Table 2.
Agreement between polysomnographic and Fitbit Charge 3™ classification of epoch-by-epoch transitions by considering the corresponding cardiac reactivity (absolute percent change in heart rate).
| … | Epoch 234 | Epoch 235 | Epoch 236 | Epoch 237 | Epoch 238 | … | |
|---|---|---|---|---|---|---|---|
| PSG epoch classification | … | Light | Light | REM | REM | Wake | … |
| PSG EBE transition (current vs. previous epoch) | … | … | Light → Light | Light → REM | REM → REM | REM → Wake | |
| Type of PSG EBE transition | … | … | Same stage | Different stages | Same stage | Different stages | … |
| FC3 epoch classification | … | Light | Light | REM | Light | Wake | … |
| FC3 EBE transition (current vs. previous epoch) | … | Light → Light | Light → REM | REM → Light | Light → Wake | ||
| EBE transition agreement | … | … | 1 | 1 | 0 | 0 | … |
| FC3 HR (bpm) | … | 50 | 52 | 54 | 49 | 58 | … |
| Absolute FC3 HR change (%) (current vs. previous epoch) | … | … | 4.00 | 3.85 | 9.26 | 18.37 | … |
PSG, Polysomnography; FC3, Fitbit Charge 3™; HR, Heart rate; REM, Rapid-Eye-Movement. EBE transitions between consecutive epochs were used to evaluate the relationship between EBE transition accuracy and FC3 HR changes to PSG EBE transitions. EBE transition agreement was considered as accurate (1) when both the current and the previous epoch were equally classified by FC3 and PSG, inaccurate (0) otherwise. The type of PSG EBE transition (same stage vs. different stages) was used as a further predictor of EBE transition agreement.