Validation of a portable monitoring system for the diagnosis of obstructive sleep apnea syndrome

Abstract

Study objective: To evaluate if a portable monitor could accurately measure the apnea-hypopnea index (AHI) in patients with a suspicion of obstructive sleep apnea (OSA).

Design: Prospective and randomized.

Setting: Sleep laboratory.

Participants: 80 participants: 70 patients with clinical OSA suspicion and 10 subjects without suspicion of OSA.

Interventions: N/A.

Measurements and results: Three-order randomized evaluations were performed: (1) STD (Stardust II) used at the participants' home (STD home), (2) STD used simultaneously with PSG in the sleep lab (STD+PSG lab), and (3) PSG performed without the STD (PSG lab). Four AHI values were generated and analyzed: (a) STD home; (b) STD from STD+PSG lab; (c) PSG from STD+PSG (named PSG+STD lab); and (d) PSG lab. Two technicians, blinded to study details, performed the analyses of all evaluations. There was a strong correlation between AHI from the STD and PSG recordings for all 4 AHI values (all correlations above 0.87). Sensitivity, specificity, and positive and negative predictive values at AHI cut-off values of 5, 15, and 30 events/hour were calculated. AHI values from the PSG lab and PSG+STD lab were compared to STD home and STD+PSG lab and showed the best results when STD and PSG were performed simultaneously. In all analyses, the area under ROC curve was at least 0.90. With multiple comparisons, diagnostic agreement was between 91% and 75%. The Bland Altman analyses showed strong agreement between AHI values from the STD and PSG recordings, especially when comparing the AHI from simultaneous STD and PSG recordings.

Conclusion: These data suggest that the STD is accurate in confirming the diagnosis of OSA where there is a suspicion of the disorder. Better agreement occurred during simultaneous recordings.

Figure 1

Pearson correlation between AHIs from STDs and PSGs in the 4 recordings. A: PSG lab vs. STD Home (r = 0.876; P < 0.0001; 95% CI = 0.81 to 0.91); B: PSG lab vs. STD+PSG lab (r = 0.873; P < 0.0001; 95% CI = 0.80 to 0.91); C: PSG+STD lab vs. STD Home (r = 0.892; P < 0.0001; 95% CI = 0.83 to 0.92); D: PSG+STD lab vs. STD+PSG lab (r = 0.892; P < 0.0001; 95% CI = 0.83 to 0.93).

Figure 2

ROC curves of different AHI cut-offs (5, 15, and 30, columns respectively) from STDs and PSGs considering the 4 recordings. A: PSG lab vs. STD Home (AUC = 0.90, 0.92, and 0.95, respectively); B: PSG lab vs. STD+PSG lab (AUC = 0.91, 0.93, and 0.95, respectively); C: PSG+STD lab vs. STD Home (AUC = 0.95, 0.95, and 0.96, respectively); D: PSG+STD lab vs. STD+PSG lab (AUC = 0.97, 0.98, and 0.98, respectively). AUC = area under the ROC curve.

Figure 3

Bland Altman analysis between AHIs from STDs and PSGs considering the 4 recordings. A: PSG lab vs. STD Home; B: PSG lab vs. STD+PSG lab; C: PSG+STD lab vs. STD Home; D: PSG+STD lab vs. STD+PSG lab.