0872 Meta-Analysis of Diagnostic Test Accuracy of Pediatric Sleep Questionnaire in the Screening of Pediatric Obstructive Sleep Apnea

Abstract Introduction In this study we aimed to study the diagnostic test accuracy (DTA) of Pediatric Sleep Questionnaire (PSQ) in the diagnosis of obstructive sleep apnea (OSA) in children. Methods Studies reporting sufficient information to construct 2x2 confusion matrices were included. The diagnostic threshold AHI from PSG was AHI >1. Bivariate random-effects model was used to calculate mean sensitivity, specificity, positive and negative likelihood ratios (+LR, -LR), diagnostic odds ratio (DOR) and their 95% confidence intervals (CIs), summary receiver-operating curve (SROC), Fagan’s nomogram and LR scattergram. Where reported, a pooled mean estimate of correlation between PSQ, and PSG-determined AHIs were computed. Results DTA meta-analyses included 4 studies. The bivariate weighted mean sensitivity and specificity pooled sensitivity, specificity, +LR, -LR and DOR of PSQ from 4 studies were 0.77 [0.62, 0.87], 0.42 [0.29, 0.55], 1.3 [1.1, 1.6], 0.56 [0.36, 0.88], 2.0 [1.0, 4.0], respectively. The area-under-the-curve in the SROC was 0.69 [0.65, 0.73]. +LR and −LR pairings were plotted a quadrant of LR scattergram that indicated the test was neither good for ruling-in nor for ruling-out OSA. With 43% prevalence of OSA in the pooled cohort of studies, Fagan’s nomogram predicted OSA probabilities of 50% when the test is positive and 30% when the test is negative. The pooled Spearman’s correlation coefficients between PSQ and PSG-determined AHIs was 0.18 (0.10; 0.25), p< 0.001, based on 8 studies. Conclusion As a screening questionnaire for pediatric sleep apnea, PSQ showed acceptable sensitivity but poor specificity. Further studies are required to assess the role of PSQ in screening for pediatric OSA. Support (if any)