P064 The Impact of Artefact-free Recording Time on the Diagnosis of Sleep Disordered Breathing

Background Overnight studies are used to diagnose sleep disordered breathing (SDB), however the minimum artefact-free recording time (AFRT) has not been established in children. Aim To determine the impact of AFRT on SDB diagnoses. Methods Patients attended overnight cardiorespiratory polygraphy/polysomnography, alongside pulse oximetry sleep studies. Respiratory parameter reports were generated using the first 4, 5, 6 and 7 hours of AFRT. Predetermined clinically relevant cut-off (CRCO) values were defined: Obstructive AHI (OAHI; CRCO≥2); Central Apnoea-Hypopnoea Index (CAHI; CRCO≥5); 3% Oxygen Desaturation Index (ODI3%; CRCO≥6); 4% Oxygen Desaturation Index (ODI4%; CRCO≥4). Studies crossing CRCO across different AFRTs were described as ‘Cases of Change’ (COC). Receiver operating characteristic (ROC) curves determined ranges at 4 hours which predicted COC across subsequent AFRTs. Results 137 children (0.39–17.98 years) were consecutively recruited. Mean OAHI, CAHI, ODI3% and ODI4% were 1.54 (σ=2.66), 1.56 (σ=3.43), 5.21 (σ=6.53) and 2.77 (σ=4.42) respectively. For children achieving 7 hours AFRT (n=103), COC from 4 hours were: OAHI≥2 =9.7% (10/103); CAHI≥5 =2.9% (3/103); ODI3%≥6 =3.7% (4/109); ODI4%≥4 =1.8% (2/109). For OAHI≥2, optimal points on ROC curves for predicting COC provided a range of 0.875 (AUC= 0.733; 50% sensitivity; 93% specificity) - 3.125 (AUC= 0.968; 100% sensitivity; 81% specificity). Conclusion Four hours AFRT yields diagnostic results in > 90% cases when commonly used cut-off criteria are applied. For OSA, ranges at 4 hours within which diagnostic change is most likely with longer periods of AFRT are provided. Consideration should be given to repeating short studies where values lie within these ranges.