Empiric antibiotic treatment and the misuse of culture results and antibiotic sensitivities in patients with community-acquired bacteraemia due to urinary tract infection.

Recent data suggests that culture isolations and susceptibility profiles are overlooked, misinterpreted and are not a major determinant in the process of selecting anti-bacterial therapy. Compliance with empiric antibiotic protocols, relationship between blood culture results, change and selection of secondary antibiotic regimens and effect on outcome and length of hospitalization were assessed prospectively in 134 episodes of community-acquired bacteraemia due to urinary tract infection in adults. Empiric antibiotic protocols were correct in 112 episodes (83.6%), excessive in 12 episodes (8.95%) and inappropriate in 10 (7.5%) episodes, with no significant difference in outcome or length of hospitalization. Excluding early deaths, the adequacy of the initial antibiotic therapy was clearly associated with mortality (four deaths out of eight episodes treated incorrectly vs. 21/117 treated appropriately [P = 0.029]), but not with length of hospitalization. Antibiotic change was theoretically required in 92/119 (77.3%) episodes (27 [29.3] incorrect regimen, 65 [70.65%] excessive regimen), but actual change was made in only 43 episodes, of which three protocols were changed from a correct to an incorrect regimen, and one patient continued to receive an incorrect regimen. Eighteen out of 34 changes in excessive protocols were still excessive. Adequacy of secondary antibiotic treatment was clearly associated with outcome (5/10 vs. 4/109 [P<0.001]). Excessive protocols were not associated with better outcome or shorter hospital stay. Change of antibiotic regimen was associated with the presence of background diseases (5.6-fold increase) and inversely with hospitalization on the urology ward (0.254), but no independent factors associated with correct or incorrect secondary regimens could be identified. Although the compliance rate with empiric protocols was satisfactory, in many bacteraemic episodes blood culture results and antibiotic sensitivity profiles are overlooked, leading to higher mortality and excessive, unjustified use of expensive and broad-spectrum antibiotics. We could not identify factors associated with this disregard of susceptibility profiles.