Chronic Fatigue in Airline Pilots
This next section is largely based on findings from a survey I conducted of around 350 pilots in a single airline. The sample captured 10% of the pilot population and was representative in terms of age, gender, aircraft fleet and rank. One problem with conventional fatigue risk management models is that they deal with short-term sleep and recovery, which affect acute fatigue levels but fail to take into consideration the endemic, steady-state chronic fatigue experienced by individuals. Most studies of chronic fatigue in the population use questionnaires originally designed to assess recovery from major illness or surgery. As yet, no specific measure of pilot chronic fatigue has been developed although there have been a few studies of pilots that have used the same measurement instrument: The FSS (Krupp et al., 1989). The FSS is a 9-item questionnaire which asks individuals to rate their current state on a 7-point scale. A mean score of 4 or greater indicates excessive fatigue.
Hockey (2013) suggests that about a third of the general population would score above this threshold.
In my survey, 74.4% of the pilots scored 4 or more on the FSS with a mean of 4.6. Pilots in Indonesia (Yuliawati et al., 2015) reported a mean FSS score of 4.66. Two studies of Portuguese pilots (Reis, Mestre & Canhao, 2013; Reis et ah, 2016) found that 84% and 90.6% respectively of pilots scored above the threshold as did 68.3% of pilots flying in the Gulf States (Aljurf et ah, 2017). Chronic fatigue, as measured by the FSS, would appear to be more than twice as prevalent in pilots as the general population.
To better understand the distribution of chronic fatigue symptoms, a smaller sample of 80 pilots completed the Oldenburg burnout inventory (OLBI) (Demerouti et ah, 2003) and the FSS. The OLBI generates two scores: disengagement and exhaustion. The sample specifically included a control group of pilot managers on the grounds that, because of their reduced flying task, they were more likely to be representative of rested pilots. Of the pilots surveyed, 10.71% met the criteria to be considered at risk of burn out on the OLBI (scoring in the upper quartile on both scales). The mean FSS scores for the rested and burnout risk groups were used to create cut-off values which were then applied to the FSS scores for the larger, original survey. This created three groups labelled rested, intermediate and burnout risk. Some 21.89% of the larger sample of pilots scored equal to or better (lower) than the ‘rested’ cut-off on the FSS while 34.31% scored equal to or worse (higher) than the burnout risk group, leaving 43.79% occupying the middle ground. The FSS cut-off for the rested group was 3.77, which is similar to the standard threshold score of 4. The burnout risk group had a cut-off of 5.14.
Respondents in the original survey also answered questions on sleep quality and daytime sleepiness. The Jenkins sleep quality (JSQ) scale (Jenkins et al., 1988) measures rates of sleep disturbance and difficulty in sleeping, an extreme case of which would be classed as insomnia. None of the rested group scored above the threshold on the JSQ whereas 2.9% of the intermediate and 6.7% of the burnout risk group showed evidence for poor sleep quality.
The Epworth daytime sleepiness scale (ESS) (Johns, 1991) captures symptoms of fatigue manifested as a tendency to doze off during the day. Figure 4.3 shows the distribution of scores for the sample of pilots. Those pilots in the burnout risk group
FIGURE 4.3 Distribution of ESS scores (% within the group).
had the highest incidence of abnormal daytime sleepiness and also had a lower rate of normal conditions. However, all three groups contained pilots with issues with daytime sleepiness, and it seems that the transition from a normal state to a chronic state starts slowly but then accelerates approaching burn out.
I said earlier that whereas acute fatigue can be mitigated through sleep, chronic fatigue could not be so easily remedied. This difference may be explained in part by the fact that recuperation is partly a function of the recovery time between periods of work but also involves having the ability to psychologically detach from work. Searle (2012) discussed this need for detachment in a study of the characteristics of crew hotels. Akerstedt et al. (2000) suggested that for workers in demanding roles, such as aircrew, a single day off after a block of workdays offered inadequate time for recovery, and 3-4days would be required after periods of disturbed circadian rhythmicity.
In addition to the FSS, crews who took part in my survey also completed the need for recovery (NFR) scale (Van Veldhoven & Broersen, 2003). The NFR has a cut-off score of 4, and 80.2% of all respondents scored above that threshold, a number - perhaps unsurprisingly - that almost matches the number of pilots reporting excessive fatigue. Table 4.3 shows the distribution of NFR scores by fatigue group. All groups show a significant recovery need but, interestingly, the intermediate and burnout risk groups suggest that recovery is a significant factor in the fatigue problem.
The sample of pilots used to construct the FSS cut-off values also completed the Whitworth inter-shift recovery scale (Winwood et al., 2005). The output is reported as a percentage (Figure 4.4), and the higher the value, the more recovery has been achieved. The difference between the three groups is statistically significant.
Inadequate recovery from the demands of work would appear to be a significant factor in the prevalence of chronic fatigue in airline pilots. Table 4.3 showed an almost universal NFR among pilots in the upper two fatigue groups, and this is reinforced by the data presented in Table 4.4. A Significant number of pilots in the sample, then, are experiencing chronic fatigue (as measured by the FSS), probably because of an inadequate recovery between the blocks of work. The persistent, debilitating effects of chronic fatigue, a physiological condition, make it a significant contributor to psychological fatigue. Furthermore, there are a group of pilots occupying a space between the ‘rested’ and ‘burnout’ who present a different challenge. From an organisational perspective, it would be interesting to know how to stem the drift towards ‘burnout risk’. At this point, I want to return to the other elements of the stress response discussed earlier - anxiety - and to try to integrate the process into the fatigue model.
Distribution of Need for Recovery
TABLE 4.4 Inter-shift Recovery