Computer analysis of oxygen levels in the blood during sleep could — by itself — provide an easy, relatively inexpensive and sufficiently reliable way to determine which children who snore habitually could benefit from a diagnosis and treatment for obstructive sleep apnea. This approach was most accurate for children with severe apnea.
Because of the scarcity of clinical sleep laboratories and certified pediatric sleep specialists — as well as the high costs, inconvenience for parents and children and the need for overnight staff — only a minority of children with sleep apnea, even in the United States and Europe, are thoroughly evaluated. The lack of resources for sleep studies is even more problematic in less developed countries.
“By simplifying the procedure and dramatically reducing the cost, we believe we can evaluate more children who are at significant risk, especially in areas where there is limited access to a pediatric sleep laboratory facility,” said the study’s senior author, David Gozal, MD, MBA, professor of pediatrics at the University of Chicago.
In the study, a multinational group of researchers describe an automated system they developed that incorporates 23 analytic features into a diagnostic neural-network algorithm. All of their data comes from a pulse oximeter, a simple device that clips onto a patient’s fingertip to measure his or her heart rate and blood-oxygen levels overnight. Their research showed that this pared-down approach compared favorably to a full sleep study. The authors estimate it could cut costs by as much as 90 to 95 percent.
The preferred routine for diagnosing sleep apnea in the U.S. is polysomnography. This is based on the use of an oximeter to record oxygen levels, but adds additional components such as brain activity monitors, eye movement assessment, cardiac signal recordings and measures of muscle tension. A video camera typically records each study. This approach, considered definitive, requires a lot of equipment and multiple staff.