The association of adequate empirical treatment and time to recovery from bacteraemic urinary tract infections: a retrospective cohort study.

Our objective was to evaluate whether patients with bacteraemic urinary tract infection (UTI) who receive inadequate empirical therapy have worse outcomes than those with adequate therapy. This was a retrospective cohort study of patients with bacteraemic UTI. The exposure variable was adequate versus inadequate empirical antibiotic therapy (AEAT versus IEAT) within 24 h of culture collection. Primary endpoint was time to cure. The primary analysis used propensity score models with inverse probability of treatment weights. A secondary Cox proportional hazards modelling approach was used to test the robustness of this finding, and to evaluate other patient and pathogen predictors of time to cure. Of 469 patients with bacteraemic UTI, 368 (78.5%) received AEAT. There was no significant difference in mortality between those receiving AEAT and those receiving IEAT (adjusted OR 0.86, 95%CI 0.47-1.58). Receipt of AEAT had no association with time to cure (HR 0.93, 95%CI 0.73-1.19, p 0.55) or time to normalization of individual clinical variables. Cox proportional hazards modelling revealed that longer time to cure was associated with liver disease (HR 0.25, 95%CI 0.08-0.76, p 0.015), prior stroke (HR 0.73, 95%CI 0.54-0.99, p 0.044), empirical receipt of piperacillin-tazobactam (HR 0.77, 95%CI 0.59-0.99, p 0.044), qSOFA score >1 (HR 0.68, 95%CI 0.55-0.84, p < 0.001), and hospital-onset UTI (HR 0.53, 95%CI 0.39-0.71, p < 0.001). In conclusion, we found no association between AEAT and time to cure for patients with bacteraemic UTI. It may be appropriate to accept a higher risk threshold when choosing empirical antibiotic regimens, even in centres with high rates of resistant uropathogens.