Antimicrobial therapeutic determinants of outcomes from septic shock among patients with cirrhosis

Abstract

It is unclear whether practice-related aspects of antimicrobial therapy contribute to the high mortality from septic shock among patients with cirrhosis. We examined the relationship between aspects of initial empiric antimicrobial therapy and mortality in patients with cirrhosis and septic shock. This was a nested cohort study within a large retrospective database of septic shock from 28 medical centers in Canada, the United States, and Saudi Arabia by the Cooperative Antimicrobial Therapy of Septic Shock Database Research Group between 1996 and 2008. We examined the impact of initial empiric antimicrobial therapeutic variables on the hospital mortality of patients with cirrhosis and septic shock. Among 635 patients with cirrhosis and septic shock, the hospital mortality was 75.6%. Inappropriate initial empiric antimicrobial therapy was administered in 155 (24.4%) patients. The median time to appropriate antimicrobial administration was 7.3 hours (interquartile range, 3.2-18.3 hours). The use of inappropriate initial antimicrobials was associated with increased mortality (adjusted odds ratio [aOR], 9.5; 95% confidence interval [CI], 4.3-20.7], as was the delay in appropriate antimicrobials (aOR for each 1 hour increase, 1.1; 95% CI, 1.1-1.2). Among patients with eligible bacterial septic shock, a single rather than two or more appropriate antimicrobials was used in 226 (72.9%) patients and was also associated with higher mortality (aOR, 1.8; 95% CI, 1.0-3.3). These findings were consistent across various clinically relevant subgroups.

Conclusion: In patients with cirrhosis and septic shock, inappropriate and delayed appropriate initial empiric antimicrobial therapy is associated with increased mortality. Monotherapy of bacterial septic shock is also associated with increased mortality. The process of selection and implementation of empiric antimicrobial therapy in this high-risk group should be restructured.

Fig 1

Association of inappropriate antimicrobial therapy (A), hours of delay in effective antimicrobial therapy (B), and use of single versus combined antimicrobial therapy (C) with hospital mortality across various subgroups of patients using multivariate analyses. The following independent variables were entered in the model: APACHE II score, MELD score, immunocompromised (versus non-immunocompromised), bacteremia (versus no bacteremia), community-acquired (versus nosocomial), and culture- positive (versus culture-negative). The results are shown as aOR and 95% CI on a logarithmic scale.

Fig 2

aOR and 95% CI of hospital mortality (on logarithmic scale) by the time from the onset of hypotension to the antimicrobial therapy in hours. Adjustments were made for the following independent variables: APACHE II score, MELD score, immunocompromised (versus non-immunocompromised), bacteremia (versus no bacteremia) community-acquired (versus nosocomial), and culture-positive (versus culture-negative).