Sleep apnea detection tools

We read with great interest the Letter to the Editor by Murray and colleagues¹ in which they first claim that, contrary to our statement in the paper, the 2017 American Academy of Sleep Medicine (AASM₂₀₁₇) “criteria”² were not designed to be used as a screening test, but rather as a physician guide tool for appropriate diagnostic test selection among patients being assessed in clinical settings. In other words, since screening tools are applied to a priori asymptomatic populations, the fact that the AASM₂₀₁₇ criteria were devised for use in symptomatic individuals would negate their designation as a screening instrument. Since the definition of “screening” implies the absence of pre-existing symptoms, we concur with the stringency of the definition when applied in such a contextual comment. However, the major aim of our study was to examine the reliability of the 4 different and widely used instruments in detecting the likelihood of obstructive sleep apnea (OSA; overall disease and at 3 different levels of severity) if any of these instruments were to be used in the clinic prior to referral to the sleep laboratory.² This aim was specifically enunciated in the article, even if the term “screening” is not stringently applicable to such objective. The reasoning behind such a study is that, in many countries, access to a sleep laboratory facility is fraught with substantial delays due to relative inaccessibility and financial costs and therefore prioritization of those individuals who are more likely to experience OSA, particularly those with more severe disease using the aforementioned instruments, may enable prioritization of the referral for the testing process (ie, in-laboratory polysomnography or other).

The second point raised by Murray et al is an important one.¹ It is clear that home sleep apnea testing (HSAT) is not universally applicable for the evaluation of all patients being seen in clinics and presenting with general sleep symptoms. However, if the clinical phenotype of these patients does not include any outlier symptom indicators suggestive of sleep diseases other than OSA—that is, the patients are “bread and butter” typical cases of OSA with a constellation of typical symptoms (excessive daytime sleepiness, witnessed apnea, hypertension)—HSAT would be consistently applicable. Under these circumstances, the 3 instruments other than the AASM₂₀₁₇ criteria being used would display improved specificity, albeit slightly (our Table 3).³ We should also point out that all our patients underwent polysomnography, and as such, the assumptions formulated regarding HSAT are theoretical rather than based on actual pragmatic testing. We therefore echo the sentiment that HSAT is not the universal solution diagnostic for “all” patients with sleep symptoms and that polysomnography should be reserved for only those with “negative” HSAT findings if the original decision for HSAT was predicated on common OSA symptomatology.

Finally, we would like to convey that the scrutiny related to AASM₂₀₁₇ consensus criteria was primarily being tested at a single center in Brazil, thereby reflecting the routine clinical practice algorithms in that specific site.³ As indicated in the Discussion of the article, such findings may not necessarily be generalizable to other clinical settings.

DISCLOSURE STATEMENT

All authors have seen and approved the manuscript. The authors report no conflicts of interest.