The validation of published utility mapping algorithms: an example of EORTC QLQ-C30 and EQ-5D in non-small cell lung cancer

Table 3 Observed vs predicted goodness of fit statistics over the range of observed EQ-5D values

Mapping	Overall N = 4382			EQ-5D < =0.5 N = 288			0.5 < EQ-5D < =0.75 N = 1147			0.75 < EQ-5D < =1 N = 2947
Mapping	O-P	MAE	RSME	O-P	MAE	RSME	O-P	MAE	RSME	O-P	MAE	RSME
Young et al. (2015) [12]	0.022	0.087	0.119	−0.092	0.207	0.264	− 0.007	0.090	0.123	0.045	0.074	0.091
Khan and Morris (2014) [13]	0.129	0.176	0.211	−0.312	0.313	0.384	0.038	0.058	0.069	0.208	0.208	0.223
Khan et al. (2016) [14]	0.122	0.178	0.219	−0.360	0.360	0.427	0.013	0.046	0.057	0.212	0.212	0.229

Key: MAE Mean absolute error, N Number of questionnaires completed, O-P Difference between mean observed and predicted EQ-5D utility values, RSME Root mean square error

ISSN: 2191-1991