Objective: Urinary tract infections (UTIs) are an important source of secondary healthcare-associated bloodstream infections (BSIs), where a potential for prevention exists. This study describes the epidemiology of BSIs secondary to a urinary source (U-BSIs) in the province of Québec and predictors of mortality.
Design: Dynamic cohort of 9,377,830 patient-days followed through a provincial voluntary surveillance program targeting all episodes of healthcare-associated BSIs occurring in acute care hospitals.
Setting: Sixty-one hospitals in Québec, followed between April 1, 2007, and March 31, 2010.
Participants: Patients admitted to participating hospitals for 48 hours or longer.
Methods: Descriptive statistics were used to summarize characteristics of U-BSIs and microorganisms involved. Wilcoxon and χ(2) tests were used to compare U-BSI episodes with other BSIs. Negative binomial regression was used to identify hospital characteristics associated with higher rates. We explored determinants of mortality using logistic regression.
Results: Of the 7,217 reported BSIs, 1,510 were U-BSIs (21%), with an annual rate of 1.4 U-BSIs per 10,000 patient-days. A urinary device was used in 71% of U-BSI episodes. Identified institutional risk factors were average length of stay, teaching status, and hospital size. Increasing hospital size was influential only in nonteaching hospitals. Age, nonhematogenous neoplasia, Staphylococcus aureus, and Foley catheters were associated with mortality at 30 days.
Conclusion: U-BSI characteristics suggest that urinary catheters may remain in patients for ease of care or because practitioners forget to remove them. Ongoing surveillance will enable hospitals to monitor trends in U-BSIs and impacts of process surveillance that will be implemented shortly.