Outcomes in consecutive hospitalized UK patients with bacteraemia or fungaemia caused by medical devices and procedures

SummaryBackgroundThere is a lack of outcome data on hospitalized patients with bacteraemia or fungaemia caused by medical devices or procedures.AimTo determine the association between death and bacteraemia or fungaemia caused by medical devices and procedures.MethodsFrom December 2012 to November 2013, demographic, clinical, and microbiological data were collected on consecutive inpatients with bacteraemia or fungaemia. Multivariate analysis, using generalized estimating equations, was used to define the association.FindingsA total of 594 bacteraemic or fungaemic episodes occurred in 500 patients. Among patients with episodes caused by medical devices or procedures, 7-day and 30-day mortality were 7/167 [4.2%; 95% confidence interval (CI): 1.7–8.4] and 12/167 (7.2%; CI: 3.8–12.2) respectively. After adjustment, the association between death and bacteraemic or fungaemic episodes related to medical devices and procedures was non-significant as 7- and 30-day mortality odds ratios (OR) were 2.86 (95% CI: 0.80–10.12) and 1.72 (95% CI: 0.71–4.16) respectively. The difference between 30-day mortality associated with Escherichia coli and Staphylococcus aureus bacteraemia demonstrated a trend towards significance [6/47 (12.8%; CI: 4.8–25.7) vs 0/24; P = 0.067]. Thirty-day mortality associated with bacteraemia or fungaemia in patients with urinary catheter infections (often E. coli-associated) was significantly higher than intravascular device-associated infections (often S. aureus-associated) [4/51 (7.8%; 95% CI: 2.2–18.8) vs 1/62 (1.6%; 95% CI: 0.0–8.7); P = 0.028].ConclusionSpecial attention is required to prevent medical device- or procedure-related bacteraemia caused by E. coli. Greater attention should be placed on preventing infections caused by urinary catheters.