Caution urged when using obstructive sleep apnoea screening questionnaires

By Rebecca Jenkins
Obstructive sleep apnoea (OSA) assessment combined with sleepiness questionnaires can reliably rule in, but not rule out, clinically relevant OSA in primary care, Australian researchers caution.

Recent Medicare changes allow GPs to directly refer patients for sleep studies if they attain certain scores on one of three OSA screening assessments in combination with cut-offs on the Epworth sleepiness scale (ESS).

However, the questionnaires’ combined performance has never been assessed in primary care, researchers wrote in the Medical Journal of Australia, prompting them to design a prospective validation study using middle-aged participants from the ongoing Tasmanian Longitudinal Health Study.

Analysing data from the 282 people with an OSA trigger symptom who completed screening assessments and type 4 sleep studies, researchers found the STOP-Bang and OSA-50 questionnaires correctly identified most participants with clinically relevant OSA (sensitivity, 81% and 86%, respectively), but they each had poor specificity (36% and 21%, respectively).

The Berlin questionnaire (BQ) had low sensitivity (65%) and low specificity (59%), the team of Melbourne-based researchers reported.

When the score from any of the questionnaires was combined with a high ESS score of 8 or more, the specificity for each tool was high (94 to 96%), but at a cost to the sensitivity (36 to 51%).

Study coauthor Associate Professor Garun Hamilton, Director of Sleep Medicine Research at Melbourne’s Monash Health, said it was important for GPs to know that these questionnaires do not rule out OSA.

‘If they have a clinical suspicion for OSA and the questionnaires are negative they should refer the patient to a sleep disorders or respiratory physician for further assessment,’ he told Medicine Today.

He also noted that the questionnaires were likely to perform even less well in the elderly or among young adults. They should not be used in children, he added.

Associate Professor Hamilton said the Medicare criteria were under planned review, but there were no other screening tools that could replace the current questionnaires.

Instead, this research has suggested adding in a decision support tool for primary care practitioners – although it will require further validation.

‘A helpful change that this paper suggests is that the threshold for a “positive” STOP-Bang questionnaire should be lowered from 4 to 3. This will improve the sensitivity for detecting clinically significant OSA (not as many missed cases) without lowering the specificity (no significant increase in false positives),’ he told Medicine Today.
Med J Aust 2019; doi: 10.5694/mja2.50145.