Length of stay and mortality associated with healthcare-associated urinary tract infections: A multi-state model

Background: The emergence of antimicrobial resistance is of particular concern with respect to urinary tract infections, since the majority of causative agents are Gram-negative bacteria. Healthcare-associated urinary tract infections (HAUTIs) are frequently associated with instrumentation of the urinary tract, specifically with indwelling catheters. Aim: To evaluate the current incidence, mortality, and length of hospital stay associated with HAUTIs. Methods: A non-concurrent cohort study design was used, conducted between January 1st, 2010 and June 30th, 2014. All patients admitted to one of the eight participating Australian hospitals and who were hospitalized for more than two days were included. The primary outcome measures were the incidence, mortality, and excess length of stay associated with HAUTIs. Findings: From 162,503 patient admissions, 1.73% [95% confidence interval (CI): 1.67–1.80] of admitted patients acquired a HAUTI. Using a multi-state model, the expected extra length of stay due to HAUTI was four days (95% CI: 3.1–5.0 days). Using a Cox regression model, infection significantly reduced the rate of discharge (hazard ratio: 0.78; 95% CI: 0.73–0.83). Women were less likely to die (0.71; 0.66–0.75), whereas older patients were more likely to die (1.40; 1.38–1.43). Death was rarer in a tertiary referral hospital compared to other hospitals, after adjusting for age and sex (0.74; 0.69–0.78). Conclusion: This study is the first to explore the burden of HAUTIs in hospitals using appropriate statistical methods in a developed country. Our study indicates that the incidence of HAUTI, in addition to its associated extra length of stay in hospital, presents a burden to the hospital system. With increasing incidence of UTI due to antimicrobial-resistant organisms, surveillance and interventions to reduce the incidence of HAUTI are required.