A sleep physician identifies alternative metrics for the assessment of the severity of obstructive sleep apnea.

By Edward D. Michaelson, MD, FACP, FCCP, FAASM

My recent article in Sleep Review addressed the many limitations of the apnea-hypopnea index (AHI) as a tool for the determination of the severity of obstructive sleep apnea (OSA) and suggested additional metrics that might improve the estimation of OSA severity.1 Most of these limitations and my suggestions were based on data commonly recorded during attended polysomnography (PSG) and limited-channel home sleep testing devices.

The limitations of the AHI are well recognized and have also been reviewed by others.2-5 A recent article suggested the collaborative development of a multidimensional score for measuring the severity of OSA.6

This article reviews several newer and/or less conventional ways of evaluating the severity of OSA using data that are not typically collected (or, in some cases, calculated) during PSG. Some of these novel techniques may also be useful for evaluating the effectiveness of OSA treatment.

A closely related issue is the development and evolution of newer technologies, data analysis, and the shrinking differences between clinical and consumer technologies, as well as the positions of the American Academy of Sleep Medicine (AASM)7 and the US Food and Drug Administration (FDA) on these issues, but these are beyond the scope of this article.

The novel approaches summarized here include point process, sleep breathing impairment index, heart rate variability, and ventilatory or hypoxic burdens.

While these alternative methods are interesting and may be more accurate and/or reproducible compared to the AHI, they are not standardized, may not be economically practical, and their full clinical usefulness has not been established. Furthermore, it has been difficult enough to get the medical community (not to mention payers or other stakeholders) up to speed even with the current, relatively simplistic, methodology for assessing the severity of OSA.

Nevertheless, these non-traditional metrics address different aspects of the adverse effects of OSA on a patient’s pathophysiology and appear to be more applicable to system-specific problems such as cardiovascular outcomes. They could be used alone or in conjunction with traditional or modified AHI.

Several Novel Alternatives

Point Process

Point process is a relatively simple, highly predictive statistical model using parameters that can be estimated from a conventional PSG, making more effective use of the hundreds of megabytes of collected data.

The model is based on a point process theory allowing moment-to-moment analysis of the associations between respiratory events and other sleep parameters being recorded, which leads to the evaluation of the temporal association between respiratory events and other parameters and simultaneous, quantitative relationships among the variables. This methodology can provide identification of phenotypes and ultimately predict and improve patient outcomes.8

Sleep Breathing Impairment Index (SBII)

The sleep breathing impairment index (SBII) accounts for respiratory events and hypoxia. It combines the degree and duration of each event-related desaturation (that is, it considers hypoxic burden) and the frequency of events, providing a more comprehensive assessment of OSA severity than AHI (which only describes event frequency). The SBII provides higher sensitivity and better predictive capability for cardiovascular disease outcomes than the AHI.9

Heart Rate Variability

Heart rate variability (HRV) represents sympathetic activity shifts in the autonomic nervous system, which can be induced by apneas or hypopneas. The index is calculated using the R-R interval from an electrocardiogram. The measurement is simple, noninvasive, and can be used to determine cardiac autonomic activity. It correlates with the AHI in patients with OSA and has been used to monitor the efficacy of OSA treatments and the detection of cardiovascular risk.

However, the relationship between AHI and HRV is difficult to quantify as HRV may vary individually by age, sex, and/or physical activity.10-13

Ventilatory or Hypoxic Burdens

These authors emphasize that OSA is a disease whose severity is defined across at least three possibly independent domains, including underlying ventilatory disturbance, hypoxic burden, and resulting arousals. They argue that the AHI is a rudimentary and partial (italics mine) way of combining the burden of OSA across these domains. They also point out that some recent novel measures of hypoxic burden (such as the total area under the respiratory-event-related desaturation curve) and the arousal burden (the cumulative duration of arousals attributable to respiratory events) have shown modest association with cardiovascular disease and all-cause mortality not captured in the AHI.

The ventilatory burden is an automated breath-by-breath single metric that determines the relationship between upper airway obstruction and all-cause and cardiovascular mortality with and without hypoxic burden and is a better predictor of all-cause and cardiovascular disease mortality. It is not dependent on the consequences of hypoxemia or arousals in OSA. Additionally, the ventilatory burden has shown a relationship to daytime sleepiness and hypertension.14

CPAP Download Data

While this article focuses on non-AHI approaches for determination of sleep apnea severity based on diagnostic studies, periodic review of data downloads from PAP (for example, CPAP, APAP, BiLevel PAP) machines can provide valuable objective information about the effectiveness of CPAP therapy—including not only adherence and mask leak but also the frequency and type of respiratory events.

Of course, this data has the same limitations as diagnostic PSG and limited-channel home sleep testing devices and does not measure oxygen saturation, though O2 saturation could be an easy addition with the plethora of available devices. Also, the algorithm for the determination of respiratory events based on CPAP download data (what is actually being measured) varies among manufacturers.

An interesting related metric of sleep depth, the Odds Ratio Product (ORP) developed by Cerebra, predicts long-term CPAP adherence in patients with obstructive sleep apnea using data obtained from diagnostic PSG. This “Adherence Index” predicts the long-term adherence to CPAP prior to the initiation of therapy and may enable health care providers to identify patients with obstructive sleep apnea who require more support, facilitating a personalized approach to management.15

Where Do We Go From Here?

The AASM Consumer and Clinical Technology Committee developed a sleep technology directory to familiarize its members with new and popular sleep devices and apps. The listings provide summaries of the capabilities and limitations of these technologies and are periodically updated.7 A nice review of the directory is covered in the AASM’s magazine Montage, including the evolution and shrinking difference between clinical and consumer technologies, novel hardware, data outputs and processing, and the use of artificial intelligence and deep learning algorithms. The review also covers the FDA’s different levels of clearance and approval.16

Finally, in my view, a “one-size-fits-all” metric for an improved and standardized single index, although possibly more accurate and clinically useful than AHI, would be insufficient and not applicable to the many variants of OSA, including not only the respiratory events and other parameters typically collected during PSG, but other variables such as age, gender, phenotypes, and comorbid medical conditions.

Any single approach would suffer from many of the same limitations of the AHI. But new metrics—such as point process, SBII, HRV, ventilatory burden, CPAP download data, the predictive Adherence Index, and others—could still be useful and more applicable to system-specific problems.

References

1. Michaelson ED. AHI for severity of sleep apnea—we can do better! Sleep Review. 2023 Jan;24(1):11-3.

2. Classification of OSA severity: more than meets the AHI. American Thoracic Society webinar. Aired 2023 Jan 27.

3. Malhotra A, Ayappa I, Ayas N, et al. Metrics of sleep apnea severity: beyond the apnea-hypopnea index. Sleep. 2021 Jul 9;44(7):zsab030.

4. Leppänen T, Myllymaa S, Kulkas A, Töyräs J. Beyond the apnea-hypopnea index: alternative diagnostic parameters and machine learning solutions for estimation of sleep apnea severity. Sleep. 2021 Sep 13;44(9):zsab134

5. Malhotra A, Gottlieb DJ. The AHI is useful but limited: how can we do better? Sleep. 2021 Sep 13;44(9):zsab150.

6. Martínez-García MÁ, Oscullo G, Gomez-Olivas JD, Gozal D. Measuring severity in OSA: the arguments for collaboratively developing a multidimensional score. J Clin Sleep Med. 2023 Oct 1;19(10):1705-7.

7. #SleepTechnology. American Academy of Sleep Medicine. Available at https://aasm.org/consumer-clinical-sleep-technology.

8. Chen S, Redline S, Eden UT, Prerau MJ. Dynamic models of obstructive sleep apnea provide robust prediction of respiratory event timing and a statistical framework for phenotype exploration. Sleep. 2022 Dec 12;45(12):zsac189.

9. Dai L, Cao W, Luo J, et al. The effectiveness of sleep breathing impairment index in assessing obstructive sleep apnea severity. J Clin Sleep Med. 2023 Feb 1;19(2):267-74.

10. Nam EC, Chun KJ, Won JY, et al. The differences between daytime and nighttime heart rate variability may usefully predict the apnea-hypopnea index in patients with obstructive sleep apnea. J Clin Sleep Med. 2022 Jun 1;18(6):1557-63.

11. Park DH, Shin CJ, Hong SC, et al. Correlation between the severity of obstructive sleep apnea and heart rate variability indices. J Korean Med Sci. 2008 Apr;23(2):226-31.

12. Ucak S, Dissanayake HU, Sutherland K, et al. Heart rate variability and obstructive sleep apnea: Current perspectives and novel technologies. J Sleep Res. 2021 Aug;30(4):e13274.

13. Karimi Moridani M. A novel clinical method for detecting obstructive sleep apnea using of nonlinear mapping. J Biomed Phys Eng. 2022 Feb 1;12(1):31-4.

14. Parekh A, Kam K, Wickramaratne S, et al. Ventilatory burden as a measure of obstructive sleep apnea severity is predictive of cardiovascular and all-cause mortality. Am J Respir Crit Care Med. 2023 Dec 1;208(11):1216-26.

15. Younes MK, Beaudin AE, Raneri JK, et al. Adherence Index: sleep depth and nocturnal hypoventilation predict long-term adherence with positive airway pressure therapy in severe obstructive sleep apnea. J Clin Sleep Med. 2022:18(8):1933–44.

16. #SleepTechnology. MONTAGE. American Academy of Sleep Medicine. 2023. (8);2.

Illustration 83159298 © Daniil Peshkov | Dreamstime.com