. 2018 Mar 27;18(4):377–393. doi: 10.1007/s10754-018-9238-z

Table 5.

Multiple regression results for mental health utilization

Variables	Frequency of hospitalization		LOS		Hospitalization cost
Variables	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
Type of hospital	0.183	0.160	0.196	0.177	0.026	0.019
Gender	0.031	0.044	0.157	0.161	0.042	0.044
Age	0.069	0.070	0.111	0.114	0.026	0.026
Marital status	− 0.219^a	− 0.220^a	− 0.599^a	− 0.609^a	− 0.177^a	− 0.179^a
Diagnosis	− 0.007	− 0.009	0.001	0.000	− 0.007	− 0.007
Residence	− 0.591	− 0.647	− 0.114	− 0.168	− 0.053	− 0.036
Admission severity	0.256	0.266	0.665	0.674	0.263^a	0.266^a
Types of insurance	0.179^a	0.076	0.123	0.116^a	0.061	0.016^a
URBMI in 2007	0.814	− 1.690	0.132	0.501	0.058	− 0.091
RR	0.095^a	0.090	0.102^a	0.094^a	0.066^a	0.064^a
RR * NCMS	–	0.329^a	–	0.121^a	–	0.011^a
RR * URBMI	–	0.036	–	0.033	–	0.003
RR * UEBMI (Control group)	–	–	–	–	–	–
R²	0.159	0.166	0.106	0.107	0.228	0.229

The coefficients could not be read in the same way because different regression models were used for each dependent variable

^aP value < 0.05