Frequency of bacteremia associated with transesophageal echocardiography in intensive care unit patients: a prospective study of 139 patients

Abstract

Objective: To determine the occurrence rate of bacteremia associated with transesophageal echocardiography in intensive care unit (ICU) patients.

Design: A prospective study of 139 patients undergoing transesophageal echocardiography.

Setting: The medical ICU of a tertiary referral teaching hospital.

Patients: One hundred thirty-nine ICU patients (mean age 58 yrs) who underwent transesophageal echocardiography.

Interventions: Blood samples were systematically drawn for aerobic and anaerobic culture at the following times: before (blood culture 1), at the end of (blood culture 2), and 30 mins after (blood culture 3) transesophageal echocardiography examinations.

Measurements and main results: The mean duration of transesophageal echocardiography was 35 mins (range 7 to 120). One hundred thirty-four patients received mechanical ventilation; 125 patients had a nasogastric tube. Fifty-one patients had one or more underlying conditions that usually justify antimicrobial prophylaxis of bacterial endocarditis before high-risk procedures. Fifty-six patients did not receive any antibiotic treatment at the time of transesophageal echocardiography. In 114 patients, the three blood cultures were negative. In six patients, transesophageal echocardiography was performed during a preexisting bacteremia. A contamination (only one positive blood culture of the three sampling times) with coagulase-negative staphylococci occurred in four patients at blood culture 1, five patients at blood culture 2, and six patients at blood culture 3. Contamination with Corynebacterium species occurred in two patients at blood culture 2. In one patient receiving cefotaxime and netilmicin, blood culture 1 was sterile and blood cultures 2 and 3 yielded coagulase-negative staphylococci. In one patient receiving no antibiotic treatment, blood culture 1 was sterile and blood cultures 2 and 3 yielded Enterococcus faecalis. None of these two patients received a specific antibiotic treatment or developed any secondary septic focus.

Conclusions: The overall frequency of bacteremia induced by transesophageal echocardiography in ICU patients was 1.4% (two of 139 patients) (95% confidence interval 0.2% to 5.1%). The frequency did not differ whether patients received antibiotics before transesophageal echocardiography (one [1.2%] of 83 patients) or not (one [1.8%] of 56 patients) (p = .96). Therefore, routine antimicrobial prophylaxis does not appear justified before transesophageal echocardiography in ICU patients.