No abstract

A notable portion of deaths in bloodstream infections (BSI) have previously been shown to occur within 2 days after taking the first positive blood culture specimen. The aim of this study was to analyse patients' characteristics and causative pathogens of BSIs, leading to early deaths in order to explore possibilities for prevention. Patients with BSI in Helsinki and Uusimaa region (population = 1.5 million) in 2007 were identified from the National Infectious Disease Register (n = 2181) and their deaths within 2 days after the first positive blood culture from the Population Information System (n = 76). Of the early fatal BSIs, 42 (55%) were community-acquired (CA-BSI) and 34 (45%) healthcare-associated (HA-BSI). Charlson comorbidity index was moderate-to-high (index ≥ 3) in 71% of HA-BSIs and 60% of CA-BSIs. The most common pathogens in CA-BSIs were Streptococcus pneumoniae (29%) and Escherichia coli (24%) and in HA-BSIs Pseudomonas aeruginosa (24%) and Staphylococcus aureus (18%). The respiratory tract (50%) was the most common focus of infection. Empiric antimicrobial treatment was more often appropriate in CA-BSIs vs HA-BSIs (81% vs 41%, p < 0.001), but treatment delays were longer in CA-BSIs. The majority of the BSI patients who died early had severe comorbidities. S. pneumoniae accounted for one third of CA-BSIs, highlighting the potential role of pneumococcal vaccines in prevention. Early recognition of BSI and its origin (CA-BSI vs HA-BSI) is crucial. Continuous surveillance data on causative microbes and resistance trends in hospitals is needed to propose guidelines for empiric antimicrobial therapy of BSIs.

Few studies covering all patient groups and specialties are available regarding the outcome of nosocomial bloodstream infections (BSI). We analyzed the role of patient characteristics and causative pathogens of nosocomial BSIs reported by the hospitals participating in national surveillance in Finland during 1999-2014, in terms of outcome, with particular interest in those leading to death within 2 days (i.e. early death). National nosocomial BSI surveillance was laboratory-based and hospital-wide. Data on nosocomial BSIs was collected by infection control nurses, and dates of death were obtained from the national population registry with linkage to national identity codes. A total of 17,767 nosocomial BSIs were identified; 557 BSIs (3%) were fatal within 2 days and 1150 (6%) within 1 week. The 1-month case fatality was 14% (2460 BSIs), and 23% of the deaths occurred within 2 days and 47% within 1 week. The patients who died early were older than those who survived > 28 days, and their BSIs were more often related to intensive care. Gram-positive bacteria caused over half of the BSIs of patients who survived, whereas gram-negative bacteria, especially Pseudomonas aeruginosa, caused more often BSIs of patients who died early, and fungi BSIs of patients who died within 1 week. A significant portion of patients with nosocomial BSIs died early, which underlines the importance of rapid recognition of BSI. Hospital-wide surveillance data of causative pathogens can be utilized when composing recommendations for empiric antimicrobial treatment in collaboration with clinicians, as well as when promoting infection prevention.