TABLE 1.
Detecting PSG-ORDI >5 events/hr | Detecting PSG-ORDI >15 events/hr | Detecting PSG-ORDI >30 events/hr | |
|
|||
Optimal cut-off (9.53) |
Optimal cut-off (12.65) |
Optimal cut-off (24.81) |
|
Sensitivity | 0.88 | 1.00 | 0.79 |
Specificity | 1.00 | 0.75 | 0.96 |
F1 | 0.94 | 0.89 | 0.86 |
BAC | 0.94 | 0.88 | 0.87 |
Positive predictive value | 1.00 | 0.80 | 0.95 |
Negative predictive value | 0.89 | 1.00 | 0.82 |
Positive likelihood ratio | Inf | 4.00 | 19.0 |
Negative likelihood ratio | 0.12 | 0.00 | 0.22 |
Youden J index | 0.88 | 0.75 | 0.75 |
BAC, balanced accuracy; MM, mandibular movement; ORDI, obstructive respiratory disturbance index; PSG, polysomnography