Table 3.

Statistical models on GOS by sex

Tests		Bangladeshi data	Bootstrap of Bangladeshi data	CRASH data
Fisher’s exact test	P-value	1		<0.001
	CI	0.427 ∼ 2.441		0.639 ∼ 0.863
	Odds	1.001		0.743
Test of proportions	P-value	0.99		<0.001
	CI	-0.174 ∼ 0.173		-0.074 ∼ -0.023
Binary logistic model	P-value	0.998	0.998	<0.001
	CI	0.453 ∼ 2.211	0.453 ∼ 2.211	0.641 ∼ 0.861
	Odds	1.001	1.001	0.743
Proportional odds model	CI	0.477 ∼ 1.682	0.477 ∼ 1.682	0.813 ∼ 1.025
	Odds	0.896	0.896	0.913
Sliding dichotomy model	P-value	0.841	0.841	0.688
	CI	0.441 ∼ 2.735	0.441 ∼ 2.736	0.816 ∼ 1.144
	Odds	1.098	1.098	0.966

The reference level for sex was ‘male’