Table V.

Univariable exact logistic regression (for variables with p<0.20) and final multivariable logistic regression results, examining the effects of predictor variables on the odds of visiting a health clinic or hospital for AGI in Iqaluit, Nunavut (September 2012 and May 2013)

Iqaluit healthcare use (September and May)
			Univariable logistic results			Final multivariable logistic results
Predictor variable		n	Odds ratio	p*	95% CI	Odds ratio	p*	95% CI
Variables in the final multivariable model
Missed traditional activities due to AGI	No	95	ref.
	Yes	28	4.591	0.003	1.479–14.440	3.822	0.012	1.180–12.463
Took OTC medication for AGI	No	68	ref.
	Yes	57	4.778	0.005	1.522–18.001	3.882	0.018	1.176–15.090
Variables considered in building the multivariable model
Vomited	No	70	ref.			–	–	–
	Yes	55	4.055	0.0071	1.348–13.870	–	–	–
Food secure	No	57	ref.			–	–	–
	Yes	68	0.360	0.041	0.122–1.044	–	–	–
Missed paid employment due to AGI	No	71	ref.			–	–	–
	Yes	51	2.629	0.077	0.912–8.050	–	–	–
Household owns a vehicle	No	47	ref.			–	–	–
	Yes	76	0.485	0.144	0.166–1.396	–	–	–

^*Score method for estimating p-values does not assume a symmetrical distribution for discrete data. p<0.05 was considered significant.