The comparative effectiveness of telehealth versus primary care and collection of urine cultures on outcome in urinary tract infection

Abstract

Studies of urinary tract infection (UTI) treatment in telehealth settings have primarily evaluated young, healthy females. Urine culture collection is less common in telehealth settings but is recommended for all patients with potentially complicated infection. The aims of this study were: (1) to compare UTI-related clinical failure between telehealth and in-person primary care settings, and (2) evaluate if urine culture (UC) collection impacted UTI-related failure in patients with risk for antibiotic resistant infection. A retrospective cohort study of outpatients diagnosed with UTI between 2019 and 2021 in the Department of Veterans Affairs system was conducted. Inclusion required a telehealth or primary care visit with UTI International Classification of Diseases-Clinical Modification 10th revision code documentation and an antibiotic dispensed. Patients with recent UTI, concurrent indications for antibiotics, or where asymptomatic bacteriuria treatment was appropriate were excluded. Treatment failure was defined as combination of a new UTI-related outpatient visit or hospitalization that occurred between 3 to 30 days after the antibiotic dispense date. Antibiotic exposure data, covariates, hospitalization, and UC history were obtained. Overlap weighting and generalized estimating equation models estimated the relative risk of failure for clinical setting and for UC collection versus no collection. There were 16,266 telehealth and 29,296 primary care patient-visits evaluated. The adjusted relative risk (±95% CI) of failure for telehealth relative to primary care was [0.87 (0.70, 1.08)]. An interaction between setting and age ≥65 [1.45 (1.12, 1.87)] indicated higher failure for elderly patients treated for UTI in telehealth. Urine culture collection was associated with increased risk of failure for patients treated in telehealth [2.06 (1.56, 2.72)]; however, an interaction between UC collection and prior antibiotic exposure ≤90 days indicated a protective effect in both telehealth [0.70 (0.53, 0.93)] and primary care [0.77 (0.60, 0.99)] settings. Overall, no difference in the clinical failure rate for UTI treatment between telehealth and primary care was observed. However, elderly patients treated for UTI in telehealth experienced higher failure relative to in-person primary care. Patients with recent prior antibiotic exposure in both settings had lower clinical failure rates when UCs were collected.

Keywords: comparative effectiveness; microbial culture; telehealth; urinary tract infection.

Figure 1.

Study flow diagram for Veterans treated for urinary tract infection in telehealth and primary care settings with and without a urine culture collected.

Figure 2.

Overlap weighting balance plot for Veterans treated for urinary tract infection in telehealth and primary care settings.

Figure 3.

Overlap weighting balance plot for Veterans treated for urinary tract infection in telehealth and primary care settings with and without a urine culture collected.