The Digital Divide: Exploring Inequities in UTI Care Utilization and Outcomes

Objectives: Urinary tract infections (UTIs) are among the most prevalent bacterial infections in the United States. The rise of virtual visits for UTIs, particularly after the COVID-19 pandemic, signifies a shift in healthcare delivery. While virtual care offers potential for broader accessibility, it may perpetuate disparities. We wanted to evaluate racial, linguistic, and socioeconomic disparities in the utilization of virtual care and outcomes for UTIs

Methods: This cross-sectional retrospective cohort study reviewed the Kaiser Permanente Southern California database from January 2015, to December 2022. Women aged >18 years who had an outpatient encounter, virtual or in-person, for a diagnosis of UTI were included. Index encounters from the emergency department (ED) or inpatient settings were excluded. Virtual visits included telephone nurse visits, telephone consultations, video calls, and e-portal interactions. Visit type, demographics, Elixhauser Comorbidity Index (ECI), median household income, Neighborhood Deprivation Index (NDI), and 30-day outcomes (hospitalizations, ED visits and related follow-up visits) were extracted. Descriptive statistics identified differences in racial, linguistic, and socioeconomic factors. Univariate and multivariate logistic regression models were implemented.

Results: There were 1,220,698 outpatient UTI encounters among 428,855 patients included (60% virtual; 40% in-person), with sample description stratified by visit type in Table 1. Patients who accessed virtual care were younger (53 vs. 58 years old), more likely to identify as White (43% vs. 40%), more likely to speak English (93% vs. 86%), less likely to have median household income under $45,000 annually (40% vs. 45%), and less likely to live in an area with a high NDI (23% vs. 28%). These associations all maintained significance in multivariate regression (table 2). While statistically significantly different, rates of 30-day hospitalization (0.7% vs. 0.8%), ED visits (1.5% vs. 2%) and follow up visits (22.7% vs 23.2%) were similar between those who accessed virtual versus in-person care (table 1). In multivariate modeling controlling for BMI and comorbidities, risk of 30-day hospitalization was higher among patients who identified as White, those with a spoken language other than English or Spanish, and those with a lower median household income (table 2). ED visit within 30 days was more common in those who identified as Black, those whose spoken language was Spanish, those with a lower median household income, and those with a higher NDI. Conversely, scheduled follow up visit within 30 days was more common among those who identified as White, those with a spoken language other than English, and those with higher income and lower NDI (table 2).

Conclusions: Virtual visits were more common among English speakers, who identified as White, with higher household income, and living in less deprived neighborhoods. After controlling for age, BMI, and comorbidities, outcomes differ by race, language, income, and NDI, underscoring importance of continued efforts to combat these inequities.