{
  "abstract": "Introduction Biomathematical Models (BMM) of fatigue in the workplace predict subjective and objective indicators of sleepiness. Such predictions rely on assumptions about sleep duration, workload and work-time in relation to circadian phase. We have reverse engineered a BMM, originally developed for the Health and Safety Executive (HSE), in order to optimise sleepiness prediction.Methods On about eight consecutive days, at the beginning and end of work times, fifty-two shift-working drivers (38.03+8.22yrs, 18 women) completed a single Karolinska Sleepiness Scale (KSS, Åkerstedt & Gillberg, 1990) either side of a 10-minute Psychomotor Vigilance Task (PVT, Dinges & Powell, 1985). HSE based its fatigue predictions on how people perform these tests, with KSS scores above 6 designated as a criterion for sleepiness (see Skeldon et al., 2016; Cleator et al., 2021).Results KSS scores across some 400 shifts distributed across 24 hours show that drivers were sleepier at the end of shifts than when they began ( figure 1).Twelve drivers never rated their post-shift sleepiness above KSS-6. For the remaining 40, HSE model predictions were generated for each shift. Predictions were significantly higher in shifts where drivers rated their sleepiness at 7 or above (9.94+9.11) than when they rated their sleepiness at 6 or below (6.35+4.33; t40=3.12, p<.005; Cohen’s d= 0.49; Wilcoxon 2.07, p<.05 ). While this confirms that the HSE model does indeed predict extreme sleepiness, this result must be treated with some caution. Figure 2 shows that the model predicts sleepiness, when drivers do not rate themselves as such, and fails to predict sleepiness when they are.Discussion The HSE model can predict higher levels of driver sleepiness, but it lacks sensitivity and specificity. We suggest that taking workers’ individual characteristics into account when attempting to predict sleepiness is an essential aspect of any attempt to model fatigue biomathematically.References Åkerstedt T, Gillberg M. Subjective and objective sleepiness in the active individual. International Journal of Neuroscience 1990;52:29–37.Cleator SF, Coutts LV, Phillips R, Turner R, Dijk DJ, Skeldon A. Fatigue, alertness and risk prediction for shift workers. bioRxiv 2021;2021-01. preprint doi: https://doi.org/10.1101/2021.01.13.426509.Dinges DF, Powell JW. Microcomputer analyses of performance on a portable, simple visual RT task during sustained operations. Behavior research methods, instruments, & computers 1985;17(6):652–655.Skeldon AC, Derks G, Dijk DJ. Modelling changes in sleep timing and duration across the lifespan: changes in circadian rhythmicity or sleep homeostasis?. Sleep Medicine Reviews 2016;28:96–107.Abstract P51 Figure 1Karolinska Sleepiness Scale ratings before and after shiftsAbstract P51 Figure 2Predicted sleepiness lacks sensitivity and specificity",
  "authors": [
    {
      "affiliations": [
        "Department of Psychology, Nottingham Trent University, Nottingham, United Kingdom",
        "‘SleepiEst’- Road Safety Trust funded research project"
      ],
      "name": "John A Groeger"
    },
    {
      "affiliations": [
        "Department of Psychology, Nottingham Trent University, Nottingham, United Kingdom",
        "‘SleepiEst’- Road Safety Trust funded research project"
      ],
      "name": "Peter Goodwin"
    },
    {
      "affiliations": [
        "Department of Psychology, Nottingham Trent University, Nottingham, United Kingdom",
        "‘SleepiEst’- Road Safety Trust funded research project"
      ],
      "name": "Fran Pilkington-Cheney"
    },
    {
      "affiliations": [
        "‘SleepiEst’- Road Safety Trust funded research project"
      ],
      "name": "Yvonne Taylor"
    }
  ],
  "title": "P51 Predicting the risk of sleepiness in shift-working drivers",
  "uid": "58726394-140a-56b1-94d8-89f82fb323b9"
}