Impact of appropriate antimicrobial therapy on patient outcomes and antimicrobial use: a sub analysis of the DIANA Study Dataset
摘要
Although inappropriate therapy has been consistently associated with adverse outcomes, the magnitude and consistency of the benefit associated with appropriate empiric therapy in critically ill patients remain uncertain. We aimed to quantify the prevalence of appropriate empiric antimicrobial therapy and evaluate its association with outcomes and antimicrobial exposure in a large international ICU cohort.
MethodsThis predefined sub-analysis of the DIANA study included adult ICU patients receiving empirical antimicrobials for suspected or confirmed bacterial infection. Only patients with microbiologically confirmed infections were analyzed, and therapy was classified as appropriate if at least one agent demonstrated in vitro activity against the identified pathogen. Associations with 28-day mortality and antimicrobial-free days were assessed using multivariable logistic and Cox regression models. Pre-specified interaction analyses explored effect modification by disease severity and diagnostic certainty.
ResultsOf 845 patients with microbiologically confirmed infections, 87.7% received appropriate empirical antimicrobial therapy. Compared with inappropriate therapy, appropriate therapy was associated with significantly lower ICU mortality and longer 28-day antimicrobial-free days and mechanical ventilation-free days. After multivariable adjustment, appropriate therapy remained independently associated with reduced 28-day mortality [adjusted odds ratio (OR) 1.83, 95% confidence interval 1.11–3.06, p = 0.02; hazard ratio (HR) 1.51, 95% CI 1.03–2.21, p = 0.035]. Effect-modification analyses demonstrated that the survival benefit of appropriate therapy was consistent across levels of diagnostic certainty and was most pronounced in patients with moderate illness severity (SOFA 3–9).
ConclusionIn critically ill ICU patients, appropriate empirical antimicrobial therapy is independently associated with reduced 28-day mortality rates.
Visual abstract