Using total bed time instead of total sleep time as denominator can significantly underestimate sleep apnoea severity, according to a Singapore study presented at the 22nd Congress of the Asian Pacific Society of Respirology (APSR) held in Sydney, Australia.

“These misclassifications could affect management decisions and cast doubts on the validity of sleep monitoring devices that do not incorporate electroencephalogram (EEG) determination of sleep,” the investigators said.

A total of 1,616 patients were analysed in this study, which excluded five sleep articles with incomplete data. [APSR 2017, abstract AP159]

The use of total bed time rather than total sleep time as denominator resulted in the classification of 10.4 percent (32/308) of patients with mild sleep apnoea as normal. Another 2.5 percent (21/843) of patients classified as either normal or mild when using total bed time as denominator actually had severe sleep apnoea.

Furthermore, 27.5 percent (39/142) of patients with mild or moderate or severe sleep apnoea were likewise classified as normal with the use of total bed time as denominator.

To analyse the effect of different denominators on the classification of sleep apnoea severity and hence demonstrate the importance of EEG determination of sleep, the investigators conducted a review of laboratory sleep studies done at National University Hospital in Singapore between January 2014 and March 2017 using SPSS software for statistical analysis.

Sleep apnoea severity was measured by the total number of apnoeic events divided by the denominator (total bed time or total sleep time). Apnoeic events included obstructive, mixed, central and hypopnea events. Total sleep time was taken as the total time the patients spent in bed.

The investigators coded sleep apnoea severity as follows: normal, X<5; mild, 5≤X<15; moderate, 15≤X<30; and severe X≥30. Then, they compared the use of total sleep time as a denominator to that of total bed time.

In another study, researchers found that using total recording time as a denominator in home sleep apnoea testing (HSAT) would result in substantial underestimation of sleep apnoea severity when applied to a large clinical cohort of diagnostic polysomnography. Phenotyping errors included the following: risk of falsely negative results (from mild to normal) and category errors (moderate or severe moving to mild or moderate severity, respectively). [J Clin Sleep Med 2017;13:551–555]

“Clinicians should recognize this underestimation limitation, which directly affects diagnostic phenotyping and thus therapeutic decisions,” researchers of the previous study said.

One of the major limitations of using HSAT is the lack of sleep staging, such that the respiratory event index is measured using the total recording time instead of the total sleep time. [J Clin Sleep Med 2017;13:551–555]