Skip to main content
. 2019 Apr 11;17:57. doi: 10.1186/s12955-019-1137-6

Table 5.

Efficiency of EQ-5D-5 L and SF-6D to detect differences of dichotomized ODI groups

Measure Dichotomies of ODI health status† N (%) Mean ± SD ES RE ROC curve
AUC 95% CI
EQ-5D-5 L 1 95 (35%) 0.85 ± 0.13 0.42 1.48 0.89* (0.85, 0.93)
2–5 177 (65%) 0.47 ± 0.34
SF-6D 1 95 (35%) 0.7 ± 0.14 0.28 1 0.82* (0.77, 0.87)
2–5 177 (65%) 0.54 ± 0.11
EQ-5D-5 L 1 and 2 193 (71%) 0.76 ± 0.18 0.47 1.4 0.93* (0.90, 0.97)
3–5 79 (29%) 0.22 ± 0.32
SF-6D 1 and 2 193 (71%) 0.64 ± 0.13 0.33 1 0.87* (0.82, 0.91)
3–5 79 (29%) 0.48 ± 0.1
EQ-5D-5 L 1–3 240 (88%) 0.69 ± 0.24 0.27 1.33 0.97* (0.94, 0.99)
4 and 5 32 (12%) −0.04 ± 0.25
SF-6D 1–3 240 (88%) 0.61 ± 0.14 0.2 1 0.91* (0.86, 0.95)
4 and 5 32 (12%) 0.43 ± 0.07

†1 to 5 represents different levels of low back pain measured by ODI from minimal disability to bedbound

*indicates that area under the ROC curve statistically significantly greater than 0.5 and that measure has discriminatory power (P < 0.001)

EQ-5D-5 L 5-level EuroQol 5-dimension, SF-6D Short Form 6-dimension, ODI Oswestry Disability Index, SD Standard deviations, ES Effect size, RE Relative efficiency, ROC Receiver Operating Characteristic, AUC Area Under Curve, CI Confidence Interval