Skip Navigation LinksHome > May 2008 - Volume 16 - Issue 3 > Duration of Treatment for Ventilator-Associated Pneumonia
Text sizing:
Infectious Diseases in Clinical Practice:
doi: 10.1097/IPC.0b013e31817588c0
Editorial Comment

Duration of Treatment for Ventilator-Associated Pneumonia

Mandell, Lionel MD, FRCPC, FRCP(LOND)

Free Access
Article Outline
Collapse Box

Author Information

Division of Infectious Diseases, Hamilton Health Sciences, McMaster University.

Address correspondence and reprint requests to Lionel Mandell, MD, FRCPC, FRCP(LOND), Henderson Hospital, 711 Concession St, 40 Wing Hamilton, Ontario, Canada L8V 1C3. E-mail:

The article by Dr Beidas asked whether nosocomial pneumonia should be treated for 8 days and questions the basis of this recommendation in the Guidelines for the Management of Adults with Hospital-Acquired Pneumonia, Ventilator Associated Pneumonia.1

Although it is certainly admirable and appropriate to question new recommendations, I believe that Dr Beidas has erred on 3 counts.

First of all, let us examine exactly what the recommendation is. Dr Beidas begins her article with a statement: "In 2005 the ATS/IDSA Guidelines for the Management of Nosocomial Pneumonia recommended a shortened course (8 days) for the treatment of non-pseudomonas ventilator-associated pneumonia (VAP)." In fact, what is actually stated is the following: "If patients receive an initially appropriate antibiotic regimen, efforts should be made to shorten the duration of therapy from the traditional 14 to 21 days to periods as short as 7 days, provided that the etiologic pathogen is not Pseudomonas aeruginosa, and that the patient has a good clinical response with resolution of clinical features of infection."

The statement by the ATS/IDSA is a carefully worded and carefully qualified one that takes into account 2 critical issues: the use of appropriate antibiotics and the clinical response of the patient. It certainly is not a blanket statement that recommends 8 days of treatment in a perfunctory manner and leaves it at that.

In the second sentence of her article, Dr Beidas then states that the recommendation was based upon a study by Chastre et al.2 This too is somewhat of a misrepresentation of the facts. Although the study by Chastre et al2 was important and ultimately served as the basis for the level 1 status accorded the new statement by the guidelines, it was by no means the only evidence for this. A discussion of the issue of duration of therapy in the guidelines cites articles by Dennesen et al3 and Luna et al.4 Both are observational studies but provided important insights into length of treatment for patients with VAP. Data from both studies suggested that, as long as appropriate antibiotic treatment has been initiated, those patients who will respond do so within several days and prolonged treatment may not be required. The study by Singh et al5 that used the Clinical Pulmonary Infection Score as a way to reduce unnecessary antibiotic use in the treatment of patients with pulmonary infiltrates in the intensive care unit was able to identify patients at lower risk who could be treated for shorter periods.

Finally, she focuses her attention on the article by Chastre et al2 and its shortcoming in terms of experimental design and methodology, and the Consolidated Standards of Reporting Trials (CONSORT) statement is cited.6 Ultimately, Dr Beidas asks that the evidence-based level grade applied to the study by Chastre et al2 be changed from a level 1 to a level 2.

The CONSORT statement is certainly an important and helpful contribution to the literature dealing with the methodology of clinical trials and represents the ideal in the reporting of a randomized controlled trial (RCT). It is an indicator of the quality of reporting in a particular RCT, and although it certainly represents a standard to which all investigators should aspire, it is not a criterion for acceptance of a trial nor does failure to fulfill the checklist of items published in the CONSORT statement constitute a basis on which to dismiss a particular RCT.

In fact, the patient flow diagram in the article by Chastre et al2 very closely follows the template designed to show the flow of participants through each stage of an RCT laid out in the CONSORT statement.

The criteria for the evidence-based grading system used in the ATS/IDSA guidelines are clearly stipulated in the guidelines, and the ATS and IDSA committee members felt that the study by Chastre et al,2 although not perfect, certainly met those criteria.

Dr Beidas raises a number of specific points as well that I will attempt to address individually. Delay in randomization of patients to third day of treatment. Concern is expressed that by being allowed to observe a patient's clinical response for 3 days before randomization, "there is an increased chance for selecting patients that may influence patient outcomes to a particular treatment group." It is not clear to me, however, why this should be the case. Chastre et al2 have specifically excluded patients with early-onset pneumonia and no antimicrobial treatment within the 15 days before infection in a deliberate attempt to exclude patients with infection caused by pathogens that are usually highly sensitive to antibiotics. The approach that they have used allows them to use patients in whom it has been possible to verify that the responsible pathogens are susceptible to the antibiotics that have been instituted. Inappropriate empirical antibiotics and their exclusions. In the article by Chastre et al,2 the authors' conclusion begins with the words "Among patient who have received appropriate initial empirical treatment." The investigators did not overstate their findings and were very specific and circumspect in their conclusions. It is also not clear to me how they could have carried out such a study without the use of this particular design. I do agree, however, that the 61 patients excluded for "other reasons" should have been explained more carefully. Variability of empirical treatment regimens. The next issue addressed is that of the antimicrobial regimens themselves. Dr Beidas states that, although more than 90% of the study patients were initially treated with a β-lactam plus either a fluoroquinolone or an aminoglycoside, only 33% and 39% of patients received these drugs by days 8 and 15, respectively. However, the objective of the study was not to test specific drug regimens but to ascertain whether 8 days of treatment was as effective as 15 days of treatment for patients with microbiologically proven VAP. Chastre et al2 specifically state in the methods section that "investigators were strongly encouraged to convert this initial regimen into a narrow spectrum therapy based on culture results." The variability in empirical treatment is of no consequence therefore. The important issue is whether appropriate treatment was given initially and if alterations in treatment based on culture results were required. Inclusion of vancomycin in empirical therapy. Neither the ATS/IDSA guidelines nor the article by Chastre et al2 recommends vancomycin as part of routine initial empirical treatment nor should they. Dr Beidas claims that the rate of MRSA superinfection and recurrence is higher in the 15-day treatment group, but this is incorrect. The superinfection rates for MRSA were 6 (28.6%) of 21 and 5 (23.8%) of 21 in the 8- and 15-day groups, respectively, but the relapse rates were somewhat higher in the 15-day group, 3 (14.3%) of 21 versus 4 (19%) of 21. It is also worth noting that the patients treated for 8 days had significantly more mean antibiotic free days (13.1 vs 8.7; P < 0.001). For MRSA specifically, the number of antibiotic free days was 12.9 for the 8-day group and 4.9 for the 15-day group. Also significant were broad-spectrum antibiotic free days (18.4 vs 15.3 days; P = 0.1) in the 2 treatment groups. It is true that overall relapse rates were higher in the 8-day treatment arm than in the 15-day treatment arm (16.8% vs 11.3%) as were relapse rates for nonfermenting gram-negative organisms (32.8% vs 19%), but some important facts relating to recurrence and relapse have been left out by Dr Beidas. Chastre et al2 discussed the issue of recurrence with nonfermenters and also showed that, in those patients with recurrent infections, multiresistant pathogens were encountered less frequently in the 8-day group than in the 15-day group (42.1% vs 62%; P = 0.04). MRSA relapse. The issue of MRSA relapse has already been addressed. Dr Beidas' assumption that the MRSA relapse rate in the 15-day treatment arm is "likely to have skewed the results in favor of the 8 day treatment course achieving parity with the 15 day treatment course" is not in keeping with the strict standards that she seems to want to impose for reporting of RCTs and the practice of evidence-based medicine. New end point. Here, Dr Beidas suggests that a single primary end point should have been used rather than multiple ones. The single end point would be "success or failure." However, there are multiple ways of defining success; which definition would she suggest? She goes on to say that the primary end point of death due to any cause measured at 28 days lacks precision. I could not disagree more. Such an outcome measurement is easy to define and to identify and represents an ideal primary outcome measure for such a study. Given our understanding of the multiple physiological derangements resulting from serious pulmonary infections and the potential for myocardial infarction and stroke to occur in such settings, all-cause mortality seems a logical extension of the concept of attributable mortality when dealing with serious pneumonias.

The problem of the lack of a gold standard for the diagnosis of pneumonia is universally recognized and plagues both clinicians and investigators. Until such time that an ideal diagnostic test becomes available, physicians will have to resort to a combination of clinical, laboratory, and radiological methods to try to determine if the patient is likely to have pneumonia and how best to treat the patient.

In summary, contrary to Dr Beidas, I would argue that we should not change the recommendation as currently worded in the ATS/IDSA Guidelines regarding duration of treatment from level 1 to level 2. The recommendation based on the article by Chastre et al very nicely fulfills the requisite criteria for a level 1 recommendation. I would also point out that Dr Biedas' call for a continued use of the "time honored 14-21 day antibiotic regimen" is suspect at best. We have learned time and again that time honored is not synonymous with proven and if the carefully worded recommendation of the guidelines is followed, it may ultimately lead to less antimicrobial resistance, fewer adverse drug reactions, and lower costs associated with treatment.

Back to Top | Article Outline


1. Guidelines for the Management of Adults with Hospital-Acquired Pneumonia, Ventilator Associated Pneumonia. American Thoracic Society Documents. Am J Respir Crit Care Med. 2005;171:388-4163.

2. Chastre J, Wolff M, Fagon JY, et al. Comparison of 8 vs 15 days of antibiotic therapy for ventilator associated pneumonia in adults: a randomized trial. JAMA. 2003;290:2588-2598.

3. Dennesen P, van der Ven A, Kessels A, et al. Resolution of infectious parameters after antimicrobial therapy in patients with ventilator-associated pneumonia. Am J Respir Crit Care Med. 2001;163:1371-1375.

4. Luna C, Blanzaco D, Neiderman M, et al. Resolution of ventilator-associated pneumonia: prospective evaluation of the clinical pulmonary infection score as an early clinical predictor of outcome. Crit Care Med. 2003;31:676-682.

5. Singh N, Rogers P, Atwood C, et al. Short-course empiric antibiotic therapy for patients with pulmonary infiltrates in the intensive care unit. A proposed solution for indiscriminate antibiotic prescription. Am J Respir Crit Care Med. 2000;162:505-511.

6. Altman D, Schulz K, Moher D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med. 2001;134:663-694.

© 2008 Lippincott Williams & Wilkins, Inc.