Key points are not available for this paper at this time.
Objectives To investigate efficacy of sleep video clips to diagnose Obstructive Sleep Apnoea (OSA). To develop a validated scoring system to predict clinically significant OSA. Methods 111 children (aged 2–12) referred for OSA diagnosis underwent lab polysomnography (PSG). Parents were shown video recording instructions, they recorded videos on night of study and in the home. Sleep videos were scored using a tool, developed by a panel of sleep consultants. It consists of a 9-point system assessing OSA-indicative behaviours (figure 1). The chief investigator scored each video with a possible score of 9. Inter-scorer reliability was assessed by a second scorer. Both scorers were blinded to the PSG score which was carried out by the senior sleep physiologist. The obstructive apnoea hypopnea index (OAHI) was used to define severity of OSA. None/mild OSA defined as OAHI 5. Median and range was used to summarise continuous variables. Chi-Square test was used to compare categorical variables such as presence of behaviours against OSA. Spearman correlation was used to compare lab and home video scores. Kappa statistic was used to compare agreement between scorers. ROC curves were used to assess optimal cut-off score to distinguish between absence and presence of clinically significant OSA. Stata version 15 was used for analysis. Results 49 children had moderate/severe PSG diagnosed OSA. 62 had no/mild OSA. Median scores were 6 and 2 for moderate/serve, and none/mild OSA, respectively. The distribution of video scores are outlined in figure 2. Kappa statistics showed substantial to fair agreement between primary and secondary scorers for the nine behaviours. Lab and home video scores correlated strongly (rs=0.805, pConclusion This tool effectively predicted the presence or absence of moderate/severe OSA. Given prolonged wait times for diagnostic sleep studies, a screening tool could assist in triaging referrals. A high-negative predictive power tool could enable 'watch-and-wait' approach. A high-positive predictive power tool could identify high-risk patients who would need more urgent sleep diagnostics.
Habibi et al. (Tue,) studied this question.