Purpose: Sepsis treatment protocols emphasize source control with empiric antibiotics and fluid resuscitation. Previous reviews have examined the impact of infection site and specific pathogens on mortality from sepsis; however, no recent review has addressed the infection site. This review focuses on the impact of infection site on hospital mortality among patients with sepsis.
Methods: The PubMed database was searched for articles from 2001 to 2014. Studies were eligible if they included (1) one or more statistical models with hospital mortality as the outcome and considered infection site for inclusion in the model and (2) adult patients with sepsis, severe sepsis, or septic shock. Data abstracted included stage of sepsis, infection site, and raw and adjusted effect estimates. Nineteen studies were included. Infection sites most studied included respiratory (n = 19), abdominal (n = 19), genitourinary (n = 18), and skin and soft tissue infections (n = 11). Several studies found a statistically significant lower mortality risk for genitourinary infections on hospital mortality when compared to respiratory infections.
Conclusion: Based on studies included in this review, the impact of infection site in patients with sepsis on hospital mortality could not be reliably estimated. Misclassification among infections and disease states remains a serious possibility in studies on this topic.
Keywords: infection; mortality; review; sepsis.