Hours worked per day (work duration); and issues about Shift work

1. All industries*

2. Agriculture

- sub sector:

Dairy & cattle farming*

Independent evidence support: Strong

Long hours of work may result in workers obtaining less than the necessary 7-8 hours of sleep and cause fatigue (Hartley et al., 1995). The relative risk of crash involvement for vehicle drivers who reported a driving time in excess of 8 hours was almost twice that for drivers who had driven fewer hours (Jones and Stein, 1987).

'Hours worked per week' was identified to be an important risk factor related to machine-related farm injuries. Dairy farms, farms with non-resident workers, and large farms were associated with an increased risk of injuries (Layde et al., 1995).

Haque and Bott (1997) stated that "...a driver working shifts may be 'out of synch' with his friends, family and the neighbourhood in general." This problem could be worse in summer due to social events like picnics and barbecues, etc.

Shift work contributes to fatigue in workers because it limits the amount of sleep that workers are able to obtain to ensure that they are able to maintain sufficient alertness when at work (Folkard and Monk, 1985). For example, human error and poor judgement, related to sleep loss and shift work during the early morning hours, were cited as contributing to the Space Shuttle Challenger incident (Folkard and Sutton, 2000).

An Indian study examined the risks of heat induced workplace incidents (4125 cases in the textile industry) and the heat tolerability of the rotating day and permanent night shift workers in hot-dry and hot-humid environment (Nag and Nag, 2001). Incident prevalence was significantly high in the summer months when the ambient temperature was high (hot-dry). The influence of hot climate in incident causation was evident from the shift-wise variations in the occurrence of incidents. The longitudinal study showed that the night workers were more vulnerable and less tolerant to heat, as compared to the rotating day workers. (Also see 'Temperature').

Level of staffing

1. All industries*

2. Agriculture, forestry & fishing*

3. Agriculture -sub sector:

Horticulture & fruit*

Independent evidence support: Moderate

Staffing levels are related to worker stress and workload, which in turn affect the likelihood of human error (e.g., Reason, 1990; Parkes and Sparkes, 1998).

Nurse staffing has been related to patient safety. Research reveals a close link between inappropriate nurse staffing levels and higher rates of unwanted outcomes for patients (CHSRF, 2006).

Level of casual labour

1. All industries

Agriculture, Forestry & Fishing*

2. Agriculture -sub sector:

Horticulture & fruit growing*

Independent evidence support: Moderate

Zierold et al. (2004) reported a survey result concerning summer work and injury among middle school students. Of the 3189 'working students' who responded to the survey, the majority were employed in informal job settings such as working for someone in a home, newspaper delivery, and working on family farms or family businesses. Overall, 18% of children reported being injured at work. Variables that were associated with injury included having a 'near miss' incident at work, having a co-worker injured, and being asked to do something dangerous.

The health and safety issues of casual, temporary and migrant labour, especially in agriculture, have been recognised by HSE^[23]

Visitors/non-workers attending work sites

1. All industries

Agriculture, Forestry & Fishing*

2. Agriculture -sub sector:

Horticulture & fruit growing*

3. Manufacturing*

Independent evidence support: Weak

Pickett et al. (1995) identified the four most common reasons for children's involvement in fatal incidents with farm tractors: inadequate supervision, permitting children to be in the area of moving or unguarded machinery, allowing children to accompany workers using farm machinery, and having children performing work related tasks inappropriate for their age. (Also see 'Industrial/occupational related evidence').

Variation in work activities

1. Agriculture, Forestry & Fishing*

2. Construction*

Independent evidence support: Weak

N/A

Work involving vehicles

1. Agriculture -Sub sector:

Horticulture & fruit growing*

2. Construction*

Independent evidence support: Moderate

In the agriculture industry in the UK, of the 489 people killed over the past 10 years (1995-2005), transport related incidents (being run over or vehicle overturns) accounted for 24% of fatalities.

Work under poor weather conditions

All industries*

See 'Environmental factors'.

Lack of recovery from fatigue

1. All industries*

2. Agriculture, Forestry & Fishing**

3. Construction*

Independent evidence support: Strong

Fatigue is directly related to human error and incident occurrence (e.g. Haworth et al., 1989; Reason, 1990).

Philip et al. (1996) surveyed 567 vehicle drivers travelling on summer vacations on a European highway for their subjective daytime sleepiness while driving and any sleep deprivation just prior to departure. 50% of the responders had a sleep restriction just prior to departure (mean -203 minutes) compared to usual total sleep time during the year; 10% had no nocturnal sleep prior to departure. Drivers younger than 30 years were significantly more acutely sleep deprived than other drivers. Economic migrants (those with low economic status) also experienced significant acute sleep restriction.

Fatigue can also be viewed as having a multifactorial origin, influenced by non-work related circumstances and personal characteristics, with a prolonged character that may affect an individual's performance and ability to function at work (Lewis and Wessely, 1992).

Fatigue can reduce the ability of the worker to process important visual and perceptive information relevant to avoiding an incident (Hsiao and Simeonov, 2001).

Fatigue and need for recovery were found to be independent risk factors for being injured in an occupational incident (Swaen et al., 2003).

Tight timescales/ deadlines

Agriculture -Sub sector:

Horticulture & fruit growing

Independent evidence support: Strong

There is a large quantity of evidence to show that increased time pressure (e.g., due to late service) and mental workload lead to stress, decreased human performance, inattention and more human error (e.g., Michon, 1985; Hancock, 1989; Reason, 1990; Schlegel, 1993; Hancock et al. 1995).

Vehicle drivers are found to be stressed and irritated in traffic congestion and tend to exhibit aggressive or unsafe behaviour when driving in those situations (Turner et al., 1975; Stokols et al., 1978; Hennessy and Wiesenthal, 1997).

Use of hazardous substances

All industries*

Independent evidence support: Weak

A British study confirmed that chemical incidents were most frequent in the summer, between 12:00 and 17:59, and least frequent between 0:00 and 5:59 (Olowokure et al., 2004). (Also see 'Time of day').

Time of day (and time into shift)

All industries***

Agriculture, Forestry & Fishing***

(Figures 23 & 24)

Independent evidence support: Strong

Research has shown truck drivers are susceptible to both sudden fatigue, due to temporary irregularities of the sleep cycle, and accumulated fatigue due to long working hours (US Congress, 1988). In fact, most road incidents occur between 4am and 7:30 am, which is well within the time span when drivers are most likely to fall asleep. There is also a crash risk during the mid-afternoon 'siesta hours' (Mitler et al., 1988).

Monthly variation of road traffic incidents in Riyadh, Saudi Arabia, between 1989 and 1993 was reported by Nofal and Saeed (1997) with reference to time of day, lighting conditions and weather conditions. Total incidents were found to be positively correlated with increased temperatures and inversely correlated with increased relative humidity and rain. Maximal incident rate was evident during the summer season particularly between 12 noon and 15:00. This period is characterized by heavy traffic and intense sunlight in their local environment.

Folkard (1997) analysed several previous studies of hours of driving and crash risk and found that there was a rise in likelihood of a crash at 2 hours into the trip before risk dropped back to starting levels at 4 hours into the trip. The likelihood of a crash then started to rise again the more hours driven until at 11 hours the risk was higher than at any previous time. This result was in agreement with another study of lorry drivers that showed that the risk of an incident was worst in the first 4 hours of a shift unless the driver worked for longer than 12 hours (Hamelin, 1987).

In order to understand time-related incident risk with shift work in Poland, Oginski et al. (2000) investigated 668 incidents in the metallurgical industry in terms of time of day, time on task, consecutive day of the shift block, day of the week, and season. The incident rate was similar on all shifts but more severe incidents happened in the night time. Somewhat more incidents occurred in the second half of the shift, in the second part of a shift block, and in summer compared with winter. There were fewer injuries at weekends.

A Swedish study demonstrated that the peak traffic incident risk was at 4:00am for the summertime, shortly after the early summer sunrise and with consistently higher night time risk than for winter driving. It was concluded that early morning driving is several times more dangerous than driving during the forenoon. Apart from alcohol the effect seems related to sleepiness, but not to darkness (Akerstedt et al., 2001).

The work-related fatal traffic injuries in New Zealand have been found to peak at around 10am, and remained relatively high from 10am till 15:00 (New Zealand Environmental and Occupational Health Research Centre, 2003).

A UK study on SPAD incidents with normalised data showed peaks in early morning (7:00-8:00), after lunch (14:30-15:00), and late evening (around 22:30) (Li and Lock, 2003).

Olowokure et al. (2004) reported that chemical incidents in the UK were most frequent in the summer, between 12:00 and 17:59, and least frequent between 0:00 and 5:59. (Also see 'Day of the week').

In a study related to road incidents in the UK, 'at blame' company car drivers were found to have two peak times for incident involvement. The first was between 8:00 and 9:00 in the morning; the second was between 8:00 and 9:00 in the evening (Clarke et al., 2005).

(Also see 'Circadian rhythms - reduced wakefulness/alertness')

Day of the week

All industries***

(Figure 26)

Independent evidence support: Strong

Williams' analysis of causes of SPADs (1977) suggested that day of the week was a relatively unimportant factor.

var der Flier and Schoonman (1988) determined that the distribution of incidents throughout the days of the week corresponds with the number of driver shifts. Their report supported the suggestion that the important factor is not the day of the week but the time when an incident occurs in relation to the individual's shift pattern.

Adcock & Sparkes (1993) determined that SPAD rates are lower at weekends than on weekdays. The day with the least amount of SPADs is Sunday, it was thought that this was because (at that time) Sunday work was voluntary and therefore the drivers working on that day were self-selected and perhaps keener. This was more likely due to reduced service levels at weekends.

A Polish study reported that in the metallurgical industry, there were fewer incidents at weekends as compared to weekdays (Oginski et al. 2000) (Also see 'Time of day').

On New Zealand's roads, most working crashes (94%) occurred between Monday and Saturday, and the incident rates on each of those days were similar. Most commuting crashes (85%) happened between Monday and Friday. A minority of working decedents (6%) or commuters (6%) were fatally injured outside the working days (New Zealand Environmental and Occupational Health Research Centre, 2003).

In a UK study, Olowokure et al. (2004) investigated whether there were temporal or seasonal patterns in the occurrence of chemical incidents recorded from January 1997 to December 2001. The data analysis showed more incidents occurred in the summer and they were more likely to occur on Thursdays and least likely on Saturdays. Incidents were most frequent between 12:00 and 17:59 and least frequent between 0:00 and 5:59.

Industrial/ occupational related evidence

See the results of detailed analysis in section 3.2

Independent evidence support: Strong

In a study to assess rates and patterns of agricultural machinery injuries in farm children over a 5-year period ending 31 March 1990, Pickett et al. (1995) identified a prominent summer peak in the occurrence of fatal injuries. The farm tractor was the machine most commonly associated with these injuries. Common reasons associated with the injury risk included: inadequate supervision, permitting children to be in the area of moving or unguarded machinery, allowing children to accompany workers using farm machinery, and having children performing work related tasks inappropriate for their age.

Rodriguez et al. (1996) identified that incidents with tractors are a major preventable morbimortality factor in rural areas for children safety.

An early study found that the industries with the highest rates of work-related death were 'Mining', 'Agriculture, Forestry and Fishing', 'Construction', and 'Transport & storage' (New Zealand Environmental and Occupational Health Research Centre, 1999).

Falls from heights were found to be one of the main concerns with migrant workers in farming industry in Greece (Alexe et al., 2003). Young migrant workers also tend to suffer severe multiple injuries from machinery. The study suggested the enforcement of safety guidance and regulations for poorly trained workers including migrants concerning farming machinery, and discouragement of risky farming activities among elderly individuals. (See also 'Tourists').

Pegula (2004) reported that in the USA, the industries with overall fatality rate (measured as per 100,000 workers from 1995 to 2001) from high to low were: mining (26.0), agriculture, forestry and fishing (23.2), construction (13.9), transportation and public utilities (12.4), wholesale trade (4.7), manufacturing (3.4), retail trade (2.8), services (2.0), and finance, insurance and real estate (1.2).

A Canadian study looked into the magnitude of both fatal and non-fatal farm machinery injuries in Alberta children and adolescents (0-17 years) between 1990 and 1997 (Lim et al., 2004). A total of 302 farm machinery injuries were recorded in this period, and of these, 14 resulted in death. ATVs were the most common cause of injury (n=76, or 25.2%), followed by tractors (n=72, 23.8%), and power take-offs (n=15, 5.0%). The predominant injury mechanism was entanglement (n=69), followed by falls from machines (n=57), and being pinned/struck by a machine (n=49). There were significantly more injuries reported during the summer and autumn than during the winter and spring. Those injured in the autumn were significantly older than those injured in the spring; and injury rates were significantly higher during the school holidays.

In a study to determine occupational facial fractures in central Switzerland during 2000-02, Eggensperger et al. (2006) found that 69% of the injuries occurred in farm and forestry workers and in construction labourers during the summertime (33%). Workers in these occupations carried a 127-fold (for farm and forestry) and a 44-fold (construction) higher risk of incurring maxillofacial injuries than did service and office workers. The incidents were most frequently caused by being struck by an object or an animal. The study called for the introduction of personalized safety measures in these high-risk occupations.

Railway related incidents in Turkey account for 213 deaths per year per 100 million passengers during 1997 and 2003 (Ozdogan et al., 2006). This quoted study evaluated the epidemiological aspects of these casualties and found that train-pedestrian incidents caused the highest number of mortality and level crossing incidents caused the highest numbers of casualties. The majority of the fatalities and injuries occurred in males and most often in the 25-60 age group. Summer time was the season with the highest number of fatalities and injuries. (Also see 'Age' and 'Gender').

Rainfall

1. All industries***

2. Cultural, recreational & other services**

3. Manufacturing**

4. Transport & storage*

Independent evidence support: Moderate

Williams (1977) investigated the possibility of the influence of rain on SPADs. Particularly the possibility that it may influence distance perception and cause misjudgement. He found that there was an approximate 60% increase in the perceived distance of a signal in 'rain conditions'.

Schandersson (1993) conducted a quantitative analysis of external factors associated with road incidents. The analysis "indicates that although heavy rainfall increases incident risk, small amounts of rain might actually decrease the incident rate compared to dry conditions. These results are not satisfactorily explained by differences in speed. Most likely there are also other aspects of driver behaviour that are important - aspects related to driver vision and behaviour."

For railway related incidents, evidence suggests that rain (especially heavy rain) does not reduce adhesion as much because it tends to wash the rails and remove the slurry (Jenks, 1997).

Wet day

1. Agriculture, Forestry & Fishing***

2. Construction**

3. Cultural, recreational & other services**

4. Manufacturing***

Independent evidence support: Moderate

In two of the nine 'accident accounts' of a work-related road traffic incidents (Clarke et al., 2005), the feature of wet day was mentioned: "It was early in the morning on a wet day in autumn....", or "It was the middle of the morning on a wet day in winter...".

Temperature

1. All industries**

2. Agriculture, Forestry & Fishing***

3. Cultural, recreational & other services***

4. Manufacturing***

Independent evidence support: Moderate

Joki (1982) reported decreases in human performance as the temperature rises. This was based on observations of both physical work and office work.

A study by Ramsey et al. (1983) of unsafe behaviours in industrial settings showed a rise in safety-related errors began at about 74°F (23°C) and increased with temperature, particularly at high workload levels.

Ramsey (1995) concluded from a review of past studies that, as long as a hot climate is not 'uncomfortable', neither light physical work nor mental task performance was affected by a warm environment. However, as the temperature rises above the comfort level, problems may arise first of a subjective nature and then of physical problems which impair workers' efficiency.

High temperature during summer has been identified as a major factor contributing to increased road incidents in Saudi Arabia (Nofal and Saeed, 1997). The authors suggested that the high temperature especially between 12 noon and 3pm might lead to increased stress and decreased performance of intellectual tasks which require considerable physical effort and motor skills. (Also see 'Time of day').

Ergonomics guidelines recommend a comfortable air temperature in summer should be between 20 and 24°C (20-21°C in winter) (Kroemer and Grandjean, 1997).

ISO 7730 (1994) recommends a summer temperature range of 72-78°F (22.2-25.6°C) and ASHRAE Standard for Thermal Comfort makes a similar recommendation.

Buxton et al (1999) reported that thermal environment was an important factor influencing driver performance. However, their measurements of the thermal environment in train cabs did not show that the environment was stressful, this may be partially due to the fact that their measurements were taken in the winter only.

Nag and Nag (2001) studied the risks of heat induced workplace incidents in Indian textile industry, and the heat tolerability of the rotating day and permanent night shift workers in hot-dry and hot-humid environment. They found that incident prevalence was significantly higher in the summer months when the ambient temperature was high (hot-dry). The longitudinal study showed that the night workers were more vulnerable and less tolerant to heat, as compared to the rotating day workers. (Also see 'Hours worked per day; and issues about shift work').

The relationship between hot weather conditions and work-related incidents in central Italy was investigated by Morabito et al. (2006) over a 6-year period (1998-03). The findings indicate that hot weather conditions might represent a risk factor for work-related incidents during summer. The peak of work-related incidents occurred on days characterized by high, but not extreme, thermal conditions. Workers were reported to change their behaviour when heat stress increases, reducing risks by adopting preventive measures. The study suggested that days with an average daytime AT (Apparent Temperature) value ranging between 24.8C° and 27.5C° were at the highest risk of work-related incidents. Such information may be useful for the development of a watch/warning system that might be used by employers for planning work activities.

Sunshine hours

1. All industries*

2. Agriculture, Forestry & Fishing***

3. Cultural, recreational & other services***

4. Manufacturing***

5. Transport & storage**

Independent evidence support: Moderate

Dray et al. (1999) found that sunlight is the primary contributing factor identified as leading to 'misread' SPADs.

Prolonged visual exposure to bright sunlight reduces the sensitivity of the eyes to visual targets with different luminance levels and delays dark adaptation (Hood and Finkelstein, 1986).

Ground frost

1. Agriculture, Forestry & Fishing***

2. Cultural, recreational & other services***

3. Transport & storage***

Independent evidence support: Weak

N/A

Strong wind

Manufacturing*

Independent evidence support: Weak

N/A

Summer months/ season

See the results of seasonal trend analysis in section 3.2

Independent evidence support: Strong

Smith et al. (1986) reported that, in Georgia, motorcycle-associated fatalities occurred most frequently during summer months, on weekends, and during afternoon and evening hours. A similar trend has also been reported by others demonstrating that more traffic incidents occurred in the summer, and more frequently at 12-18 hours driving (Beyaztas and Alagozlu, 2002).

Abel and Welte (1987) reported an analysis of monthly, seasonal and day-of-the-week trends for four types of violent deaths (suicide, homicide, traffic incident and miscellaneous accident) which occurred in Erie County, New York, between 1973 and 1983. Of a total of 3787 violent deaths, there was no seasonal trend noted for any type of death except traffic deaths, which increased during the summer. There was a general trend for increased deaths on weekends for each type of death except suicide. The authors speculate that this culturally entrained temporal variable may be related to increased alcohol consumption during weekends.

Niino et al. (1995) reported a Japanese study to assess the prevalence and circumstances of falls by season among the elderly living in a rural community. Interview surveys were conducted among 1321 elderly individuals (aged 65+) every three months between 1992 and 1993. The results showed the prevalence of falls (i.e., the rate of those surveyed with falls) were 7.4% in summer, 5.9% in autumn, 6.5% in winter and 6.7 in spring. However, the seasonal variation was not statistically significant. Also in each season, there was no significant difference between genders. The seasonal differences in rate of falls were considered to be due to the climate change of the area.

Rodriguez et al. (1996) reported a review of a small sample (11) of serious childhood injuries related to tractor incidents in Chile. The majority of the incidents occurred in males and during the summer months. The authors suggested that incidents with tractors are major preventable morbimortality factor in rural areas. (Also see 'Industrial/occupational related evidence').

In a study to examine the patterns of seasonal variation in mortality in Moscow between 1993 and 1995, McKee et al. (1998) found a marked summer increase in accidental deaths among young people, especially from incidents and other deaths associated with alcohol consumption.

Wick et al. (2006) reported that in Australia, most accidental deaths in adults due to electrocution occurred in late spring and summer, accounting for 64% of the total cases investigated over a 30-year period from 1973 to 2002. The lowest number of accidental deaths occurred in winter and early spring. This report also showed a significantly higher rate of electrocution deaths among males, with a summer predominance of fatal incidents, most likely due to increased outdoor activities in better weather conditions.

In a PhD Thesis, Moore (2006) reported that all LCE (Loss of Control Events) and the LCE with serious injury outcomes in New Zealand farms have high summer and early spring peaks. The research also showed that over half of all LCE, not just the January ones, take place on hard and dry ground; and about a third take place in the muddy wet slippery conditions. The author stated that quadbikes work best in soft ground conditions where the contract area between tyre and ground is high. They loose traction more easily on hard ground especially if the surface has been made slick by a shower.

Tourists

1. Agriculture, Forestry & Fishing*

2. Construction*

3. Cultural, recreational & other services*

4. Transport & storage**

Independent evidence support: Moderate

Young migrant workers were found to be more likely to have incidents (e.g. falls from heights, injured by machinery) in the agriculture industry in Greece (Alexe et al., 2003). (Also see 'Industrial/occupational related evidence').

School holiday

1. Agriculture, Forestry & Fishing*

2. Manufacturing*

Independent evidence support: Moderate

In a Canadian study looking into the magnitude of both fatal and non-fatal farm machinery injuries in Alberta children and adolescents (0-17 years) from 1990 to 1997, along with other findings, Lim et al. (2004) found that machinery injury rates were significantly higher during the school holidays than during the study period. (Also see 'Industrial/occupational related evidence').

Public holiday

Agriculture, Forestry & Fishing*

Independent evidence support: Moderate

Farmer and Williams (2005) examined fatal vehicle crash deaths in the USA from 1986 to 2002 by time of day, day of week, month, and season, and to explore why some days of the year tend to experience a relatively high number of fatalities. Their analysis revealed that summer and fall months have more crash deaths than winter and spring, largely due to increased vehicle travel. July 4 (Independence Day) has more crash deaths on average than any other day of the year, with a relatively high number of deaths involving alcohol. January 1 (New Year's Day) has more pedestrian crash deaths on average, also due to alcohol impairment. On other days the high fatality rates are likely due to increases in holiday or recreational travel. (Also see 'Alcohol').

Daylight saving

Agriculture, Forestry & Fishing*

Independent evidence support: Moderate

Whittaker (1996) reported the effect of British Summer Time (BST) on road traffic incident casualties over a 10-year period (1983-93), by comparing incident data (4185 casualties in total) from before and after the onset of BST. The results indicated that the onset of BST in spring was associated with reductions in casualty numbers of 6% in the morning and 11% in the evening. The change back to Greenwich Mean Time (GMT) in autumn produced a reduction (6%) in casualties in the lighter mornings. The darker evenings were associated with significant increases in casualties (4%). There was an overall net reduction in casualty numbers when the analysed period of BST were compared to those during GMT.

In the summer of 1980 for the first time clocks in the Federal Republic of Germany were advanced 1h ahead of Central European Time (CET), which had been in use until then. A study was conducted to examine a sample of a total of 1070 accident patients who had incidents on data pairs taken from the moths of May 1979 and May 1980, before and after the introduction of the CEST, Central European Summer Time. A statistically significant increase in incident frequency between 7:30pm and 5:30am was found for the year in 1980 as compared to 1979. This observed increase in incidents was considered to be related to the change in routine for the adaptation to daylight saving time (Pfaff and Weber, 1982). That study also noted that the influence of CEST apparently exceeds a short adjustment phase, and suggested further studies to investigate a possible correlation between daylight saving time and an increased risk of incidents.

Whittington (1981) reported that most fatal motorcycle incidents occurred in the summer. (Also see 'Alcohol').

Household

Cultural, recreational & other services*

Independent evidence support: Weak

Those with low economic status (such as economic migrants) have been reported to experience significant acute sleep restriction during summer, which contributes to lack of recovery from fatigue (Philip et al., 1996). (Also see 'Lack of recovery from fatigue').

Consumer price index

Cultural, recreational & other services*

Independent evidence support: Weak

N/A

Effect of alcohol

1. All industries*

2. Agriculture, Forestry & Fishing***

Independent evidence support: Strong

Alcohol leaves the body at a rate of 7-10g/hour (1 unit). If an individual imbibes 6 pints of premium strength beer (12 units) during an evening (from 20:00), then ethanol will still be detectable in their blood at 11:00 (13 hours later). (Morgan and Ritson, 2003).

Whittington (1981) analysed 51 fatal motorcycle incidents and found that fatal injuries were frequently associated with alcohol consumption and excessive speed. In addition, all those victims were male, and most incidents occurred in the summer.

McLellan et al. (1990) carried out a 3-year study to determine the demographics, injury severity, and alcohol positivity of motor vehicle crash victims. The study found a marked seasonal variation in BAC (blood alcohol concentration) positivity, with 46.1% of drivers positive during the summer months.

Gilchrist (1990) reported that 20% of SPAD offenders were physically or psychologically unfit at the time of the incident.

In a study examining fatal vehicle crash deaths in the USA from 1986 to 2002, Farmer and Williams (2005) reported that July 4 (Independence Day) has more crash deaths on average than any other day of the year, with a relatively high number of deaths involving alcohol. January 1 (New Year's Day) has more pedestrian crash deaths on average, also due to alcohol impairment. (Also see 'Public holiday').

Alcohol consumption, its consequences as a seasonal phenomenon in Estonia, and the social and environmental factors affecting this seasonal change were studied by Silm and Ahas (2005) through a survey among 87 university students. As expected, the peak period of beer and wine consumption is in the summer. These were in line with the seasonal variability of traffic incidents and the frequency of medical treatment. The authors also noted that such a seasonal trend is influenced by, or interlinked with, some natural factors such as temperature and humidity, and some social factors such as celebrations and vacations.

Tobacco/ smoking

Agriculture, Forestry & Fishing*

Independent evidence support: Moderate

Smoking was significantly related to the risk of being injured in an occupational incident (Swaen et al., 2003).

Smoking is regarded as a confounding factor (with alcohol consumption) for raised blood pressure and violent death (Scottish Intercollegiate Guidelines Network, 2003).

Use of mobile phone while working

Burns et al. (2002) conducted a study comparing the danger of driving a car whilst using a mobile phone with that of driving under the influence of alcohol. They found that even the use of a hands-free phone is more dangerous than driving with more than the UK legal limit of alcohol (80mg/100ml) in the bloodstream.

Age

1. All industries*

2. Agriculture, Forestry & Fishing*

Independent evidence support: Strong

Zhang et al. (1998) examined the patterns of fatal motor vehicle traffic crashes (MVTC) by age group (16-24, 25-64, 65+) among Canadian drivers between 1984 and 1993. Compared to the middle-aged group as a reference population, young drivers demonstrated excess risk for risk-taking behaviours and conditions, alcohol and illicit drug use, speeding, non-use of seat belts, fatigue, falling asleep, and inexperience. They also had higher risks during the summer. Excess risk among elderly drivers was noted for their medical and physical conditions, inattention, and improper control actions. The study demonstrated a need for preventive efforts that incorporate age-specific strategies.

A previous study on work-related fatal injuries in New Zealand found that there was a dramatic increase in the rate of work-related injury death for those over the age of 65 years (New Zealand Environmental and Occupational Health Research Centre, 1999).

In the Netherlands, van Rossum et al. (2001) conducted a 25-year follow-up study of 19019 male civil servants aged 40-69 years to determine the seasonal effect on all-cause and cause-specific mortality and to identify high-risk groups. The results showed a seasonal all-cause mortality, with the ratio of highest mortality rate during winter versus lowest rate during summer of 1.22 (95% CI: 1.1-1.3), largely due to the seasonal nature of ischaemic heart disease, and with older age groups. (Also see 'Gender', and 'Industrial/occupational related evidence').

A study in the US confirmed that "Regardless of the type of workers, the fatality rate of workers aged 55 and older is greater than the fatality rate of workers aged 16 to 54" (Pegula, 2004).

Numerous issues on age-related changes in human capabilities especially in vision and some cognitive functions have been reviewed in Li (2004). As is commonly understood, ageing has been related to various types of functional changes, normally towards a capability reduction. This has many implications for work safety of the elderly workers.

Gender

All industries

Majority with male (88.1%) and minority with female (8.3%)

Independent evidence support: Strong

A previous study investigating work-related fatal injuries in New Zealand reported that the majority of deaths were of males, and their rate of death was more than 30 times that of females (New Zealand Environmental and Occupational Health Research Centre, 1999).

The involvement in work-related road traffic incidents was found to be heavily biased towards male drivers/workers, with a male to female ratio of 14.5:1 (Clarke et al., 2003).

Wick et al. (2006) reported a significantly higher rate of electrocution deaths among males (M:F=63:3), with a summer predominance of fatal incidents. (Also see 'Summer months/ season').

Ozdogan et al. (2006) reported that most railway related fatalities and injuries in Turkey were in males, and most of the victims were aged between 25-60 years. (Also see 'Industrial/occupational related evidence').

Employment status

All industries

EMP: 56.6%

SE: 23.2%

BS: 15.2%

Independent evidence support: Strong

The majority of work-related deaths in New Zealand during 1985-1994 were of employees, but the rate of death was also consistently higher for self-employed persons. (New Zealand Environmental and Occupational Health Research Centre, 1999).

From 1995 to 2001 in the USA, the wage and salary workers suffered more than 3 times as many fatal occupational injuries as did self-employed workers. However, there were 9 times as many workers in the wage and salary group than in the self-employed group. When fatality rates are compared, the self-employed workers were 2.7 times more likely to be victims of fatal work injuries than their wage/salary counterparts (Pegula, 2004).

In a study carried out in England and Wales during periods of four years around the 1981, 1991, and 2001 censuses, Edwards et al. (2006) found that the overall rates of death from injury and poisoning in children have fallen in England and Wales over the past 20 years, except for rates in children in families in which no adult is in paid employment.

Other personal or medical issues

Anderson et al. (2004) examined if there are diurnal, weekly or seasonal variations in the occurrence of stroke by a population-based study in Auckland. They found that strokes were less likely to occur during the summer and autumn than in the winter or spring. They also noted an increase in the stroke occurrence in the late morning.

A Japanese study found that stroke occurred most frequently in summer (26.0%), followed by autumn (25.8%), winter (25.3%), and spring (22.9%). No differences in age were observed among the four seasons, but stroke in men were more frequently observed in summer compared to other seasons. (Ogata et al., 2004).

Circadian rhythms -Reduced wakefulness (alertness)

Researchers have found that 'the neural processes controlling alertness and sleep produce an increased sleep tendency and diminished capacity to function during certain early morning hours (2-7am) and to a lesser degree, during a period in the mid-afternoon (2-5pm), whether or not we have slept' (Mitler et al., 1988).

It is well established that the body temperature rhythm is intimately coupled to cyclic changes in arousal such that there is a more pronounced tendency towards sleepiness at certain times of day. During these periods (2-7am, 2-5pm), we are most likely to 'nod off' unintentionally. It is during these times that drivers are most likely to be 'asleep at the wheel', having 'microsleeps, which may result in inattention, forgetfulness and other performance lapses (Mitler et al., 1988; Pheasant, 1991).

(Also see 'Time of day').

Worker return to duty after a holiday break

N/A

A variety of incidents, such as road incidents, have been linked with post-holiday fatigue and lack of sleep, along with recovery from increased alcohol intake. It has been suggested that people take 'power naps' (a short 20 minute rest) during the workday to aid recovery from holidays, as this can increase productivity, and reduce errors & incidents.

Within the nuclear industry, Lewis and Swaim (1986) found that shift workers have shown greater impairments at shift handover following a block of 7-8 consecutive days off, compared with a block of 4 consecutive rest days.

Wharf (1993) reported that safety risks were higher for the first shift back after time off, although it was unclear whether these were related to the length of time absent.

Gilchrist (1990) found that when driving certain units for the first time or after a long break, drivers were more likely to have a SPAD.

In a study of air traffic controllers, Becker and Milke (1998) suggested that the ability to handle simultaneous visual and auditory information, or to return to a task after a break, is critical to task performance, and this is the sort of cognitive function most affected by age.

Based on the rail incident data recorded from 1998 to 2002, Li (2003) and Li and Lock (2003) reported that train drivers tend to have 191% more SPADs after a period of annual leave than predicted by chance alone.

A more recent study by Gibson et al. (2006) found an increased risk of SPADs when the drivers having only a single day's break; a decrease in SPAD risk for breaks between 3 and 7 days; and an increase in risk following a break of at least 7 days or more.