Editorials: Point and Counterpoint |

Rebuttal From Dr Rapoport FREE TO VIEW

David M. Rapoport, MD, FCCP
Author and Funding Information

FINANCIAL/NONFINANCIAL DISCLOSURES: D. M. R. receives patent royalties through New York University from Fisher & Paykel Healthcare Limited and Sefam Medical, Ltd for CPAP modifications; consulting fees and grant support from Fisher & Paykel Healthcare Limited; and consulting fees from BioMarin Pharmaceutical Inc, Morphy, Inc, and Jazz Pharmaceuticals plc.

CORRESPONDENCE TO: David M. Rapoport, MD, FCCP, Sleep Disorders Center, Bellevue Hospital, 462 First Ave, Room 7N2, New York, NY 10016

Copyright 2016, American College of Chest Physicians. All Rights Reserved.

Chest. 2016;149(1):19-20. doi:10.1378/chest.15-1320
Text Size: A A A
Published online

In his counterpoint editorial, Dr Punjabi points out intrinsic limitations of the apnea-hypopnea index (AHI) concept. The cycle rate of events does not capture physiologic severity of each event (eg, quantitative flow reduction, event duration) or biologic impact (eg, degree of desaturation, sympathetic activation); the temporal distribution of events is not captured (eg, clusters of apneas and hypopneas are counted equally to widely separated individual events); and other essential aspects of sleep are ignored (eg, total duration of sleep determines overall exposure to sleep-disordered breathing). Thus, he argues that AHI is an overly simplistic representation of the complex physiology of a 7-h polysomnogram recording.

These criticisms are numerical tautologies. These conceptual limitations of the AHI as a marker of physiologic information cannot be denied in concept, and we definitely should explore better metrics to correlate physiologic and clinical outcomes. However, to date (somewhat surprisingly), none of the attempts to improve on the AHI has produced improved results. This may change in the future, but an alternative conclusion is that the AHI is as good as anything because severity of the physiologic process is not the relevant end point, as if differential susceptibility plays a role. In the latter case, one can preserve the utility of the AHI by restricting its use to two extremes of severity: defining severe disease and identifying normal subjects who do not have sleep-disordered breathing. Other metrics (as suggested by Dr Punjabi) can then be explored but primarily in patients exhibiting the midregion of AHI severity.

This approach refines the approach to sleep-disordered breathing by moving away from the idea of a linear spectrum of severity. I propose that the AHI be used in the following way:

  • 1.

    An inclusive (ie, very sensitive) AHI, with a relatively low threshold, likely defines individuals who are normal (ie, without demonstrable health consequences of their upper-airway physiology). If forced to give such a definition today pending further research, perhaps < 10/h with arousals lie within normal limits.

  • 2.

    A different and more restrictively defined AHI that is above a different and higher threshold likely identifies individuals with clear sleep-disordered breathing. On the basis of the literature quoted earlier, > 30/h with 4% desaturation might be a place to start (today) in defining this group.

  • 3.

    For individuals neither normal nor having clear sleep-disordered breathing, the AHI may be less useful for the reasons suggested by Dr Punjabi. For these subjects only (I estimate approximately one-third of those being evaluated for sleep-disordered breathing are in this category), I would advocate that either refinements of analysis (eg, rapid eye movement or supine AHI, peak AHI) or entirely new metrics are needed in research and clinical practice for purposes that are not quite severity related (ie, to decide which individuals have a sufficiently increased likelihood of consequences that they should be treated and which of these have findings on their polysomnogram that can or cannot be used to explain clinical complaints in epidemiologic studies and to evaluate the effect of therapeutic interventions). This search for other metrics, useful only in the gray area of AHI severity, should replace the current reliance on clinical symptoms for these subjects.

I believe that many thoughtful clinicians follow a similar thought process rather than simply use the AHI as a linear metric of severity. What is needed today is not to throw out the AHI entirely but, rather, to demote it from the level of “gold-standard” as a severity metric and use it rationally.


Punjabi N.M. . Counterpoint: is the apnea-hypopnea index the best way to quantify the severity of sleep-disordered breathing? No. Chest. 2016;149:16-19 [PubMed]journal




Punjabi N.M. . Counterpoint: is the apnea-hypopnea index the best way to quantify the severity of sleep-disordered breathing? No. Chest. 2016;149:16-19 [PubMed]journal
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543