Impact of the rural pipeline in medical education: practice locations of recently graduated family physicians in Ontario

Table 3 Significant logistic regression models for family physicians’ practice location by undergraduate or postgraduate medical school

Variables*	OR	95% CI	p value	Nagelkerke R ²	Percentage of cases correctly classified: model with constant, final model
Rural (vs. urban) practice location in Ontario				0.093	87.9, 87.9
NOSM UG	2.57	1.21–5.44	0.014
Age at graduation	1.11	1.04–1.18	0.001
Northern (vs. southern) Ontario practice location				0.547	87.9, 94.4
NOSM PG	57.88	18.21–183.98	<0.001
UG × PG	3.20	0.73–14.11	0.124
Gender	1.89	0.87–4.09	0.106

*Backwards elimination (conditional method) logistic regression started with gender, age at graduation, UG school (NOSM or not), PG school (NOSM or not), and the UG × PG interaction with probability for elimination set at 0.15 (n = 535). Final models are significant at p < 0.001

ISSN: 1478-4491