O2173: Neighbourhood Marginalization and Rates of Virtual Pediatric Emergency Department Visits

Background: Virtual care has the potential to improve access to urgent pediatric care, but it is unclear if marginalized populations face inequitable access. Our institution created a virtual pediatric Emergency Department (ED) in May 2020. The objective of this study is to compare rates of virtual pediatric ED visits from neighbourhoods with different levels of marginalization, and to determine if this association varies based on distance to the hospital.

Methods: This is a repeated cross-sectional study of rates of virtual ED visits per census dissemination area within 100 km of our institution from May through December 2020 for patients < 18 years of age with a home address in our province. The Ontario Marginalization Index was used to apply quintiles of marginalization to each dissemination area according to four dimensions: material deprivation, ethnic concentration, residential instability, and dependency. Analyses were conducted using negative binomial generalized linear mixed-effects models with random intercept. Univariate regression was followed by models adjusted for distance to the hospital.

Results: There were 2,925 virtual ED visits from 1,077 census dissemination areas during the study period. Compared to the lowest quintile of material deprivation (least deprived), there were lower incidence rate ratios (IRR) of virtual ED visits for the third (0.69, 95% CI 0.58 – 0.82), fourth (0.67, 95% CI 0.56 – 0.79), and fifth (0.44, 95% CI 0.36 – 0.52) quintiles. This effect remained the same when adjusted for distance to the hospital. Compared to the lowest quintile of ethnic concentration (least ethnically diverse), there were higher IRR of virtual ED visits for the second (1.40, 95% CI 1.14 – 1.72), third (1.91, 95% CI 1.57 – 2.32), fourth (1.90, 95% CI 1.57 – 2.32), and fifth (1.64, 95% CI 1.36 – 1.99) quintiles. However, there were no significant differences in IRR of virtual ED visits for the higher quintiles of ethnic concentration compared to the lowest when adjusted for distance to the hospital. There were no significant differences in IRR of virtual ED visits for different quintiles of residential instability in either univariate or adjusted analyses. Compared to the lowest quintile of dependency, there were slightly lower IRR of virtual ED visits for the second (0.84, 95% CI 0.71 – 0.99), fourth (0.82, 95% CI 0.69 – 0.98), and fifth quintiles (0.62, 95% CI 0.51 – 0.75). There were no differences between quintiles of dependency after adjusting for distance to the hospital.

Conclusion: Neighbourhoods with higher material deprivation (lower incomes, higher rates of poverty) had lower rates of virtual ED visits compared to wealthier neighbourhoods, even when adjusting for distance to the hospital. Further research is needed to identify and eliminate barriers to use of virtual pediatric emergency care so populations with material deprivation can equitably access this resource.