Table 2

Predictive performance of candidate OLS-mapping models in the validation cohort (n=1753)

Model	MAE	RMSE	Adjusted R²
Total score‡	0.115	0.151	0.298
Total score+total score²	0.119	0.152	0.306
Dichotomised, all individual item scores	0.117	0.155	0.285
All individual item scores§	0.111	0.150	0.324
Most parsimonious model based on individual item scores	0.112	0.150	0.323
Most intuitive model based on individual item scores	0.112	0.150	0.319

‡In acknowledgement that not all who wish to use a mapping algorithm will have access to individual item score data, the total score algorithm which did not perform as well as the preferred individual MSWS-12 item-score equation is provided here (EQ-5D=−0.003*^† (transformed MSWS-12 total score)+0.894 (constant)); note that an asterisk (*) indicates the item was multiplied by/multiplication operator and a dagger (^†) denotes a p value <0.05.
§Preferred model.
MAE, MSWS-12, 12-item Multiple Sclerosis Walking Scale; mean absolute error; OLS, ordinary least squares; RMSE, root mean square error.