Table 4 Model performance for predicting NEI-VFQ-25 composite score based on EQ-5D-5Â L utility values
| Â | Model 1Â A Tobit Regression with Log of Composite Score | Model 2Â A Tobit Regression | Model 3Â A OLS with log of Composite Score | Model 4Â A OLS | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Coef. | p-value | 95% CIs | Coef. | p-value | 95% CIs | Coef. | p-value | 95% CIs | Coef. | p-value | 95% CIs | ||
Utility value | 0.852 | 0.000 | (0.698, 1.005) | 28.356 | 0.000 | (21.7, 35.01) | 0.842 | 0.000 | (0.688, 0.996) | 28.435 | 0.000 | (21.73, 35.14) | |
Vision in better eye | Mild | -0.041 | 0.457 | (-0.15, 0.068) | -4.570 | 0.058 | (-9.3, 0.16) | -0.042 | 0.451 | (-0.151, 0.068) | -4.565 | 0.061 | (-9.33, 0.2) |
Moderate | -0.352 | 0.000 | (-0.456, -0.248) | -27.045 | 0.000 | (-31.57, -22.52) | -0.354 | 0.000 | (-0.458, -0.249) | -27.029 | 0.000 | (-31.59, -22.47) | |
Severe | -0.751 | 0.000 | (-0.903, -0.599) | -46.385 | 0.000 | (-52.99, -39.78) | -0.756 | 0.000 | (-0.909, -0.603) | -46.346 | 0.000 | (-53, -39.69) | |
Blind | -1.198 | 0.000 | (-1.354, -1.043) | -54.687 | 0.000 | (-61.45, -47.93) | -1.199 | 0.000 | (-1.355, -1.042) | -54.479 | 0.000 | (-61.28, -47.68) | |
Constant | 3.680 | 0.000 | (3.521, 3.838) | 63.027 | 0.000 | (56.17, 69.88) | 3.689 | 0.000 | (3.53, 3.847) | 62.954 | 0.000 | (56.05, 69.86) | |
R Square | 0.735 | 0.164 | 0.8012 | 0.7966 | |||||||||
Adjusted R Square | Â | Â | 0.7978 | 0.7931 | |||||||||
AIC | 188.55 | 2441.5 | 179.0656 | 2443.744 | |||||||||
BIC | 214.57 | 2467.426 | 201.2883 | 2465.966 | |||||||||
MAE | 0.2058 | 10.26743 | 0.207 | 10.27025 | |||||||||
RMSE (95% CI) | 0.3195 (-0.0035, 0.037) | 13.92907 (-1.562, 1.608) | 0.3197 (-0.036, 0.036) | 13.92893 (-1.585, 1.585) | |||||||||
Breuch- Pagan | Â | Â | 0.0510 | 0.0883 | |||||||||
VIF | 1.44–2.27 | ||||||||||||