Preventing Ventilator-Associated Pneumonia: Looking Beyond the Bundles

Abstract

TO THE EDITOR—We read with interest the article by Bouadma et al [1] on prevention of ventilator-associated pneumonia (VAP) [2]. In their article, Bouadma et al [1] described the effect of a “bundle” approach for VAP prevention [3]. The methods that they used to implement these interventions included a multifaceted behavioral approach, education, and feedback on performance, among others [3]. This article indicates that the bundled approach to prevention improves the opportunity for success, even when compliance with each individual intervention is not optimal. VAP prevention bundles are now an important part of VAP prevention guidelines [4]. There are several limitations in this article that deserved to be acknowledged. The generalizability of the findings to the United States is limited. First, in the United States, the most commonly used VAP surveillance definition is the Centers for Disease Control and Prevention definition, which is based on clinical criteria and may be more inclusive than the definition used in the Bouadma study [1], which required microbiological confirmation [5, 6]. It is unclear whether bronchoscopic confirmation was attempted every time that a patient presented with signs and symptoms suggestive of VAP, or how the receipt of prior antimicrobial agents changed the cut off values mentioned by the authors in describing their methodology. Furthermore, Bouadma et al [1] did not describe previous antibiotic use. In addition, the authors did not address the possible presence of tracheobronchitis or other pulmonary infections, which may have influenced the diagnosis of VAP [1]. The authors also included organisms not commonly associated with VAP, such as coagulase-negative Staphylococcus, Enterococcus species, and fungus. Given that microbiological confirmation was required for a VAP diagnosis, these organisms should have been excluded. The authors preferred the use of proton pump inhibitors (PPIs) for the prevention of stress ulcers, which have been associated with increased rates of VAP. The proportion of patients who received PPI should have been mentioned [1, 4]. Regarding the reported median improvement in duration of hospital stay, it would be interesting to know whether the difference remains significant when the mean duration of hospital stay is compared between time periods [1]. Finally, one of the conclusions described by the authors was that “Our results strongly suggest that the goal of eliminating VAP from the ICU may be unreasonable” [1]. We concur with the authors that the long-term elimination of all VAP cases is unreasonable; however, this study did not implement a comparison intervention, such as one evaluating the use of silver-impregnated endotracheal tubes (ETT), ETT with subglottic secretion suction ports, better-engineered endotracheal cuffs, or even “selective” gastrointestinal tract decontamination [7, 8]. We argue that the success of VAP prevention intervention bundles requires the implementation of all of the interventions concurrently, but in the Bouadma study, individual intervention compliance was mostly <90% [1]. Additionally, there was a high patient-staff ratio in the study [1]. Low nurse-to-patient ratios have been associated with elevated rates of hospital-acquired infections [9]. We conclude that a lack of evidence does not mean negative evidence. Additional studies should be performed to determine whether additional interventions could further contribute to the success of bundled VAP prevention [7].