0873 Meta-Analysis of Diagnostic Test Performance of Type 3 Home Sleep Apnea Tests and Peripheral Arterial Tonometry Studies in Diagnosis of Pediatric OSA

Abstract Introduction The gold standard for the diagnosis of obstructive sleep apnea (OSA) in children is polysomnography (PSG). In this study, we aimed to study the diagnostic test accuracy (DTA) of type-III home sleep apnea tests (HSAT) and peripheral arterial tonometry (PAT) studies in children with OSA. Methods Studies reporting sufficient information to construct 2x2 confusion matrices were included. For type-III HSATs, bivariate weighted mean sensitivity, specificity, positive and negative likelihood ratios (+LR and -LR) were estimated using PSG apnea-hypopnea index (AHI) cut-offs >1, > 5 and > 10. For PAT studies, a univariate model was used for summary estimates. Concordance between PSG and type-III HSATs in OSA severity classification was assessed with Cohen’s kappa. Results DTA meta-analyses for type-III HSATs and PAT studies included 6 and 2 studies, respectively. With the PSG AHI cut-offs of > 1/hour, > 5/hour and > 10/hour, mean sensitivity and specificity of type-III HSATs were 0.85 [0.66, 0.94] and 0.33 [0.03, 0.89], 0.88 [0.63, 0.97] and 0.93 [0.66, 0.99], and 0.65 [0.47, 0.80] and 1.00 [0.17, 1.00], respectively. The +LRs and -LRs using PSG AHI cut-offs of > 1/hour, > 5/hour and > 10/hour were 1.3 [0.5, 3.3] and 0.46 [0.05, 4.05], 12.4 [2.1, 72.6] and 0.13 [0.04, 0.46], and 197.9 [0.1, 308670.3] and 0.35 [0.21, 0.58], respectively. Pooled Cohen’s kappa for concordance between type-III HSATs and PSGs for mild, moderate and severe severity of OSA were 0.36 (SE=0.19), 0.14 (SE=0.34), and 0.64 (SE=0.15), respectively. DTA meta-analysis for PAT studies using PSG AHI cut-off of > 5/hour, showed univariate pooled sensitivity and specificity of 0.94 [0.81, 0.98], and 0.95 [0.51, 0.99], respectively. Conclusion Diagnostic accuracy of type-III HSATs increases with increasing PSG AHI diagnostic thresholds, indicating that the test is only good for ruling-in but not for ruling-out OSA in children with high pre-test probability of severe OSA. Substantial agreement was noted between type-III HSATs and PSG in classifying OSA as severe but not in classifying OSA as mild or moderate. DTA meta-analysis on PAT studies showing high sensitivity and specificity, albeit based only on 2 studies, needs to be evaluated further with more studies. Support (if any)