Prevention of Intensive Care Unit-acquired Pneumonia

Michael Klompas, MD, MPH


Semin Respir Crit Care Med. 2019;40(4):548-557. 

In This Article

Oral Care With Chlorhexidine

Oral care with chlorhexidine has come under scrutiny in recent years because of a series of studies associating oral chlorhexidine with a possible increased risk of mortality and ventilator-associated events.[84] This signal has been noted on meta-analyses of randomized trials,[73,85] observational analyses of associations between ventilator bundle components and outcomes,[31,86,87] and in one hospital-wide observational analysis of prescribed medications.[88] In addition, doubt has been cast on whether oral care with chlorhexidine truly prevents VAP. While meta-analyses of randomized trials have reported that oral care with chlorhexidine lowers VAP rates by approximately 30%, this signal is only evident in open-label studies.[85] If one restricts the meta-analysis to double-blind studies, the signal is diminished and no longer significant (yet another reminder of the risk of bias in VAP prevention studies and the importance of looking at objective outcomes for corollary evidence of benefit).[85]

Despite the array of independent signals suggesting possible harm, there are important limitations to the existing evidence. First, no single randomized trial has reported an association between oral care with chlorhexidine and higher mortality rates. Second, the meta-analyses that reported a possible increase in mortality combined studies performed in different populations with different preparations of chlorhexidine. Third, all observational studies are at risk of confounding. Fourth, the hospital-wide analysis suggesting an association between chlorhexidine and higher mortality rates specifically did not find this association amongst ventilated patients and may not have adequately controlled for confounding by indication insofar as oral chlorhexidine was selectively prescribed for frail and dependent patients.[88] Fifth, the mechanism by which oral care with chlorhexidine may increase ventilator-associated events and mortality risk is unclear. Investigators speculate that it may be because some patients aspirate the antiseptic which in turn might precipitate acute lung injury and the acute respiratory distress syndrome; however, none of the randomized trials of chlorhexidine have specifically evaluated this hypothesis.[84] Note that the recent publication of a large cluster randomized trial of different oral decontamination strategies (the R-GNOSIS trial) failed to resolve the question of oral chlorhexidine's safety.[82] The trial reported that 6 months of routine care with 2% oral chlorhexidine was associated with an adjusted 28-day mortality hazard ratio of 1.13 (95% CI: 0.68–1.88) compared with baseline care; however, baseline care in 11 of the 13 study ICUs included oral care with 0.12% chlorhexidine. Thus the trial provided a comparison of high-concentration chlorhexidine versus low-concentration chlorhexidine but not a comparison of oral chlorhexidine versus placebo. The study did document a high rate of oral mucosal reactions to 2% oral chlorhexidine but these resolved when the investigators switched to a 1% chlorhexidine preparation.[89]

At the end of the day, the data suggesting that oral care with chlorhexidine may pose a safety risk are far from certain. At the same time, however, the evidence that oral care with chlorhexidine prevents VAP is equally tenuous (no signal in the vast majority of individual trials, no signal on meta-analysis of double-blinded studies).[85] Faced with these two uncertainties, the better part of valor is to follow the precautionary principal: it is best for now to remove chlorhexidine from oral care regimens given the absence of clear evidence of benefit and the possible suggestion of harm.[90] Unfortunately, this recommendation does leave hospitals in a quandary regarding whether and what to use instead of chlorhexidine. There is no clear answer to this question since chlorhexidine is by far the best studied oral antiseptic in ventilated patients. There are very little data on other oral antiseptics and some concern that aspiration triggering acute lung injury could be a class effect for oral antiseptics rather than an isolated effect of chlorhexidine alone. Indeed, a randomized trial of one possible alterative (povidone-iodine) documented higher rates of acute respiratory distress syndrome in patients randomized to povidone-iodine versus placebo.[91] For the time being, toothbrushing and oral care with sterile water alone may be the most prudent course until more data are available.