Early Gastrostomy Decreases Ventilator-Associated Pneumonia in Brain-Injured Patients

Abstract & Commentary

By David J. Pierson, MD, Editor, Director of Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle, is Editor for Critical Care Alert.

Synopsis: In a randomized, controlled trial, patients with stroke or head injury who required mechanical ventilation were less likely to develop ventilator-associated pneumonia if they underwent early percutaneous gastrostomy for nutritional support than if they continued to be fed via nasogastric tube.

Source: Kostadima E, et al. Early gastrostomy reduces the rate of ventilator-associated pneumonia in stroke or head injury patients. Eur Respir J. 2005;26:106-111.

This clinical trial from Larissa University Hospital in Greece tested the hypothesis that enteral nutrition via gastrostomy would reduce the incidence of ventilator-associated pneumonia (VAP) as compared to feeding via nasogastric tube. Forty one patients without known pulmonary disease who presented to the ICU with either stroke (25 patients) or head injury (16 patients), who had Glasgow Coma Scale scores less than 6 and required mechanical ventilation, were randomized to receive either a nasogastric tube (21 patients) or a percutaneous gastrostomy (20 patients) within 24 hours of admission. All patients had intracranial pressure monitoring devices and all received ceftriaxone 2 g every 12 h.

VAP was defined as a new and persistent pulmonary infiltrate on chest radiograph, plus at least 2 of the following: temperature > 38.3°C or an increase of > 1°C; leukocyte count > 10,000 and an increase of > 25% from baseline, or < 5000 and a decrease of > 25% from baseline; and purulent tracheal aspirate with > 25 neutrophils per high-power field on Gram stain. Bronchoalveolar lavage fluid colony counts > 10,000 was used to confirm the diagnosis in some but not all patients and was not a requirement for the diagnosis of VAP. Data on the development of VAP were recorded only during the first 3 weeks after intubation.

The 2 patient groups were well matched in terms of demographics, diagnosis, APACHE II scores, development of sinus opacification on CT, and selected comorbidities. Overall duration of mechanical ventilation and length of ICU stay were 37 and 38 days, respectively, in both groups, and mortality (4 and 6 patients in gastrostomy and control groups, respectively) also did not differ. However, VAP developed in 2 patients (10%) in the gastrostomy group as compared to 8 patients (38%) in the nasogastric tube group (P = 0.036), and this difference persisted after excluding those who did not complete the study because of extubation or death prior to 3 weeks.


VAP is the most lethal nosocomial infection in ICU patients,1 and its connection with the gastrointestinal tract is an area of intense current interest.2 Several studies have established the association between nasogastric tubes and VAP, and this study by Kostadima and colleagues supports the notion that not having a tube traversing the upper and lower esophageal sphincters is desirable in this respect. Other measures shown in randomized controlled trials to reduce the incidence of VAP include elevation of the head of the bed to 45 degrees, changing ventilator circuits as infrequently as possible, and avoidance of intubation through the use of noninvasive ventilation.

This study is relatively small. We are not told the Glasgow Coma Scale scores of the 2 patient groups, a potentially important omission if they indicated, for example, that the patients fed via nasogastric tubes had more severely impaired neurological function than those who received gastrostomies. However, the design and execution of the study appear sound in other respects. Whether its findings in severely brain-injured patients are generalizable to other ICU patients is unknown, and additional investigations of this approach to preventing VAP will be important.

Although no important complications from percutaneous gastrostomy tube insertion were noted in this study, placement of these tubes in critically ill patients is not always so benign, and I have seen serious and even fatal complications resulting from the procedure. Of note, in keeping with the findings of studies of several other interventions to reduce the incidence of VAP, no differences in ICU length of stay or hospital mortality were found despite a significant reduction in the number of patients who developed VAP.


  1. Chastre J. Conference summary: ventilator-associated pneumonia. Respir Care. 2005;50:975-983.
  2. Kallet RH, Quinn TE. The gastrointestinal tract and ventilator-associated pneumonia. Respir Care. 2005;50:910-921.