Issues in computerized polysomnography.

Reliable, properly tested, computerized systems for recording and scoring polysomnographic data would be welcome assistants in sleep laboratories. They potentially could save time, effort, paper, storage and cost. There exists, however, questions of competence and confidence. The nature and scope of published literature fails to convince many cautious sleep specialists that the cost-benefit ratio is favorable. Nonetheless, computerized systems for polysomnography are here to stay and are improving. Most of us represent past, present or future consumers of these products. The market will adjust to our level of knowledge and demands. We can adopt a variety of perspectives; however, the three basic modes of use should guide our thinking. For clarity, it is helpful to dissect 1) recording issues, 2) scoring issues and 3) monitoring issues from one another. Also, we should insist on greater statistical sophistication in testing trials. Manufacturers face a dilemma. Sleep disorders medicine has few true standards and many idiosyncratic practices. The expertise of consultants vary and manufacturers may be in no position to judge their collaborators. We desperately need guidelines to meet both manufacturers' and clinicians' expectations for testing and using computerized polysomnography.