Diagnosing Obstructive Sleep Apnea: Do At-Home Testing Devices Measure Up?

A middle-age woman with high blood pressure and epilepsy gained 50 lb over the course of 2 years. What was the best approach to diagnosing her problem? The solution was to sleep on it, said Nancy Collop, MD, and Scott Hoff, MD, of Emory University in Atlanta, in JAMA.

On physical examination, the patient's blood pressure was 182/97 mm Hg, oxygen saturation was 98%, and body mass index (BMI) was 51.3; other findings were unremarkable.

When asked about her sleep quality, the patient noted that since her weight gain began, she had been snoring loudly, and occasionally awakened feeling short of breath; nevertheless, she had no reason to suspect she had sleep apnea. She admitted to feeling tired during the day, but denied having headaches or excessive daytime sleepiness.

Lab test results were notable for a serum bicarbonate level of 24 mEq/L (reference range 23-29 mEq/L). Suspecting sleep apnea, the clinicians advised the patient to test herself using a home testing device for a night.

The test showed that of 364 minutes of valid sleep time, the patient had 51.7 apnea-hypopnea index (AHI) events per hour, findings consistent with severe OSA (30 or more events/hour). Her AHI was substantially worse while she was supine (94.2 events/hour) compared with non-supine (43.7 events/hour), Collop and Hoff noted.

Based on the sleep study findings, the team started the patient on continuous positive airway pressure (CPAP) therapy, using a machine with autotitrating pressure set at 6 to 16 cm H2O.

A few days after beginning to use the device, the patient reported a significant improvement in her daytime fatigue. After about 6 weeks of use, data downloaded from the machine showed she had used CPAP on 23 of 30 nights monthly, for an average of 4.5 hours per night; this had resulted in a decrease to 1.2 AHI events per hour.

Discussion

Collop and Hoff noted that OSA affects about 17% of women and 34% of men in the U.S., and is most often diagnosed with a sleep study using either home- or laboratory-based sleep testing.

"The risk of OSA can be assessed based on medical history and pretest prediction models, such as STOP-Bang," they wrote. STOP-Bang is a brief 2-minute screening questionnaire that quantifies risk based on four self-reported criteria (STOP: snoring, tiredness, observed apnea, and high blood pressure) and four demographic items (Bang: BMI, age, neck circumference, and gender).

Collop and Hoff also said that the 2017 American Academy of Sleep Medicine (AASM) clinical practice guideline, which recommends use of home sleep apnea testing or polysomnography among adults with daytime sleepiness or unrefreshing sleep, or other signs and symptoms that raise suspicion of moderate to severe OSA, may be helpful in risk assessment.

Research has suggested that since AHI events occur during sleep, many people may not be aware that they have OSA. In fact, "it has been estimated that up to 80% of individuals with moderate to severe OSA may remain undiagnosed and, more alarmingly, untreated," noted the authors of one commentary.

The prevalence of OSA is substantially higher in some population subgroups compared with the general population, according to the AASM. For example, patients being evaluated for bariatric surgery have an estimated prevalence of 70% to 80%, and patients who have had a transient ischemic attack or stroke have an estimated prevalence of 60% to 70%.

As for determining diagnosis, polysomnography is advised over home testing for patients with the following conditions:

• Significant cardiorespiratory disease
• Possible respiratory muscle weakness due to neuromuscular disease
• Awake hypoventilation
• Long-term opioid medication use
• History of stroke or severe insomnia

"Compared with polysomnography, home sleep apnea testing devices monitor various physiological signals but do not provide a direct assessment of sleep because they do not incorporate electroencephalography, electro-oculography, and surface electromyography," Collop and Hoff wrote.

"Commonly used home sleep apnea testing devices (type 3) measure at least 4 physiological parameters, typically 2 respiratory variables (airflow and thoracic movement), a cardiac variable (electrocardiographic or heart rate), and pulse oximetry," they added. "Respiratory events are recorded if airflow amplitude and oxygen saturation both decrease during a segment of sleep."

Sleep apnea may also be assessed at home with type 4 devices, which use peripheral arterial tonometry (PAT), "a watch-sized computer worn on the wrist that measures the pulsatile arterial waveform associated with cardiac contractions," they noted. "Respiratory events are recorded during episodes of oxyhemoglobin desaturation, heart rate acceleration, and dampening of arterial pulsation amplitude."

Collop and Hoff cited a meta-analysis of 19 studies comparing type 3 home sleep apnea testing devices with polysomnography in adults with suspected sleep-disordered breathing, which showed that home sleep apnea testing had a sensitivity of 93% and a specificity of 60% if the AHI was five or more events per hour, and a sensitivity of 79% and a specificity of 90% if the AHI was 30 or more events per hour.

In addition, a meta-analysis of 20 studies reported a diagnostic accuracy of 61% in classification of OSA severity comparing home sleep apnea testing and polysomnography. Another meta-analysis of PAT-based home sleep apnea testing that included 13 trials reported a sensitivity of 96% and a specificity of 44% for patients with an AHI of five or more events per hour, and a sensitivity of 80% and a specificity of 90% for patients with an AHI of 30 or more events per hour.

According to the 2022 Medicare fee schedule, reimbursement for home sleep apnea tests is $93.44 for type 3 devices and $164.03 for PAT-based devices.

"Although polysomnography provides a comprehensive assessment of sleep, it is more expensive than a home sleep apnea testing device and requires travel to a sleep center, which is inconvenient for some patients," Collop and Hoff concluded.

Comment