Table 4.

Association Between Asian American Enclaves and Geographic Healthcare Accessibility (CA, FL, NJ, NY, and TX), 2010

Model	Prevalence ratio (95% CI)	Marginal effects at the means (95% CI)
Asian American enclave (bivariable)	1.45 (1.38, 1.53) **	0.10 (0.08, 0.11) **
Asian American enclave (multivariable)	1.23 (1.17, 1.29) **	0.05 (0.03, 0.06) **
Multivariable by enclave trajectory
Never enclave	1.00 (ref)
Persistent enclave	1.27 (1.20, 1.34) **	0.05 (0.04, 0.07) **
Emergent enclave	1.17 (1.09, 1.27) **	0.04 (0.02, 0.05) **
Former enclave	1.13 (1.02, 1.26) *	0.03 (0.01, 0.05) *

Note: Boldface indicates statistical significance

^*p<0.05 and

^**p<0.01.

Poisson regression models with robust variance estimation were used to estimate prevalence ratios. Models included an indicator for state to account for within-state differences in healthcare accessibility, and the reference was nonenclave census tracts. Multivariable models adjusted for percentage poverty, metropolitan RUCA classification, population density, residential mobility, the proportion of uninsured individuals, crime, vehicle access, and the underlying age structure. Enclave trajectory was derived on the basis of Asian American enclave classifications in the years 2000 and 2010. Never enclaves were not classified as enclaves in either year, persistent enclaves were classified as enclaves in both years, emergent enclaves were classified as enclaves in 2010 but not in 2000, and former enclaves were classified as enclaves in 2000 but not in 2010.

CA, California; FL, Florida; NJ, New Jersey; NY, New York; RUCA, Rural-Urban Commuting Area; TX, Texas.