diagnosis as well as other clinical and laboratory features consistent with pneumonia. All events were reviewed in duplicate, independently; differences were resolved by consensus. This adjudication process has been shown to be rigorous and has been used successfully in two other multicenter trials.11 ,14
For patients undergoing bronchoscopy for the evaluation of a clinical suspicion of pneumonia, we administered the questionnaire examining physician assessment of the diagnosis and management plan prior to bronchoscopy (prebronchoscopy) and once the microbiology results of bronchoscopy were available (postbronchoscopy). The prebronchoscopy and postbronchoscopy questionnaire asked physicians (ICU attendings or ICU fellows) to estimate the likelihood that the patient had VAP, their confidence in their assessment of the diagnosis of VAP, antibiotics they prescribed, and their overall level of comfort with the patient's management plan. The questionnaire was pretested by seven intensivists; feedback included format changes, such as written anchors for all questions and boxes to check to denote antibiotics administered.
Because of a variety of provider, patient, or institutional conditions,
some patients with a clinical suspicion of pneumonia did not undergo
bronchoscopy with PBC or BAL. From this cohort, we assembled a
nonconsecutive control group. In order to compare before and after
antibiotic changes with the group that had bronchoscopy, all patients
who had a clinical suspicion of pneumonia and did not undergo
bronchoscopy and who remained in the ICU for 
5 days were included
in the control group. The questionnaire was not administered to
physicians looking after patients in the control group; rather,
antibiotics were recorded the day of the clinical suspicion of
pneumonia (preassessment) and 5 days later (postassessment). This time
interval is consistent with the median time interval between before and
after questionnaires (5 days) in the group that underwent bronchoscopy.
Patients from both groups were followed up prospectively and information on antibiotic use and clinical outcomes was recorded. Decisions about patient management (whether BAL or PBC was performed) and antibiotic use were made by the attending physicians. Individual antibiotics were assessed on the questionnaire; however, for analysis and presentation of this study, we categorized antibiotics into groups as shown in Table 2 . We did not record reasons why antibiotics were prescribed or changed or whether antibiotic administration was appropriate or inappropriate. We combined the results of PBC with results of BAL given that patient outcomes appear to be similar regardless of which invasive diagnostic method is used.15 Ethics approval was obtained from the local institutional review board of each participating site and informed consent was obtained from the patient's next-of-kin.
|
2 test. In addition, we separately
analyzed the results of the subgroup of patients who had "no
growth" on PBC or BAL culture. Duration of mechanical ventilation and
ICU and hospital length of stay were compared using the Wilcoxon Rank
Sum Test. We considered a p value of < 0.05 to be statistically
significant. |
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
|
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our study, 93% of patients with a clinical suspicion of VAP were already receiving antibiotics at time of bronchoscopy. Given the multicenter nature of our study, this likely reflects clinical practice in university-affiliated ICUs in Canada but may not generalize to other settings. This has significant implications in interpreting the results of PBC or BAL. Using postmortem examinations as the reference standard to make the diagnosis of pneumonia, previous studies have documented the high incidence of false-positive and false-negative results in patients receiving antibiotics who underwent PBC or BAL.7 ,8 ,17 ,18 In patients already receiving antibiotics, if there is minimal correlation among clinical findings, microbiological data, and histopathologic condition of the lung, the utility of histopathologic study as the gold standard for the diagnosis of pneumonia is questionable. In another study, using adjudicators to determine presence or absence of pneumonia, investigators have shown that recent changes to antibiotics affect microbial growth on cultures.19 To maintain accuracy, it has been suggested that the diagnostic thresholds for PBC and BAL should be reduced.19 However, from a patient outcomes perspective, the question still remains: Does a management strategy that combines clinical judgment with results from invasive diagnostic tests improve patient outcomes compared with clinical judgment alone (regardless of which gold standard is used)?
In the management of VAP, there are few controlled trials evaluating a strategy based on information derived from invasive diagnostic techniques compared with clinical judgment. In the absence of compelling data from large randomized trials, a reasonable, next-best approach is to examine the impact of invasive diagnostic techniques on health-care workers, patient management, and patient outcome using an observational study design.20 In this study, we demonstrated that PBC or BAL increases physician confidence in the diagnosis and management of VAP. On average, physicians were more comfortable with their diagnosis and management plan as a consequence of having information from bronchoscopy with PBC or BAL. We found that most patients (72% of patients who underwent bronchoscopy and 76% of control patients) had a change in antibiotic therapy in the 5 days following the initiation of empiric therapy. This agrees with other reports that document empiric therapy based on clinical judgment in patients suspected of having pneumonia was often inadequate.21 ,22 In our observational study, we documented a change in therapy; however, we were not able to comment on the "appropriateness" or "inappropriateness" of antibiotic prescriptions. We did not intentionally influence antibiotic prescribing patterns but assumed that antibiotic therapy was changed on the basis of diagnostic test results or changes in the clinical condition of the patient.
In a controlled study, we have demonstrated that narrowing the spectrum of antibiotics and/or discontinuing antibiotic therapy occurs more often with the use of invasive diagnostic tests for VAP. While both groups experienced a similar rate of antibiotic changes, patients who underwent PBC or BAL ended up receiving fewer antibiotics and a greater proportion had their antibiotic therapy discontinued all together. Furthermore, compared with empiric antibiotic therapy initiated at baseline, patients in the group who underwent bronchoscopy ultimately received fewer prescriptions for third-generation cephalosporins while patients in the control group received a trend toward an increase in broad-spectrum antibiotics. Most of the time, antibiotic therapy was not discontinued in patients with no growth on PBC/BAL suggesting that patients either had other reasons to be receiving antibiotics or that physicians were not comfortable withholding antibiotic therapy in patients with suspected VAP and a negative PBC/BAL result. In the subgroup of 34 patients with "no growth" on culture, while the diagnosis of pneumonia was judged to be less likely, physicians were not more confident in their diagnosis following bronchoscopy, perhaps due to the diagnostic uncertainty in patients already receiving antibiotics. Nevertheless, a potential advantage of using invasive techniques is that antibiotic therapy may be tailored to the results of PBC or BAL. Reducing unnecessary broad-spectrum antibiotic therapy has important clinical implications as it may minimize the emergence of resistant microorganisms in the ICU and reduce antibiotic costs.23
In our study, patients who underwent bronchoscopy with PBC and BAL had a similar duration of mechanical ventilation and ICU stay and a lower mortality rate compared with patient who did not undergo bronchoscopy. Given that patients were not randomized to invasive or noninvasive diagnostic approaches, we caution making strong inferences from these findings. Differences in clinical outcomes observed in this study might be due to differences in patient populations or other management strategies (other than bronchoscopy) that we did not measure.
How do our findings compare with other published studies examining outcomes associated with bronchoscopy with PBC or BAL? Bonten and colleagues24 studied 138 patients with a clinical suspicion of pneumonia who underwent bronchoscopy with PBC or BAL or both. At the time of or following bronchoscopy, 74 (53%) patients were placed on a regimen of empiric antibiotics. Based on the microbiological results of invasive tests, VAP was proved in 72 cases (52%). In those patients with a confirmed diagnosis of pneumonia, empiric therapy was modified in 14 of 40 (35%) cases. In 11 of these 14 cases, the spectrum of antibiotics was narrowed based on results of bronchoscopy. In the group that did not have pneumonia (based on negative test results) but were started on a regimen of empiric antibiotics, antibiotic therapy was discontinued within 48 h in 17 of 34 (50%) patients. There were no differences in mortality rates between these different subgroups. The lack of a control group in this study limits the inferences one can derive from this study as antibiotic changes may be due to other factors than the results of invasive tests.
Sanchez-Nieto and colleagues25 randomized 51 ICU patients with a clinical suspicion of pneumonia to undergo quantitative endotracheal aspiration with bronchoscopy and PBC and BAL (group A) or quantitative endotracheal aspiration alone (group B). Thirty-nine (76%) patients from both groups were receiving antibiotics prior to sampling. Initial antibiotic treatment was changed in 42% of patients in group A and 16% of patients in group B (p < 0.05). Duration of mechanical ventilation, ICU length of stay, and mortality were similar between the two groups. A similar study of 34 patients, recently reported in abstract form, found no clinical benefit to invasive diagnostic techniques.26 Consistent with our observations, these preliminary studies suggest that invasive diagnostic techniques result in more frequent changes to empiric antibiotics and greater ability to narrow the spectrum or discontinue antibiotic therapy.
Are changes to antibiotic prescriptions based on results of invasive diagnostic tests, like those observed in our study, likely to improve patient outcomes? Some investigators have suggested that most prescriptions based on empiric therapy are inadequate compared with the results of PBC or BAL.20 ,21 ,27 Theoretically, changing antibiotic therapy on the basis of PBC or BAL results could result in improved patient outcomes since adequate or appropriate use of antibiotics may be associated with a good prognosis in the setting of VAP.28 ,29 However, Luna and colleagues27 demonstrated that changes to antibiotic therapy based on BAL data may not influence outcome. In a single center, they followed up 132 patients with a clinical suspicion of pneumonia and described antibiotic use prior to and after BAL results were available. Sixty-five patients had a "positive" BAL. Of those 65 patients, only 16 (25%) were receiving adequate antibiotic therapy prior to bronchoscopy. The mortality rate in those patients receiving initial, inadequate therapy was extremely high (31/34, 91%) compared with those receiving adequate therapy (6/16, 38%). When initial antibiotic therapy (prebronchoscopy) was considered adequate, the mortality rate was lower (6/16, 38%) compared with cases in which initial antibiotic therapy was considered inadequate or no antibiotics were given (40/49, 81.6%). When the BAL results were available, the mortality rate of patients receiving adequate therapy (21/37, 57%) was similar to patients who continued to receive inadequate therapy (2/5, 40%). This last comparison may be biased in that patients receiving inadequate therapy after BAL continued to receive inadequate therapy because they were responding clinically and radiographically. In addition, 23 patients died before the results of BAL were available. In a "less sick" patient population, the results of invasive tests might prove to be more useful, but this requires further testing.
Even if invasive diagnostic tests do not alter patient outcomes, there may be an economic rationale for advocating the use of bronchoscopic techniques in the workup of patients with VAP. In some settings, the costs associated with bronchoscopy and special cultures may be offset by antibiotic savings.30
In summary, reasons favoring invasive diagnostic techniques to evaluate a patient with suspected VAP include the following: (1) they may provide a more accurate assessment of the pathogens and their sensitivity patterns allowing for tailored antibiotic administration; (2) they probably increase the likelihood of making a diagnosis; (3) they increase the confidence and comfort level of health-care workers in treating patients with suspected VAP; and (4) they may provide information that may change patient management and improve patient outcomes. Reasons not to use invasive diagnostic techniques include the following: (1) the accuracy of invasive diagnostic techniques is questionable in patients receiving prior antibiotics7 ,8 ,17 ,18 ; (2) bronchoscopy may transiently worsen patient's status (cardiac arrhythmias, hypoxemia, bleeding, etc), although several studies of mechanically ventilated patients suggest that the incidence of such complications is quite low31 ; (3) an invasive approach to diagnosing VAP may increase costs of caring for critically ill patients, although the incremental costs are probably not substantial4 and may be offset by antibiotic savings29 ; and (4) while patient management may change based on results from invasive tests, data suggesting that these changes lead to an improvement in patient outcome are lacking.27 Currently, we believe there are insufficient data to generate strong clinical recommendations. As clinical experience grows and new evidence emerges from randomized trials, practitioners and policy makers continue to weigh the benefits, risks, and costs of bronchoscopic diagnostic tests in patients with suspected VAP.
|
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Drs. Heyland and Cook are Career Scientists with the Ontario Ministry of Health. This study was funded in part by grants from the Ontario Thoracic Society, Medical Research Council of Canada, Hoescht Marion Rousell, and the Boston Scientific Corporation.
Correspondence to: Daren K. Heyland, MD, MSc, Angada 3, Kingston General Hospital, 76 Stuart St, Kingston, Ontario, Canada K7L 2V7; e-mail: dkh2@post.queensu.ca
Abbreviations: MOD = multiple organ dysfunction; PBC = protected brush catheter; VAP = ventilator-associated pneumonia
Received for publication May 27, 1998. Accepted for publication October 11, 1998.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  The Canadian Critical Care Trials Group A Randomized Trial of Diagnostic Techniques for Ventilator-Associated Pneumonia N. Engl. J. Med., December 21, 2006; 355(25): 2619 - 2630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Koenig and J. D. Truwit Ventilator-Associated Pneumonia: Diagnosis, Treatment, and Prevention Clin. Microbiol. Rev., October 1, 2006; 19(4): 637 - 657. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Fujitani and V. L. Yu Diagnosis of Ventilator-Associated Pneumonia: Focus on Nonbronchoscopic Techniques (Nonbronchoscopic Bronchoalveolar Lavage, Including Mini-BAL, Blinded Protected Specimen Brush, and Blinded Bronchial Sampling) and Endotracheal Aspirates. J Intensive Care Med, January 1, 2006; 21(1): 17 - 21. [Abstract] [PDF]
|
|
|
|
 |
|
 M. Kollef, D. Pittet, M. Sanchez Garcia, J. Chastre, J.-Y. Fagon, M. Bonten, R. Hyzy, T. R. Fleming, H. Fuchs, L. Bellm, et al. A Randomized Double-Blind Trial of Iseganan in Prevention of Ventilator-associated Pneumonia Am. J. Respir. Crit. Care Med., January 1, 2006; 173(1): 91 - 97. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Guidelines for the Management of Adults with Hospital-acquired, Ventilator-associated, and Healthcare-associated Pneumonia Am. J. Respir. Crit. Care Med., February 15, 2005; 171(4): 388 - 416. [Full Text] [PDF]
|
|
|
|
|
|
D. E. Ost, C. S. Hall, G. Joseph, C. Ginocchio, S. Condon, E. Kao, M. LaRusso, R. Itzla, and A. M. Fein Decision Analysis of Antibiotic and Diagnostic Strategies in Ventilator-associated Pneumonia Am. J. Respir. Crit. Care Med., November 1, 2003; 168(9): 1060 - 1067. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Spain Ventilator-Associated Pneumonia and Surgical Patients Chest, May 1, 2002; 121(5): 1390 - 1391. [Full Text] [PDF]
|
|
|
|
|
|
J. Chastre and J.-Y. Fagon Ventilator-associated Pneumonia Am. J. Respir. Crit. Care Med., April 1, 2002; 165(7): 867 - 903. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. A. Balk Is a Silver Coating a Silver Lining? Chest, March 1, 2002; 121(3): 682 - 683. [Full Text] [PDF]
|
|
|
|
|
|
J. C. Pereira Gomes, W. L. Pedreira Jr., E. M. P. A. Araujo, F. G. Soriano, E. M. Negri, L. Antonangelo, and I. Tadeu Velasco Impact of BAL in the Management of Pneumonia With Treatment Failure : Positivity of BAL Culture Under Antibiotic Therapy Chest, December 1, 2000; 118(6): 1739 - 1746. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Morehead and S. J. Pinto Ventilator-Associated Pneumonia Arch Intern Med, July 10, 2000; 160(13): 1926 - 1936. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-Y. Fagon, J. Chastre, M. Wolff, C. Gervais, S. Parer-Aubas, F. Stephan, T. Similowski, A. Mercat, J.-L. Diehl, J.-P. Sollet, et al. Invasive and Noninvasive Strategies for Management of Suspected Ventilator-Associated Pneumonia: A Randomized Trial Ann Intern Med, April 18, 2000; 132(8): 621 - 630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Cook and W. J. Sibbald The promise and the paradox of technology in the intensive care unit Can. Med. Assoc. J., November 1, 1999; 161(9): 1118 - 1119. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
