Journal of the New Zealand Medical Association, 16-December-2005, Vol 118 No 1227
Work-related fatal traffic crashes in New Zealand:
Bronwen McNoe, John Langley, Anne-Marie Feyer
The annual rate of acute unintentional fatal injury in New Zealand is 26.0 per 100,000 person per year.1 About half of these fatalities occur as the result of motor vehicle traffic injury.1 Work-related injury also contributes substantially to the overall injury burden. In New Zealand, two studies have attempted to describe the extent of work-related fatalities, neither have included traffic fatalities.2,3 The more recent, the Work-Related Fatal Injury Study, determined the rate of fatalities due to work-related activity was 5.03 per 100,000 workers.3
We have identified two population-based studies that report on work-related traffic fatalities. The first was in Australia and identified that 37% of work-related fatal injury was due to motor vehicle traffic injury.4 The second study was in the United States and found that roadway crashes accounted for more than 23% of workplace fatalities.5 When bystanders to work (a person who is killed as the direct result of someone else’s work activity) are included, these figures are likely to increase further. Current priorities and strategies for injury prevention in New Zealand are established without a good understanding of work-related traffic injury.
Currently agencies responsible for routine injury data collection in New Zealand cannot provide accurate estimates of the number of work-related traffic crashes or fatalities in New Zealand. New Zealand Police do not identify in their Traffic Crash Report whether crashes were work-related or not. Crashes involving commercial vehicles are identified, but some of these commercial vehicles may have been used for non-work activity at the time of the crash.
The notifications database maintained by Occupational Safety and Health (OSH) and the national compensated claims database maintained by the Accident Compensation Corporation (ACC) do not provide an adequate base from which to estimate the extent of the problem because of reporting and coverage issues.6
If a similar situation exists in New Zealand to the United States and Australia, where the largest single category of work-related fatalities is traffic injury, then the absence of accurate information on the number and circumstances is a major barrier to developing priorities for injury prevention.
The aim of this study was to identify and describe all work-related traffic vehicle fatalities that occurred on a public road in New Zealand between 1985 and 1998 inclusive. This paper describes the methods used in the study and gives a synopsis of the overall findings obtained. Other papers provide an analysis of bystander deaths as well as an international comparison of work-related traffic fatalities between New Zealand, Australia, and the United States.
The methodology adopted in this study closely followed that in a similar Australian study4 so as to allow meaningful comparisons. A work-related traffic case was defined as a person who suffered a traumatic death that occurred in New Zealand and involved a traffic vehicle, to which workplace exposures contributed as a necessary factor to the death, and which can be attributed to those exposures. A traffic vehicle was defined operationally as a conveyance in which, any person or property may be transported on a public road. In other words, this study was not confined to incidents involving motor vehicles but also includes transportation such as push bikes. Work-related activity encompassed most aspects of work, including working, commuting, activity during a recess period (e.g. lunch break), at an employer sponsored social function, and during training or a non-work period if the incident arose because of work.
Three sources of data were used to identify potential deaths due to a “traffic crash” that occurred in New Zealand between 1985 and 1998. The New Zealand Health Information Service (NZHIS) maintains a record of all injury deaths within New Zealand. Those records classified with an external cause code (E-code) of traffic incident or similar (E810-E829, E846-E848, E919) were identified as potential cases.
The Land Transport Safety Authority (LTSA) maintains a database that contains the official police records of reported motor vehicle traffic crashes in New Zealand. All motor vehicle traffic fatalities in that database were initially identified as potential cases. The Accident Compensation Corporation administers New Zealand’s Accident Compensation Scheme, which provides no-fault accident insurance for all New Zealand citizens, residents, and temporary visitors to New Zealand. All work-related fatal traffic crashes resulting in Accident Compensation Corporation entitlement claims were identified from the Accident Compensation Corporation motor vehicle account as potential cases. The death must have occurred between 1 January 1985 and 31 December 1998. Exclusions from the study included those aged over 85, suicide deaths, domestic violence, persons performing home duties, criminal activity as work, and delay of greater than 1 year between the incident and death.
The absence of a work-related identifier in any of the three data sources referred to above necessitated the use of an alternate source of information to identify work-related fatalities. For this study, the primary source of data to determine work-relatedness for this study was from coronial inquest files. Certain deaths, including unnatural deaths are notifiable to the police and the coroner. The role of the coroner in these cases is to establish the cause and manner of death by way of a coronial inquest. Although work-relatedness of an incident is not required, in many instances information recorded within the coronial file can be used to make a determination of work-relatedness.
From the three source databases, 12,519 potential cases were identified. These were electronically matched by the victims’ name to the coronial register, which contains a list of decedents name, date of birth, and coronial inquest number. The commercially available data linkage software Automatch7 was used for the matching process. Cases that remained unlinked were hand matched, where possible, with a coronial number. The matching process produced 10,993 coronial numbers (88%). It is likely that some duplication existed within the remaining 1,526 files given that three data sources were used. Of the 10,993 names, coronial files were identified for 10,809 (98%) of potential cases.
Each identified file was requested in a random sequence from Coronial Services at the Department of Justice where coronial files are stored. Demographic details (name, age, sex, date of death) from the source databases were checked against the coronial file to ensure a correct match had been made.
Due to variation and inaccuracies in the format and spelling of names, both in the databases and on the coronial register, the matching process was, at times, imprecise. Up to three attempts were made to match the correct coronial file. Of the potential cases identified, 98% were correctly matched to a coronial file number. A further 15 cases that had not been identified by the source databases were identified from within coronial files reviewed. Each potential case identified was reviewed for work-relatedness. Deaths were classified as definitely work-related, possibly work-related, definitely not work-related, and indeterminate. The files for all cases identified as definitely or possibly work-related were photocopied and transferred to the study headquarters for further assessment.
Victims were classified as one of the following: worker, commuter, and bystander to work.
Definition of a worker:
“Persons who work for pay, profit or payment in kind, in a job, business or on a farm, and persons who worked without pay in a family business or on a farm”. This includes employees, employers and self-employed persons, working full time, part-time and ad-hoc hours.
“Persons who worked in an official volunteer capacity for an organisation”.
“Students, who comprised any person who was studying and whose death was the result of an incident, that occurred during school time, while they were performing a task directly connected with their course”.
Definition of a commuter:
“People who satisfied the workers definition, but died as a result of an incident that occurred while travelling directly from home to work, work to home, or between two jobs. If the incident occurred while the person was travelling in the course of their work duties, the person was not classified as a commuter but as a worker”.
Definition of a bystander:
“All persons who were killed directly as a result of someone else’s work activity, even though the deceased was not working at the time”.
Cases were coded in random order by three coders over a 5-month period. Demographic details of the deceased, circumstances surrounding the injury, and work details (such as industry in which fatality occurred) were some of the details recorded on worker or commuter fatalities. Because of the large number of bystanders and the lack of detail in the coronial files on the working component of the crash, bystanders were simply counted, the type of vehicle involved in the crash recorded, and (when available) the active contribution of working activity to the crash noted.Variables pertinent to the analysis presented here include: gender, age, occupation (coded according to the New Zealand Standard Classification of Occupations 19968) and industry (coded according to the Australian and New Zealand Standard Industrial Classification 19969).
Inter- and intra-rater reliability was assessed for both case determination and the coding of coronial files using a sample of cases. For case review, there was 97% agreement, both between coders and by coders over time. High levels of agreement also existed for the coding of coronial files, for most variables being higher than 90%. The two exceptions to this were for industry and occupational coding. The lower level of agreement between coders for industry and occupation (80–86%) was of concern. To ensure a higher level of reliability, every work-related coronial file was thus re-reviewed for occupation and industry, and corrections made where necessary.
Data were analysed using Statistical Procedures for the Social Sciences (SPSSx) software. Population rates per 100,000 person years and 95% confidence intervals were calculated (assuming a Poisson distribution). For the purposes of deriving rates, census data were used to approximate the number of persons of working age who were exposed to the road traffic environment, while at work, or when going to or from work. It was assumed that everyone in the working population could be exposed.
Estimates of the exposed population were available by gender, age group, ethnicity, occupation, and industry. Census data were available for 1986, 1991, 1996, and 2001, with the population in the intervening years being determined by interpolation using a linear function for the inter-census years to span the time frame 1985 to 1998 inclusive. The number of workers in each level were summed over time to give an estimate of person-years at risk.
The total number of coronial files reviewed was 10,809 for which 4034 files (37%) contained insufficient detail to make a work-related determination.
This study identified 234 incidents resulting in 241 deaths where the deceased was engaged in work-related activity on a public road, and a further 183 incidents resulting in 192 deaths where the deceased was commuting to or from work on a public road. In addition, although not engaged in work themselves, 1447 people died in the process of another person’s work activity on a public road (bystanders). This paper describes worker and commuter fatalities only.
The overall rate of work-related traffic fatalities of workers and commuters on a public road, between 1985 and 1998, was 2.01 per 100,000 workers per year. This was comprised of decedents working 1.11 and commuting 0.89 per 100,000 workers per year. No obvious trend in either category was apparent over the period studied.
The majority of the work-related traffic fatalities were male, for both decedents working (93%) and commuting (80%). The highest number of worker deaths occurred in the 25–34 year age group (n=64, 27%). The highest number of commuter deaths was in the 15–24 year age group (n=68, 35%). Nearly three-quarters of worker deaths (n=172, 71%) occurred between age 20 and 49. Nearly three-quarters of the commuter deaths occurred under age 40 (n=138, 72%). The mean age was substantially lower for commuter fatalities (33 years) than for worker fatalities (39 years).
The age-specific rates differed for worker and commuter fatalities (Figure 1). For workers, the rate of fatal injuries increased with increasing age group, with the highest rate observed in the 65–84 year age group. For commuters, the rate of fatalities declined with increasing age group, with the highest rate observed in the 15–24 year age group.
The only information relevant to the nature of the work, which was consistently available in the coronial files was that on the industry and occupational group of the decedents.
For decedents working, the highest number and rate of fatal injuries occurred in the transport and storage industry (10.1 per 100,000 workers per year) which accounted for about 40% of the fatalities (Table 1). The sub grouping road freight transportation had a relatively high number of workers killed (n=83), with a correspondingly high rate (33.7).
Although the number of fatalities was small, high rates were observed in milk vending (19.4), road and bridge construction (8.8), postal and courier services (6.5), plumbing services (6.3), and taxi and other road passenger transport (excluding bus transportation) (6.2).
Commuting fatalities were more evenly spread across a number of occupations. The highest numbers of fatalities occurred in manufacturing (n=36) as well as agriculture, farming, and fishing (n=33) (Table 1). The highest rates were observed in livestock farming (7.6), log saw milling (7.6), automotive repair and service (4.5), services to agriculture (4.2), plumbing services (4.2), horticulture and fruit growing (2.6), and meat processing (2.4).
Figure 1. Age-specific rates (95% confidence interval) for fatal work-related traffic injuries: 1985–1998
The pattern observed in occupational groupings was similar to that of industry (Table 2). For the worker group, the highest number (with rates approximately 17 times the population overall) occurred in the plant and machine operators and assemblers, most notably to drivers and mobile machinery operators (111 deaths) with a rate of 21.8 per 100,000 workers (Table 1). Within this group high rates were observed in heavy truck or tanker drivers (28.6), light truck and van drivers (15.1), and taxi drivers (12.4). Agricultural/earthmoving and equipment operators also had high rates (9.1) particularly ‘earthmoving machinery operators’ (14.1) and roading/paving machine operators (38.6). Other occupations that had significantly higher rates of fatal injuries than the total working population were couriers and deliverers (10.7), and commercial travellers and sales representatives (4.9).
For commuters, the highest numbers of fatalities occurred in the plant and machine operators and assemblers and the agricultural forestry and fishing occupations. In the plant and machine operators, the deaths were predominantly to stationary machine operators and assemblers (20 deaths) and in the agricultural industry to market orientated agricultural and fishery workers (34 deaths) and market orientated animal producers (17 deaths).
The relative importance of work-related traffic crashes to other work-related injury is important in terms of prioritising areas for injury prevention interventions. Previous comparable work estimated was that approximately 75 deaths (excluding traffic crashes) occurred due to work-related activity each year.3,10 In the present series we estimated an average of 31 (worker and commuter deaths) per year. In total then we estimate there were an average of 106 work-related deaths per year. Therefore, an estimate of the contribution of work-related traffic crash (worker and commuter deaths) to the annual overall burden of work-related fatal injury was 29%. This estimate ranged between 24% and 40% for particular years in the time period studied.
At an average of 31 deaths per year, work-related traffic crashes represent the single largest category of work-related death in New Zealand. Even if commuters are excluded from the estimate, the average is approximately 17 deaths per year and this still represents a major contributor to the annual burden of work-related death of approximately 100 per year. Because of the difficulties associated with case ascertainment, the estimates produced are likely to be a conservative, particularly for commuting incidents that are more difficult to identify from coronial files than worker fatalities. This in part is due to the lack of uniformity of recording systems for coronial findings throughout New Zealand.11Direct comparisons of these results with findings elsewhere need to be treated with caution because of differences between national collections including different case definitions and case ascertainment procedures. One of the major differences is the inclusion in this study of commuter fatalities. In broad terms, however, similar patterns are observed in this study to those found elsewhere in the developed world namely that traffic injury generally forms the largest category of work-related death5,12–15 and the most common industry in which traffic injuries occur is road transportation 5,14,16–19.
Work-related traffic fatalities were found to be predominantly male for decedents working and commuting. This large portion of male deaths may be due to a number of factors. Firstly, it may reflect exposure. Males are undertaking nearly twice the number of commuting journeys as females16. Males also typically dominate many of the occupations involving professional driving. Secondly, a bias may exist in the data, where the work component is more explicitly stated in the coronial file for males than for females. Finally, particularly in occupations such as forestry and fishery (where workers are predominantly male), it seems likely that commuting is undertaken at riskier times of the day or for longer periods. The highest number of commuting fatalities occurred in the agricultural forestry and fishing industry and in the manufacturing industry. This may reflect that these workers are commuting longer distances at riskier times of the day. In these industries, many workers start work early in the morning and will therefore travel to work at their circadian low points and may well be at risk for fatigue-related crashes.
Rates of work-related traffic injury in this study were relatively constant between ages 15 and 64 but rose markedly over the age of 65. However, the confidence intervals around the later estimates suggest it should be viewed with caution. In terms of raw numbers, approximately two-thirds of working decedents were under the age of 45.
The lower numbers of occupational fatalities observed amongst older workers in this study may reflect the age distribution of driving occupations. As professional drivers age, they may be likely to move from the road to static workplace-based positions that are less physically demanding, thus decreasing their exposure to road incidents. For example, a cohort study conducted on professional drivers in the Scandinavian countries demonstrated that over a 10-year period, about one-quarter of the cohort had changed tasks within their trade (for example, from bus driver to supervisor). A further quarter had left their trade altogether.17
Age may have a detrimental effect on risk of occupational injury in that some particular tasks, such as some physically demanding activities and continuous rapid information processing, decline in older age groups. Conversely, age may afford the opportunity for greater experience on the road, which will benefit tasks that require experience and refined skills.5,18
The number and rate of commuting incidents was highest in the 15–24 year age group and declined with increasing age. Nearly two-thirds (64.1%) of decedents commuting were under the age of 44. This may reflect driver behaviour in those age groups. Young people always feature prominently in crash statistics whether they are commuting or not. It may be that young people travel longer distances because they enjoy driving, because they need to travel further to obtain work, or because of where they live, all of which would increase their exposure.
There is likely to be an overestimate of risk based on populations. Data on the number of driver hours exposed are not available but it is likely that a risk estimate based on these denominator would be different to that based on a population. For example, truck drivers will spend more hours on the road than plumbers. In New Zealand, currently there is no routine surveillance of work-related traffic injury nor even any means of systematically and simply identifying events. New Zealand’s changing employment climate—including the increasing casualisation of the labour force, variable working hours, and multiple job handling—may effect the rate of work-related injury on the road, but current data collection processes will be unable to detect this. Therefore, it is important that relevant agencies have a good understanding of this problem so that they can accurately and effectively identify and prioritise areas for work-related injury prevention.
This study has enabled an estimate of the burden of work-related traffic injury to New Zealand but because of delays in the availability of coronial files the information provided is not particularly timely and was very resource intensive to produce given the number of coronial files that needed to be manually reviewed. An added problem was that the information contained in coronial files often does not provide much (if any) information on the work-relatedness of the fatal incident.
Since all injury deaths are investigated by the coroner, the coronial system provides an obvious means of determining whether a death was work-related. A Bill (Coroners Bill) is currently before the New Zealand Parliament that proposes a major reform to the coronial system.
This provides an excellent opportunity to put in place uniform systems for identifying important classes of death such as work-related deaths and capturing key features of these deaths in a uniform manner.
Author information: Bronwen McNoe, Assistant Research Fellow; John D Langley, Professor and Director, Injury Prevention Research Unit, Department of Preventive & Social Medicine, Dunedin School of Medicine, University of Otago, Dunedin;
Anne-Marie Feyer, Director, Health Risk, Management Practice, Price Waterhouse Pty Ltd, Sydney, Australia
Acknowledgements: This research was supported by the Health Research Council of New Zealand. The assistance provided by the Department of Courts, in particular Craig Leahy (Coroners Clerk), was also invaluable. In addition, we gratefully acknowledge our colleagues on the study team including Shaun Stephenson (Biostatistician); Nicola Dow (Data Manager); Dorothy Begg (Research Adviser); Jane Bishop, Rebbecca Lilley and Patricia Virtue (Junior Research Fellows); and James Falconer (Research Assistant).
Correspondence: John Langley, Injury Prevention Research Unit, University of Otago, PO Box 913, Dunedin. Fax: (03) 479 8337; email: firstname.lastname@example.org
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads)
Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals