University of Groningen
Sleep and fatigue offshore
Riethmeister, Vanessa
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2019
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Riethmeister, V. (2019). Sleep and fatigue offshore. University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
ABSTRACT
Objectives This study examined daily scores of fatigue and circadian rhythm markers over two-week offshore day-shift periods.
Methods A prospective cohort study among N = 60 offshore day-shift workers working two-week offshore shifts was conducted. Offshore day-shifts lasted from 07:00 – 19:00 h. Fatigue was measured objectively with pre- and post-shift scores of the 3-minute psychomotor vigilance tasks (PVT-B) parameters (reaction times, number of lapses, errors and false starts) and subjectively with pre- and post-shift Karolinska Sleepiness Scale (KSS) ratings. Evening saliva samples were collected on offshore days 2,7 and 13 to measure circadian rhythm markers such as dim-light melatonin onset times and cortisol. Generalized and linear mixed model analyses were used to examine daily fatigue scores over time.
Results Complete data from N = 42 offshore day-shift workers was analysed. Daily parameters of objective fatigue, PVT-B scores (reaction times, average number of lapses, errors and false starts), remained stable over the course of the two-week offshore day-shifts. Daily subjective post-shift fatigue scores significantly increased over the course of the two-week offshore shifts. Each day offshore was associated with an increased post-shift subjective fatigue score of 0.06 points (95%CI: 0.03 – 0.09 p < .001). No significant statistical differences in subjective pre-shift fatigue scores were found. Neither a circadian rhythm phase shift of melatonin nor an effect on the pattern and levels of evening cortisol was found.
Conclusion Daily parameters of objective fatigue scores remained stable over the course of the two-week offshore day-shifts. Daily subjective post-shift fatigue scores significantly increased over the course of the two-week offshore shifts. No significant changes in circadian rhythm markers were found. Increased post-shift fatigue scores, especially during the last days of an offshore shift, should be considered and managed in (offshore) fatigue risk management programs and fatigue risk prediction models.
4
83 83 INTRODUCTION
Fatigue has been identified to be among the most important health and safety risk factors in the offshore oil and gas industry.1,2 Some of the major offshore and industry disasters have been linked to human error, and more specifically fatigue.3,4 In industrial settings the terms fatigue and sleepiness are often used interchangeably although conceptual differences exist. Fatigue has been associated with impaired task performance resulting from physical or psychological strain, whereas sleepiness has been associated with the neurobiological need to sleep, resulting from physiological wake and sleep drives.5 Although the causes of fatigue and sleepiness may vary, the consequences are similar. Both fatigue and sleepiness may cause mental and physical performance impairments, which can increase the likelihood of health and safety incidents. In this article, we will use the term ‘fatigue’ to refer to both fatigue and sleepiness constructs.
The offshore work conditions (e.g. remote shift work, limited light exposure, long periods of consecutive work days and 12-h shifts) are likely to predispose offshore workers to higher degrees of fatigue and fatigue-related work injuries and accidents. Previous research showed that shift work, long daily work hours and working more than 50 h per week increases the risk of work injuries related to poor sleep quality.6-8 In the offshore environment, poor sleep quality,9-11 short sleep periods,10-12 and high fatigue scores13,14 have been identified.
Although a few studies on fatigue have been conducted in the offshore industry, not much is known on daily fatigue scores during offshore shifts. The majority of existing offshore studies has focused on the effect of shift work (night- and swing-shifts) on fatigue, circadian rhythm adjustments as well as health and safety outcomes.15-17 Yet, only a few studies have also identified fatigue and sleep problems in offshore day-shift workers.10,11 Ignoring offshore day-shift workers from occupational health and safety studies is of concern, as offshore day-day-shift workers represent the largest workforce in the oil and gas industry. In addition, most existing offshore fatigue and sleep studies have not used longitudinal, repeated measures designs and have for the most part been conducted using small sample sizes15,18 employing primarily self-reported measures.10,12,15,18
An important factor in preventing fatigue is a consolidated period of sleep. A consolidated sleep phase is possible when people have the opportunity to sleep long enough at the optimal circadian phase.19 The high prevalence of fatigue offshore may be caused by a multitude of factors related to the nature of offshore work affecting sleep and circadian rhythms. For example, combinations of long offshore work periods, 12-h work days, early start times and offshore workers evening behaviours and routines, can change the phase angles between sleep and circadian rhythms, resulting in circadian rhythm phase shifts even in day-shift workers. These circadian rhythm phase shifts may negatively impact offshore workers sleep timing, duration and quality as well as daily recovery processes and fatigue in return.
A more thorough investigation of daily fatigue scores over two-week offshore day-shifts is needed to better understand the course of fatigue over time spent offshore and the possible association with circadian rhythm markers. Moreover, detailed knowledge on daily fatigue scores over offshore day-shift periods will help to improve the existing fatigue risk prediction models. Thus, the aims of this study were to examine daily fatigue scores and changes in circadian rhythm markers over the course of two-week offshore day-shift periods.
MATERIALS AND METHOD Study population & design
A longitudinal observational study was conducted in N = 60 offshore day-shift workers on four offshore platforms located in the Dutch Continental Shelf. This study is part of a larger investigation on sleep and fatigue parameters across full offshore day-shift rotations, including both work and leave periods. The present study concerns the two-week offshore work period. The offshore work schedule consisted of two weeks of 12-h workdays. The day-shifts lasted from 07:00 – 19:00 o’clock. During the study period, no overtime was officially requested nor recorded. Break times can vary between platforms. In this study, no data on break times was assessed. All offshore workers, contractors and permanent staff, working 14-day rotations on one of the four selected platforms, between February to June 2015, were invited to participate in the study. Offshore workers were excluded from the study if they performed any night-shifts during the study period. (Figure 1) Study participation was voluntarily. Informed consent was obtained from all participants. Ethical approval for the study was granted from the Medical Ethics Committee of the University Medical Center Groningen, The Netherlands (reference number: M14.165646).
Figure 1. Flowchart of study sample.
Final sample size N = 42
Night-shift: N = 8 Volunteers
N = 60
Total sample size N = 50 Dropout x Shift changes N = 2 x Personal reasons N = 4 x Contract ended N = 2 x Forgot study N = 2 equipment
4
85 85 Procedure
The study was conducted between February and June 2015. Due to operational and logistic constraints, it was not possible for the investigators to be present on the offshore platforms for the conduction of the study. On each platform, a study supervisor was allocated to implement the study protocol. The implementation was mainly executed by the offshore medic or first aider. In addition, elaborate briefing sessions with the offshore platforms and the participating offshore workers were held to ensure correct conduction of the study tasks. Also, individualized study material (i.e. actigraphs) and procedures were sent to the offshore workers including detailed daily study timelines. One week before the start of the offshore shift period, all participating offshore workers received an electronic link, via e-mail, for a baseline questionnaire on demographic, work and health variables. In addition, offshore workers started to wear an actigraph (MotionWatch8®, Camntech) one week before the offshore work period commenced until one week after the offshore work period ended. This study concerns only the offshore work period.
Twice a day electronic questionnaires were used for the assessment of fatigue. Every morning before shift start and every evening post-shift, offshore workers received a personal electronic link, via e-mail, to the fatigue questionnaire. An online questionnaire portal was used to access, monitor and store fatigue recordings. Psychomotor vigilance task (PVT-B) testing took place in the accommodation block of the offshore platforms. Offshore workers were instructed to find a quiet room to complete the task. The online portal of the PVT-B app provider (Pulsar Informatics; Joggle Research®) was used to access, monitor and store PVT-B recordings in real-time. On days with saliva sampling, daily reminders were sent to the offshore platforms and offshore workers. Offshore workers were instructed to collect hourly saliva samples from 19:00 h until bedtime. Exposure to light, consumption of food and beverages other than water were forbidden during this period. All salivary samples were stored in the offshore freezers (-20°C) until study sampling was completed. Upon study completion, all samples were sent to the laboratory of the University Medical Center Groningen, Groningen, The Netherlands for analyses.
Measurements
Fatigue
Reaction times and accurateness on simple reaction time tasks, objective proxies for fatigue, were measured before and after each shift with the 3-minute IPad app version of the
psychomotor vigilance tsk (PVT-B) (Pulsar Informatics; Joggle Research®).20,21 Each platform
was equipped with a maximum of four IPads which were shared among offshore workers and were stored in the common room areas of the living quarters. The PVT-B has been successfully implemented in workplace settings to measure subcomponents of fatigue such as: alertness,
sleepiness, and neurobehavioral performance.20,22,23 Due to operational constraints, no
were investigated: mean valid reaction times in milliseconds (reaction times > 100 ms), number of false starts (responses without a stimulus or reaction times < 100 ms), errors (pressing the wrong button or failing to release the button for 3 s or longer), and lapses ;ƌĞĂĐƚŝŽŶƚŝŵĞƐш 355 ms). The lapse threshold was adopted from suggested PVT-B scoring algorithms for 3-min versions.20
Self-reported sleepiness, from now on referred to as subjective fatigue, was assessed pre- and post-shift (07:00 & 19:00 h) using the Karolinska Sleepiness Scale (KSS).24 The KSS consists of a nine-point Likert scale, rating sleepiness with (1) extremely alert, (2) very alert, (3) alert, (4) rather alert, (5) neither alert or sleepy, (6) some signs of sleepiness, (7) sleepy, but no effort to keep awake, (8) sleepy, some effort to keep awake (9) very sleepy, great effort to keep awake, fighting sleep.25,26 Higher KSS scores indicate higher subjective fatigue.
Circadian rhythm markers
Circadian rhythm markers were calculated based on salivary cortisol and melatonin levels in the evening. Saliva was collected using Salivettes® cotton rolls (Sarstedt) on three offshore days (day 2, 7 and 13) following the methodology of Harris and colleagues.15 Both, the pattern of evening cortisol concentration on the three investigated offshore days and the average cortisol concentration in the evening were used as a measure of circadian rhythmicity over the course of the two-week offshore shifts. Dim light melatonin onset (DLMO) times were measured as phase markers of endogenous time. Cortisol and melatonin rhythms have been shown to be reliable circadian markers27,28 and have been used in industrial settings to measure circadian disruption as a result of shift schedules and occupational exposures.15,29
Saliva samples were analysed using liquid chromatography in combination with isotope dilution mass spectrometry, as described elsewhere.30,31 The functional sensitivity was 200 pmol/L for cortisol and 3.0 pmol/L for melatonin. Salivary melatonin scores were used to obtain DLMO times. DLMO was defined as the moment at which the melatonin rhythm crossed the 11.01 pmol/L concentration by linearly interpolating the raw values around the 11.01 pmol/L concentration at the rising part of the curve, or that was within the first hour of extrapolation.32
Covariates
Covariates included platform location, self-reported chronotype and time in bed (TIB). Offshore platforms varied in size, age, location, equipment and culture. Therefore, platform location was used as a covariate in all linear mixed models to adjust for possible cluster effects. Self-reported chronotype was included in all mixed models with circadian rhythm markers to adjust for offshore workers chronobiological inclination. Chronotype was defined as the midsleep point on days off-work corrected for sleep on working days and was assessed at
4
87 87
baseline with the Munich Chronotype Questionnaire.33,34 The later the mean midsleep sleep-corrected time, the later the chronotype. The official scoring protocol was used.35
TIB was assessed using daily actigraphy recordings from the MotionWatch8®, Camntech. The MotionWatch8® is a wrist-worn actigraph, worn on the non-dominant hand, which has been shown to be a valid and reliable measure for sleep parameters, using tri-axial sensors data.36 Generated data consisted of 1-minute epochs.
Statistical analysis
Generalized and linear mixed model analyses were conducted to examine daily fatigue scores and changes in circadian rhythm markers over the two-week offshore day-shifts. Linear mixed models were used for continuous outcomes (KSS scores, reaction times, cortisol and DLMO times). Log transformations were performed for pre- and post-shift reaction times and cortisol concentrations as the data was log-normally distributed. We chose to dichotomize the PVT-B variables: number of false starts, errors and lapses, because of the high number of zeros in these variables. For these outcome variables, we used generalized linear mixed models. PVT-B test observations were excluded if they showed reaction times higher than 1000 ms (1x), or if they had more than 20 lapses (7x), false start (2x) or errors (2x). Inclusion of random intercepts and/or slopes was based on lowest Aikaikes Information Criteria (AIC) values. Time (offshore days) was entered as a continuous variable and PVT-B scores (reaction times, number of lapses, number of errors, number of false starts), subjective fatigue (KSS), evening cortisol concentrations (timing and level) and DLMO times were entered as dependent variables. For both pre- and post-shift PVT-B and KSS scores, separate (generalized) linear mixed models were performed. All models were adjusted for platform location by entering platform as a fixed effect. Analyses with cortisol levels were additionally adjusted for self-reported chronotype. In addition, all mixed model analyses were repeated with adding TIB as an extra explanatory variable to assess whether TIB could (partly) explain possible trends over time (see appendix).
For circadian rhythm analyses, the average log-transformed cortisol concentration in the evening (main effect of days) and the pattern of average log-transformed evening cortisol concentration across the three sampling days (interaction between day and time of day) were investigated. The effect of time of day on average log transformed evening cortisol concentrations was fitted adding a quadratic function of time of day. For DLMO, the change in DLMO times across the three sampling days was examined. Missing values were assumed to be missing at random. Under this assumption, linear mixed models provide valid estimates, without data imputation.37 All analyses were performed using SPSS version 23.
RESULTS
Sample characteristics
The final sample consisted of N = 42 (70%) male offshore workers, after the exclusion of N = 8 offshore workers who worked night-shifts during the study period. The spread of offshore workers across the platforms was as follows: Platform 1 (N = 15), platform 2 (N = 13), platform 3 (N = 9) and platform 4 (N = 5). About half of the offshore workers were contractors (N = 20, 48%). Complete data was obtained for PVT-B, KSS and circadian rhythm phase markers for N = 35, N = 42 and N = 39, respectively. Sample characteristics are displayed in Table 1.
Table 1. Sample characteristics of the final study sample (N = 42).
Mean SD Range
Age (years) 42 12.1 21 – 63
Job tenure (years) 6.6 6.2 1 – 27
Mean midsleep sleep-corrected (hh:mm)* 02:59 00:37 02:00 – 04:17 Time in Bed (hh:mm)** 07:32 0:58 04:10 – 10:58 *Mean midsleep sleep-corrected represents offshore workers chronotype.
**Time in Bed (TIB) is based on N = 41 offshore workers du to sickness of one offshore worker who left the platform after 5 days offshore.
Fatigue
Daily parameters of objective fatigue, PVT-B scores (reaction times, average number of lapses, errors and false starts), remained stable over the course of the two-week offshore day-shifts (Figure 2, Appendix Table A1). Daily post-shift subjective fatigue (KSS) scores significantly increased over the two-week offshore day-shifts. Each day offshore was associated with an increase in post-shift KSS score of 0.06 points (95%CI: 0.03 – 0.09, p < .001). Adjustment for TIB, did not substantially change the results (Appendix Table A1). No significant changes of pre-shift subjective fatigue scores over the course of the two-week offshore day-shifts were found (Figure 3).
Circadian rhythm markers
In total, N = 243 saliva samples were analysed. Four cortisol concentration measurements scored zero and were assigned the next smallest cortisol concentration value for the linear mixed model analyses. Absolute cortisol concentration values were not normally distributed. The absolute median cortisol concentration value was 0.73 pMol/l (IQR = 1.79) on day 2, 0.76 pMol/l (IQR = 1.82) on day 7 and 0.78 pMol/l (IQR = 1.90) on day 13, respectively. There was a significant main effect of time (b = -0.29, 95%CI: -0.40 – -0.18, p < .001) and the quadratic
function time2 (b = 0.07, 95%CI: 0.00 – 0.14, p = .036) on the pattern of average log-
transformed cortisol concentration in the evening. Cortisol decreases over time, but this pattern over time was not significantly different between the three study days (p-value for
4
89 89 interaction = 0.433) (Figure 4, graph 1). No significant main effect of study day for average log-transformed evening cortisol concentration was found (p = .250). Mean predicted DLMO times were at 20:08 (0:58) for day 2, 20:08 (0:59) for day 7 and 20:02 (1:03) for day 13 (Figure 4, graph 2). The timing of DLMO did not differ significantly between the three study days (p = .832).
Figure 2. Outcomes of the pre- and post-shift objective fatigue metrics, obtained from the psychomotor vigilance
task (PVT-B), over the course of the two-week offshore day-shift periods. Means, standard errors and linear prediction lines are plotted for (1) reaction times, (2) number of lapses, (3) number of errors, (4) and number of false starts. 200 220 240 260 280 300 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean reaction time s (m s) Offshore days 0 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean n u mber of lapses Offshore days 0 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean n u mber of errors Offshore days 0 1 2 3 4 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean n u mber of false starts Offshore days
Figure 3. Mean pre- and post-shift subjective fatigue scores with 95% confidence intervals, measured with the
Karolinska Sleepiness Scale (KSS), and trendlines over the course of the two-week offshore shifts. Please note truncation of Y-axis.
Figure 4. Outcomes of the circadian rhythm marker analyses. Graph (1) depicts mean salivary cortisol levels and
standard errors during the three sampling days (offshore day 2, 7 and 13). Graph (2) depicts dim-light melatonin onset (DLMO) times and standard errors across the three sampling days (offshore days 2, 7 and 13).
19:40 19:47 19:54 20:01 20:09 20:16 20:23 20:30 2 7 13 DLMO time s (hh:m m ) Offshore days DLMO -1.00 -0.80 -0.60 -0.40 -0.20 0.00 0.20 0.40 0.60 19:00 20:00 21:00 22:00 Mean ln cortisol con centration Time (hh:mm)
Day 2 Day 7 Day 13
2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean KSS Offshore days
Morning Evening Linear (Morning) Linear (Evening)
9.0
4
91 91 DISCUSSION
Daily parameters of objective fatigue, PVT-B scores (reaction times, average number of lapses, errors and false starts), remained stable over the course of the two-week offshore day-shifts. Post-shift subjective fatigue scores significantly increased over the course of the two-week offshore day-shifts. In addition, no circadian rhythm phase shift was found over the course of the two-week offshore day-shifts.
Fatigue
Daily objective fatigue (PVT-B) scores remained stable over the course of the two-week offshore shift periods. A potentially learning effect may be excluded as an explanation for this
finding, as the PVT-B has been found to only have minor learning effects.38 Participants on one
offshore platform reported that the PVT-B task turned into a competition in which offshore workers kept scores on who had the fastest reaction time. This competitive PVT-B task environment might have positively influenced the motivation and alertness levels of offshore workers and reduced reaction times, accordingly. In general, reaction times were fast
compared to other working populations23 and our scores align with previous reaction times
found among other offshore workers.15
Post-shift subjective fatigue scores, significantly increased over the course of the two-week offshore day-shifts. It is important to note that although the daily increases of subjective fatigue seemed low, they might be relevant in practice. According to the Karolinska Sleepiness
Scale (KSS) developers, daytime KSS ratings usually range between 3 and 4.39 Our findings
indicate slightly higher daytime KSS ratings between 4 and 5. Even though a daily increase of 0.06 on the KSS is small, the average offshore worker already reports higher KSS scores and the additional accumulation of KSS scores over time may pose potential additional risks. Akerstedt et al (2011) found that the relative risk of accident involvement doubles once KSS
scores are equal to 5.40 Hence, the increase in post-shift KSS scores over the course of the
two-week offshore day-shifts is a relevant finding for fatigue risk management programs in offshore operations.
On average, post-shift subjective fatigue scores were higher compared to pre-shift subjective fatigue scores. Over the two-week offshore shifts, pre-shift subjective fatigue scores remained fairly constant. A possible explanation for the elevated post-shift fatigue scores might be that the ability to regulate the daily subjective fatigue scores decreases over the two-week offshore shifts. This could be due to homeostatic depletion processes like depleted energy resources or accumulating sleep debts during the two-week offshore shifts, preventing daily
subjective fatigue recovery.41 More research is needed to investigate the underlying causes
Another interesting observation is the peak of post-shift subjective fatigue scores on day ten (about three quarters of the offshore shift) rather than day fourteen (the end of the offshore shift). Earlier, Waage et al (2012) investigated the course of subjective fatigue among three different offshore shift schedules (day-, night- and swing-shifts).18 The authors present a figure of the mean subjective fatigue (KSS) values over the two-week offshore period and visual inspection of their data reveals a peak in subjective fatigue scores on day eleven for day-shift workers (Figure 2, p.69). A potential explanation of this finding could be linked to psychosocial/motivational motives. In long-term space, military and Antarctica expeditions a related construct of significant psychological changes in mood and performance due to monotony, boredom and restricted social contacts has been termed the ‘third-quarter phenomenon’.42,43 During this period expedition staff experienced more conflicts, anxiety, and demotivation. The third-quarter phenomenon most likely results from the realization that the expedition is only just more than half-way completed, and a rather long period of work and isolation still awaits. Although, to our knowledge, never been directly linked to subjective fatigue, psychosocial changes during day ten/eleven of a two-week offshore shift might account for the increased post-shift fatigue scores in this period. Future research needs to investigate the links between sleep and fatigue and the third-quarter phenomenon in more detail and whether this phenomenon has potential implications for fatigue risk management programs.
Differential findings between objective and subjective fatigue scores have previously been noted and explained by different underlying theoretical constructs.44 Subjective fatigue ratings have been viewed as warning signs, signalling the brain that insufficient sleep was obtained and that preventative actions need to be initiated.44 Only when the subjective warning signals are ignored objective fatigue indications are noticeable. Thus, subjective fatigue perceptions occur before objective fatigue performance deteriorations. More research on the differential findings between objective and subjective fatigue is needed to further confirm and validate the findings.
Circadian rhythm markers
The course of evening cortisol levels, the average values of evening cortisol concentration, and DLMO times did not differ between the three investigated offshore days (days 2, 7 and 13). Cortisol levels decreased on each of the investigated evenings in line with normal 24 h-cortisol rhythms.45 In addition, salivary evening cortisol concentrations of offshore workers were comparable to normal reference values of other populations, ranging from 0.5 to 9.9 nmol/L.46-48
No significant differences in DLMO times were found across the three investigated offshore days. These results imply that living and working offshore for two weeks, in a 12-h day-shift schedule, did not induce a circadian rhythm phase shift of melatonin nor cortisol, nor did it
4
93 93
induce a change in the level of evening cortisol concentration. This finding is in line with previous circadian rhythm investigations offshore, which showed that mainly night-shift workers showed circadian misalignment.15,49,50 Considering these findings, it seems unlikely that the increase of fatigue scores of offshore day-shift workers over the course of two-week offshore shifts is the result of changes in timing of circadian rhythms or is related to elevated cortisol levels. More research is needed to further investigate these relationships and underlying mechanisms to verify our results.
Strengths and limitations
A strength of this study is that multiple daily fatigue measurements were taken, over the course of two-week offshore shifts, in a real-life work setting in a relatively large cohort of offshore workers compared to previous offshore studies, e.g. Harris et al. (2010) and Waage et al. (2012).15,18 In addition, both objective and subjective measures of fatigue and circadian rhythms were used. Previous offshore studies included only limited measurement points and mostly either objective or subjective parameters to investigate sleep and fatigue parameters offshore.18 Another strength concerns the fact that this study included offshore workers from several offshore platforms and countries.
Study limitations mainly concern operational and logistic constraints, which could have led to information bias. For instance, the principal investigator was not able to be on the platforms for data collection as the study took place on all four platforms simultaneously. As a result, measurements were conducted under supervision of the appointed offshore study supervisor, who had to assure the correct application of study procedures and measurements. We cannot exclude e.g. that the competitive PVT-B task environment on one of the offshore platforms might have led to biased information. However, as the competition was only reported on one platform, and only a limited number of offshore workers participated in the study at a given time, the risk of bias is probably low. Other operational constraints (i.e. weather conditions, flight times, handovers) prohibited the conduction of the B on day 1. In the analyses, PVT-B estimates for the missing first offshore day predicted. Also, due to operational constraints, no cortisol awakening response could be measured and our circadian rhythm sampling days were limited to the evening hours of three days. Furthermore, we cannot rule out residual confounding. Finally, the analyses included many statistical tests. Analyses were not adjusted for alpha levels, because this would reduce the power of the analyses. This means that the change over time in subjective post-shift fatigue might be the result of a type one error.
Implications
Elevated post-shift subjective fatigue scores may increase the risk of health and safety incidents among offshore workers. Being fatigued at the end of a shift is likely to increase the odds of making mistakes and/or errors, which could adversely impact the health and safety of offshore workers. Offshore work schedules should consider the potential adverse
consequences of elevated post-shift fatigue scores (especially towards the end of a two-week offshore shift rotation) and plan work tasks accordingly. For instance, if possible, safety critical tasks should be avoided towards the end of a shift or appropriate additional fatigue mitigation/proofing strategies should be put in place (i.e. buddy system/extra monitoring).51 Findings of this study might be applicable to other extended shift work environments in which workers accumulate fatigue over a prolonged period of time on shift. In shift work environments, road safety is a critical safety concern as increased subjective fatigue has been linked to an increased accident risks.40 Thus, both offshore and general shift work employers and employees, should focus more on post-shift fatigue risks and potential consequences, especially towards the last days of an (offshore) shift rotation.
More research on the antecedents and consequences of fatigue development offshore is needed to improve fatigue risk management programs and policies. In particular, daily recovery processes and day-on-shift effects on fatigue parameters pose interesting research areas and should be further investigated. In addition, health and safety incidents reports should be analysed on day-on-shift and time-of-day effects and monitored accordingly to investigate whether post-shift fatigue levels are a potential causal factor for incidents. Further analyses of the health and safety incident reports is likely to support the improvement of current fatigue risk management programs and policies. Moreover, the differential results of the objective PVT-B and subjective KSS fatigue ratings should be further investigated. Finally, the focus of this study was on day-shift workers, as this is a large, increasing and often neglected group of offshore shift workers. Therefore, the effects of long, remote, consecutive day-shifts need to be explored to manage fatigue risks associated with offshore day-shifts appropriately.
Conclusions
Daily parameters of objective fatigue, PVT-B scores, remained stable over the course of the two-week offshore day-shifts. Post-shift subjective fatigue significantly increased over the course of two-week offshore day-shifts. No circadian rhythm phase shift was found over the course of the two-week offshore day-shifts. Increased post-shift fatigue scores, especially during the last days of an offshore shift, should be considered and managed in (offshore) fatigue risk management programs and fatigue risk prediction models. Further research is needed to confirm our findings on daily fatigue scores and circadian rhythm markers during offshore shifts.
4
95 95 REFERENCES
[1] Ross JK. Offshore industry shift work—Health and social considerations. Occup Med. 2009;59(5):310-5.
[2] Parkes KR. Sleep patterns of offshore day-workers in relation to overtime work and age. Appl Ergon. 2015;48:232-9.
[3] U.S. Chemical Safety and Hazard Investigation Board. Investigation report volume 3 drilling rig explosion and fire at the macondo well. 2016; Report no. 2010-10-I-OS.
[4] U.S. Chemical Safety and Hazard Investigation Board. Investigation report - refinery explosion and fire. 2007; Report no. 2005-04-I-TX.
[5] Sadeghniiat-Haghighi K, Yazdi Z. Fatigue management in the workplace. Ind Psychiatry J. 2015;24(1):12-7.
[6] Uehli K, Miedinger D, Bingisser R, Duerr S, Holsboer-Trachsler E, Maier S, Mehta AJ, Mueller R, Schindler C, Zogg S, Künzli N, Leuppi JD. Sleep quality and the risk of work injury: A swiss case-control study. J Sleep Res. 2014;23(5):545-53.
[7] Folkard S, Tucker P. Shift work, safety and productivity. Occup Med-Oxford. 2003;53(2):95-101.
[8] Arlinghaus A, Lombardi DA, Willetts JL, Folkard S, Christiani DC. A structural equation modeling approach to fatigue-related risk factors for occupational injury. Am J Epidemiol. 2012;176(7):597-607.
[9] Parkes KR. Sleep patterns, shiftwork, and individual-differences - a comparison of onshore and offshore control-room operators. Ergonomics. 1994;37(5):827-44.
[10] Parkes KR. Age and work environment characteristics in relation to sleep: Additive, interactive and curvilinear effects. Appl Ergon. 2016;54:41-50.
[11] Menezes M, Pires M, Benedito-Silva A, Tufik S. Sleep parameters among offshore workers: An initial assessment in the campos basin, rio de janeiro, brazil. Chronobiol Int. 2004;21(6):889-97.
[12] Parkes KR. Age, smoking, and negative affectivity as predictors of sleep patterns among shiftworkers in two environments. J Occup Health Psychol. 2002;7(2):156-73.
[13] Riethmeister V, Brouwer S, van der Klink J, Bültmann U. Work, eat and sleep: Towards a healthy ageing at work program offshore. BMC Public Health. 2016;16:134.
[14] Parkes KR. Human factors, shift work and alertness in the offshore oil industry. part I A survey of onshore and offshore control room operators. part II alertness, sleep and cognitive performance. London, England: HSMO. 1993; Report no. OTH92-389.
[15] Harris A, Waage S, Ursin H, Hansen AM, Bjorvatn B, Eriksen HR. Cortisol, reaction time test and health among offshore shift workers. Psychoneuroendocrinology. 2010;35(9):1339-47.
[16] Barnes R, Forbes MJ, Arendt J. Shift type and season affect adaptation of the 6-sulphatoxymelatonin rhythm in offshore oil rig workers. Neurosci Lett. 1998;252(3):179-82.
[17] Bjorvatn B, Stangenes K, Oyane N, Forberg K, Lowden A, Holsten F, Akerstedt T. Subjective and objective measures of adaptation and readaptation to night work on an oil rig in the north sea. Sleep. 2006;29(6):821-9.
[18] Waage S, Harris A, Pallesen S, Saksvik IB, Moen BE, Bjorvatn B. Subjective and objective sleepiness among oil rig workers during three different shift schedules. Sleep Med. 2012;13(1):64-72.
[19] Dijk DJ, Czeisler CA. Paradoxical timing of the circadian-rhythm of sleep propensity serves to consolidate sleep and wakefulness in humans. Neurosci Lett. 1994;166(1):63-8.
[20] Basner M, Mollicone D, Dinges DF. Validity and sensitivity of a brief psychomotor vigilance test (PVT-B) to total and partial sleep deprivation. Acta Astronaut. 2011;69(11-12):949-59.
[21] Grant DA, Honn KA, Layton ME, Riedy SM, Van Dongen HPA. 3-minute smartphone-based and tablet-smartphone-based psychomotor vigilance test for the assessment of reduced alterness due to sleep deprivation. Behav Res Methods. 2017:49(3):1020-9.
[22] Lamond N, Dawson D, Roach GD. Fatigue assessment in the field: Validation of a hand-held electronic psychomotor vigilance task. Aviat Space Environ Med. 2005;76(5):486-9.
4
97 97
[23] Shattuck NL, Matsangas P. Psychomotor vigilance performance predicted by epworth sleepiness scale scores in an operational setting with the united states navy. J Sleep Res. 2015;24(2):174-80.
[24] Akerstedt T, Gillberg M. Subjective and objective sleepiness in the active individual. Int J Neurosci. 1990;52(1-2):29-37.
[25] Akerstedt T, Anund A, Axelsson J, Kecklund G. Subjective sleepiness is a sensitive indicator of insufficient sleep and impaired waking function. J Sleep Res. 2014;23(3):240-52.
[26] Kaida K, Takahashi M, Akerstedt T, Nakata A, Otsuka Y, Haratani T, Fukasawa K. Validation of the karolinska sleepiness scale against performance and EEG variables. Clin Neurophysiol. 2006;117(7):1574-81.
[27] Voultsios A, Kennaway DJ, Dawson D. Salivary melatonin as a circadian phase marker: Validation and comparison to plasma melatonin. J Biol Rhythms. 1997;12(5):457-66.
[28] Groeschl M. Current status of salivary hormone analysis. Clin Chem. 2008;54(11):1759-69.
[29] Ferguson SA, Kennaway DJ, Baker A, Lamond N, Dawson D. Sleep and circadian rhythms in mining operators: Limited evidence of adaptation to night shifts. Appl Ergon. 2012;43(4):695-701.
[30] Booij SH, Bos EH, Bouwmans MEJ, van Faassen M, Kema IP, Oldehinkel AJ, de Jonge P. Cortisol and alpha-amylase secretion patterns between and within depressed and non-depressed individuals. Plos One. 2015;10(7):e0131002.
[31] Bouwmans MEJ, Bos EH, Booij SH, van Faassen M, Oldehinkel AJ, de Jonge P. Intra- and inter-individual variability of longitudinal daytime melatonin secretion patterns in depressed and non-depressed individuals. Chronobiol Int. 2015;32(3):441-6.
[32] Benloucif S, Burgess HJ, Klerman EB, Lewy AJ, Middleton B, Murphy PJ, Parry BL, Revell VL. Measuring melatonin in humans. J Clin Sleep Med. 2008;4(1):66-9.
[33] Zavada A, Gordijn MCM, Beersma DGM, Daan S, Roenneberg T. Comparison of the munich chronotype questionnaire with the horne-ostberg's morningness-eveningness score. Chronobiol Int. 2005;22(2):267-78.
[34] Roenneberg T, Wirz-Justice A, Merrow M. Life between clocks: Daily temporal patterns of human chronotypes. J Biol Rhythms. 2003;18(1):80-90.
[35] Juda M, Vetter C, Roenneberg T. The munich chronotype questionnaire for shift-workers (MCTQ(shift)). J Biol Rhythms. 2013;28(2):130-40.
[36] Elbaz M, Yauy K, Metlaine A, Martoni M, Leger D. Validation of a new actigraph motion watch versus polysomnography on 70 healthy and suspected sleep-disordered subjects. J Sleep Res. 2012;21:218.
[37] Twisk J, de Boer M, de Vente W, Heymans M. Multiple imputation of missing values was not necessary before performing a longitudinal mixed-model analysis. J Clin Epidemiol. 2013;66(9):1022-8.
[38] Dinges D, Pack F, Williams K, Gillen K, Powell J, Ott G, Aptowicz C, Pack A. Cumulative sleepiness, mood disturbance, and psychomotor vigilance performance decrements during a week of sleep restricted to 4-5 hours per night. Sleep. 1997;20(4):267-77.
[39] Akerstedt T, Kecklund g, Gillberg M, Lowden A, Bjorvatn B. Subjective sleepiness in shift work. In: Touitou Y, editor. Biological clocks: Mechanisms and applications. Amsterdam, The Netherlands: Elsevier. 1998;235.
[40] Akerstedt T, Philip P, Capelli A, Kecklund G. Sleep loss and accidents-work hours, life style, and sleep pathology. Human Sleep and Cognition, Pt II: Clinical and Applied Research. 2011;190:169-88.
[41] Raslear TG, Hursh SR, Van Dongen HPA. Predicting cognitive impairment and accident risk. Human Sleep and Cognition, Pt II: Clinical and Applied Research. 2011;190:155-67.
[42] Kanas N, Manzey D. Basic issues of human adaptation to space flight. In: Wertz JR, editor. Space psychology and psychiatry. 2nd ed. California, USA: Microcosm Press and Springer. 2008;15-40.
[43] Bechtel RB, Berning A. The third-quarter phenomenon: Do people experience discomfort after stress has passed? In: Harrison AA, Clearwater YA, McKay CP, editors. From antarctica to outer space. New York, USA: Springer, New York, NY. 1991.
[44] Dijk D. Subjective sleepiness: An undervalued early sign of insufficient sleep? J Sleep Res. 2014;23(3):241.
4
99 99
[45] Chan S, Debono M. Replication of cortisol circadian rhythm: New advances in hydrocortisone replacement therapy. Ther Adv Endocrinol Metab. 2010;1(3):129-38.
[46] Cheung IN, Zee PC, Shalman D, Malkani RG, Kang J, Reid KJ. Morning and evening blue-enriched light exposure alters metabolic function in normal weight adults. Plos One. 2016;11(5):e0155601.
[47] Most EIS, Scheltens P, Van Someren EJW. Prevention of depression and sleep disturbances in elderly with memory-problems by activation of the biological clock with light - a randomized clinical trial. Trials. 2010;11:19.
[48] Aardal E, Holm A. Cortisol in saliva - reference ranges and relation to cortisol in serum. Eur J Clin Chem Clin Biochem. 1995;33(12):927-32.
[49] Merkus SL, Holte KA, Huysmans MA, Hansen AM, van de Ven PM, van Mechelen W, van der Beek AJ. Neuroendocrine recovery after 2-week 12-h day and night shifts: An 11-day follow-up. Int Arch Occup Environ Health. 2015;88(2):247-57.
[50] Fossum IN, Bjorvatn B, Waage S, Pallesen S. Effects of shift and night work in the offshore petroleum industry: A systematic review. Ind Health. 2013;51(5):530-44.
[51] Dawson D, Chapman J, Thomas MJW. Fatigue-proofing: A new approach to reducing fatigue-related risk using the principles of error management. Sleep Med Rev. 2012;16(2):167-75.
APP E NDIX Table A1 . Gen eralized linear (G LMM ) an d linear mixed mo del (LMM) re
sults for pre
- an d post -shift subjecti v e fati gue (KSS) s core s, react ion times, averag e numbers of psychomotor v igi lan ce te sts ( PVT -B ) error s, fal se
starts and lapses o
v er th e cou rse o f tw o -we ek off shore d a y -shi fts adju sted an d u n ad ju
sted for time
i n b e d (TIB). Unad ju sted scor es Ad ju sted scor es for TIB N Estimate 95% CI p -v a lu e N Estimate 95% CI p -v a lu e Karolinska Sle e pine ss S core (K SS) Pre-shift 537 .01 (-.02 – .04) .476 467 .01 (-.01 – .04) .306 Post-shift 540 .06 (.03 – .09 ) <.001 463 .05 (.02 – .08 ) .004 Reaction tim e (R T)* Pre-shift 374 1.00 (.99 – 1.0 0) .142 330 1.00 (.99 – 1.0 0) .045 Post-shift 382 1.00 (1.00 – 1. 00) .635 334 1.00 (.99 – 1.0 0) .329 Errors (E )* Pre-shift 375 1.07 (.97 – 1.1 8) .199 331 1.08 (.97 – 1.2 0) .141 Post-shift 380 1.03 (.95 – 1.1 1) .494 332 1.01 (.93 – 1.1 0) .766 False Starts (F S)* Pre-shift 375 1.04 (.95 – 1.1 5) .357 331 1.08 (.97 – 1.1 9) .154 Post-shift 380 1.03 (.95 – 1.1 1) .523 332 1.02 (.93 – 1.1 1) .687 Lapses (L)* Pre-shift 373 1.04 (.96 – 1.1 3) .342 329 1.07 (.98 – 1.1 7) .141 Post-shift 377 1.02 (.93 – 1.1 1) .692 329 1.05 (.96 – 1.1 6) .261 In th e columns, 95% con fidence interv al s an d p -v a lu es of si
gnificance are dis
p lay e d. A ll models wer e ad ju sted for plat form loca tion an d time in bed (T IB). Subjecti v e fatigu e was measur
ed with the Karoli
nska Sl eepin e ss Scal e (KS S) hi gher scor es in dicate h igh er su bjectiv e fatigue. *Log-nor mally d is trib u ted val u es w e re b ack -tran sfor med.
4
101 101
