Risk factors and clinical impact of multidrug resistance in healthcare-associated bacteraemic urinary tract infections: a post-hoc analysis of a multicentre prospective cohort in Spain

Abstract

Background: The global burden associated with antimicrobial resistance is of increasing concern. Aim: To evaluate risk factors associated with multidrug-resistant (MDR) infection and its clinical impact in a cohort of patients with healthcare-associated bacteraemic urinary tract infections (BUTIs). Methods: This was a prospective, multicentre, post-hoc analysis of patients with healthcare-associated-BUTI (ITUBRAS-2). The primary outcome was MDR profile. Secondary outcomes were clinical response (at 48-72 h and at hospital discharge) and length of hospital stay from onset of BUTI. Logistic regression was used to evaluate variables associated with MDR profile and clinical response. Length of hospital stay was evaluated using multivariate median regression. Findings: In all, 443 episodes were included, of which 271 (61.17%) were classified as expressing an MDR profile. In univariate analysis, MDR profile was associated with E. coli episodes (odds ratio (OR): 3.13; 95% confidence interval (CI): 2.11-4.69, P < 0.001) and the extensively drug-resistant (XDR) pattern with P. aeruginosa aetiology (7.84; 2.37-25.95; P = 0.001). MDR was independently associated with prior use of fluoroquinolones (adjusted OR: 2.43; 95% CI: 1.25-4.69), cephalosporins (2.14; 1.35-3.41), and imipenem or meropenem (2.08; 1.03-4.20) but not with prior ertapenem. In terms of outcomes, MDR profile was not associated with lower frequency of clinical cure, but was associated with longer hospital stay. Conclusion: MDR profile was independently associated with prior use of fluoroquinolones, cephalosporins, imipenem, and meropenem, but not with prior ertapenem. MDR-BUTI episodes were not associated with worse clinical cure, although they were independently associated with longer duration of hospital stay.