OBJECTIVES: Obstructive Sleep Apnea Syndrome (OSAS) is major cause of morbidity & mortality in developing countries. It is estimated that 82% of men & 92% of women with moderate‑to‑severe sleep apnea have not been diagnosed. Therefore, aim of this study is to assess the diagnostic accuracy of Level I versus Level III sleep study device at home as well as at lab for the diagnosis of OSA in adult patient.

METHODS:Meta-analysis of comparative studies of level III versus level I sleep tests in adults with suspected OSA were done. We searched 3 research databases (Cochrane, PubMed, & Embase) from Jan 2007 for studies that reported on diagnostic accuracy parameters. Bivariate mixed-effects logistic regression model used to estimate summary of diagnostic accuracy parameters using Revman 5.3. However, STATA 12 were used to estimate the pooled diagnostic accuracy for mild, moderate and severe OSA.

RESULTS: 12 studies involving a total of 2570 patients were included in the meta analysis. Summary sensitivity of level III for mild, moderate & severe OSA was 89%, 79% & 72% in home diagnosed participants and 97%, 96% & 94% for test conducted at lab. Whereas, specificity of level III for mild, moderate & severe OSA was 71%, 87% & 45% for home and 85%, 89% & 88% at lab. Positive predictive value(PPV) for mild, moderate & severe OSA were 89%, 89% & 57% respectively at home. At lab PPV were 91%, 86% & 77%.Negative predictive value(NPV) for mild, moderate & severe OSA for level III were 72%, 78% & 61% respectively at lab. Whereas, NPV at were 94%, 95% & 96% respectively.

CONCLUSIONS: Level III devices scored well for sensitivity, and specificity with a trade-off of increasing specificity and decreasing sensitivity as disease severity increased from mild to severe. However, Level III device showed good sensitivity at lab.