Practice variation in treatment of suspected asymptomatic bacteriuria prior to cardiac surgery: are there differences in postoperative outcome? A retrospective cohort study

OBJECTIVES: It is unclear whether postoperative infections can be prevented by treating asymptomatic bacteriuria, or whether, on the other hand, such treatment will increase the risk of more serious infection by pathogenic bacteria different from the ones causing bacteriuria. This study aimed to support future treatment decisions for preoperative cardiothoracic surgery patients with asymptomatic bacteriuria, by examining current preoperative practice, in relation to postoperative outcome. METHODS: A retrospective cohort study was conducted. All patients who underwent cardiothoracic surgery in 2011-2013 using extracorporeal circulation in St. Antonius Hospital Nieuwegein, and who preoperatively had nitrituria and/or leucocyturia were included. Exclusion criteria were C-reactive protein level higher than 10 mg/l, emergency surgery, critical preoperative state and/or antibiotic treatment because of other infections. Outcomes were postoperative infections and length of stay. Furthermore, we compared culture results of preoperative urine with postoperative infection sites in order to study the hypothesis of haematogenous spread. RESULTS: One thousand and two patients with leucocyturia or nitrituria were eligible, of whom 3.9% had been treated with antibiotics preoperatively (AB+). Of the 96.1% of patients who had not been treated (AB-), 8.3% had an infection postoperatively, compared with 5.1% in the treatment (AB+) group. This was not statistically significant {odds ratio, corrected for EuroSCORE, 0.53 [95% confidence interval (CI) 0.12-2.24, P = 0.39]}. Length of stay, corrected for EuroSCORE, between the treated (AB+) and the non-treated (AB-) group did not differ, with a hazard ratio of 1.05 (95% CI 0.63-1.75, P = 0.85). As regards bacterial culture results, none of patients not treated with antibiotics preoperatively (AB-) seemed to have a postoperative infection due to haematogenous spread of bacteria from the urinary tract present preoperatively. CONCLUSIONS: The risk of haematogenous spread of bacteria seems to be non-existent in this large cohort of non-treated patients, under our local clinical practice. Based on this current, best available evidence, it seems therefore safe not to treat patients with asymptomatic bacteriuria prior to cardiothoracic surgery. This could also imply that it is safe not to perform routine preoperative urine testing.