We have read with interest the article by Etienne et al. in which the authors evaluate the value of blood cultures (BC) in patients with acute prostatitis (AP) (2). Since our group has a particular interest in these patients, which are included in a database, we would like to make some comments regarding the article.First of all, as pointed out by Etienne et al. in a previous article from which the patients of the aforementioned article were recruited, the diagnostic criteria for AP are not well defined (1). Although it has been well demonstrated with different diagnostic methods, such as measurement of the serum prostate-specific antigen (3) and the use of transrectal prostatic ultrasonography (3) and leukocyte scintography (5), that the prostate is the organ most frequently involved in males with febrile urinary tract infection (UTI), in our clinical experience, this is not always the case, even for patients with a painful prostate palpation. Therefore, we prefer to use the more generic term "men with febrile UTI" at least until more agreement in the definition of AP exists.The issue of whether BC could be useful in UTI has been, among others, evaluated by Velasco et al., who concluded that BC provided no useful information for clinical management of acute pyelonephritis, but the study excluded women with complicated UTI and men (4). Thus, Etienne et al. are the first to evaluate the diagnostic and prognostic value of BC for patients with AP. In their study, they found that 21% of the patients had positive BC and that these patients contributed to microbiological diagnosis in 5% of the cases. After reviewing our database, we have observed that 20 out of 89 (22%) male patients with febrile UTI, from which BC were drawn, had positive BC, which contributed to microbiological diagnosis in 4 cases (4.5%) (positive BC and negative urine cultures), results that are comparable to those seen by Etienne et al. More important, 3 of the microorganisms isolated from BC that contributed to the diagnosis (2 Escherichia coli isolates and 1 Enterococcus faecalis isolate) were resistant to quinolones and 1 (Morganella morganii) was, in addition, resistant to cefuroxime and to amoxicillin-clavulanic acid. As the degrees of antimicrobial resistance in the cases in which BC contributed to microbiological diagnosis were not reported by Etienne et al., we wonder if this was also the case. A positive answer would give additional arguments for the recommendation of drawing blood cultures from patients with AP.Regarding the prognostic value of BC, as pointed by Etienne et al., the greater duration of fever in patients with AP and positive BC could reflect the existence of a more severe disease, although the retrospective nature of the analysis limits their conclusions. The measurement of hemodynamic parameters, for instance, mean arterial pressure, could give additional information about the clinical situations of these patients and whether the existence of positive BC is a potential marker of the severity of the infectious episode.
