| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology and Intensive Care, Hotel-Dieu Hospital, Lyon France 69288
Address correspondence and reprint requests to Bernard Allaouchiche, MD, Department of Anesthesiology and Intensive Care, Hotel-Dieu Hospital, 1, place de l’hôpital, 69288 Lyon cedex 02, France. Address e-mail to allaouch{at}univ-lyon1.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
103 CFU/mL). A final diagnosis of VAP was established in 67 patients and there was no infection in 49 cases. Regarding detection of bacteria using the Gram stain, we found a sensitivity of 76.2%, a specificity of 100%, a positive predictive value of 100% and a negative predictive value of 75.4%. There was a good agreement with the final diagnosis (kappa statistic 0.73; concordance 86.2%). The degree of qualitative agreement between Gram stain and quantitative cultures was analyzed in the VAP group: the correlation was complete in 39% (26 of 67 VAP), partial in 28% (19 of 67 VAP) and there was no correlation in 33% (22 of 67 VAP). We conclude that despite its overall "good agreement," the Gram stain is of limited use for the rapid diagnosis of VAP and unreliable for the early adaptation of empirical antimicrobial therapy when using the noninvasive PBAL procedure.
Implications: We investigated the reliability of direct examination versus quantitative cultures in the early diagnosis of ventilator-associated pneumonia using the protected bronchoalveolar lavage in 104 patients. Regarding detection of bacteria using the Gram stain, we found low sensitivity and negative predictive value and high specificity and positive predictive value.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
38.5°C), purulent tracheal aspirates, leukocytosis (12,000 cells/mm3), and new or persistent radiographic lung infiltrates unrelated to cardiogenic causes. All samples were performed in patients who had not received antimicrobial therapy during the previous 3 days. No patient received topical prophylactic antibiotics. We performed 116 mini-BAL using the PBAL catheter technique (CombicathTM; Plastimed, St. Leu La Foret, France). During the procedure, 100% oxygen was administered and patients were sedated with IV narcotics and opioids. Topical anesthesia was not used. Cardiovascular and oxygen saturation monitoring was performed during the entire procedure. The tracheal sputum was aspirated and collected before introducing the protected catheter. The PBAL was inserted using a previously described technique (3). After the blinded introduction in the bronchial system, the inner catheter was advanced until resistance was encountered and 20 mL of sterile saline were administered. The fluid was then withdrawn by hand suction into the infusion syringe. When at least 2 mL of fluid had been sterilely retrieved, the entire catheter was removed. The entire sampling procedure lasted <2 min and was performed by the same physician. All samples were sent within 15 min to the laboratory for cytological and bacteriological examination. Gram stains were performed using standard methods by the same technologist.
Aliquots from the original suspension (0.2-mL) were dropped into a cytospin and centrifuged at 300g for 10 min. Slides were Gram stained and were examined at high magnification (100x) results. The possible presence of microorganisms was analyzed on 10 to 50 fields and classified according to the Gram stain morphology.
The fluid was diluted to obtain concentrations of 10-1, 10-3, and 10-5. The samples were then plated onto Petri dishes: Colombia agar, chocolate agar, trypticase soy, McConkey and Sabouraud agar. Bacterial colonies were counted and identified using conventional techniques.
The final diagnosis of pneumonia was based on positive results of PBAL culture (cutoff 103 cfu/mL). VAP was excluded if the following criteria were fulfilled: negative or nonsignificant growth in culture of PBAL and full recovery without antimicrobial therapy, or diagnosis of another disease of the chest accounting for the chest radiograph abnormality. Pulmonary densities erroneously defined as "having recently appeared" after first examination were retrospectively reclassified by careful reviewing of historical radiographs before results of Gram stains. The therapeutic strategy based on the results of the determination of Gram stain in PBAL fluid was to give early empirical antimicrobial therapy if a microorganism was seen on Gram stain slides and if there was evidence of septic shock or severe hypoxemia. According to our bacterial ecology, patients were initially treated with ceftazidime and vancomycin and secondarily adapted to the quantitative culture results.
Demographic data are expressed as mean (± SD). Sensitivity, specificity, positive predictive values (PPV) and negative predictive values were estimated using standard formulas. The degree of concordance between Gram stains and quantitative cultures was established using the Cohen-Kappa coefficient. Kappa values more than 0.81 were considered to indicate "very good agreement," values of 0.60–0.81 "good agreement," values of 0.41–0.60 "moderate agreement," values of 0.21–0.40 "fair agreement," and values <0.20 "poor agreement." To assess the correlation between PBAL Gram stains and PBAL quantitative cultures, results were divided into three categories: total correlation if each Gram stain morphotype present/or absent grew/or not at significant concentration (i.e., 103 cfu/mL), absent correlation if each Gram stain morphotype present did not grow in significant concentration, and partial correlation if part of the Gram stain morphotypes present grew in significant concentration.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The mean duration of mechanical ventilation was 13 (±3) days. The primary indications for ventilator support were postoperative respiratory failure (n = 47), exacerbation of chronic obstructive pulmonary disease (n = 38), severe sepsis (n = 12), multiple organ failure (n = 4), and acute pancreatitis (n = 3).
According to significant positive quantitative cultures, the diagnosis of VAP was established in 67 cases; 21 patients were infected by Gram-negative bacilli, 17 by Gram-positive cocci, and polymicrobial growth was seen in 43% (29 of 67). There was no bacterial pneumonia in 49 cases. During or after the sampling procedure, no major hemodynamic changes, pneumothorax, or hemorrhage were observed. The mini-BAL effluent showed less than or equal to 1% squamous epithelial cells in all patients.
Bacteria identified on Gram staining were determined in 51 of the 67 VAP, whereas none of the 49 controls. Thus, when comparing the final diagnosis with the presence or absence of bacteria on Gram stain, sensitivity, specificity, PPV, and negative predictive values were respectively 76.2%, 100%, 100%, and 75.4%. There was good agreement between the final diagnosis and Gram stain (kappa statistic 0.73; concordance 86.2%).
When assessing the degree of qualitative agreement (total, partial, or absence of correlation) between Gram stain and quantitative cultures, the correlation was complete in 39% (26 of 67 VAP), partial in 28% (19 of 67 VAP), and absent in 33% (22 of 67 VAP) ( Table 1).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The diagnosis of VAP in critically ill patients is a crucial challenge for ICU physicians. Nowadays, the reference method is the pathologic examination and/or culture of lung biopsy specimens, which cannot be performed routinely in intubated patients. Hence, numerous methods for the diagnosis of VAP have been developed and are currently used, but all are imperfect because of their lack of sensitivity or specificity (6). Furthermore, results of quantitative cultures are not available until 24 to 72 hours after the procedure and potentially contribute to the frequent rate of morbidity and mortality (6–10). Luna et al. (11) showed that adequate empirical antibiotics reduced the mortality rate when compared with inadequate or absent chemotherapy. The intent of this work was not to assess the usefulness of PBAL direct examination for diagnosing VAP but to solve the problem of delay by prospectively investigating the correlation between Gram staining and quantitative cultures of PBAL.
For bacteria, the Gram stain is the most frequently used procedure and provides morphological information that can be used in the empirical selection of antibiotics for therapy (12). In the literature, few prospective studies have investigated the correlation between the Gram stain and quantitative cultures of mini-BAL fluid (4,13,14). In all studies microbiological analyses were performed on BAL obtained by bronchoscopic methods and/or nonbronchoscopic methods such as plugged telescoping catheter or protected specimen brushes ( Table 2). Solé-Violán et al. (15) found a correlation between Gram staining and both protected specimen brushes and BAL culture results but the presence or absence of antibiotic therapy is unknown. Meduri et al. (8) found low sensitivity of Gram staining of PBAL and protected specimen brushes, prompting many authors to use microscopic analysis of BAL quantitative cultures. For Aubas et al. (16), the presence of bacteria on a Gram stain was significantly more frequent in the pneumonia group (28 patients) with low sensitivity and specificity but the presence or absence of antibiotic therapy is unknown.
|
There was good agreement between the final diagnosis and Gram staining (kappa statistic 0.73; concordance 86.2%). However, these results had limited accuracy when used to select adequate antibiotic therapy; in the VAP group, the correlation between the Gram stain and PBAL quantitative cultures was complete in 39% (26 of 67 VAP), partial in 28% (19 of 67 VAP), and absent in 33% (22 of 67 VAP). The results of this study support that the reliability of Gram staining is variable, dependent on the result of the Gram staining. Actually, Gram-negative stains from PBAL specimens were highly predictive of Gram-negative cultures, whereas Gram-positive stains were poor indicators of the culture results (Table 1).
Other authors have reported discrepancies between the Gram stain and quantitative cultures of pulmonary samples (17–30) (Table 2). Papazian et al. (4) support quantitative assessment of intracellular organisms on bronchial blind sampling to separate VAP and non-VAP patients. These findings do not agree with our results. An explanation could be the method of establishing VAP diagnosis. The pneumonia diagnosis was based on histologic examination of transbronchial biopsy that has not been well established (17). Kollef et al. (18) reported poor diagnostic agreement between BAL fluid Gram stain results and microbiologically confirmed Gram-negative VAP and suggested that the concentration of endotoxin in BAL fluid (>5 EU/mL) could be an acceptable adjunct for the rapid diagnosis of Gram-negative VAP, but its cost-effectiveness has not been determined. Croce et al. (19) found that the Gram staining of BAL effluent correlated poorly with quantitative cultures and was not reliable for empirical therapy. Nonetheless, in this study, the high quantitative threshold (>105 cfu/mL) used to establish diagnosis of VAP could explain these differences with our results. Namias et al. (20) studied the role of Gram stains of the tracheal aspirates in guiding antimicrobial therapy. As in our study, they found variable correlation regardless of the culture results: high predictive value for the Gram-negative bacilli and low predictive value for the Gram-positive cocci. They reported that errors could come from overdecolorization or underdecolorization during the Gram stain procedure.
In our study, PBAL was considered the reference method. This simplified and safe technique has been advocated as a potentially better alternative compared with bronchoscopic procedures because of its minimal invasiveness, wide availability, and relative inexpensiveness (5,13). However, the "gold standard" technique for evaluating the diagnosis of VAP is problematic. Actually, one study reported disagreement between histology and bacteriological cultures. Corley et al. (21) performed a postmortem open lung biopsy in all ICU patients who underwent two weeks of mechanical ventilation. Four different pathologists analyzed the slides. The prevalence of VAP varied from 18% to 38%. Kirtland et al. (22) evaluated histological, microbiological, and clinical criteria in the recognition of VAP in 39 patients who died while mechanically ventilated. They found neither bacterial density from the airway quantitative culture nor the bacterial density from quantitative culture of lung tissue accurately distinguished the histological VAP and non-VAP groups. Solé-Violán et al. (23) compared bronchoscopic diagnostic techniques with histological findings in organ donors without suspected pneumonia immediately after death. Seven of the nine donors without clinical criteria of VAP and not receiving antibiotic therapy showed histological features of VAP.
In summary, the Gram stain has limited value for the rapid diagnosis of VAP from PBAL fluid particularly when bacteria were not seen on the direct examination. Furthermore, the Gram stain is not reliable for the early adaptation of empirical chemotherapy. In case of VAP, other techniques are warranted to improve the choice of early antibiotic treatment.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
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]
|
|
|
|
 |
|
 D. Church, S. Elsayed, O. Reid, B. Winston, and R. Lindsay Burn Wound Infections Clin. Microbiol. Rev., April 1, 2006; 19(2): 403 - 434. [Abstract] [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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
