Ambulatory Care Fragmentation, Emergency Department Visits, and Race: a Nationwide Cohort Study in the U.S.

Abstract

It is unclear whether highly fragmented ambulatory care (i.e., care spread across multiple providers without a dominant provider) increases the risk of an emergency department (ED) visit. Whether any such association varies with race is unknown. We sought to determine whether highly fragmented ambulatory care increases the risk of an ED visit, overall and by race. We analyzed data for 14,361 participants ≥ 65 years old from the nationwide prospective REasons for Geographic and Racial Differences in Stroke (REGARDS) cohort study, linked to Medicare claims (2003-2016). We defined high fragmentation as a reversed Bice-Boxerman Index ≥ 0.85 (≥ 75^th percentile). We used Poisson models to determine the association between fragmentation (as a time-varying exposure) and ED visits, overall and stratified by race, adjusting for demographics, medical conditions, medications, health behaviors, psychosocial variables, and physiologic variables. The average participant was 70.5 years old; 53% were female, and 33% were Black individuals. Participants with high fragmentation had a median of 9 visits to 6 providers, with 29% of visits by the most frequently seen provider; participants with low fragmentation had a median of 7 visits to 3 providers, with 50% of visits by the most frequently seen provider. Overall, high fragmentation was associated with more ED visits than low fragmentation (adjusted risk ratio [aRR] 1.31, 95% confidence interval [CI] 1.29, 1.34). The magnitude of this association was larger among Black (aRR 1.48, 95% CI 1.44, 1.53) than White participants (aRR 1.23, 95% CI 1.20, 1.25). Highly fragmented ambulatory care was an independent predictor of ED visits, especially among Black individuals.