For decades, the diagnostic pathway for a patient with suspected obstructive sleep apnea (OSA) has followed a predictable sequence. The patient gets tested in a sleep lab overnight, a technician scores the data, and the patient receives an apnea-hypopnea index (AHI) score that quantifies how often they stopped breathing per hour of sleep. If that number is high enough, the patient walks away with a continuous positive airway pressure therapy (CPAP) prescription. That linear protocol has generally made sense because CPAP was, for the most part, the only therapy available to patients.
However, recent years have ushered in a plethora of new sleep apnea treatments. Zepbound became the first pharmacologic therapy approved by the FDA for OSA in December 2024. Apnimed recently published Phase 3 trial results for its oral candidate targeting OSA’s neuromuscular root cause, and the company expects a potential Prescription Drug User Fee Act (PDUFA) target action date early next year. Hypoglossal nerve stimulation has matured into an established surgical option, and oral appliances are becoming more mainstream alternatives for appropriate candidates. While CPAP remains a cornerstone therapy for many patients, it’s no longer the only option. This is good news for the estimated 84 million people in the U.S. living with OSA.
That change presents both an opportunity and a problem for clinicians. Matching the right patient to the right therapy now requires understanding what’s driving the apnea, not just how often it occurs. That puts pressure on a diagnostic system that, for most patients, still hasn’t moved past a single overnight test in a sleep lab, a single AHI number, and a workforce of lab technicians that is increasingly stretched thin. Closing that gap means rethinking what sleep apnea testing identifies and where it happens.
What AHI misses
The limitations of the current standard become clear in the clinic itself. Consider two patients who walk into a sleep clinic and receive the same diagnosis of moderate OSA, with an AHI of 22. Their condition may look identical on paper, but the underlying cause could be vastly different. One patient’s apnea may be driven primarily by an anatomically collapsible upper airway, while the others may be related to obesity, with excess tissue around the neck and upper airway contributing to nighttime airway collapse.
This distinction wasn’t always significant when CPAP was the default treatment for both patients, but now it matters a great deal. The first patient may be a candidate for an oral appliance or hypoglossal nerve stimulation, while the other may benefit most from a GLP-1 therapy such as Zepbound. Those treatment decisions can’t always be made based on AHI alone.
What’s missing from the current diagnostic standard is information about the underlying physiology that produced the AHI in the first place. The signals exist to capture that information, including measurements of airflow, respiratory effort, and oxygenation throughout the night. Yet the diagnostic standard has not been updated to require them or to interpret them as a clinical input rather than a technical detail.
The bottleneck between patients and treatment
Beyond diagnostic standards, there’s also the issue of access. Sleep labs evolved from research environments before becoming the clinical standard for diagnosis, as recognition of sleep-disordered breathing grew in the latter half of the 20th century. That history is visible in how labs operate today, with one patient per bed per night, scored by a sleep technologist, and reviewed by a sleep physician. This caps the number of patients who can move through the system at a time, regardless of how many need evaluation.
The workforce supporting these labs is stretched even thinner than the infrastructure. The ratio of Americans to board-certified sleep medicine specialists is more than 43,000:1, and millions of Americans live in counties without a board-certified sleep specialist. Wait times for an initial specialist appointment are often months long, and the wait for testing can add even more time. For a condition associated with cardiovascular disease, stroke, and metabolic dysfunction, those delays can be consequential.
We should expect that capacity gap to widen. There is growing public awareness of OSA and other sleep issues thanks to the rise of consumer wearables and a fresh set of therapy options that are bringing more new patients into the diagnostic funnel. The system won’t be able to scale to meet that demand by building more sleep labs. It can meet it only by embracing diagnostic work in the home environment, without losing the respiratory signals that make a clinically useful diagnosis possible in the first place.
OSA needs a diagnostic upgrade
Personalized sleep medicine won’t be possible without personalized diagnostics. That principle has held in other specialties like oncology and rheumatology for years, and it’s now making its way into sleep medicine. Endotyping (the characterization of the specific physiological mechanism driving a given patient’s apnea) is the diagnostic upgrade most needed in sleep right now.
Where the AHI captures how often breathing is disrupted, endotyping captures why. Some patients have a breathing control system that overcorrects in response to changes in blood gases. Some have a low arousal threshold and wake from minor disturbances. Others have an anatomically collapsible upper airway. Many patients have a combination of these traits to a certain degree, and measuring each makes a difference in how patients are treated.
Endotyping requires data that the standard diagnostic process doesn’t surface. Physiological measurements of airflow, respiratory effort, and oxygenation need to be recorded and captured across multiple nights, not averaged into a single number. A patient’s sleep apnea severity can change from night to night, and a single overnight test can miss or misclassify cases where the underlying mechanism would have been clearer with more data.
The same diagnostic capability supports something the sleep medicine field also needs: ongoing measurement of how well a given therapy is working for a given patient. Sleep apnea has historically been managed as a one-time diagnosis followed by a long-term prescription, with limited feedback about whether the treatment is producing the expected effect. With multiple therapies now in play, patients and clinicians need a way to assess progress, adjust doses, and combine or switch approaches when results fall short. The diagnostic tools that enable endotyping at the front end are the same tools that enable longitudinal measurement throughout the course of care.
Where we go from here
Encouragingly, the gap between treatment innovation and diagnostic capability is closing. The physiological signals that enable endotyping are well understood, and the technology to capture them outside a sleep lab is growing. The work ahead of us is now largely operational. We must advance diagnostic standards so they include data beyond AHI, expand access to testing options outside the lab, and reframe OSA care as ongoing and requiring continued management.
For patients, this shift should lead to faster diagnoses, fewer mismatched treatments, a more accurate measurement of whether a therapy is working, and options to course-correct early when a therapy isn’t a good fit. For clinicians, it means more comprehensive information and a way to manage sleep apnea more like other chronic conditions.
Sleep medicine is finally moving toward more personalized care, where the rest of medicine has been progressing for a generation. New therapies may be the most visible part of that shift, but the diagnostic system behind sleep medicine will determine how we deliver on the promise of these therapies.
Photo: Roos Koole, Getty Images
Amir Reuveny is the CEO and co-founder of Wesper, a leading provider of advanced home sleep apnea testing (HSAT). Before founding Wesper, Dr. Reuveny was a postdoctoral fellow at Cornell Tech University, investigating commercial applications for flexible electronics. Amir received his Ph.D. in Electrical Engineering and Information Systems from the University of Tokyo and holds a double major degree with honors from Technion, Israel.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
