CR 189: Development of measures of fatigue: using an alcohol comparison to validate the effects of fatigue on performance (2000)
The aim of this study was to develop a range of measures that could be used in evaluating the effectiveness of different work-rest schedules for managing fatigue. To be useful for this purpose, tests need to have demonstrated sensitivity to fatigue and be able to be applied in working environments. In addition, to be useful, tests are needed that provide results that can be interpreted in terms of the person's relative capacity compared to a recognised standard for safety.
A range of eight tests were selected from the research literature based on evidence that they are affected by fatigue. The tests included Simple Reaction Time, Unstable Tracking, a Dual Task, the Mackworth Clock Vigilance Test, Symbol Digit Coding, Visual Search, Sequential Spatial Memory and Logical Reasoning. To investigate the effects of fatigue on these tests, subjects were kept awake for 28 hours and tested at regular intervals. As a comparison, the same subjects were also administered varying doses of alcohol up to 0.1% blood alcohol concentration (BAC) and tested after each dose using the same tests as for the sleep deprivation condition. An alcohol comparison was used as there are legal and community-accepted standards for alcohol use when driving which are based on the effects of alcohol on performance. Performance on these tests following levels of alcohol that are known to be too high for safe performance were compared with test performance after increasing amounts of sleep deprivation. It should then be possible to match the effects of sleep deprivation on performance with the effects of alcohol as they can be directly compared on the same tests.
The study involved 20 long-haul truck drivers (Drivers) and 19 people who were not employed as long haul drivers (Controls). Long haul drivers were compared to people who are not long haul drivers as there is evidence from previous research that long distance drivers may have, or may develop, a higher ability to overcome performance effects of fatigue.
Drivers and controls first completed a three hour practice session in the afternoon of the first day. On the second day they participated in either the alcohol or sleep deprivation condition. For the alcohol condition, subjects were tested every hour from 8am to 12pm . The first test was a baseline test and the next four tests occurred around 30 minutes after subjects were administered doses of alcohol aimed to produce increments of 0.025% BAC in time for each performance test session. The aim was to do performance tests at no alcohol, 0.025% BAC, 0.05% BAC, 0.075% BAC and 0.1% BAC. For the sleep deprivation condition, testing occurred every hour from 8am for the next four hours, then every other hour for the next 23 hours with the last test at 9am the following morning.
The results showed the following:
Performance effects of sleep deprivation were seen for most tests, in particular, those tests involving passive attention, such as the Mackworth Clock Vigilance Test, or a relatively difficult discrimination, such as the Simple Reaction Time Test.
Performance effects were seen due to alcohol for all tests with the strongest effects for Simple Reaction Time and the Mackworth Clock Vigilance Task. Unstable Tracking, the Dual Task, Symbol Digit Coding, and Sequential Spatial Memory also showed reliable effects.
The effects of sleep deprivation and alcohol were not the same for all tests. Sleep deprivation had no effect on performance on any measure of either the Visual Search test or the Logical Reasoning test whereas alcohol reduced performance accuracy markedly on both tests but did not affect the speed measures.
Some tests showed evidence of a circadian rhythm effect on performance. In particular, Simple Reaction Time, the Dual Task, the Mackworth Clock Vigilance Task, and the Symbol Digit Coding Task showed performance trends consistent with the expected drop in performance in the early hours of the morning and in the early afternoon. With the exception of the number of correct responses on the Mackworth Clock Vigilance Task, these circadian effects were only ever apparent on measures of response speed, and not response accuracy. The remaining tests showed no evidence of being affected by circadian influences.
Drivers showed different patterns of performance compared to controls. Drivers were slower but more accurate than controls in performance of the Symbol Digit test, suggesting that drivers took a more conservative approach to performance of this test.
Using the legal limits for alcohol as a standard, performance deficits equivalent to 0.05% BAC were seen after 17 to 19 hours of sleep deprivation on most tests corresponding to between around 10.30pm and just after midnight in this study. Levels equivalent to 0.1% BAC were predicted to occur after 18 to 20 hours awake which in this study occurred between 11.30pm and 1.30am.
For many people, 17 to 19 hours without sleep is the upper limit of their waking period. Where safety-sensitive activities are being carried out at this time, or where wakefulness continues beyond this period, people are likely to be at increased risk.
In conclusion, this study has demonstrated which tests are most sensitive to sleep deprivation and fatigue. The study has also provided a basis for making judgements about the extent of performance impairment on these tests that is likely to compromise performance capacity and, as a consequence, to compromise safety. Further research is needed to determine how these changes in performance capacity are related to crash risk. Nevertheless, based on the previous research on increased crash risk with increasing alcohol consumption over 0.05%BAC, the findings of this study suggest that performance impairments at the level of produced by 0.05%BAC or 0.1%BAC but due to sleep deprivation instead of alcohol may well have a similar effect on crash risk. As a result, this study has therefore established a set of tests that can be used in evaluations of alternative compliance schedules for managing fatigue.
Download Complete Document: Fatig_Alc [PDF: 547 KB]
Type: Research and Analysis Report
Sub Type: Consultant Report
Author(s): A Williamson, A-M Feyer, R Friswell, S Finlay Brown
ISBN: 0 642 25579 2
Topics: Alcohol, Fatigue, Methodology
Publication Date: 01/06/00