Worldwide variation in Pseudomonas associated ventilator associated pneumonia. A meta-regression

Abstract

Purpose: The objective here is to define the extent and possible reasons for geographic variation in Pseudomonas aeruginosa associated with Ventilator Associated Pneumonia (VAP).

Methods: A random effects meta-regression model of Pseudomonas associated VAP incidence within the published literature was undertaken incorporating group level factors such trauma admission, year of publication and use of bronchoscopic sampling towards VAP diagnosis.

Results: Pseudomonas associated VAP incidence was reported in 162 studies from seven worldwide regions published over 30 years. The highest incidence is among reports from the Middle East and Mediterranean being respectively 6.8; 5.2-9.0 (mean; 95% CI) and 6.9; 5.4-8.8 per 1000 mechanical ventilation (MV) days, versus that from North American ICU's (3.7; 2.3-5.9). In a meta-regression model, the variation in the incidence of Pseudomonas associated VAP was not significantly associated with bronchoscopic sampling in the diagnosis of VAP (p = 0.12) nor with admission to a trauma ICU (p = 0.13).

Conclusion: Pseudomonas associated VAP incidence among reports from six geographic regions worldwide varies by less than twofold with some decline by year of publication. Trauma ICU admission is a significant factor underlying variations in incidence of VAP overall but not Pseudomonas associated VAP.

Keywords: Geographic variation; Intensive care unit; Pseudomonas aeruginosa; Ventilator associated pneumonia.