Great Barrier Reef and Torres Strait
- Great Barrier Reef pilotage fatigue risk assessment
- Fatigue study on coastal pilots
- The work practices of marine pilots
- Work schedules of Great Barrier Reef Pilots
- Impact on wives and families
- Work and Sleep patterns
- Implications for fatigue management
- Information for Interest
- Restricted access
On Tour analyses of the work and rest patterns of Great Barrier Reef pilots: implications for fatigue management
Results: ALERTNESS AND FATIGUE DURING BRIDGEWORK
9.0 Measures of Alertness and Fatigue during Bridgework
Ratings of alertness and the presence of fatigue symptoms were recorded to assess performance decrements during bridge periods. Both measures were highly relevant to the present investigation given that previous research had found high correlations between self reports of alertness and/or sleep loss and measures of performance (Gillberg, et al., 1994). The onset of fatigue symptoms such as slow eye movements correspond with alertness ratings in the lower end of the scale (Akerstedt & Folkard, 1995).
9.1 Alertness Ratings during Bridgework
One aim of the logbooks was to explore any fluctuations in alertness during the course of bridgework. Alertness was scored on a visual analogue scale (range: very sleepy, to very alert; score 1-9). Pilots average alertness during bridge periods ranged from 6.2-7.7, that is, towards the higher end of the scale (Figure 9.0). There was no significant shipping route effect on alertness during bridge periods (Table 9.0), and alertness did not fluctuate significantly across the duration of bridge periods.
Given that earlier results identified that work on Hydrographers Passage was less demanding it is not surprising that alertness ratings were marginally higher during bridge periods on Hydrographers Passage compared with the other two shipping routes (Figure 9.0). The overall trend in alertness during bridge periods was distinctly different between the shorter routes (Hydrographers Passage and the GNE Channel) and the longer Inner Route (Figure 9.0). However, at the three hour point on both shipping routes, that is, at the end of the bridge period on the longer route, and midway through the longer bridge periods (6-7 hours) there was a decrease in alertness (Figure 9.0).
To further examine the fatigue issue the percentage of bridge periods with minimum alertness ratings were assessed. Alertness ratings of 3 or less are associated with sleepiness. Three percent of bridge periods on Hydrographers Passage met this criteria with the figure rising to 8 and 12 percent for the Inner Route and GNE Channel, respectively (Figure 9.1). The lower percent of bridge periods associated with minimum alertness on Hydrographers Passage is also consistent with the less demanding work patterns in this region. The data suggested that the percent of bridge periods with minimum alertness was higher for GNE Channel assignments; however, there was insufficient data collected on this route to confirm this statistically (Table 9.0).
Figure 9.0 Fluctuations in alertness during bridgework, by shipping route
Figure 9.1 Mean percent of bridge periods with alertness ratings < 3, by shipping route.
Table 9.0 Analysis of alertness ratings during bridge periods, by shipping route (1)
|
Effect |
Post hoc results |
Mean (sem) |
F Statistics |
p-value |
|---|---|---|---|---|
| Alertness ratings for the beginning of bridge periods
Inner Route Hydrographers Passage GNE Channel |
n/a n/a n/a |
7.01 (0.05) 7.73 (0.14) 6.75 (0.32) |
2.11 |
= 0.120 |
| Alertness ratings for the middle of bridge periods
Inner Route Hydrographers Passage GNE Channel |
n/a n/a n/a |
6.90 (0.06) 7.16 (0.20) 6.75 (0.30) |
1.25 |
= 0.288 |
| Alertness ratings for the end of bridge periods
Inner Route Hydrographers Passage GNE Channel |
n/a n/a n/a |
6.46 (0.06) 7.29 (0.21) 6.21 (0.37) |
0.80 |
= 0.449 |
| Percent of bridge periods with alertness ratings < 3
Inner Route Hydrographers Passage GNE Channel |
n/a n/a n/a |
7.9 (0.13) 2.6 (0.23) 12.5 (0.41) |
0.25 |
= 0.78 |
- Results of full two-way Analysis of Variance (ANOVA) model. p < 0.01 considered statistically significant. n/a = Post hoc testing not performed when main effects not significant
9.2 Number of Fatigue Symptoms during Bridgework
To further assess the presence of fatigue during bridgework the number of fatigue symptoms were calculated. The number of fatigue symptoms during bridge periods ranged from 1 on Hydrographers Passage to 3 on both the Inner Route and GNE Channel (Figure 9.2). The significant shipping route effect was related to the greater number of symptoms experienced during bridge periods on the Inner Route and GNE Channel than on Hydrographers Passage (Table 9.1). This finding is consistent with the work assignment characteristics indicating work on these routes is more demanding than work on Hydrographers Passage and the slightly higher bridge alertness displayed for this route (Section 9.1).
Respondents were also required to rate the degree to which fatigue symptoms were experienced on a seven point Likert scale ranging from ‘not at all’ (scale = 1) to ‘very much’ (scale =7). The sum of the scores (range 10-70) was calculated, a higher score indicating symptoms were experienced to greater degree.
As Table 9.2 indicates the sum of the scores ranged between 12 and 17 and were at the lower end of the range. The significant shipping route effect related to the higher scores reported for the Inner Route and GNE Channel. This is also consistent with greater demands on pilots in these regions and the number of fatigue symptoms reported above.
Figure 9.2 Mean number of fatigue symptoms experienced during bridge periods, by shipping route.
Table 9.1 Analysis of fatigue symptoms during bridge periods, by shipping route (1)
|
Effect |
Post hoc results (2) |
Mean (sem) |
F Statistics |
p-value |
|---|---|---|---|---|
| No. of symptoms of fatigue/10
Inner Route Hydrographers Passage GNE Channel |
1 2 1 |
3.00 (0.09) 1.13 (0.21) 3.25 (0.67) |
6.33 |
=0.001 |
| Sum of scores of fatigue symptoms (Range 10-70)
Inner Route Hydrographers Passage GNE Channel |
1 2 1 |
14.41 (0.21) 11.78 (0.35) 16.91 (1.64) |
4.46 |
=0.001 |
- Results of full two-way Analysis of Variance (ANOVA) model. p < 0.01 considered statistically significant.
- Results of Tukey’s Studentised Range test for post-hoc differences (Type 1 Error Rate = .01) p < 0.01 for differences between shipping routes from post-hoc
9.3 Percent of Bridge Periods with one of more Fatigue Symptoms
From the number of fatigue symptoms presented above, the percentage of bridge periods with one or more symptoms of fatigue was calculated. On the Inner Route and Hydrographers Passage, 58 percent of bridge periods were associated with one or more fatigue symptoms with the percentage rising to 83 for the GNE Channel (Figure 9.3). The mean scores suggested that the percentage of bridge periods within this classification of symptoms was higher for the GNE Channel, however, there was inadequate data collected on this route for the result to be confirmed statistically.
The similar percentage of bridge periods on the Inner Route and Hydrographers showing one or more fatigue symptoms is unexpected given that significantly more fatigue symptoms were reported for the Inner Route than Hydrographers Passage (3 vs 1). However, a closer inspection of this result showed that the similar percentage of bridge periods meeting this criteria is due to the fewer assignments recorded on Hydrographers and thus resulting in a higher percentage. The presence of fatigue symptoms in spite of high alertness ratings suggested that alertness was slightly over rated by the respondents.
Figure 9.3 Mean percent of bridge periods associated with one or more symptoms of fatigue, by shipping route
Table 9.2 Analysis of the percent of bridge periods with one or more symptoms of fatigue, by shipping route (1)
|
Effect |
Post hoc results |
Mean (sem) |
F Statistics |
p-value |
|---|---|---|---|---|
| Percent of bridge periods with one or more fatigue symptoms
Inner Route Hydrographers Passage GNE Channel |
n/a n/a n/a |
58.48 (1.56) 58.97 (7.98) 83.33 (7.77) |
1.29 |
= 0.275 |
- Results of full two-way Analysis of Variance (ANOVA) model. p < 0.01 considered statistically significant. n/a = Post hoc testing not performed when main effects not significant.
10.0 Factors associated with high Fatigue levels and low alertness levels on the Bridge
10.1 Predictors of Fatigue, Stress and Alertness levels on the Bridge
The results in the preceding sections of this report based on the analyses of logbooks, and findings from other data sources, particularly the work schedules (Parker et al., Report No 2, 1998) indicated the presence of several key measures associated with decreased alertness and increased fatigue. In light of these findings an initial step was taken to predict fatigue, stress and alertness during bridge work from several key variables identified from work and break patterns. Thus the aim of this section was to identify features of work and break patterns which are correlated with high fatigue and low alertness.
To explore this issue a series of multiple linear regression models were used to assess contributions to variation in mean overall fatigue, stress, and minimum alertness measures of pilot data reflecting ashore and at-sea variables. After accounting for individual pilot variation, five incremental levels of modelling were attempted. Each level evaluated a set of data at a higher level of data refinement. Each additional level of modelling considered the percentage of additional variation in fatigue, stress, and alertness accounted for by a number of predictor variables associated with work and break patterns (Table 10.0). The multiple regression models considered individual subject average ratings of fatigue and stress over all time spent on the bridge and considered the minimum alertness level during the whole bridge period.
As Table 10.0 shows, individual differences in pilots accounted for substantial variation in the fatigue, stress and alertness scores. This individual pilot variability may be composed of physiological and other personal and environmental differences not considered in this study, which restricted itself to work- and break-related determinants of fatigue, stress, and alertness on the bridge.
The percentage of variation accounted for by individual differences may not be related to genuine differences in fatigue. There are two alternative explanations for the high of variance attributed to differences between pilots. The first is that pilots may differ in the use of rating scales in terms of central tendency and dispersion, thus these subjective rating scales cannot be used to make comparisons across, rather than within individuals. The second relates to negative effect evidence in the literature, that is, some individuals complain about everything, others about nothing.
After accounting for pilot differences, the two major contributing factors to all three outcomes were the duration of breaks preceding assignments and the duration of the actual assignment.
Break duration before an assignment was confirmed to substantially contribute to fatigue, stress and alertness during bridgwork. It is notable that this factor accounted for relatively more of the remaining variance in alertness than to either fatigue or stress (Table 10.0). This finding provided further evidence of the importance of rest break duration prior to assignments to ensure optimal alertness on the bridge.
Work assignment duration was the other major contributor to fatigue, stress and alertness on the bridge (Table 10.0). This factor accounted for 14 and 23 percent of the remaining variation in fatigue and stress, respectively; and 4 percent of the remaining variation in alertness. A qualitative assessment suggests that the time spent on assignment is a relatively greater contributor to stress on the bridge; whereas the duration of the break before an assignment contributed to a greater degree (13.3 percent) of the remaining variation in bridge alertness. As Table 10.0 indicates about 90 percent of the variation in all three outcomes was explained by accounting for individual pilot differences, total break and work assignment duration.
None of the other variables involving travel, bridgework or sleep at sea and ashore accounted for more than 2.5 percent of the remaining variation. That travel contributed only slightly, but marginally more to fatigue and stress is consistent with earlier results relating to the impact of travel on assignment stress. The more direct effects of travel may have been diluted due to the adjustment for travel in work assignment and break duration. There was also a suggestion that the total duration of bridge periods per assignment and the number of critical hours spent on the bridge contributed relatively more to alertness than to stress or fatigue (Table 10.0).
Results of the modelling procedures indicated that if additional data were collected to enhance the present reporting system for evaluating fatigue, stress and alertness on the bridge, it should at least recognise the individual differences in pilots’ thresholds for coping with fatigue, stress, and alertness, and include data on work assignment and break duration and the total duration of bridge periods during an assignment and during critical hours. Future research should consider further refinement of these models to include more detailed temporal aspects of work and break patterns to enable changes in fatigue, stress and alertness during bridgework to be predicted for each additional hour of duty. Additionally, further development of the models will also consider that the potential non-linear relationship between fatigue/alertness and both break and work assignment duration. It is likely that the multiple regression modelling based on linear relationships may have underestimated the contribution to the variance in the outcomes of work assignment and break duration.
Table 10.0 Percentage of variation in overall fatigue, stress, and minimum alertness accounted for by measuring various levels of data about work and break patterns based on Inner Route data.
|
FATIGUE |
STRESS |
ALERTNESS |
|
|---|---|---|---|
| Percent of variation accounted for by : | |||
| Pilot differences | 75.2 |
63.6 |
72.5 |
| Duration of preceding break | 82.2 |
69.5 |
85.8 |
| (an increase of) | (7.0) |
(5.9) |
(13.3) |
| PLUS | |||
| Total travel hours to/from ship | 83.5 |
71.5 |
85.8 |
| (an increase of ) | (1.3) |
(2.0) |
(0.0) |
| PLUS | |||
| Duration of work assignment | 97.2 |
94.8 |
90.1 |
| (an increase of) | (13.7) |
(23.3) |
(4.3) |
| PLUS | |||
| Duration of pilotage time on assignment | 97.3 |
94.9 |
90.2 |
| (an increase of) | (0.1) |
(0.1) |
(0.1) |
| PLUS | |||
| Duration of bridge periods per assignment | 97.9 |
96.6 |
92.6 |
| (an increase of) | (0.6) |
(1.7) |
(2.4) |
| PLUS | |||
| Average duration of bridge periods per assignment | 98.3 |
97.1 |
93.7 |
| (an increase of ) | (0.4) |
(0.5) |
(1.1) |
| PLUS | |||
| Average number of bridge periods per 24 hours | 98.6 |
97.2 |
93.9 |
| (an increase of) | (0.3) |
(0.1) |
(0.2) |
| PLUS | |||
| Total duration of time in bed during assignment | 98.3 |
97.4 |
93.9 |
| (an increase of) | (-0.3) |
(0.2) |
(0.0) |
| PLUS | |||
| Number of critical hours spent on the bridge | 98.8 |
97.5 |
95.3 |
| (an increase of) | (0.5) |
(0.1) |
(1.4) |
| PLUS | |||
| Total duration of sleep during preceding break | 99.1 |
97.6 |
95.7 |
| (an increase of) | (0.3) |
(0.1) |
(0.4) |
| PLUS | |||
| Total duration of non-critical sleep during break | 99.6 |
98.0 |
95.7 |
| (an increase of) | (0.5) |
(0.4) |
(0.0) |
11.0 Discussion
Fatigue is a complex phenomenon that is difficult to define (Brown, 1994; Cameron, 1974; Parker et al.,1995) and generally presents as feelings of weariness and/or an aversion to continue with the activity. It encompasses both physical and psychological components and may be acute, chronic or both in nature. Many factors can contribute to the development of a fatigued state including poor general health, anxiety, apprehension, inadequate sleep and or poor nutrition. Inappropriate workloads, in either physical or mental tasks, or inappropriate work to rest ratios can lower a persons level of arousal (Grandjean, 1970; Parker et al., 1995) and exposure to specific environmental conditions such as noise, vibration and extreme temperatures may also induce fatigue. These factors may act independently or in combination with one another and while the above list is by no means exhaustive, it identifies many of the circumstances associated with the work of GBR pilots (Parker et al., 1997).
An earlier survey of the health, stress and fatigue of Australian seafarers (Parker et al., 1997) identified a number of factors, such as reduced length and poor quality of sleep at sea, which have the potential to induce fatigue. Relatively high levels of stress and some evidence of cardiovascular risk factors reported by pilots in the survey provided further evidence of the potential for fatigue in this group. Recent changes in the maritime industry and particularly the introduction of competition in pilotage operations were also seen as significant sources of stress and as having the potential to impact on safety. These findings coupled with increasing recognition of the role of fatigue in vessel incidents on the Barrier Reef provided the background to the initiation of a series of research projects designed to investigate those aspects of the work practices of GBR pilots that may potentiate fatigue.
Following an extensive review of the literature related to this research a retrospective analysis of work and rest schedules over an 18 month period was conducted. The analysis was based on information reported by pilots working for the 3 pilotage companies operating in the Great Barrier Reef/ Torres Strait region (Parker et al., Report No 2, 1998). The results indicated that pilotage work was characterised by irregular work and break patterns with work-related travel impacting significantly on assignment and break time. Significant differences between the companies were evident in the number of work assignments performed and in all companies workloads increased markedly during the months between July and December. Moreover, a considerable number of situations were identified where personnel undertook greater workloads and shorter breaks than depicted by average company figures.
These findings were enhanced by a more detailed survey of the general work practices of GBR pilots and their perception of aspects of their work and lifestyle which may have a bearing on fatigue (Parker et al., Report No 4, 1998). The findings indicated that the compromised sleep patterns at sea occurred as a function of the irregularity of the work and rest schedules, relatively long periods of sustained work, often at night and the necessity to work and sleep at times incompatible with the normal biological rhythms of the body. Pilots experienced fatigue particularly at the end of assignments and identified boredom, lack of sleep and high workloads as important contributing factors. The potential for a decrement in performance was further substantiated by difficulty experienced by some pilots in maintaining concentration and attention.
The final phase of the research which is the basis of this report aimed to investigate the relationship between specific work and sleep characteristics and measures of alertness during bridgework. This required pilots to record information during their work assignments and during break periods, in a log book which was similar to that used in studies of mariners in the US and UK (Sanquist et al., 1996; Seafarers International Research Centre, 1996). The logbook was modified following extensive consultation with the Pilot Advisory Group, to take account of the specific work characteristics of GBR pilots. The information included the extent and quality of sleep at sea and ashore, industry specific measures relating to the difficulty of the assignment and pilot ratings of alertness. Such ratings of alertness have been shown to have a significant relationship to performance (Gillberg et al., 1994) and this form of data collection frequently used in field settings, was considered to cause minimal disturbance of the pilot’s work schedule. As the 3 routes in which pilots operate are associated with different work schedules and degrees of difficulty, the analyses were designed to contrast any differences which may exist between the work and rest patterns as a function of the routes.
11.1 Demographics
Pilotage services in the Barrier Reef region are supplied by a workforce of 58 pilots operating in three competing pilotage companies. The pilots represented a relatively older (mean age 53.2 + 1.64 years) stable, and highly experienced workforce. They had an average of 36 years of maritime service and 9.5 years in Barrier Reef pilotage. From the data sources it was evident that the size of the workforce and fluctuations in work availability, particularly the seasonal impact, placed demands on company personnel to effectively manage a small workforce in busy and quiet shipping periods (Parker et al., Report No 2, 1998). These situations can contribute to work overload or underload conditions depending on the effectiveness of work allocation procedures.
11.2 Work Assignment Characteristics
The general work characteristics derived from the logbook data were consistent with the basic descriptions of work assignments obtained from the work schedule analysis (Parker et al., Report No 2, 1998). This was particularly evident in terms of the duration of travel to and from ships, the greater amount of travel on GNE Channel assignments, the significantly longer Inner Route assignments, and the irregularity of starting times of work and breaks across all three shipping routes.
The duration of work assignments varied from 54 hours on the Inner route to 16 and 14 hours on Hydrographers Passage and the GNE Channel, respectively. Ninety-five percent of this time was spent on pilotage duties while operating on the Inner route and GNE Channel in contrast to 67 percent on Hydrographers Passage.
The irregularity of Great Barrier Reef pilotage was consistent with work patterns in other pilotage groups (Shipley & Cook, 1980; Berger, 1984; de Vries-Grierer, 1982; Sparks, 1992; States British Columbia Oil Spill Task Force, 1997). Although comparisons with other pilotage groups enabled the general characteristics of work patterns to be evaluated, direct comparisons are limited due to the unique geographical and operational features encountered by the various pilotage organisations. For instance, the physical size of the Great Barrier Reef pilotage region requires extended work assignments which averaged 54 hours on the Inner Route. In contrast, pilotage assignments undertaken by UK pilots (Shipley, 1978) were relatively shorter and were classified as short haul (2.5 hours or less) or long haul (more than 2.5 hours).
The retrospective analysis of work schedules indicated that travel significantly increased the duration of work assignments and the amount of night work, and reduced the duration of breaks (Parker et al., Report No 2, 1998). Notably, travel to and from the ship was the highest rated of all the factors contributing to stress and/or fatigue on the three shipping routes. The logbook findings confirmed these results and provided additional evidence for the inclusion of work-related travel in the present work schedule system for the monitoring of fatigue.
11.3 Duration and Timing of Bridgwork
It was clear from the logbooks that the characteristics of bridgework differed significantly across the three shipping routes. On the shorter routes, pilots spent 85 to 90 percent of pilotage time of 8-9 hours on the bridge in a single period, whereas on the longer route, 51 percent of the 49 hours of pilotage time was undertaken during multiple bridge periods averaging 3 hours. By comparison, UK pilots involved in long haul pilotage undertook 5.5 hours of bridge periods (Shipley, 1978), while a typical assignment performed by Port Phillip pilots involved 10 hours of navigation (Berger, 1984). Given that, on a tour of duty, the majority of Reef pilots alternate between the three routes, stable patterns of bridgework are not experienced. In this aspect, pilotage work differs considerably from conventional maritime watchkeeping systems with set hours on and off duty and with some limited disruption which may occur during port calls.
In the case of the GBR pilots, both long and short duration bridge periods may be factors in accident risk. Research results have shown that the relative risk of accidents increases as the number of hours on shift increases with a peak after a 13 hour shift (Folkard, 1995). However, the same author has also demonstrated that relative accident risk after 2-3 hours is as great as a 10 hour shift, that is a shorter duration shift may be riskier than a longer period. One explanation for this may be related to the time-dependent increase in mental lapses that are not compensated early in the shift by fatigue reducing strategies, such as physical movement or caffeine consumption (Sanquist et al., 1996). The present data revealed that during bridge periods alertness dropped marginally at the three-hour point, which coincided with the end of a short bridge period, and the middle of a longer bridge period. This finding tends to support the notion put forward by Folkard (1995) of an increased relative accident risk after 3 hours due to a drop in alertness at this point.
The effects on fatigue of irregular working hours appears to be markedly exacerbated when a proportion of this work occurs during night hours. In the case of GBR pilots, between 37 and 50 percent of bridge periods occurred during night hours which were defined as between 1818-0525 hrs. Approximately 30 percent of bridge periods were undertaken during the critical hours (defined as 2300-0600hrs), a period which has been shown to coincide with reduced alertness, sub-optimal performance and higher levels of fatigue (Costa, 1993; Meijman et al., 1993; Totterdell et al., 1995; Akerstedt, 1995; Luna et al., 1997 ). Questionnaire responses indicted that a considerable number of pilots reported feeling vulnerable to performance decrements during early morning hours (Parker et al., Report No 4, 1998). Additionally, a greater risk of accidents is associated with night work, especially in the early morning hours (Brown, 1994; Folkard, 1997; Mittler et al., 1988). This relationship is demonstrated by the higher frequency of shipping incidents during these hours in the Great Barrier Reef (Filor, 1998) and other navigational regions and transport industries (Sanquist et al., 1996; McCallum et al., 1996).
11.4 Alertness Ratings
On average, pilots rated their level of alertness at between 6 and 7 which is positioned towards the upper end of the rating scale. This level of alertness was comparable with ratings of mariners involved in day shift duties (Sanquist et al., 1996). Pilots also indicated that up to12 percent of bridge periods were associated with alertness levels consistent with sleepiness, and up to 58 percent of time on the bridge was associated with fatigue symptoms. These findings were inconsistent with the relatively high alertness ratings, and suggest that the pilots’ estimates of alertness may have been slightly over rated. By comparison, an investigation of merchant marine personnel indicated 11percent of work periods were associated with minimum alertness and 28 percent of work periods displayed one or more fatigue symptoms (Sanquist et al.,1996).
The significance of maintaining optimal alertness levels during GBR pilotage is reinforced by reference to vessel accident /incident statistics in this region between 1985 and 1997. Of the 35 reported incidents, all areas of the reef in which GBR pilots operate were represented. This emphasises the high prevalence of potential hazards such as extensive reef networks, shoal water zones, narrow channels and variable tidal conditions throughout the pilotage region (AMSA, 1993; AMSA, 1996). This situation supports the need to sustain optimal alertness levels during bridgework throughout the whole pilotage region.
Several characteristics of bridgework, in particular, the duration, the percentage of bridge periods at night and during critical hours, low alertness levels, and the presence of fatigue symptoms, indicated likely reductions in work performance and thus increased accident risk. There was evidence to suggest that these factors are likely to be more problematic on Inner Route pilotages and with personnel who undertake a greater number of work assignments and experience shorter breaks than shown by average company figures (Parker et al., Report No 2, 1998).
11.5 Sleep Patterns at Sea
Many authors have reported reduced quantity and quality of sleep and a greater incidence of sleep disorders among individuals involved in work outside normal working hours (Berger, 1983; Griffiths, 1993; Harma, 1993; Parkes, 1994). This can be attributed to the fact that sleep taken during the daytime opposes circadian rhythms (Griffith, 1993; Rutenfranz et al., 1988; Scott & Ladou, 1990) and is often disrupted by social and environmental factors, such as activity, noise and heat. As a consequence, there tends to be an accumulative sleep deficit and increased levels of fatigue when successive night shifts are performed (Griffiths, 1993; Parkes,1994; Scott & Ladou, 1990).The disrupted or inadequate sleep experienced by shiftworkers may have ramifications with respect to work performance and tasks with a high vigilance component are particularly sensitive to sleep loss. Performance deterioration as a function of sleep loss tends to be further exacerbated when a cumulative sleep debt is incurred and as a consequence safety may be seriously compromised (Tilley et al., 1982).
Logbook data provided the first opportunity to record additional details with respect to the patterns and quality of pilots’ sleep while at sea and ashore. Sleep patterns at sea differed according to: the particular shipping route; the duration of sleep at sea; the duration of each sleep period; and the number of sleep periods. For example, logbook analysis revealed total sleep time per 24 hours ranged from 0.6 hours for the GNE Channel to 4 and 5.25 hours for Hydrographers Passage and the Inner Route, respectively. Sleep was also fragmented, particularly on the Inner Route, with an average of three sleep periods taken during each 24 hours. By comparison with other maritime studies the sleep duration of the present pilot group was shorter, with an average of 6.6 and 7.5 hours daily being reported for US and European mariners, respectively (Sanquist et al.,1996; Rutenfranz and others (1988). These group differences are likely related to the shorter periods of time pilots spend onboard vessels and the on call nature of pilotage work. The fatiguing effects of the relatively short and fragmented sleep at sea, may be further exacerbated by the finding that up to 50 percent of this sleep is outside optimal hours (2200-0800) on the Inner Route and GNE Channel and thus may be inferior in terms of its recuperative value.
Other findings indicative of a fatigued state were demonstrated by the fact that approximately 40 percent of sleep periods at sea and ashore displayed latencies of less than 5 minutes, and almost all of the time in bed was spent with little loss of sleep due to awakenings.
Another aspect of sleep which has the potential to impact on fatigue and performance is the daily sleep debt which is defined as the difference between the hours of sleep ashore and at sea. The daily sleep debt of 3 hours incurred by pilots is greater than that found in other Australian seafarers (Parker et al, unpublished data) and US merchant marine personnel (Sanquist et al., 1996). However, when assessing the impact of the sleep debt, it should be noted that pilots return ashore during assignment breaks following work at sea lasting between 14-16 hours on the shorter routes and 54 hours on the longer route. In contrast, other seafarers in Australian and international fleets tend to stay at sea for up to 8 weeks and thus may incur the sleep debt for markedly longer periods.
Nevertheless, previous study results have indicated that even a reduction in sleep of 1.5-2 hours for one night is associated with decreased human performance (Gillberg, 1995). Such sleep deprivation may be characterised by a slower response speed to new and previously encountered stimuli (Dinges,1992; McCarthy & Waters, 1997), an increased tendency for false positive responding (Dinges, 1992), memory problems and slowed responses to unexpected events (Dinges & Kribbs, 1991). From questionnaire responses during the present project, pilots considered that performance decrements manifested as difficulty in concentrating, maintaining attention and memory problems (Parker et al., Report No 4, 1998). Such performance decrements could potentially have serious implications on piloting performance and may jeopardise ship safety.
While the acute sleep debt incurred during one assignment may be partially eliminated by the recuperative sleep ashore during breaks, there is the potential for chronic sleep debt to occur when breaks ashore do not provide sufficient recuperative sleep. In terms of vigilance levels and safe navigation, several studies have shown that chronic sleep deprivation reduces the ability to direct attention to the task at hand (Blagrove, et al., 1995; Herscovitch & Broughton, 1981). In the present pilot group sleep deprivation could be best described as intermittent, however, the consequences of chronic sleep deprivation may still be experienced at a relative level.
Pilots themselves acknowledged the difficulties associated with sleep at sea. Pilot commentaries indicated that opportunities for sleep at sea during Great Barrier Reef pilotage work varied greatly and were dependent on the shipping route, the ships schedule, prevailing conditions and bridge team competency.
Several suggestions from pilots for improving sleep opportunities at sea related to the charting of additional regions to bypass difficult navigational sections of shipping routes and the extension of navigational aids throughout the entire region (Parker et al., Report No 4, 1998).
11.6 Breaks between Assignments
In most situations, fatigue can be alleviated by an adequate period of rest between work assignments. However, if the rest period is insufficient to allow effective recuperation and alertness, fatigue will accumulate and manifest in a chronic form (Cameron, 1974, Grandjean, 1970). Schedules that employ limited rest time tend to result in cumulative fatigue as the rest period cannot completely offset the fatigue acquired during the work. While the adaptive nature of humans may enable performance to be maintained at a satisfactory level for some time, during periods of increased stress individuals who persist with these schedules are usually unable to cope and performance decrements become evident.
The impact of shipping route duration on work and sleep at sea is addressed in a basic way by the present guidelines for minimum rest breaks between consecutive pilotage assignments performed in the Great Barrier Reef-Torres Strait region. These guidelines state that prior to any work assignment, a minimum of 12 consecutive hours of rest, excluding travel must be taken by the pilot, except when the Inner Route passage is to be piloted, in which case at least 24 consecutive hours of rest, excluding travel is required (AMSA, 1997).
Closer examination of activity during assignment breaks indicated breaks averaged between 35 and 72 hours. While average break figures appeared to indicate that break duration was in keeping with guidelines for rest breaks some concerns were identified. Our extensive analysis of 4310 work assignments (Parker et al., Report No 2, 1998) and confirmed by logbook data, showed that between 5 and 10 percent of breaks did not conform to present guidelines for rest breaks. In addition, breaks between assignments began at all hours during the 24 hour cycle, thus displacing a proportion of sleep from the normal circadian cycle.
11.7 Sleep Ashore
Sleep ashore averaged from 9 to 11 hours per 24 hours, and was taken in one sleep period of approximately 7 hours with an additional shorter sleep of 2-3 hours. Pilots reported that they were able to sleep with relative ease considering the timing irregularities and the frequent change of location (Parker et al., Report No 4, 1998). However, it is unlikely that pilots have undergone any adaptation to irregular work and sleep patterns despite the many years of pilotage service. Within the literature there is considerable evidence which indicates there is little, if any, circadian adaptation to work schedules involving frequent changes of work times (Colquhoun 1985; Costa 1993; Luna 1997; Monk & Folkard 1992). In terms of the timing of sleep ashore, between 20 and 30 percent of sleep was taken outside the optimal sleep hours of 2200-0800 hours and considerable evidence exists to indicate that such sleep has diminished recuperative value (Akerstedt, 1995; Folkard, 1996; Tilley et al., 1982; Folkard & Barton, 1993; Kecklund et al., 1997). Logbook data also revealed that 60 percent of assignment breaks were spent either in pilot accommodation, hotels/motels. Therefore, exogenous factors such as poor sleeping facilities, outside noise and activity, increasing temperatures and natural sunlight may further compromise sleep when taken out of regular sleeping hours (Akerstedt, 1995; Rutenfranz et al., 1988).
Recovery values of sleep following a period of normal wakefulness have been estimated as a function of different lengths of sleep. For instance, 8 hours of sleep provides 100 percent recovery, whereas 2.8 hours of sleep gives only 67 percent recovery. However, he also indicated that these values may exaggerate the exponential nature of recovery, that is, recovery may be more linear (Simon Folkard, Personal Communication, June, 1998).
A significant proportion of the evidence related to breaks and sleep at sea and ashore suggested that sleep was likely to be of poor quality. Therefore, the finding that pilots reported relatively high sleep quality at sea and ashore was slightly unexpected and inconsistent with previous reports on this group which indicated that the majority of pilots experienced poor to very poor sleep quality at sea (Parker et al., 1997). By comparison, pilots sleep quality ratings were in the range reported by merchant marines involved in daywork, and higher than those reported by watchkeepers (Sanquist et al.,1996). Thus the present group may have slightly overrated sleep quality which is not unusual in studies involving subject assessments, particularly in occupational settings.
11.8 Relationship between work and sleep patterns and fatigue
A major aim of the logbook analysis was to identify key factors associated with the work and rest patterns that are correlated with fatigue and decreased alertness on the bridge. To this end a number of statistical models were constructed as an initial step towards the development of a formula to predict fatigue, stress and alertness on the bridge.
The results of the modelling process indicated that pilot differences in stress, fatigue and alertness, work assignment and break duration accounted for over 90 percent of the variance in fatigue, stress and alertness levels during bridge work. It was notable that break duration prior to assignments accounted for relatively more of the variance in alertness than in fatigue or stress. In contrast, the duration of the assignment accounted for relatively more of the variance in stress and fatigue than in alertness. That a considerable proportion of work assignment and pilotage time is spent on the bridge in the Barrier Reef region suggests that the increased fatigue is a result of not only the long hours but also exposure to other workplace stressors during these prolonged periods of work (Spurgeon et al.,1997).
The substantial variance in fatigue, stress and alertness attributed to by individual pilot differences may reflect actual differences in pilots’ thresholds to cope with these three factors. Alternatively this finding may simply be due to limitations in the use of rating scales to make comparisons across groups or to the ‘negative effect’ phenomenon identified in the literature, that is, some individuals complain about everything – others complain about nothing. While a cautionary approach should be taken to the significance of individual pilot differences, there is evidence in the literature to suggest the existence of several factors which may influence individual thresholds in coping with stress and fatigue. For example, basic differences in physiological measures of health and fitness and personality differences could be expected to influence tolerance to various aspects of pilotage work. Poor levels of physical fitness have been shown to reduce an individual’s ability to handle the stress associated with irregular hours (Harma, 1993).
During an investigation of UK pilots it was noted that individuals who found it difficult to relax, and whose biological systems did not tolerate work outside normal hours, could be considered ‘vulnerable’ and hence display a lower threshold for coping with the highly variable work situations (Shipley, 1978). It is also generally recognised that increasing age has adverse effects on tolerance to shiftwork as it brings about changes not only in the sleep/awake cycle but also tends to result in earlier phasing of circadian functions, that is, people become more ‘morning oriented’ as they age (Gander et al., 1993, Harma, 1993, Parkes, 1994). Interestingly, the present pilots who had an average age of 53 years considered themselves more morning than evening oriented (Parker et al., Report No 4, 1998).
The modelling procedures identified a number of key items which are currently included or could be included in the reporting procedures to enhance the value of the data with respect to the prediction of fatigue in pilots These items comprise, work assignment and break duration data, total duration of bridge periods, time on the bridge during critical hours, (that is, between 2300-0600 hrs) and travel data. These models could be further refined to include more detailed temporal aspects of work and sleep patterns which would enable changes in fatigue, stress and alertness during bridgework to be predicted for each additional hour of duty on the bridge.
Although preliminary, the current modelling is consistent with and complementary to previous approaches to exploratory modelling of alertness and fatigue in other maritime studies. For instance, Sanquist et al. (1996) reported that total sleep quality was the main predictor of alertness during a work period. On the other hand, modelling based on the outcomes of casualty investigations, found sleep and work duration in the past 72 hours and reports of fatigue symptoms correctly classified 80 percent of casualty cases in terms of the presence of fatigue (McCallum et al., 1996). Another model, which was developed to evaluate the impact of reduced sleep on alertness (Akerstedt & Folkard, 1995) could be extremely useful in the maritime setting with further refinement to include the restorative value of sleep across the 24 hour cycle and the impact of major restricted sleep periods (Sanquist et al., 1996).
11.9 Other factors contributing to fatigue
A common theme throughout the various phases of this research was the perception by pilots and their wives/partners that the introduction of competition had a negative impact on financial security, safety and domestic situations (Parker et al., Reports No 3 & 4, 1998). It is acknowledged by pilots that many aspects of pilotage such as the irregularity of work and breaks, the percentage of work at night and during critical hours are present whether or not a single, or multi provider situation prevails. However, there is evidence from comments by pilots and their partners that commercial interests and the desire to maintain income levels results in the need to undertake increased numbers of assignments and a reduction in break periods. As evidenced earlier in this report these increased loads are likely to compound the potential to develop fatigue. Although there is very limited information related to the impact of competition on fatigue in mariners there is a strong feeling among pilots in Australia and overseas that competition may compromise safety, particularly when pilots are expected to exercise independent judgment and resist pressures which are inconsistent with the interests of safety (Sparks,1998, Cash, 1998).
In summary, the various phases of this project represent the first comprehensive examination of the work practices of GBR pilots and the likely impact of these practices on the development of fatigue. The project was initiated following an earlier survey of the Health, Stress and Fatigue of Australian Seafarers which identified a number of concerns regarding fatigue in this population. These included the reduced quantity and quality of sleep while at sea, and an older workforce with some evidence of increased cardiovascular risk factors. The potential for a major environmental catastrophe in the Great Barrier Reef region and the increasing recognition of the importance of human factors and fatigue in both vessel and personnel accidents also provided an important stimulus for the implementation of this research. The research was also conducted against a background of significant change in the maritime industry and introduction of competitiveness and increased commercial pressures in shipping generally, and pilotage in particular.
The results from all phases of the research have shown that the work practices of pilots are characterised by several features relating to both work assignments and bridge and sleep periods at sea and sleep ashore that have the potential to increase fatigue levels. These include the irregularity of work and displaced work and sleep from normal circadian cycles, a considerable percentage of bridge work during night and early morning hours, short and fragmented sleep at sea with a considerable percentage of sleep at sea and ashore taken in opposition to normal physiological sleeping times for optimal recuperative value. Furthermore, the fluctuations in workloads during particular time periods and the situations where personnel undertook considerably greater workloads and shorter breaks than average company figures further exacerbated the fatigue potential related to the general and specific features of pilotage work.
The research has provided extensive documentation of the work practices of GBR pilots and some evaluation of the nature and extent of fatigue associated with these practices. The findings provide a basis for the establishment of enhanced fatigue monitoring and management procedures and their implementation at a regulatory, company and individual level.
last updated: October 1998