![[ Visit the Fifth Decennial online ]](images/banner.gif)
Friday, March 19, 2010
Grand Hall (Hyatt Regency Atlanta)
Background: E. coli is the most common cause of bacteriuria and urinary tract infections & may lead to bacteremia. It is unclear which bacteriuric patients should have blood cultures drawn (blood cx) & if identifying bacteremia alters outcome.
Objective: To compare patients (pts) with E. coli bacteriuria based upon whether blood cultures were obtained, and subsequent patient outcomes.
Methods: We conducted a 7-week prospective cohort study of adult, hospitalized patients with E. coli bacteriuria at a tertiary-care hospital. Blood cx were at the discretion of the treating physician & had to occur within ±1 day of the bacteriuria. Asymptomatic bacteriuria was defined as absence of urinary symptoms, & pyelonephritis as flank pain/tenderness and/or fever.
Results: 113 pts had E. coli bacteriuria. The median age was 59.2 (±19.7) years; 83 (74%) were female. 81 (72%) had community-acquired bacteriuria. 40 (35%) had asymptomatic bacteriuria, 12 (11%) had cystitis, 41 (36%) had pyelonephritis, and 20 (18%) were unable to report urinary symptoms (e.g., altered mental status, intubation). 16 (14%) pts had catheter-associated bacteriuria. In 54 (48%) pts a blood cx was obtained and 9 of 54 (17%) were bacteremic.
Bacteriuric pts who had blood cx drawn were more likely to be male [22/54 (41%) with blood cultures vs. 8/59 (14%) without blood cultures; p=0.001], have underlying malignancy [12 (22%) vs. 5 (9%); p=0.04], pyelonephritis [29 (54%) vs. 12 (29%); p<0.001], fever [28 (52%) vs. 12 (20%); p<0.001], sepsis [36 (67%) vs. 19 (32%); p<0.001], and sepsis-induced hypotension [10 (19%) vs. 2 (3%); p<0.01]. There were no differences in age (p=0.4), diabetes (p=0.5), renal insufficiency (p=0.9), Charlson comorbidity score (p=0.2), and time to appropriate antibiotics (median 6 vs. 9 hours, p=0.4). Of the 29 pts with pyelonephritis in whom blood cx were taken, 7 (24%) were positive. Blood cx were obtained in 10 (19%) pts with asymptomatic bacteriuria; none were positive. 18 (16%) of 113 pts received no effective antibiotic therapy during the hospital admission & this did not differ whether blood cx were obtained or not (p=0.8). Length of hospital stay after bacteriuria was longer in pts who had blood cx taken [median 4.5 (range 0-36) vs. 3.1 days (0.2-13.9); p=0.04]. In-hospital mortality was similar whether blood cx were taken or not [6 (13%) vs. 2 (4%); p=0.1].
Conclusions: Blood cultures were more frequently drawn in patients with pyelonephritis and in febrile or septic patients, but no difference in mortality could be detected in this small sample. Blood cultures were more frequently drawn in male patients. Blood cultures in patients with asymptomatic bacteriuria had a very low yield and should not be obtained.
Objective: To compare patients (pts) with E. coli bacteriuria based upon whether blood cultures were obtained, and subsequent patient outcomes.
Methods: We conducted a 7-week prospective cohort study of adult, hospitalized patients with E. coli bacteriuria at a tertiary-care hospital. Blood cx were at the discretion of the treating physician & had to occur within ±1 day of the bacteriuria. Asymptomatic bacteriuria was defined as absence of urinary symptoms, & pyelonephritis as flank pain/tenderness and/or fever.
Results: 113 pts had E. coli bacteriuria. The median age was 59.2 (±19.7) years; 83 (74%) were female. 81 (72%) had community-acquired bacteriuria. 40 (35%) had asymptomatic bacteriuria, 12 (11%) had cystitis, 41 (36%) had pyelonephritis, and 20 (18%) were unable to report urinary symptoms (e.g., altered mental status, intubation). 16 (14%) pts had catheter-associated bacteriuria. In 54 (48%) pts a blood cx was obtained and 9 of 54 (17%) were bacteremic.
Bacteriuric pts who had blood cx drawn were more likely to be male [22/54 (41%) with blood cultures vs. 8/59 (14%) without blood cultures; p=0.001], have underlying malignancy [12 (22%) vs. 5 (9%); p=0.04], pyelonephritis [29 (54%) vs. 12 (29%); p<0.001], fever [28 (52%) vs. 12 (20%); p<0.001], sepsis [36 (67%) vs. 19 (32%); p<0.001], and sepsis-induced hypotension [10 (19%) vs. 2 (3%); p<0.01]. There were no differences in age (p=0.4), diabetes (p=0.5), renal insufficiency (p=0.9), Charlson comorbidity score (p=0.2), and time to appropriate antibiotics (median 6 vs. 9 hours, p=0.4). Of the 29 pts with pyelonephritis in whom blood cx were taken, 7 (24%) were positive. Blood cx were obtained in 10 (19%) pts with asymptomatic bacteriuria; none were positive. 18 (16%) of 113 pts received no effective antibiotic therapy during the hospital admission & this did not differ whether blood cx were obtained or not (p=0.8). Length of hospital stay after bacteriuria was longer in pts who had blood cx taken [median 4.5 (range 0-36) vs. 3.1 days (0.2-13.9); p=0.04]. In-hospital mortality was similar whether blood cx were taken or not [6 (13%) vs. 2 (4%); p=0.1].
Conclusions: Blood cultures were more frequently drawn in patients with pyelonephritis and in febrile or septic patients, but no difference in mortality could be detected in this small sample. Blood cultures were more frequently drawn in male patients. Blood cultures in patients with asymptomatic bacteriuria had a very low yield and should not be obtained.