Guidance, not screening

We read the report by Duarte et al with interest.¹ The authors evaluated the performance of the 2017 American Academy of Sleep Medicine “criteria” (AASM 2017) as a screening tool against 3 validated screening instruments—that is, the neck circumference, obesity, snoring, age, sex (NoSAS), snoring tiredness, observed apnea and high blood pressure-BMI, age, neck circumference, gender (STOP-Bang), and gender, obesity, age, and loud snoring (GOAL) questionnaires—and concluded that the AASM 2017 was a not a useful screening tool for obstructive sleep apnea (OSA).²

Screening, by definition, involves the application of a test or instrument to detect a potential condition in individuals without known signs or symptoms of that condition.³ In general, the goal of screening is to find disease at an earlier stage so that treatment is more effective or efficient. To achieve this goal, screening tests often have better sensitivities than specificities.⁴ This type of trade-off was explored in development of an enhanced 2-stage application of the screening test STOP-Bang.⁵ The AASM 2017 “criteria” were not designed to be used as a screening test but rather as a tool for clinicians to guide appropriate diagnostic test selection in patients who have been evaluated and a decision has been rendered to proceed with testing for OSA.²

As the authors point out, there are significant access and cost challenges in obtaining in-laboratory polysomnography (PSG), while home sleep apnea testing (HSAT) may be more widely available. However, HSAT is not appropriate for every patient. The AASM 2017 recommendations serve as a guide to “triage” the more appropriate patients (uncomplicated and suspected of moderate to severe OSA) to testing via HSAT, thereby reserving PSG for those cases in which HSAT would be inadequate but who could go on for PSG. Viewed in that light, the observation that “The AASM₂₀₁₇ criteria, unlike the GOAL, STOP-Bang, and NoSAS instruments, always preferentially performed better as far as specificity at the expense of lower sensitivity” would seem to make the use of those criteria quite appropriate (the authors’ Table 2).¹

Finally, the authors in their Discussion point out: “… implementation of this approach would undoubtedly identify more selectively those with moderate-to-severe OSA but at the expense of discarding many patients with OSA.”¹ The paper suggests that those who do not meet those criteria go on for PSG, but the authors dismiss this as it would be “potentially valuable only when there is unrestricted access to sleep studies but not of pragmatic value in a highly prevalent and often underdiagnosed disease such as OSA.”¹ One could take the approach to do HSAT in everyone without a contraindication, and then repeat the test when negative. However, this may lead to delays and frustration, and may even cost more in the long run.^6,7

Overall, the purpose of these AASM 2017 recommendations is distinctly different than the aforementioned screening questionnaires and thus evaluating side-by-side performance of these tools would not be comparing “apples to apples.” We agree with the authors that the AASM 2017 recommendations are not suitable for screening; rather, we advocate their use for the intended purpose of identifying for whom HSAT is an appropriate alternative to PSG.

DISCLOSURE STATEMENT

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