Comparison of Full Versus Short Induced-Sleep Polysomnography for the Diagnosis of Sleep Apnea

Objectives/Hypothesis: Polysomnography (PSG) is the gold-standard method for diagnosing obstructive sleep apnea (OSA). However, the gap between demand and capacity in performing PSG is a major health-care problem. We sought to validate a short day-time induced sleep for the diagnosis of OSA. Study Design: Prospective diagnostic method validation. Methods: We studied 25 consecutive patients referred to the sleep laboratory and 15 healthy volunteers. All subjects were evaluated by means of full overnight PSG (Full-PSG) and short day-time induced-sleep PSG (Induced-PSG). Sleep was monitored during both procedures (Embla, 16 channels). Sleep was induced by slow intravenous drip infusion of midazolam. Results: The population studied (N = 40) was 60% male (mean age, 42 +/- 10 years; body mass index, 29 +/- 6.5 kg/m(2)). Sleep was successfully induced in all subjects, and no complications were observed (midazolam doses, 6.2 +/- 3.8 mg; time of induced sleep 41.5 +/- 18.9 minutes). The apnea-hypopnea index (AHI) and minimal oxygen saturation during Full-PSG versus Induced-PSG were similar: median AHI (with 25%-75% interquartile range) was 13 (3-35) events per hour versus 17 (4-36) events per hour, and median oxygen saturation was 84% (75-90) versus 85% (76-92); P = .89 and P = .53, respectively. The majority of the respiratory events during induced sleep were obstructive and similar to those observed during Full-PSG. AHI and lowest oxygen saturation during Induced-PSG correlated significantly with Full-PSG (r = 0.67 and r = 0.77, respectively). Sensitivity and specificity for the diagnosis of OSA (AHI > 15 events per hour) by Induced-PSG were 0.83 and 0.72, respectively. Conclusions: Induced-PSG by midazolam during the day is safe and correlates with Full-PSG; it therefore is a promising alternative method in the diagnosis of OSA