Non-pulmonary infections but not specific pathogens are associated with increased risk of AKI in septic shock

Abstract

Introduction: Little is known regarding the relationship between the anatomic infection site and etiologic pathogen with the occurrence of acute kidney injury (AKI) in severe infections. We set out to determine the association between the site of infection, type of pathogen in septic shock and occurrence of AKI.

Methods: Using a large, international multicenter database that included data from 28 academic and community hospitals, we retrospectively analyzed adult (age >18 years) cases of septic shock occurring between January 1996 and December 2008. Early acute kidney injury (AKI) was classified by the RIFLE criteria at or within 24 h of shock diagnosis. Multivariate logistic regression was used to determine the association between the infection site/microbial pathogen and occurrence of AKI. Analyses were adjusted for demographics, illness severity, comorbidities and intensive care unit interventions (partial adjustment) ± site of infection and microbial pathogen (full adjustment).

Results: After exclusions, 4,493 cases from potentially eligible patients in the database were included in the analytic cohort of whom 3,298 (73.4 %) experienced AKI. Patients with AKI were older (p < 0.0001), had a higher mean Acute Physiology and Chronic Health Evaluation score (p < 0.0001), and had greater laboratory and hemodynamic abnormalities. The most common site of infection among septic shock patients with AKI was the lung (34.5 %), followed by gastrointestinal (GI) (26.2 %) and urinary (15.3 %) sources. Likewise, the most common infecting organism among septic shock patients with AKI was E. coli (23.9 %) followed by S. aureus (GI) (16.1 %) and other enterobacteriaceae (15.7 %). There was a large degree of variability in the occurrence of AKI based on the site of infection and the pathogen in unadjusted analysis (p < 0.0001), which persisted with partial (excluding infection site and microbial pathogen grouping) adjustment (p < 0.0001). Fully adjusted multivariate analysis showed significant variations in AKI only in relation to the anatomic source of infection, with non-pulmonary infections having higher risk than pulmonary infections. The pathogen group/pathogen had no significant independent impact on AKI.

Conclusion: This study demonstrates that the presence of septic AKI varies significantly based on the site of infection but not the type of causative organism.