Development of a predictive model for integrated medical and long-term care resource consumption based on health behaviour: application of healthcare big data of patients with circulatory diseases

Table 4 Evaluation of prediction models by multiple regression analysis

Index		Partial regression coefficient	Standard partial regression coefficient	F value	p value	SE	VIF
Broad adherence
1	Secondary prevention	-0.048	-0.241	3,820.6	< 0.001	0.001	1.07
2	Rehabilitation intensity	-0.250	-0.210	2,740.7	< 0.001	0.005	1.13
3	Guidance	-0.057	-0.144	1,413.1	< 0.001	0.002	1.03
4	PDC	-0.057	-0.075	366.3	< 0.001	0.003	1.10
5	Overlapping outpatient visits	0.028	0.053	116.4	< 0.001	0.003	1.67
6	Overlapping clinical laboratory and physiological tests	0.012	0.091	343.1	< 0.001	0.001	1.70
7	Medical attendance	0.001	0.261	4,460.5	< 0.001	0.005	1.08
8	Generic drug rate index	-0.019	-0.016	17.7	< 0.001	0.004	1.04
Age		0.009	0.032	56.6	< 0.001	0.001	1.25
Sex		-0.509	-0.086	518.8	< 0.001	0.022	1.01
Follow-up period		0.051	0.254	3,207.9	< 0.001	0.001	1.41
Constant term		3.421		1,249.7	< 0.001	0.105

Abbreviations: PDC proportion of days covered, SE standard error, VIF variance inflation factor

ISSN: 1741-7015