Table 5.

Results from conditional logistic regression model: disabled child study sample (n = 11) and parent study sample (n = 30)

Attribute	Disabled child sample (n = 11)					Parent sample (n = 30)
	β-coefficient	95 % bootstrapped CIa	P value	MRS valuesb (cost)	95 % bootstrapped CIa	β-coefficient	95 % bootstrapped CIa	P value	MRS valuesb,c (cost)	95 % bootstrapped CIa
Comprehensiveness of wheelchair assessment	1.4247*	1.4153 to 2.0824	0.009	£153.19	£133.20 to £182.53	1.5329*	1.4507 to 2.1633	<0.0001	£547.46	£353.38 to £1435.45
Cost contribution for wheelchair	−0.0093*	−0.0138 to −0.0089	0.019	–	–	−0.0028	−0.0060 to 0.0005	0.092	–	–
Level of training provided by service	0.0306	−0.1955 to 0.2858	0.924	–	–	−0.1557	−0.4002 to 0.0311	0.371	–	–
Waiting time for delivery of wheelchair	−0.9221*	−1.4086 to −0.8442	0.041	£99.15	£81.93 to £121.32	−1.3699*	−1.9859 to −1.3104	0.000	£489.25	£313.29 to £1326.78
Frequency of wheelchair reviews	0.0364	−0.0022 to 0.0749	0.519	–	–	−0.0390	−0.0813 to 0.0032	<0.0001	–	–
Number of observations = 88						Number of observations = 240
Number of individuals = 11						Number of individuals = 30
Log likelihood function = −26.64						Log likelihood function = −64.51
Log likelihood ratio (5) = 33.85						Log likelihood ratio (5) = 114.86

*Significant attribute [P < 0.05]

^a95 % confidence intervals generated using non-parametric bootstrapping (5000 replications)

^bMarginal rate of substitution values = β-coefficient for attribute/β-coefficient for cost attribute

^cThough the cost contribution attribute was not significant to parents (P = 0.092 [>0.05]), everything being equal, parents preferred lower cost contribution; the parents’ MRS values were calculated using the cost contribution attribute as the denominator to show how parents trade-off the cost contribution attribute against the other attributes. This allowed comparison with the disabled child sample MRS values