Table 4.

Regression models of variables as predictors of use of inpatient care among 24,973 hospitalized psychiatric patients in Victoria, Australia, 8,879 of whom were conditionally released between 1990 and 2000

	Total inpatient daysa			Inpatient days per inpatient episodeb			Inpatient days per 100 days in carec
Independent variable	Bd	SE	p	Bd	SE	p	Bd	SE	p
Conditional release group membership	14.85	4.39	.001	−8.32	2.55	<.001	−7.76	.49	<.001
Total community treatment days	−.04	.01	.004	−.03	.01	<.001	.01	.00	<.001
Interaction of involuntary outpatient commitment and community treatment days received per community care episode	.53	.08	<.001	.19	.04	<.001	−.03	.01	<.001
Gender	19.79	3.34	<.001	11.05	1.94	<.001	1.15	.37	.002
Age	1.44	.10	<.001	1.00	.06	<.001	.13	.01	<.001
Schizophrenia versus other disorders	28.31	4.73	<.001	17.78	2.75	<.001	−18.24	.53	<.001
Major affective disorder versus other disorders	−8.84	6.07	.145	−4.71	3.52	.181	−13.33	.68	<.001
Dementia versus other disorders	91.33	7.11	<.001	57.79	4.13	<.001	−4.83	.80	<.001
Paranoia and other psychosis	−32.95	11.14	.003	−13.27	6.47	.04	−10.61	1.25	<.001
Time from first contact with the mental health system to study’s end	.01	.00	<.001	.00	.00	<.001	−.00	.00	<.001
Mean year of episode commencement	−7.20	.73	<.001	−3.95	.43	<.001	.14	.08	.096
Number of inpatient episodes	5.68	.50	<.001	−2.00	.29	<.001	−.85	.06	<.001

^aModel summary: R=.27; R²=.07; adjusted R²=.07; N=24,973; F=166.73, df=12 and 24,960, p<.001

^bModel summary: R=.24; R²=.06; adjusted R²=.06; N=24,973; F=130.10, df=12 and 24,960, p<.001

^cModel summary: R=.41; R²=.17; adjusted R²=.17; N=24,973; F=425.89, df=12 and 24,960, p<.001

^dUnstandardized regression coefficient