A validation study of an esophageal probe-based polygraph against polysomnography in obstructive sleep apnea

Study objectives The aim of this study was to validate the automatically scored results of an esophageal probe-based polygraph system (ApneaGraph (R) Spiro) against manually scored polysomnography (Nox A1, PSG) results. We compared the apnea-hypopnea index, oxygen saturation index, and respiratory disturbance index of the devices. Methods Consenting patients, referred for obstructive sleep apnea workup, were tested simultaneously with the ApneaGraph (R) Spiro and Nox A1 (R) polysomnograph. Each participant made one set of simultaneous registrations for one night. PSG results were scored independently. Apnea-hypopnea index, oxygen desaturation index, and respiratory disturbance index were compared using Pearson's correlation and scatter plots. Sensitivity, specificity, and positive likelihood ratio of all indices at 5, 15, and 30 were calculated. Results A total of 83 participants had successful registrations. The apnea-hypopnea index showed sensitivity of 0.83, specificity of 0.95, and a positive likelihood ratio of 5.11 at an index cutoff of 15. At a cutoff of 30, the positive likelihood ratio rose to 31.43. The respiratory disturbance index showed high sensitivity (> 0.9) at all cutoffs, but specificity was below 0.5 at all cutoffs. Scatterplots revealed overestimation in mild OSA and underestimation in severe OSA for all three indices. Conclusions The ApneaGraph (R) Spiro performed acceptably when OSA was defined by an AHI of 15. The equipment overestimated mild OSA and underestimated severe OSA, compared to the PSG.