Association between Source of Infection and Hospital Mortality in Patients Who Have Septic Shock

• Published in: American journal of respiratory and critical care medicine 2014-05, Vol.189 (10), p.1204-1213
• Main Authors: LELIGDOWICZ, Aleksandra, DODEK, Peter M, NORENA, Monica, WONG, Hubert, KUMAR, Aseem, KUMAR, Anand
• Format: Article
• Language: English
• Subjects: Abscesses; Aged; Anesthesia. Intensive care medicine. Transfusions. Cell therapy and gene therapy; Anti-Infective Agents - therapeutic use; Antimicrobial agents; Biological and medical sciences; Canada - epidemiology; Cardiovascular disease; Cohort analysis; Cohort Studies; Critical Illness - mortality; Cross Infection - diagnosis; Cross Infection - drug therapy; Cross Infection - microbiology; Cross Infection - mortality; Emergency and intensive care: infection, septic shock; Female; Gram-positive bacteria; Hospital Mortality; Hospitals; Humans; Hypotension; Intensive care; Intensive care medicine; Intensive Care Units; Ischemia; Male; Medical sciences; Middle Aged; Mortality; Nosocomial infections; Pathogens; Peritonitis; Physiology; Pneumology; Retrospective Studies; Risk Factors; Saudi Arabia - epidemiology; Sepsis; Shock, Septic - diagnosis; Shock, Septic - drug therapy; Shock, Septic - microbiology; Shock, Septic - mortality; United States - epidemiology; Urinary tract diseases; Urinary tract infections; Urogenital system; Variables
• ISSN: 1073-449X; 1535-4970
• DOI: 10.1164/rccm.201310-1875oc

Mortality caused by septic shock may be determined by a systemic inflammatory response, independent of the inciting infection, but it may also be influenced by the anatomic source of infection. To determine the association between the anatomic source of infection and hospital mortality in critically ill patients who have septic shock. This was a retrospective, multicenter cohort study of 7,974 patients who had septic shock in 29 academic and community intensive care units in Canada, the United States, and Saudi Arabia from January 1989 to May 2008. Subjects were assigned 1 of 20 anatomic sources of infection based on clinical diagnosis and/or isolation of pathogens. The primary outcome was hospital mortality. Overall crude hospital mortality was 52% (21-85% across sources of infection). Variation in mortality remained after adjusting for year of admission, geographic source of admission, age, sex, comorbidities, community- versus hospital-acquired infection, and organism type. The source of infection with the highest standardized hospital mortality was ischemic bowel (75%); the lowest was obstructive uropathy-associated urinary tract infection (26%). Residual variation in adjusted hospital mortality was not explained by Acute Physiology and Chronic Health Evaluation II score, number of Day 1 organ failures, bacteremia, appropriateness of empiric antimicrobials, or adjunct therapies. In patients who received appropriate antimicrobials after onset of hypotension, source of infection was associated with death after adjustment for both predisposing and downstream factors. Anatomic source of infection should be considered in future trial designs and analyses, and in development of prognostic scoring systems.