Measurement variability in sleep disorders medicine: the Victorian experience.

Background: Surveys of laboratories in North America have documented significant diversity in the working definitions used for reporting respiratory events in sleep studies.

Aim: To assess sources of variability in the measurement of sleep-disordered breathing (as defined by the Apnoea-Hypopnoea Index) between sleep laboratories in Victoria, Australia.

Methods: A self-complete written questionnaire was constructed following literature review and interviews with staff at three separate sleep laboratories. The survey was sent to all laboratories listed in Victoria by the Australasian Sleep Association. The first part of the survey related to the type of equipment used to record sleep and other variables during overnight polysomnography and the second part related to the definitions and methods used to report results.

Results: Seventeen out of 18 laboratories returned the surveys. There were variations identified in the types of sensors used to measure particular signals. There were also inconsistencies identified in the criteria used to score arousals, apnoeas and hypopnoeas by different laboratories. The variability was greatest for hypopnoea definitions.

Conclusions: There is considerable variation in the methods used to measure and define sleep-disordered breathing between sleep laboratories in Victoria. The extent to which these variations influence the comparability of reported results between laboratories requires further evaluation. The survey findings may assist the process of developing and implementing local guidelines for the performance and reporting of polysomnography.