Study Objectives: A decrease in REM time during polysomnography (PSG) in patients with obstructive sleep apnea (OSA) can result in underestimation of apnea hypopnea index (AHI). We propose adjusting AHI to normalized REM% in subjects with REM% ≤15% to avoid under diagnosis of OSA. Methods: All children who completed diagnostic PSG from 2016 to 2023 with REM% of ≤ 15% of total TST were selected for adjustment. AHI Adjustment was done by multiplying AHI by a normalization factor (20%)/REM%). Changes in OSA diagnosis and severity were evaluated before vs after adjustment. Intra class comparison and Bland-Altman plots were used to evaluate agreement level of adjusted AHI vs non-adjusted AHI with REM AHI as the reference. P value <0.05 was significant. Results: Of 389 children reviewed, only 79 (20%) children had low REM% of ≤ 15%. Median (range) age was 12.8(0.9-18) years with Male/female ratio 2.3/1. Mean (SD) sleep efficiency was 64.7% (12.3). Mean (SD) REM% was 10.5% (3.4). Median AHI (range) before AHI adjustment was 1.7(0-44) events/hour vs 4.1 (0-74.4) events/hour after AHI adjustment (P<0.001). Adjusted AHI had better agreement with REM- AHI (ICC=0.691; 95% CI: 0.58,0.80) than non-Adjusted AHI (ICC=0.485; 95% CI: 0.39,0.58). AHI adjustment changed diagnosis from normal to mild OSA in 12 (15%) patients, from mild to moderate OSA in 7(9%) patients, and from moderate to severe OSA in 9 (11%) patients (p=0.044). Conclusions: Adjusting AHI in patients with low REM% to a normalized REM% can help avoid underdiagnosis of OSA and/or underestimation of its severity.
Keywords: AHI; OSA; REM AHI; adjusted AHI; apnea-hypopnea index; obstructive sleep apnea; sleep-disordered breathing.
© 2025 American Academy of Sleep Medicine.