American Journal of Respiratory and Critical Care Medicine

Hundreds of articles have been published describing, contrasting, and debating the merits of various techniques and strategies to diagnose ventilator-associated pneumonia (VAP). Inaccuracies of clinical criteria for VAP diagnosis have stimulated this work. Since an accurate diagnosis is critical to analysis of risk factors, prevention strategies, and treatment, uncertainty regarding diagnostic accuracy has left the whole area of VAP in a state of flux. The study by Sanchez-Nieto and coworkers (1) published in this issue of the American Journal of Respiratory and Critical Care Medicine describes a landmark study in the literature of VAP.

Two aspects of the study by Sanchez-Nieto and colleagues (1) are unique and, as such, raise the research of diagnosis of VAP to a new level. Randomization of management based on two diagnostic strategies—quantitative cultures of bronchoalveolar lavage (BAL) and/or protected specimen brush (PSB) versus quantitative endotracheal aspirates—is the first laudatory step taken by these researchers. Previously, operating characteristics of the diagnostic techniques was the primary focus (2-4). A range of sensitivities and specificities has been found, even using the same technique in similar populations. These studies assume that the “truth,” i.e., actual presence or absence of pneumonia, can be known and the results of each test can be checked against this “truth.”

One of the clear messages of the literature published to date is that this truth is not available. Even autopsy histology has been found to have inaccuracies. VAP has been demonstrated to exist as focal areas within a lobe, which may not extend to the pleural surface and can be missed by the peripheral sampling performed in some studies (3, 5). VAP may be preceded by a purulent bronchiolitis, which has quantitative culture growth at a level thought to be suggestive of pneumonia (5). Conversely, normal host response may sterilize the lung by the time sampling is performed, leaving behind histologic pneumonia without positive cultures, even in the absence of antibiotics (3, 5). Even different pathologists reviewing the same specimens cannot always agree upon the presence of VAP. Equilibrating results in the dead or dying to those obtained at the time of first clinical suspicion in a viable patient is also problematic.

The time has come for a moratorium on further sensitivity and specificity studies of VAP diagnostic techniques. Most have operating characteristics equivalent to frequently ordered tests, such as ventilation perfusion scans and CT staging of bronchogenic carcinoma (6). Clinicians rarely question the use of these tests, even to make critical, potentially life-threatening decisions, despite accuracy and cost that are similar to BAL or PSB. Research now needs to refocus on how (and if) the information obtained by alternative diagnostic strategies for VAP directly impacts patient care, rather than assuming that different test results lead to different patient management. Only prospective randomized studies, such as that of Sanchez-Nieto and colleagues (1), can accomplish this.

Research must also focus on improved early diagnosis. Several studies (7-9), including that of Sanchez-Nieto (1), have found that inappropriate initial antibiotic therapy is one of the most significant risk factors for death in VAP. Since culture results are not available at the time VAP is first suspected, initial antibiotic choices are empiric no matter which culture technique is used. Whether antibiotic modifications once culture and sensitivity results return will salvage these patients is unclear (9). To broaden initial empiric antibiotic coverage to include all the potential causative organisms will clearly have financial implications, if not significant effects on selection for resistance and drug toxicity (8). A test that accurately excludes VAP could minimize the adverse effects of this therapeutic strategy.

The second notable aspect of this study is the prospective use of mortality as an endpoint. Unless a diagnostic tool leads to a clinically relevant, objectively measurable change in outcome, the only consideration in choosing between alternative diagnostic strategies is cost. Outcomes relevant to VAP include not only mortality, but also organ failure (especially prolonged mechanical ventilation), and development of superinfection, either pneumonia or extrapulmonary. Any cost/benefit or cost effectiveness analysis must also include these factors. Differences in antibiotic therapy based on the results of diagnostic tests should impact one of these outcomes, most likely cost, and should not be considered an independent outcome parameter.

However, outcome studies are much more difficult to perform. Mortality in VAP is the result of a complex interaction between the adequacy of host defense, virulence of the pathogen, and adequacy of antibiotic therapy. By allowing modification of inappropriate antibiotic therapy, an accurate diagnosis is only the first step in attempts to impact mortality. Even if the diagnosis is accurate, mortality may be unaffected if subsequent antibiotic therapy is ineffective (8) or the patient is too ill to mount an adequate host response. Similar confounding factors will affect any of the other measurements of outcome.

Despite the laudable design, the study of Sanchez-Nieto (1) does not settle the original question. The study was designed to focus principally on mortality. Despite randomization, important differences in factors pertinent to mortality of VAP existed between the groups. The incidence of infection with Pseudomonas and Acinetobacter, with clearly higher crude and possibly higher attributable mortality rates than other causes of VAP (10, 11), was significantly different between the two groups (p < 0.02). Inappropriate initial antibiotic therapy was also significantly more common in the invasive group, disproportionately among the patients with Pseudomonas VAP where inappropriate therapy is most likely to affect mortality. The slightly higher mortality in the invasive diagnosis group is most likely explained by these two factors, rather than any issue related to diagnosis. In addition, differences in mortality based on diagnostic strategy can never be demonstrated unless the disease itself affects mortality. Baker and colleagues (12) have previously suggested that VAP in trauma patients has no attributable mortality, at least by case control methodology. Trauma accounted for > 50% of causes for admission in the study of Sanchez-Nieto (1), significantly diluting the possibility of demonstrating mortality differences based on diagnosis. Unfortunately, the authors chose to prematurely close the study, leaving inadequate patient numbers for multivariate analysis.

A major flaw in the study of Sanchez-Nieto (1) was continuation of antibiotics in all patients despite negative cultures. This management strategy neutralizes one of the most significant potential advantages of any diagnostic test in patients with a clinical suspicion of VAP. The adverse consequence of this strategy is illustrated by a recent Bayesian analysis suggesting more deaths if patients were treated with antibiotics based only on clinical suspicion of VAP than if antibiotics were withheld (13). Other studies have suggested that antibiotics can be safely stopped in patients with negative quantitative cultures (14), despite the false negative cultures mentioned by the authors. Continued treatment of culture-negative cases precludes any meaningful cost analysis since the main cost benefit of any type of quantitative culture technique is avoidance or discontinuation of antibiotics in culture-negative cases (14). Close linkage of antibiotic management to culture results is mandatory when comparison of usual practice and quantitative cultures is finally studied. Since nonquantitative tracheal aspirate cultures consistently overdiagnose the presence of pneumonia, this linkage should include stopping antibiotics in patients with culture growth below threshold.

Even if the study showed a definitive difference between the strategies, the results would not be pertinent to the majority of North American physicians. Therapy based on clinical suspicion and either nonquantitative cultures of tracheal aspirates or epidemiologic data remains the diagnostic strategy of the overwhelming majority of physicians. Two debates regarding diagnosis of VAP are occurring concurrently—(1) does any quantitative culture technique lead to improved outcome compared to clinical criteria and nonquantitative cultures and (2) which quantitative culture technique is better? The authors have really only studied the second question. Their results add further support to the suggestion that results of one quantitative technique are similar to those of another quantitative technique, in agreement with prior case/control and cohort studies. However, inability to perform quantitative cultures is the most frequently stated reason invasive procedures are not used to diagnose VAP (15).

The first question is most pertinent for North American physicians. Sanchez-Nieto's results (1) cannot be extrapolated to nonquantitative tracheal aspirate cultures. A separate prospective randomized study is required to prove whether diagnosis based on quantitative or nonquantitative cultures leads to better outcome. The pilot study of Sanchez-Nieto and coworkers (1) has demonstrated some of the difficulties and important considerations in performing this critical study. Hopefully, they have now set future research regarding VAP diagnosis on a course that will ultimately settle these issues.

Richard G. Wunderink

Methodist Hospitals of Memphis

University of Tennessee

Memphis, Tennessee

1. Sanchez-Nieto J. M., Torres A., Garcia-Cordoba F., El-Ebiary M., Carrillo A., Ruiz J., Nunez M. L., Niederman M.Impact of invasive and non-invasive quantitative culture sampling on outcome of ventilator-associated pneumonia: a pilot study. Am. J. Respir. Crit. Care Med1571997371376
2. El-Ebiary M., Torres A., Gonzalez J., De La J., Bellacasa, Garcia C., Jimenez de Anta M., Ferrer M., Rodriguez-Roisin R.Quantitative cultures of endotracheal aspirates for the diagnosis of ventilator-associated pneumonia. Am. Rev. Respir. Dis148199715521557
3. Marquette C. H., Copin M. C., Wallet F., Neviere R., Saulnier F., Mathieu D., Durocher A., Ramon P., Tonnel A. B.Diagnostic tests for pneumonia in ventilated patients: prospective evaluation of diagnostic accuracy using histology as a diagnostic gold standard. Am. J. Respir. Crit. Care Med151199518781888
4. Chastre J., Fagon J. Y., Bornet-Lecsco M., Calvat S., Dombret M. C., Khani R., Basset F., Gibert C.Evaluation of bronchoscopic techniques for the diagnosis of nosocomial pneumonia. Am. J. Respir. Crit. Care Med1521995231240
5. Rouby J. J., DeLassale E. M., Poete P., Nicolas M. H., Bodin L., Jarlier V., Le Charpentier Y., Grosset J., Viars P.Nosocomial bronchopneumonia in the critically ill: histologic and bacteriologic aspects. Am. Rev. Respir. Dis148199210591066
6. Baker A. M., Bowton D. L., Haponick E. F.Decision making in nosocomial pneumonia: an analytic approach to the interpretation of quantitative bronchoscopic cultures. Chest10719958595
7. Torres A., Aznar R., Gatell J. M., Jimenez P., Bonzalez J., Ferrer A., Celis R.Incidence, risk, and prognosis factors of nosocomial pneumonia in mechanically ventilated patients. Am. Rev. Respir. Dis1421990523528
8. Alvarez-Lerma F.Modification of empiric antibiotic treatment in patients with pneumonia acquired in the intensive care unit. Intensive Care Med221996387394
9. Luna C. M., Vujacich P., Niederman M. S., Vay C., Gherardi C., Matera J., Jolly E. C.Impact of BAL data on the therapy and outcome of ventilator-associated pneumonia. Chest1111997676685
10. Fagon J. Y., Chastre J., Hance A. J., Montravers P., Novara A., Gibert C.Nosocomial pneumonia in ventilated patients: a cohort study evaluating attributable mortality and hospital stay. Am. J. Med941993281299
11. Kollef M. H., Silver P., Murphy D. M., Trovillion E.The effect of late-onset ventilator-associated pneumonia in determining patient mortality. Chest108199516551662
12. Baker A. M., Meredith J. W., Haponik E. F.Pneumonia in intubated trauma patients: microbiology and outcomes. Am. J. Respir. Crit. Care Med1531996343349
13. Sterling T. R., Ho E. J., Brehm W. T., Kirkpatrick M. B.Diagnosis and treatment of ventilator-associated pneumonia—impact on survival. Chest110199610251034
14. Croce M. A., Fabian T. C., Shaw B., Stewart R. M., Pritchard F. E., Minard G., Kudsk K. A., Baselski V. S.Analysis of charges associated with diagnosis of nosocomial pneumonia: can routine bronchoscopy be justified? J. Trauma371994721727
15. Haponik E., Mehta A.Pulmonary fellows' perspectives about ventilator-associated pneumonia (VAP) (abstract). Am. J. Respir. Crit. Care Med1551997A507


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record