Correlation Between Oxygen Desaturation Index Measured by Overnight Oximetry and Apnea-Hypopnea Index Measured by...

Background The most common sleep-disordered breathing (SBD) is obstructive sleep apnea (OSA), which affects nearly a billion adults worldwide. OSA is characterized by frequent episodes of a complete (apnea) or partial (hypopnea) collapse of the upper airway, mainly the oropharyngeal tract during sleep, with a consequent cessation or reduction of the airflow. Undiagnosed and untreated OSA can lead to cognitive and neurobehavioral dysfunction, cardiovascular comorbidities, inability to concentrate, memory impairment, and mood changes like irritability and depression, with a remarkable effect on quality of life. Polysomnography (PSG) remains the gold standard for diagnosis; however, it is expensive, less accessible, and has a long waiting period to get it done. Home nocturnal oximetry, measuring the oxygen desaturation index (ODI), is a promising alternative for diagnosis.  Materials and methods A prospective observational study was conducted over two years, involving patients with a high pre-test probability of OSA. Patients underwent both home nocturnal oximetry and in-lab PSG. Statistical analysis was performed using the Spearman correlation, the Wilcoxon signed-rank test, and the receiver operating characteristic (ROC) curves to compare ODI and apnea-hypopnea index (AHI) values. Results Out of 67 participants, 43 (64%) had abnormal home oximetry results (ODI>5) and 62 (92.5%) had abnormal PSG (AHI>5). There was a significant correlation between ODI and AHI (r=0.734, p=0.001). No significant difference was found between ODI values from home oximetry and in-lab PSG. The sensitivity, specificity, and diagnostic accuracy of home oximetry in detecting OSA (ODI>5) were 69%, 100%, and 71%, respectively. The ROC analysis demonstrated an area under the curve (AUC) of 0.98 for predicting AHI>5. Conclusion Home nocturnal oximetry correlates well with level 1 PSG and provides a reliable diagnostic tool for OSA, especially in resource-limited settings. However, further validation is required to improve its sensitivity when ruling out OSA.