Having a family doctor was associated with lower utilization of hospital-based health services

Table 5 Effects of primary care doctor choice on use of health services during the last episode of illness

Logistic regression results		RFD vs. NRD†	RFD vs. ORD†	ORD vs. NRD†	Other significant independent variables (p < 0.05)‡
Logistic regression results	Overall significance	Odds ratio (95% C.I.)
Any primary care doctor consultation	0.012	*2.486	*1.342	*1.853	Drinking, regular exercise, occupation, marital status
		(2.053,3.010)	(1.074,1.676)	(1.509,2.275)	Drinking, regular exercise, occupation, marital status
Emergency service visit	<0.001	*0.479	*0.624	0.768	Age, general health, long-term medication, household monthly income, district
		(0.330,0.695)	(0.411,0.949)	(0.536,1.098)
Hospital admission	0.032	*0.458	*0.514	0.891	Age, long-term medication
		(0.267,0.788)	(0.284,0.932)	(0.540,1.470)	Age, long-term medication

Notes: RFD = Regular family doctor group; ORD = Other regular doctor group; NRD = No regular doctor group.
*Statistically significant (p < 0.05) difference.
†as reference category for doctor choice groups in logistic regression, Odds ratio <1 (less likely than the reference category), Odds ratio >1 (more likely than the reference category).
‡Adjustment of confounding factors including socio-demographics, health status, chronic morbidity and lifestyle (Additional file 1).

ISSN: 1472-6963