The implementation of sepsis bundles on the outcome of patients with severe sepsis or septic shock in intensive care units

Background: The goal of the study was to implement sepsis bundles and examine the effect on patients with severe sepsis or septic shock in intensive care units (ICUs). Methods: All patients with severe sepsis or septic shock admitted to the 13-bed ICU were included. Sepsis bundles were implemented within 24 hours after admission. The implementation of sepsis bundles was categorized into preintervention (January to April 2010), education (July to October 2010), operational (November to December 2010), and postintervention (January to April 2011) phases. Comparison of bundle compliance and outcome between each phase were examined. We also found mortality predictors between preintervention and postintervention phases. Results: There were 164 patients included in the study. Compared with the preintervention phase, the bundle compliance of each phase (education, operation, and postintervention separately) was higher (43.3%, 84.6%, and 79.2%, respectively, vs. 20.0%, p < 0.05), the hospital mortality was lower (10.0%, 23.1%, and 24.5%, respectively, vs. 43.6%, p < 0.05). Under multivariate analyses, the predictors for mortality between the preintervention and postintervention phases were: lactate at ICU (odds ratio [OR] 2.212), urinary tract infection (OR 0.026), and postintervention (OR 0.239). Conclusion: Implementation of modified sepsis bundles was successful in changing sepsis treatment behavior and was associated with a substantial reduction in hospital mortality and trends of decreased hospital expenditure. Factors improved hospital mortality, as lower lactate levels at ICU, urinary tract infection, and postintervention. The proposed intervention is generally applicable to achieve similar improvements.