Table 1. Studies selected after full text screening.
| Author | Study type | Number of Patients | Does it compare with PSG1? | Does it show the recording time (in minutes)? | Does it show the time (in minutes) spent in supine position? | Results |
|---|---|---|---|---|---|---|
| Mello et al. 17 | Cohort study | 478 patients | Yes | Yes: 398 (355–434) for PSG1 and 423 (360–467) for PM3 | Yes: 41.0 (16.4–66.7) for PSG1 and 33.9 (15.8–51.1) for PM3 | PSG1 may increase time spent in the supine position and overestimate OSA severity |
| Wittine et al. 22 | Cohort study | 129 patients | No | Yes: 475 (446–479) | No | At least 300 minutes of recording are required for significant test accuracy |
| Cheliout- Heraut et al. 24 | Clinical trial | 90 patients suspected of OSA | Yes | seven recordings were excluded due to a total sleep time < 5 hours | No | Automatic event detection based on mandibular movement has important potential |
| Yin et al. 27 | Clinical trial | 44 patients | Yes | 436,3±87,7 FOR pm3, 426,3 ±71,2 FOR PSG1 | Yes: 62.7 ± 26.4% for PSG1 and 51.3 ± 30.0% for PM3 | At least 390 minutes required in the PM3, longer supine time in the PSG1. |
| Planès et al. | Clinical trial | 50 patients | Yes | Yes: 346 ± 62 (105–438) | No | PM3 is suitable for coronary artery disease patients |
| Driver et al. 25 | Clinical trial | 73 patients suspected of OSA | No | 239 (62-409) for PSG1 and 379 (58-467) for Pm3 | No | Low registration time reduced the agreement between PSG1 and PM3 |
| Reichert et al. 26 | Clinical trial | 51 patients | Yes | 37.95 ± 25.10% FOR PSG1 AND 33.96 ± 22.61% FOR PSG3 | No | High agreement, high negative predictive value |
| Kingshott et al. 28 | Clinical trial | 16 patients | No | Yes: 382 ± 118 | Yes: 37.95 ± 25.10% for PSG1 | Better sleep quality on PM3, no difference in objective sleep on the next day |
| Masa et al. 29 | Clinical trial | 348 patients | Yes | 442.5 ± 42.8 for PSG1 and 428.3 ± 81.9 for PM3 | No | Lower AHI in PM3 by recording time significantly longer than sleep time |
| Ng et al. 23 | Clinical trial | 90 patients | Yes | 402.7 ± 74.5 FOR PSG1 AND 477.3 ± 87.1 FOR PM3 | Yes: 73 ± 0.25% | High sensitivity, specificity, and negative predictive value |
| Hernández-Bendezú et al. 30 | Cohort study | 70 patients | No | Yes: 465 (276–546) | Yes: 76.3 (0.6–464.5) | PM3 with good signal quality even with patients moving around the city after PM3 has been connected |
| Hsu et al. 31 | Cohort study | 215 hypertensive patients | No | No | 56,6 ± 29,8% | The time spent in the supine position is a predictor of OSA. |
| Oliveira et al. 32 | Clinical trial | 26 patients with COPD | Yes | No | No | High loss of PM3 signal suggests difficulty using nasal cannula |
| Guerrero et al. 33 | Clinical trial | 56 patients | Yes | Yes: 380.2 ± 66.0 for PSG1 and 401.9 ± 59.6 for PM3 | Yes: 45.6 ± 24.2% for PSG1 and 46.6 ± 10.3% for PM3 | PM3 together with detailed medical evaluation can be used in patients without high pretest probability, time in the supine position much shorter in PM3 |
| Gjevre et al. 34 | Clinical trial | 47 patients suspected of OSA | Yes | No | 33.96 ± 22.61% for PM3 and | No significant difference in time in the supine position |
| Vonk et al. 35 | Cohort study | POSA and non-apneic snoring patients | No | No | 43.1% during the PSG1 night phase compared with 28.6% for PSG3 | Using the PSG1 apparatus leads to an increase in the percentage of supine sleeping position causing an overestimation of OSA severity |
| Levendowski et al. 36 | Clinical trial | 37 patients | Yes | No | No | Results showed less variation in the night-to-night AHI in PM3 than in PSG1 |
| Ng et al. 37 | Clinical trial | 50 patients | Yes | Yes: 402.7 ± 74.5 for PSG1 and 477.3 ± 87.1 for PSG3 | Significant test sensitivity and specificity |
Abbreviations: AHI, apnea-hypopnea index; COPD, chronic obstructive pulmonary disease; OSA, obstructive sleep apnea; PM3, level 3 portable monitor; POSA, obstructive sleep apnea; PSG1, type-1 polysomnography; PSG3, type-3 polysomnography.