Can high-cost spending in the community signal admission to hospital? A dynamic modeling study for urgent and elective cardiovascular patients

Table 3 Extended Survival Models for High Cost Community-Based Care: Primary Care, Specialist, Home Care, Laboratory Services and Emergency Department Care as Time Varying Predictors of Hospital Admission

	Beta	SE	Chi-square	P	HR
Survival Model – ELECTIVE Hospitalization
Monthly High cost GP	−0.12121	0.10581	1.3123	0.2520	0.886
Monthly High cost Specialist	2.20546	0.05964	1367.4911	< 0.0001	9.074
Monthly High cost homecare	−0.59350	0.16195	13.4294	0.0002	0.552
Monthly High cost laboratory	−0.04933	0.11268	0.1916	0.6616	0.952
Monthly High cost emergency department	−0.24214	0.10776	5.0490	0.0246	0.785
Survival Model – URGENT Hospitalization
Monthly High cost GP	0.14998	0.04485	11.185	0.0008	1.162
Monthly High cost Specialist	0.3736	0.04485	72.3877	< 0.0001	1.456
Monthly High cost homecare	0.32515	0.04056	64.2785	< 0.0001	1.384
Monthly High cost Laboratory	0.18537	0.04889	14.3784	< 0.0001	1.204
Monthly High cost emergency department	0.94101	0.03895	583.8090	< 0.0001	2.563

Other variables in the models included sex, age, rural/urban, ADG score, SES, GP Care Model (FFS, Family Health Team, Other), baseline CVD condition (AMI, arrhythmia, CHF, coronary syndrome, stroke), and baseline co-morbidities (diabetes, hypertension, COPD, renal failure, mood disorders)

ISSN: 1472-6963