Physical accessibility of primary care facilities for people with disabilities: a cross-sectional survey in 31 countries

Table 3 Linear multilevel regression analysis of physical accessibility of GP practices in 31 countries. Coefficient (standard error) (for empty model and model with practice/location variables: n_countries = 31; n_practices = 5865)

Fixed effects	Empty model	Model with practice/location variables
Constant	0.59 (0.030)	0.58 (0.029)
*Practice/location*
Shared practice (yes = 1)		0.011 (0.003)**
Age of the GP		−0.0005 (0.0001)**
Urbanisation (ref. = inner city)
Suburbs/towns		0.015 (0.003)**
Mixed urban-rural		0.018 (0.004)**
Rural		0.024 (0.004)**
Social deprivation of practice population		0.001 (0.002)
*Country* ^a
Years since ratification UN Convention (n_countries = 31; n_practices = 5865)		−0.007 (0.014)
Accessibility regulation index (n_countries = 31; n_practices = 5865)		0.034 (0.036)
Centralised healthcare system (n_countries = 31; n_practices = 5865)		− 0.001 (0.061)
Gross Domestic Product per head (n_countries = 28; n_practices = 4922)		0.000 (0.000)
Years social democratic government (n_countries = 27; n_practices = 4655)		0.018 (0.010)*
Random effects
Practice level variance	0.008 (0.0001)	0.008 (0.0001)
Country level variance	0.027 (0.007)	0.021 (0.0058) ^b
ICC	0.779	0.728

* p < 0.10
** p < 0.01
^a Country level variables were included one at a time
^b In the model with Years social democratic government as independent variable at country level

ISSN: 1472-6963