Oral Decontamination and Ventilator-Associated Pneumonia

Abstract & Commentary

By Uday Nanavaty, MD, Assistant Director, AICU, St Agnes Hospital, Baltimore, is Associate Editor for Critical Care Alert.

Dr. Nanavaty reports no financial relationships related to this field of study.

Synopsis: Aspiration of oropharyngeal secretions is considered one of the mechanisms of development of VAP. In this randomized trial of two oral decontamination medications, the authors show a significant reduction in the incidence of VAP.

Source: Koeman M, et al. Oral decontamination with chlorhexidine reduces the incidence of ventilator-associated pneumonia. Am J Respir Crit Care Med. 2006;173:1348-1355.

In this randomized-controlled trial, Koeman and associates wanted to determine the effect of two oral decontaminating gels, chlorhexidine 2% (CHX) or chlorhexidine 2% plus colistin 2% (CHX/COL), compared to a placebo, on the development of ventilator-associated pneumonia (VAP) in patients requiring more than 48 hours of mechanical ventilation. The study used an uncommon statistical technique called sequential survival analysis whereby the data were monitored throughout the study and positive (or negative) conclusions could be drawn at any time certain end points were reached without having to go through the full recruitment of all patients as anticipated by traditional statistical calculations. The authors had expected that if the therapies halved the incidence of VAP they would need more than 500 patients with two active treatment arms and one placebo arm; instead they were able to terminate the study at 385 patients since they met all the criteria of successful trial.

They enrolled all patients in their surgical or mixed ICUs (all in The Netherlands) who were older than 18 years of age, non-pregnant, and expected to need mechanical ventilation for more than 48 hours. Immunocompromised patients as well as patients whose physical condition did not allow the oral application of the study medications were excluded from the study. Within 24 hours of intubation, either the study medication or placebo gel was applied; using a gloved finger, 2 inches of paste was applied to the buccal mucosa every 6 hours. In patients in whom VAP was clinically suspected, it was defined as a new and persistent lung infiltrate on X-ray, associated with at least 3 of the following 4 criteria: rectal temperature > 38°C or less than 35.5°C; blood leukocytosis (> 10,000/mL ) and/or left shift or leukopenia (< 3000/mL); purulent tracheal aspirate; and, a positive semiquantitative culture from tracheal aspirate (at least 10,000 cfu/mL). All cases of VAP thus defined were assessed by blinded investigators and the investigators agreed with the diagnosis of VAP in all cases.

The trial was stopped after enrollment of 385 patients as the rate of VAP in the CHX group was found to be significantly lower as compared to placebo. Of 130 patients in the placebo group, 23 developed VAP (18%), compared to 13 cases (10%) out of 127 patients in the CHX group and 16 cases (13%) out of 128 patients in the CHX/COL group. The rate of VAP was statistically significantly lower in the CHX group and approaching statistical significance in the CHX/COL group. The colonization rate was reduced by both CHX and CHX/COL treatments. Interestingly enough, in spite of the lower rate of VAP, ICU length of stay and overall mortality were not reduced by either CHX or CHX/COL treatment. The trial, however, was not designed to detect a difference in either of these outcomes. Total numbers of days with antibiotics were also similar between placebo and CHX groups, whereas patients in the CHX/COL group needed fewer days of antibiotics as compared to placebo group.

CHX was effective in reducing oral colonization with Gram-positive organisms, whereas CHX/COL was effective in reducing oral colonization with both Gram-positive as well as Gram-negative organisms. Tracheal colonization rates were similar to placebo in the CHX group whereas they were lower in the CHX/COL group in the 4-8 day period. Subsequently, in the 9-12 day period, tracheal colonization rates were similar in all three groups. Average ventilator days were statistically similar in all groups (mean of 6.95 days in placebo, 9.16 days in CHX, and 8.52 days in CHX/COL).


In this study, application of 2% CHX gel to the buccal mucosa four times a day resulted in a significant reduction in VAP. Surprisingly, with a similar number of patients, the effects of application of CHX/COL gel did not reach statistical significance. The intervention in this study—application of gel to the buccal mucosa—is fairly innocuous and had a rather dramatic effect on rate of VAP. The authors have used a fairly standard definition of VAP and the three blinded intensivists agreed on all cases of VAP diagnosis. Routine measures such as keeping the head of bed elevated to 30° were employed. With a drop of nearly 50% in the incidence of VAP, this intervention seems something that most centers may consider adapting. There are several caveats however prior to making a relatively large change in practice.

VAP in this study was defined with quantitative culture. If oral secretions eventually trickle down the ET tube and result in VAP, application of treatment to buccal mucosa may certainly result in CHX and or COL entering the airway. Thus the secretions would be less likely to grow bacteria, even if true infection existed. Thus, the criteria of quantitative sputum culture may be in favor of the study drugs. We have no information on how many cases were suspected of suffering from VAP. Mild elevation of temperature, leukocytosis and sputum purulence are not too unusual in the ICU setting and hence, often the culture becomes the key finding in diagnosis. Also, in spite of a 50% reduction in the incidence of VAP, there was no impact on ventilator days, ICU length of stay, or mortality. Although none of these were primary end points, it seems that the VAP cases had almost no attributable morbidity.

Reduction in the rate of VAP is a must for most US hospitals since the introduction of ventilator bundles. How we diagnose VAP often varies. Depending on how high one sets the bar, the rate of VAP may be surprisingly low and, in that case, the 50% reduction in the VAP rate that this seemingly simple intervention promises may not be observed.