John F McNamara1, Patrick N A Harris2, Mark D Chatfield3, David L Paterson4. 1. University of Queensland Centre for Clinical Research, Building 71/918 Royal Brisbane & Women's Hospital Campus, Herston, 4029, Brisbane, QLD, Australia; The Prince Charles Hospital, 627 Rode Road, Chermside, 4032, Brisbane, QLD, Australia. Electronic address: mcnamarajfg@gmail.com. 2. University of Queensland Centre for Clinical Research, Building 71/918 Royal Brisbane & Women's Hospital Campus, Herston, 4029, Brisbane, QLD, Australia; Central Microbiology, Pathology Queensland, Royal Brisbane and Women's Hospital, Cnr Butterfield St and Bowen Bridge Rd, Herston, 4029, QLD, Australia. 3. University of Queensland Centre for Clinical Research, Building 71/918 Royal Brisbane & Women's Hospital Campus, Herston, 4029, Brisbane, QLD, Australia. 4. University of Queensland Centre for Clinical Research, Building 71/918 Royal Brisbane & Women's Hospital Campus, Herston, 4029, Brisbane, QLD, Australia; Department of Infectious Diseases, Royal Brisbane and Women's Hospital, Cnr Butterfield St and Bowen Bridge Rd, Herston, 4029, QLD, Australia.
Abstract
OBJECTIVES: Understand the long-term mortality, risk of readmission for sepsis and cause of death following a gram-negative bloodstream infection (GN-BSI). METHODS: This was a propensity-matched study using data linkage of Queensland hospital data, Australia. GN-BSIs were collected from 2005 to 2010 and matched 1:1 to hospital admissions without BSI for age, gender, year of culture collection, frequency of admissions in the prior year and Charlson-Deyo Comorbidity score and each comorbidity within the Charlson-Deyo score. Readmissions for sepsis, mortality and causes of death were evaluated. RESULTS: Cases of GN-BSI were propensity-matched 1:1 to culture-negative hospital admissions (n = 14016). Readmissions for sepsis were higher in the GN-BSI cohort from 91 to 365 days (P < 0.001) and in the four subsequent years (P < 0.001). The five-year survival in the GN-BSI cohort was 52% versus 65% in the culture-negative cases (P < 0.001). Infection was only a common underlying cause of death within the first 90 days. Sepsis was the most common contributing cause of death (CCOD) for the two years following index culture in the GN-BSI cohort. CONCLUSIONS: Compared to a similarly vulnerable group of hospital attendees, GN-BSI had higher mortality and demonstrated a persistent long-term risk of readmission for sepsis and sepsis as a CCOD.
OBJECTIVES: Understand the long-term mortality, risk of readmission for sepsis and cause of death following a gram-negative bloodstream infection (GN-BSI). METHODS: This was a propensity-matched study using data linkage of Queensland hospital data, Australia. GN-BSIs were collected from 2005 to 2010 and matched 1:1 to hospital admissions without BSI for age, gender, year of culture collection, frequency of admissions in the prior year and Charlson-Deyo Comorbidity score and each comorbidity within the Charlson-Deyo score. Readmissions for sepsis, mortality and causes of death were evaluated. RESULTS: Cases of GN-BSI were propensity-matched 1:1 to culture-negative hospital admissions (n = 14016). Readmissions for sepsis were higher in the GN-BSI cohort from 91 to 365 days (P < 0.001) and in the four subsequent years (P < 0.001). The five-year survival in the GN-BSI cohort was 52% versus 65% in the culture-negative cases (P < 0.001). Infection was only a common underlying cause of death within the first 90 days. Sepsis was the most common contributing cause of death (CCOD) for the two years following index culture in the GN-BSI cohort. CONCLUSIONS: Compared to a similarly vulnerable group of hospital attendees, GN-BSI had higher mortality and demonstrated a persistent long-term risk of readmission for sepsis and sepsis as a CCOD.