Validity and clinical utility of sleep laboratory criteria for insomnia.

The sensitivity, specificity, and positive predictive value of two proposed sleep laboratory criteria for the diagnosis of insomnia were evaluated in 375 adults with a primary complaint of insomnia and 150 noninsomniac controls. The two criteria used resulted in either low sensitivity and moderately strong specificity or high sensitivity and low specificity and, accordingly, in both cases weak positive predictive values (diagnostic accuracy), both for one night and multiple nights of recordings. Further, an empirically optimized criterion also resulted in an unsatisfactory diagnostic accuracy. Finally, the optimized MMPI criteria were superior to optimized sleep criteria in differentiating insomniacs from controls. In conclusion, sleep laboratory recordings provide little relevant information for confirming or excluding the presence of insomnia.