Early bloodstream infection after cardiopulmonary bypass: Frequency rate, risk factors, and implications

Objective: To determine the incidence, predisposing factors, and outcome of early bloodstream infection after cardiopulmonary bypass. Design: A case control study. Setting: A 54-bed cardiac surgical intensive care in a tertiary referral center. Patients: Patients from a 30-month period with preoperative hospital stay of <48 hrs and subsequent bloodstream infection within 96 hrs of cardiopulmonary bypass were included in a case group. The control group consisted of patients who had cardiac surgery on the same day as the case group. Measurements and Main Results: Patient demographics, history of comorbidity, preoperative laboratory testing, details of surgery, transfusion requirement, inotropic infusions, hemodynamics, and arterial blood gases on admission to intensive care were compared in the two groups. Measures of outcome were duration of mechanical ventilation and intensive care stay, serum creatinine on the first postoperative day, highest creatinine and bilirubin concentrations, and hospital mortality. During the study period, 7,928 patients had cardiac surgery. Sixteen (0.2%) patients had early bloodstream infection; the control group consisted of 95 patients. Thirteen of the patients with bloodstream infection had Gram-negative bacilli on blood culture, two had Candida species, and two had Gram-positive bacteria. On multivariate logistic regression analysis, greater prevalence of preoperative pulmonary hypertension (odds ratio 9; 95% confidence interval 2 to 41.8; p =.004), diabetes (adds ratio 4.6; 95% confidence interval 1.4 to 15.8; p =.01), number of blood products transfused (odds ratio 1.09; 95% confidence interval 1.04 to 1.17; p =.005), and infusion of inotropes (odds ratio 4.7; 95% confidence interval 1.3 to 16.4; p =.02) or vasopressors (odds ratio 4.1; 95% confidence interval 1.3 to 15.6; p =.02) were associated with postoperative bloodstream infection. Early bloodstream infection was associated with significantly prolonged duration of mechanical ventilation (117.2 +/- 21.5 vs. 18 +/- 8.8 hrs; p=.0001), intensive care stay (213 +/- 27.5 vs. 53 +/- 11.3 hrs; p <.0001), greater creatinine concentrations on the first postoperative day (1.6 +/- 0.1 vs. 1.2 +/- 0.04 mg/dL; p=.0002), greater maximum creatinine concentration (2.4 +/- 0.2 vs. 1.3 +/- 0.1 mg/dL; p<.0001), and greater maximum bilirubin concentration (4.7 +/- 0.6 vs. 1.3 +/- 0.2 mg/dL; p <.0001) when compared with the control group. Five (32%) of 16 bacteremic patients died vs. none of the 95 control patients (p <.0001). Conclusions: Early bloodstream infection after cardiac surgery is uncommon and involves predominantly Gram negative bacteria. The risk factors associated with bloodstream infection were preoperative morbidity and more complex surgery. Bloodstream infection was associated with a significantly adverse impact on outcome after cardiac surgery.