From KIDSCREEN-10 to CHU9D: creating a unique mapping algorithm for application in economic evaluation

Table 4 Mapping equations from KIDSCREEN-10 index to Child Health Utility 9D utility scores

	Ordinary least squares estimator		Censored least absolute deviations estimator		MM-estimator		Generalised linear model
	Coeff.	SE^†	Coeff.	SE^†	Coeff.	SE^†	Coeff.	SE^†
Model 1
KS	0.043515	0.005291*	0.046580	0.006828*	0.049504	0.006682*	0.092650	0.008747*
KS²	−0.000334	0.000053*	−0.000359	0.000072*	−0.000384	0.000070*
Constant	−0.435412	0.129225*	−0.510120	0.160989*	−0.593052	0.157245*	−2.472760	0.359525*
Model 2
KS_I1	0.035797	0.008005*	0.059820	0.009940*	0.037867	0.010995*	0.296834	0.042888*
KS_I2	0.017943	0.007725*			0.023085	0.009292*
KS_I3	0.037163	0.008005*	0.039315	0.011111*	0.037192	0.009329*	0.331778	0.040524*
KS_I4	0.022713	0.006543*	0.027291	0.010421*	0.021284	0.007952*
KS_I9	0.016046	0.007037*			0.024877	0.008434*
KS_I10	0.027138	0.008991*	0.060152	0.010321*	0.022256	0.010361*	0.300356	0.041449*
Constant	0.250215	0.029866*	0.156848	0.053203*	0.222655	0.034914*	−1.735730	0.167557*

^†Heteroskedasticity robust standard errors (SE). *significant at 5%. For generalised linear model, binomial family and logit link were used.
KS – the KIDSCREEN-10 index; KS_I1 – “fit and well”, KS_I2 – “energy”, KS_I3 – “sad”, KS_I4 – “lonely”, KS_I9 – “got on well at school”, KS_I10 – “been able to pay attention”.

ISSN: 1477-7525