Table 3.

Logistic regression results of the main demands for community-based care services

Variables	The elderly care hotline	Building health archives	On-call nursing and doctor visits	Regular medical examinations	Sporting fitness
	OR	OR	OR	OR	OR
Sociodemographic characteristics
Gender (Male)	1.308	0.577	1.230	0.396*	1.224
Educational level	1.116	1.327	1.391*	1.068	1.004
Health status	1.141	1.257	0.545**	0.918	2.545***
Self-care ability	0.710	0.415**	0.643	0.488	0.912
Career (other)	1.000	1.000	1.000	1.000	1.000
Daily laborer	0.381	0.518	0.167**	0.762	1.126
Public institution employee	0.445	0.599	0.098**	0.58	0.327
State-owned company employee	0.387	0.372	0.170*	0.320	0.243*
Private company employee	0.775	0.946	0.075***	0.619	0.487
Family structure
Living status (Three generations living together)	1.000	1.000	1.000	1.000	1.000
With spouse	2.701*	0.722	0.322*	2.297	3.221*
With children	2.986*	1.029	0.695	4.011*	3.349*
Economic characteristics
Income source (other)	1.000	1.000	1.000	1.000	1.000
Retirement pension and endowment insurance	1.057	6.030**	3.385	8.878*	0.363
Money from children or relatives	0.736	8.729**	4.476*	1.393	0.338
Personal labor income	0.432	2.074	11.666*	4.061	0.144*
Elderly care intention
Elderly care intention (Community-based care facilities)	1.000	1.000	1.000	1.000	1.000
Private homes	0.255***	0.408*	0.648	0.687	1.114
Long-term care institution	1.462	0.032**	0.194	0.386	0.377
Constant	2.427	0.000	75.722*	10.360	0.074
Thepvalue of Hosmer and Lemeshow Test	0.098	0.186	0.276	0.243	0.210
Percentage of correct prediction	68%	77%	71%	84%	71%

1.The category in parentheses for each variable is the reference group; 2. *** p < 0.001, ** p < 0.01, * p < 0.05; 3. Table S1 in Additional file 2 describes confidence interval data of these five models