A new method for assessing the risks of drowsiness while driving

Abstract

The drowsiness of drivers is believed to be a major cause of road crashes, but there is no standardized method for determining how drowsy a driver is or was at a particular time. This report describes how drowsiness, measured on a new scale (the Johns Drowsiness Scale or JDS) based on a weighted combination of ocular variables measured by infrared reflectance oculography, was related to performance failures in two different kinds of reaction-time (RT) tests as well as during simulated driving tests. 31 healthy volunteers performed simple (SRT) and choice (CRT) RT tests for 15 minutes with and without sleep deprivation for 27–33 hours. Their drowsiness was measured as a JDS score (0–10) each minute. In a separate experiment, 15 healthy young adults simulated driving in a car for about 70 minutes when alert and when sleep-deprived while their drowsiness was also measured. After sleep deprivation, RTs increased and errors of omission (failure to respond within 2 seconds from the start of the stimulus) occurred more frequently in both the SRT and CRT tests, the frequency of driving off the road increased, and JDS scores also increased significantly. The risk per minute for each subject of making an error of omission in SRT tests and of driving “off road” in a car simulator increased progressively with JDS scores (p < 0.001). The risk of performance failure in the drowsy state is associated with high JDS scores, mainly reflecting changes in the characteristics of eyelid movements during blinks at the time. This could form the basis of a new method for continuously assessing the risks of drowsiness while driving.