Table 4.
Multinomial logistic modeling: predicting EQ-5D index scores from SF-36
| Model 11 a | Model 12 b | Model 13 c | Model 14 d | |||||
|---|---|---|---|---|---|---|---|---|
| Derivation set | ||||||||
| R2 | 0.452–0.611 | 0.445–0.607 | 0.455–0.615 | 0.421–0.576 | ||||
| MAE | 0.088 | 0.121 | 0.084 | 0.099 | ||||
| AE > 0.05 (%) | 32.8 | 49.8 | 32.2 | 34.3 | ||||
| AE > 0.1 (%) | 25.0 | 45.5 | 23.7 | 26.0 | ||||
| RMSE | 0.297 | 0.347 | 0.290 | 0.315 | ||||
| Predicted EQ-5D index | ||||||||
| Mean (SD) | 0.826 | (0.279) | 0.731 | 0.310 | 0.826 | (0.279) | 0.829 | (0.185) |
| Min/max | −0.171 | /1.000 | −0.171 | 1.000 | −0.171 | /1.000 | −0.171 | /1.000 |
| Internal validation set | ||||||||
| MAE | 0.097 | 0.119 | 0.092 | 0.101 | ||||
| AE > 0.05(%) | 35.9 | 48.5 | 37.2 | 35.9 | ||||
| AE > 0.1(%) | 28.7 | 43.2 | 28.9 | 28.1 | ||||
| RMSE | 0.312 | 0.344 | 0.304 | 0.315 | ||||
| Predicted EQ-5D index | ||||||||
| Mean (SD) | 0.825 | (0.288) | 0.732 | 0.313 | 0.821 | (0.292) | 0.828 | (0.292) |
| Min/max | −0.171 | /1.000 | −0.171 | 1.000 | −0.171 | /1.000 | −0.171 | /1.000 |
| External validation set | ||||||||
| MAE | 0.085 | 0.125 | 0.075 | 0.084 | ||||
| AE > 0.05 (%) | 47.2 | 60.2 | 45.5 | 44.7 | ||||
| AE > 0.1 (%) | 26.0 | 52.0 | 26.0 | 26.8 | ||||
| RMSE | 0.292 | 0.354 | 0.273 | 0.290 | ||||
| Predicted EQ-5D index | ||||||||
| Mean (SD) | 0.883 | (0.164) | 0.753 | 0.204 | 0.892 | (0.131) | 0.877 | (0.170) |
| Min/max | −0.171 | /1.000 | −0.171 | 1.000 | 0.151 | /1.000 | −0.171 | /1.000 |
MAE, mean absolute error; AE, absolute error ; RMSE, root mean squared error.
aIndependent variables: PF, RP, BP, GH, VT, SF, RE, MH.
bIndependent variables: PF, BP, SF, RE, MH.
cIndependent variables: PF, BP, GH, SF, RE, MH, PF squared, SF squared, RE squared.
dIndependent variables: PCS, MCS, PCS × PCS, MCS × MCS, PCS × MCS.