Longitudinal wearable monitoring suggests clinicians may be over-relying on PAP-derived metrics when assessing treatment success. A sleep center founder recommends a different type of remote patient monitoring solution.
By Jerald H. Simmons, MD
(Disclosure: Simmons is the founding director of REST Technologies and Comprehensive Sleep Medicine Associates; the data presented reflect clinical experience within his practice.)
Over the past five years, our team at Comprehensive Sleep Medicine Associates, with four locations throughout Houston and Austin, Texas, has gained extensive experience implementing nightly remote patient monitoring (RPM) of obstructive sleep apnea (OSA) patients. We achieved this using the REST Tracker system, developed by REST Technologies, a plug-and-play software platform that leverages longitudinal physiologic data and artificial intelligence-assisted review, to facilitate RPM for the practice. The REST Tracker platform leverages data obtained from SleepImage’s home sleep testing solution. TheSleepImage (SI) system is Food and Drug Administration (FDA)-cleared for the diagnosis and management of OSA using, at its core, cardiopulmonary coupling analysis from data collected using a ring oximeter.
The SI system generates sleep-related metrics, mainly the sAHI3% and sAHI4%. These serve as surrogates for the two standard apnea hypopnea indexes (AHIs) clinically used to describe sleep apnea severity. The sAHI4% is less sensitive than the sAHI3%, analogous to the American Academy of Sleep Medicine AHI definitions based on 1A versus 1B hypopnea criteria. This methodology received FDA clearance in 2019 (510[k] K182618) for home sleep apnea testing (HSAT) use in diagnosing and managing sleep disordered breathing.
Comprehensive Sleep Medicine Associates now manages OSA patients across treatment modalities—including positive airway pressure (PAP), hypoglossal nerve stimulation (Inspire), oral appliance therapy, and others—in a treatment-agnostic manner. As early as 2022, we provided reports of the clinical feasibility of using this approach.1
What we have found over these five years is that RPM using SI’s FDA-cleared wearable technology for OSA RPM can address major limitations inherent in systems that rely exclusively on PAP-derived data for OSA metrics.
Most RPM platforms for OSA rely on an AHI-type metric exported from PAP devices and derived from airflow signals. In this article, we describe Comprehensive Sleep Medicine Associates’ RPM protocol, using the REST Tracker, outline its advantages over PAP-derived indices, and illustrate its clinical impact through representative cases.
FDA-Cleared Versus Consumer Sleep Rings
A key differentiator of SI-based analysis compared with many consumer rings is its continuous sampling of the plethysmography signal. Sampling at 100 Hz throughout the night provides a continuous plethysmograph signal by the SI method, detecting beat-to-beat changes in pulse wave amplitude modulated by respiratory movements. From these data, a plethysmography-derived respiratory signal can be extracted along with heart rate variability (HRV) changes calculated from beat-to-beat peak waveform analysis. Combining respiratory and HRV signals enables cardiopulmonary coupling analysis from the data obtained by the ring oximeter. During apnea, normal respiratory-related pulse amplitude oscillations disappear, allowing respiratory events to be identified.
In contrast to FDA-cleared systems designed for respiratory event detection, many consumer sleep rings are not engineered to provide continuous high-frequency plethysmographic sampling throughout the night. To optimize battery performance and user convenience, these devices often rely on intermittent bursts of higher sampling interspersed with lower-resolution recording, using interpolation to estimate physiologic data between higher sampled intervals. While this approach may be sufficient for general wellness tracking, it limits the ability to reliably detect respiratory events that must persist for at least 10 seconds to meet clinical scoring criteria. Consequently, the respiratory metrics generated by such consumer devices do not offer the signal fidelity required for diagnostic determination and monitoring of OSA.
While the SI system, and home testing devices in general, lower the barrier to perform HSAT, it does not replace in-laboratory polysomnography. In our experience, in-lab polysomnography, when performed properly, has higher sensitivity than the SI system and all HSAT devices.
AHI Discrepancies
Shortly after initiating nightly monitoring in 2021, we observed frequent discrepancies between
SI-derived metrics and PAP-reported AHIs. PAP device data were imported into the REST Tracker, making these differences even easier to appreciate. We subsequently analyzed 7,835 nights of simultaneous monitoring with the SI ring and PAP, including only nights with at least 80% agreement in recording time.
The results, presented at SLEEP 2025, were striking. PAP-AHI was 4.2 events (+/-6.7), compared with a mean sAHI4% of 5.3 (+/-5.7) and a mean sAHI3% of 10.7 (+/-8.5). Paired t-tests demonstrated statistically significant differences for both comparisons (p < 0.001). Bland-Altman analysis showed mean differences of 6.47 and 1.17 for sAHI3% and sAHI4% versus PAP-AHI, respectively, with wide 95% limits of agreement, indicating variability and lack of equivalence.2
When making management decisions, clinicians unfortunately often treat the PAP-reported AHI as equivalent to in-laboratory polysomnographic AHI. When PAP downloads from the machine show good adherence and a low AHI, therapy is typically considered optimized. Consequently, when patients report persistent symptoms—such as nonrestorative sleep, daytime fatigue, or frequent awakenings—residual OSA may be overlooked if the PAP-AHI remains below 5 events/hour.
Clinical Cases
At Comprehensive Sleep Medicine Associates, REST Tracker data has guided management in numerous clinical scenarios that would have been missed using PAP-derived data alone. Longitudinal monitoring has also highlighted substantial night-to-night variability in OSA severity, which short-term studies cannot capture. When modifying therapy, extended monitoring is often necessary to identify contributing factors and confirm treatment response.3 Below are two clinical examples where SI sAHI metrics with the REST Tracker assisted clinical interventions that would not have been possible by PAP RPM monitoring alone.
Clinical case 1. A 78-year-old woman with OSA, previously stable on CPAP, took a vacation to New Mexico at elevations above 5,000 feet. She developed increased fatigue and malaise. She was flagged on the REST Tracker dashboard due to worsening sAHI and oxygen saturation metrics. Her PAP-reported AHI remained within normal limits throughout.
On contacting the patient, we learned of her high-altitude exposure. Her sAHI3% ranged from 30 to 40 events/hour, and time below 88% oxygen saturation exceeded two hours per night—all while the PAP-AHI metric remained below 5 events per hour.
PAP pressures were adjusted with partial improvement. Ultimately, supplemental oxygen via a home concentrator was added. With these interventions, the SI output parameters normalized, and her symptoms improved.
Clinical case 2. A 74-year-old man with OSA treated with combination Inspire therapy and PAP had been stable, according to REST Tracker monitoring. He later developed intermittent worsening of OSA detected only by the sAHI metrics while PAP-reported AHI remained normal. During focused follow-up, he reported binge alcohol consumption on nights corresponding to elevated sAHI. Without the ongoing REST Tracker SI-based output metrics being monitored, these episodic exacerbations would have gone unrecognized by PAP data monitoring alone.
Over the past five years, we have encountered many similar scenarios that would not have been detected without this RPM approach.
Importantly, the method is agnostic to treatment modality. We routinely use REST Tracker in patients treated with hypoglossal nerve stimulation, often avoiding in-laboratory titration by adjusting settings based on longitudinal REST Tracker SI-derived metrics. We also collaborate with dentists to assist in oral appliance titration, monitoring responses over weeks rather than relying on isolated follow-up studies.
Future Management of OSA
Longitudinal monitoring has also enabled exploration of RPM in at-risk populations. REST Technologies was awarded a National Institutes of Health grant to monitor obese pregnant women without a prior OSA diagnosis throughout pregnancy to detect the development of OSA. Monitoring began at the beginning of the second trimester and continued until delivery.
Results, presented at SLEEP 2025, demonstrated that nearly all participants had worsening sleep apnea metrics over time, with mean sAHI3% increasing from 5.01 (±2.79) early in pregnancy to 9.35 (±5.74) in late pregnancy (p < 0.001). This approach may ultimately support routine prenatal monitoring, guiding intervention to reduce adverse outcomes associated with OSA, such as gestational hypertension and diabetes, as well as pre-term labor.4
In conclusion, while RPM is gaining acceptance in sleep medicine, not all RPM approaches are equivalent. Because OSA exhibits marked night-to-night variability in many patients, longitudinal monitoring provides a more accurate foundation for management than isolated studies, and the REST Tracker SI approach is superior to the PAP-derived metrics alone. REST Technologies is focusing on developing the REST Tracker to streamline clinic workflow. For RPM to be sustainable in clinical practice, it must integrate seamlessly into existing operations without overburdening staff.
At Comprehensive Sleep Medicine Associates, the REST Tracker is now used routinely in patient care, reflecting an emerging model for the future of sleep medicine.
Jerald H. Simmons, MD, is triple board-certified in neurology, sleep, and clinical neurophysiology. Prior to founding Rest Technologies, Comprehensive Sleep Medicine Associates, and the Sleep Education Consortium, he trained in sleep medicine at Stanford University and was on faculty in the department of neurology at UCLA in the 1990s. REST Technologies will register with the Food and Drug Administration later this year and begin making the REST Tracker available to other sleep centers. Requests for additional information and REST Tracker updates are available at www.resttechnologies.com/rest-tracker.
References
1. Simmons J, Sadeghian H. 0091 The practicality of implementing nightly remote patient monitoring (RPM) of OSA patients in clinical practice. Sleep. 2022 June;45(suppl 1):A41.
2. Simmons J, Sadeghian S, Lavender M, et al. 0709 Comparison of tracking OSA from PAP machine data vs cardiopulmonary coupling based assessment obtained from nightly ring monitoring. Sleep. 2025 May 2025;48(suppl 1):A308–9.
3. Simmons J, Lavender M, Lopez S, Sadeghian S. 0634 A remote-patient-monitoring (RPM) system using FDA cleared wearable technology for OSA management overcomes deficiencies found in PAP AHI RPM data. Sleep. May 2025;48(suppl 1):A277.
4. Simmons J, Thomas R, Saade G, et al. 0694 Remote patient monitoring (RPM) of obese obstetric patients to identify onset of OSA using the REST Tracker: Monitoring at risk populations. Sleep. 2025 May. 48(suppl 1):A302.
Excellent work, and thank you Dr. Simmons for your leadership in always advancing our thinking to improve outcomes for patients.