World-wide variation in incidence of Acinetobacter associated ventilator associated pneumonia: a meta-regression

Abstract

BACKGROUND: Acinetobacter species such as Acinetobacter baumanii are of increasing concern in association with ventilator associated pneumonia (VAP). In the ICU, Acinetobacter infections are known to be subject to seasonal variation but the extent of geographic variation is unclear. The objective here is to define the extent and possible reasons for geographic variation for Acinetobacter associated VAP whether or not these isolates are reported as Acinetobacter baumanii. METHODS: A meta-regression model of VAP associated Acinetobacter incidence within the published literature was undertaken using random effects methods. This model incorporated group level factors such as proportion of trauma admissions, year of publication and reporting practices for Acinetobacter infection. RESULTS: The search identified 117 studies from seven worldwide regions over 29 years. There is significant variation in Acinetobacter species associated VAP incidence among seven world-wide regions. The highest incidence is amongst reports from the Middle East (mean; 95 % confidence interval; 8.8; 6 · 2-12 · 7 per 1000 mechanical ventilation days) versus that from North American ICU's (1 · 2; 0 · 8-2 · 1). There is a similar geographic related disparity in incidence among studies reporting specifically as Acinetobacter baumanii. The incidence in ICU's with a majority of admission being for trauma is >2.5 times that of other ICU's. CONCLUSION: There is greater than fivefold variation in Acinetobacter associated VAP among reports from various geographic regions worldwide. This variation is not explainable by variations in rates of VAP overall, admissions for trauma, publication year or Acinetobacter reporting practices as group level variables.