Table 3

Results from the discrete choice experiment conditional logistic regression

Attribute	Level	OR	P value	95% CI	Conditional attribute importance
Life expectancy with transplant	5 years	0.56	<0.001	0.49 to 0.62	21%
	10 years	1.10	<0.001	1.00 to 1.24
	15 years	1.61	<0.001	1.42 to 1.85
Quality of life with transplant	No/slight improvement	0.66	<0.001	0.49 to 0.62	16%
Quality of life with transplant	Significant improvement	1.51	<0.001	1.38 to 1.64	16%
Time waiting	3 years	0.76	<0.001	0.68 to 0.87	11%
	5 years	0.99	0.698	0.87 to 1.10
	8 years	1.33	<0.001	1.18 to 1.50
Life expectancy without transplant	3 years	0.86	0.011	0.76 to 0.97	5%
	2 years	1.07	0.312	0.95 to 1.17
	1 year	1.09	0.134	0.97 to 1.25
Adherence to medical regimen	Sometimes	0.57	<0.001	0.50 to 0.64	21%
	Often	1.07	0.329	0.94 to 1.19
	Always	1.64	<0.001	1.46 to 1.88
Age	60 years	0.87	0.010	0.78 to 0.97	5%
	40 years	1.01	0.746	0.91 to 1.14
	20 years	1.13	0.036	1.01 to 1.27
Social support	No	0.60	<0.001	0.54 to 0.68	20%
Social support	Yes	1.68	<0.001	1.50 to 1.86	20%

Modelled with effects coding. N = 497. Pseudo R2 = 0.2406. Attribute importance estimates are shown as percentages and convey the overall influence each attribute had in respondents’ profile choice, conditioned on the particular attributes and levels in the survey, with higher numbers conveying more relative influence.