Table 3. Mixed linear models and Bland-Altman tests for comparison between EBUS and histologic measurements.
| Mixed linear model | Bland-Altman test (EBUS-histology) | |||||||
|---|---|---|---|---|---|---|---|---|
| Parameter | N | p-value | R2 | Difference mean (SD) | Mean mean (SD) | Slope 1 | Intercept 2 | Systematic bias |
| Perimeter [mm] | 101 | <0.0001 | 0.20 | -5.04 (6.99) | 22.67 (8.52) | -0.42, p<0.0001 | 4.46, p = 0.02 | yes |
| Diameter [mm] | 101 | <0.0001 | 0.30 | 0.14 (1.51) | 4.7 (2.15) | -0.15, p = 0.04 | 0.87, p = 0.03 | yes |
| Lumen area [mm2] | 21 | <0.0001 | 0.80 | -4.59 (15.98) | 34.07 (42.67) | -0.15, p = 0.08 | 0.44, p<0.98 | no |
| AASM*L2 area [mm2] | 98 | <0.0001 | 0.31 | -0.0534 (1.843) | 3.3263 (2.99) | -0.01, p = 0.083 | 0.005, p<0.99 | no |
| AASM/Pi*L2 area/Pi [mm] | 98 | 0.001 | 0.16 | 0.0312 (0.0634) | 0.1323 (0.06) | 0.27, p = 0.02 | -0.004, p<0.78 | yes |
| AASM/Pi2*L2 area/Pi2 [–] | 98 | 0.45 | 0.030 | 0.0029 (0.0038) | 0.0062 (0.002) | 0.93, p<0.0001 | -0.003, p = 0.006 | yes |
1 p values refer to the null hypothesis that the slope = 0.
2 p values refer to the null hypothesis that the intercept = 0.