Background Despite high antibiotic prescribing rates for urinary tract infection (UTI) at the end of life, the evidence suggests little or no symptomatic benefit in >50% of patients. This leads to concerns around the rigour underpinning UTI diagnosis in hospice inpatient settings and the lack of an applicable evidence base, with clear but conflicting antimicrobial guidelines in other populations.

Methodology Two matched retrospective audits of hospice inpatients over six-month periods in 2014 and 2016. Notes were analysed for symptoms consistent with a UTI, clinical investigations, results and management against local antimicrobial prescribing guidelines and checked against the corresponding microbiology laboratory database. To compare any findings, the audit was extended in 2016 to include one month of community patients in their last 30 days of life.

Results The inpatient UTI incidence was 11.4% in 2014 (n=33/290), 11.3% in 2016 (n=25/222) and 10.4% for community patients in 2016 (n=10/96). Correct management of patients with positive urine cultures increased from 56% to 100%. Correctly not prescribing antibiotics (when bacteraemia without symptoms e.g. catheterised patients) increased from 38% to 75%. The percentage of patients on antibiotics at death was 1% (n=3) and 1.4% (n=3).

Conclusions Incidence of UTI at the end of life, at 10%–11%, remained consistent over time and across setting to suggest reliability. There was a marked improvement in appropriate and targeted antibiotic therapy; qualitative analysis showed improved rigour in assessment of key symptoms, and more targeted investigations and antibiotic therapy (e.g. Ertapenem, Fosfomycin). It appeared that a UTI was associated with a poorer prognosis and delayed discharge. Further research is needed, particularly around the symptom benefits of patients receiving antibiotics for UTIs at the very end of life.