Diagnosis of urinary tract infection in the emergency department – a real infection or an alarming indication for comprehensive geriatric assessment? A retrospective follow-up study

Abstract

Purpose To investigate whether urinary tract infection (UTI) diagnoses assigned in the emergency department (ED) were true UTI infections, characterize hospitalized UTI patients, and analyze the 12-month mortality and institutionalization rates. Methods A retrospective study among older adults (75+) who were admitted to the ED in 2017 and diagnosed UTI by the ED physician. Patients were categorized into four groups based on the level of UTI diagnosis and the follow-up treatment location: discharged cystitis, hospitalized cystitis, discharged pyelonephritis, and hospitalized pyelonephritis. Group differences were analyzed using Chi-squared, Fisher´s exact, or Mann-Whitney U tests. Logistic regression adjused for confounders was used to compare symptoms. Kaplan-Meier and Cox regression analyses were applied to asses mortality and institutionalization during the 12-month follow-up. Results Of 658 ED-diagnosed UTIs, only 354 (54%) had significant bacterial growth in urine cultures. Nevertheless, antibiotics were prescribed in 98% of cases. Institutionalization was significantly higher among hospitalized (31%) than among discharged (11%) cystitis patients (log-rank test, p < 0.001); age, Carlson Comorbidity Index and frailty adjusted HR=2.72, 95% CI 1.62–4.55, p < 0.001). Interestingly, hospitalized cystitis patients had higher institutionalization rates (31%) than hospitalized pyelonephritis patients (15%) (log-rank test, p = 0.003; age and gender adjusted HR=0.52, 95% CI 0.31–0.87, p = 0.014). Conclusions In ED UTIs are frequently overdiagnosed in older adults, often masking more complex underlying conditions. Because a hospitalization for cystitis serves as a major red flag for institutionalization within a year, these patients require an urgent comprehensive geriatric assessment rather than just a standard infection diagnosis.