The benefit of appropriate empirical antibiotic treatment in patients with bloodstream infection

Objectives. To test whether empirical antibiotic treatment that matches the in vitro susceptibility of the pathogen (appropriate treatment) improves survival in patients with bloodstream infections; and to measure the improvement. Design. Observational, prospective cohort study. Setting. University hospital in Israel. Subjects. All patients with bloodstream infections detected during 1988-94. Interventions. None. Main outcome measures. In-hospital fatality rate and length of hospitalization. Results. Out of 2158 patients given appropriate empirical antibiotic treatment, 4 36 (20%) died, compared with 432 of 1255 patients (34%) given inappropriate treatment (P = 0.0001). The median durations of hospital stay for patients who survived were 9 days for patients given appropriate treatment and 11 days for patients given inappropriate treatment. For patients who died, the median durations were 5 and 4 days, respectively (P < 0.05), for both comparisons. In a stratified analysis, fatality was higher in patients given inappropriate treatment than in those given appropriate treatment in all strata but two: patients with infections caused by streptococci other than Streptococcus gr. A and Streptoccocus pneumoniae (odds ratio (OR) of 1.0, 95% confidence interval (95% CI) 0.4-2.5); and hypothermic patients (OR = 0.9, 95% CI = 0.3- 2.4). Even in patients with septic shock, inappropriate empirical treatment was associated with higher fatality rate (OR = 1.6, 95% CI = 1.0-2.7). The highest benefit associated with appropriate treatment was observed in paediatric patients (OR = 5.1, 95% CI = 2.4-10.7); intra-abdominal infections (OR = 3.8, 95% CI = 2.0-7.1); infections of the skin and soft tissues (OR = 3.1, 95% CI = 1.8-5.6); and infections caused by Klebsiella pneumoniae (OR = 3.0, 95% CI = 1.7-5.1) and S. pneumoniae (OR = 2.6, 95% C = 1.1-5.9). On a multivariable logistic regression analysis, the contribution of inappropriate empirical treatment to fatality was independent of other risk factors (multivariable adjusted OR = 1.6, 95% CI = 1.3-1.9). Conclusion. Appropriate empirical antibiotic treatment was associated with a significant reduction in fatality in patients with bloodstream infection.