BACKGROUND/AIMS: Vascular access-related bloodstream infection (BSI) is frequent among patients undergoing hemodialysis increasing their morbidity and mortality, but its occurrence across various dialysis centre types is not known. The aims of this study were to describe the incidence rates and assess the variability in BSI risk between dialysis centre types and other centre-level variables. METHODS: We conducted a retrospective cohort study of 621 patients initiating hemodialysis in 7 Canadian dialysis centres. Cox regression models, where access type was continuously updated, were used to identify predictors of BSI occurrence. RESULTS: During follow-up of the cohort (median age 68.1 years, 41.7% female, and 76.7% initiating with a central venous catheter, CVC), 73 patients had a BSI (rate: 0.21/1000 person-days). The BSI risk was not different in First Nation units (adjusted relative risk: 0.47, 95% confidence interval: 0.06-3.72) and teaching hospitals (1.33, 0.70-2.54) compared to community hospitals. No other centre-related variables were associated with the risk of BSI. CONCLUSION: We did not find differences in the BSI risk among dialysis unit types, or any other centre-related variables. The rates of BSI in our population were lower than those observed in other settings, but the high proportion of patients using CVCs is concerning. Copyright 2009 S. Karger AG, Basel.
BACKGROUND/AIMS: Vascular access-related bloodstream infection (BSI) is frequent among patients undergoing hemodialysis increasing their morbidity and mortality, but its occurrence across various dialysis centre types is not known. The aims of this study were to describe the incidence rates and assess the variability in BSI risk between dialysis centre types and other centre-level variables. METHODS: We conducted a retrospective cohort study of 621 patients initiating hemodialysis in 7 Canadian dialysis centres. Cox regression models, where access type was continuously updated, were used to identify predictors of BSI occurrence. RESULTS: During follow-up of the cohort (median age 68.1 years, 41.7% female, and 76.7% initiating with a central venous catheter, CVC), 73 patients had a BSI (rate: 0.21/1000 person-days). The BSI risk was not different in First Nation units (adjusted relative risk: 0.47, 95% confidence interval: 0.06-3.72) and teaching hospitals (1.33, 0.70-2.54) compared to community hospitals. No other centre-related variables were associated with the risk of BSI. CONCLUSION: We did not find differences in the BSI risk among dialysis unit types, or any other centre-related variables. The rates of BSI in our population were lower than those observed in other settings, but the high proportion of patients using CVCs is concerning. Copyright 2009 S. Karger AG, Basel.