Sleep Deprivation and Performance of Residents

To the Editor. —  We are writing to express surprise at the conclusions that Deaconson et al¹ draw from their study entitled "Sleep Deprivation and Resident Performance."There are a number of good reasons why the particular performance battery chosen by the authors might not be sensitive enough to detect sleep-deprivation effects. Most notable are the following: (1) The extremely brief duration of the component tests (eg, one minute for the Baddeley Grammatical Reasoning Test). The literature has shown longduration tests to be the ones sensitive to sleep-deprivation effects. (2) The time of day when tested (between 6 AM and 8 AM), which coincides with the "alerting" phase of the circadian system (biologic clock); more major effects might have appeared had testing been between 2 PM and 3 PM, which coincides with the "postlunch dip." (3) The probable (though unreported) morning intake of caffeine, which would certainly have improved