Work-related mild-moderate traumatic brain injury and the construction industry

Transcription

1 Work 39 (2011) DOI /WOR IOS Press Work-related mild-moderate traumatic brain injury and the construction industry Margaret Liu a,wenliwei a, Jennifer Fergenbaum b, Paul Comper b and Angela Colantonio a,b, a Department of Occupational Science and Occupational Therapy, University of Toronto, Toronto, Canada b Toronto Rehabilitation Institute, Toronto, Canada Received 20 October 2009 Accepted 30 December 2009 Abstract. Background: Consequences of traumatic brain injury underscore the need to study high-risk groups. Few studies have investigated work-related traumatic brain injuries (WrTBIs) in the construction industry. Objective: To examine WrTBIs in Ontario for the construction industry compared to other industries. Methods: A retrospective study of individuals who sustained a WrTBI and had a clinical assessment as an outpatient at a hospital-based referral centre. Data were collected for a number of factors including demographic, injury and occupation and were analyzed according to the Person-Environment-Occupation (PEO) model. Participants: 435 individuals who sustained a WrTBI. Results: There were 19.1% in the construction industry, 80.9% in other industries. Compared to other industries, individuals in the construction industry were more likely to be male, to not have attained post-secondary education, and experience multiple traumas. WrTBIs in the construction industry were commonly due to elevated work. The construction occupations involved included skilled workers and general labourers, and compared to other industries, WrTBIs occurred most often for those employed for a short duration in the construction industry. Conclusions: Construction industry workers experience serious WrTBIs that are amenable to prevention. Use of the PEO model increased our understanding of WrTBIs in the construction industry. Keywords: Person-Environment-Occupation model, occupation, prevention, workplace safety 1. Introduction The construction industry may experience a disproportionate burden of occupational injuries due to the nature of their work environment [1]. For those employed in the construction industry, fatal and non-fatal injuries associated with crane use [2], suspended equipment [3], roofing [4], and nail gun use [5], have been identified as occupational hazards. Construction workers are also more likely to be hospitalized due to trau- Address for correspondence: Angela Colantonio, Department of Occupational Science and Occupational Therapy, Faculty of Medicine, University of Toronto, Rehabilitation Sciences Building, 500 University Ave., Toronto, ON, M5G 1V7, Canada. Tel.: ; angela.colantonio@utoronto.ca. matic brain injuries [6], and to experience fatal head injuries [7,8] and non-fatal injuries [9,10] compared to workers in some other industries. The construction industry ranked third, with regards to the rate of acute work-related traumatic brain injuries among males resulting in death, after other primary and agricultural industries [8], and second, with regards to the number of non-fatal work-related head injuries, after public administration and defense [10]. Falls have emerged as a common injury mechanism of work-related traumatic brain injury in the construction industry [1,6,11,12]. Only one study had detailed information on the factors involved in occupational injuries in the construction industry. This study showed that factors such as safety culture, project and risk management, and construction design and processes were leading factors, followed by /11/$ IOS Press and the authors. All rights reserved

2 284 M. Liu et al. / Traumatic brain injury and the construction industry the worker and the work team, equipment, the workplace, and materials [13]. Therefore, the objective of this study was to examine the characteristics of Ontario workers employed in the construction industry who sustained work-related traumatic brain injuries (WrTBIs) using the Person- Environment-Occupation (PEO) model of occupational performance [14], and compare these data to data for individuals employed in other industries. The PEO model was used in this study as a theoretical framework to describe the factors that may influence the occurrence of WrTBIs. In applying the PEO model, the person component represents demographic/clinical characteristics of the worker; the environment component refers to the work-environment including the physical, social and institutional environments; and the occupation component is paid work/productivity. These three interrelated components which are represented schematically by three interlocking circles interact with one another, with the outcome being occupational performance. Occupational performance, which is work performance, is a worker s ability to effectively perform their paid work in their work-environment; therefore, work performance is the dynamic relationship between the worker, the work-environment and their work. Increased overlap between the three PEO components results in satisfactory occupational performance. The usefulness of the PEO model for this study is that it does not focus solely on addressing the level of the person, but identifies modifiable risk factors for prevention by also addressing the environment and the occupation. To prevent further WrTBIs in Ontario s industry sectors, it is important to explore the worker, work-environment and work that contribute to the occurrence of injuries. Based on previous studies [6,8], we hypothesized that the PEO profile for the construction industry would reveal more serious causes and consequences of WrTBIs compared to other industries. 2. Methods 2.1. Study population and design This is an exploratory study, a case series, conducted on consecutive individuals who were referred to Neurology Services at the Toronto Rehabilitation Institute by the Workplace Safety and Insurance Board (WSIB). Since 1998, Neurology Services has had a contractual agreement with WSIB to provide expert diagnostic opinions, recommendations for treatment and continued case management for workers who have or are suspected to have sustained a head injury at work and who have not returned to work after a six week period. The majority of head injury referrals were mild [15]. Data on pre-injury demographic, injury-related, and occupation and work-related characteristics including detailed previous medical histories were collected for 526 individuals by retrospective chart review. Eligible individuals included those who met the following criteria: 1) were older than 15 years of age at the time of injury, 2) were injured at work, 3) had sustained at least a head injury, and 4) were discharged between the years of 1998 to After excluding individuals who did not sustain at least a head injury, there were 435 individuals eligible for the study (n = 128 in 1998/99, n = 141 in 2000, and n = 166 in 2001). The study was approved by the Toronto Rehabilitation Institute Research Ethics Board Data collection Data were abstracted from medical charts using an abstraction tool that was developed and pilot tested for this specific population. The percent agreement was 85% and the unweighted kappa statistic was 0.55 for most variables, indicating fair to good inter-rater reliability [16]. Medical records included information such as: client program status form, discharge/case conference summary report, neuropsychologicalreport, occupational therapy and physiotherapy reports, and WSIB employer s report of injury/disease. The length of the abstraction process on average was one hour. The information on person abstracted included date of birth, gender, marital status, highest level of education, first language, and immigrant status. A previous medical history was considered a person characteristic. Whether the injured worker had a previous medical history was abstracted in detail and included ear/nose or throat conditions, cardiovascular disease excluding stroke, diabetes or other endocrine disorders, cancer, sleep apnea, respiratory conditions, eye or sight conditions, arthritis, depression, anxiety and its related disorders, schizophrenia, substance abuse, other mental health disorders not otherwise specified, epilepsy, a previous brain injury, stroke, headaches or migraines, and other neurological disorders not otherwise specified. An additional other category was used to ascertain information on previous medical conditions that were not otherwise listed. Injury-related information was also classified as a person characteristic. The information abstracted included type of injury, defined

3 M. Liu et al. / Traumatic brain injury and the construction industry 285 as head exclusive, head injury with multiple traumas such as spine or neck, multiple traumas only (e.g. without a head injury), or other injuries, and whether an emergency response the day the injury occurred was required, as a proxy for injury severity. The environment information abstracted included the agent that made direct contact with the head, contributing factors, and the time of the injury. The mechanism of injury was also considered an environment characteristic and included information abstracted about the events of the incident. The occupation and work information abstracted included occupation, duration of employment, whether the injured was represented by a trade union, if their work schedule changed weekly, and employment status when working defined as fullor part-time. Whether the injured individual was working at the time of admission or attempted a return to work was considered an occupation characteristic. Information on industry was abstracted from the WSIB employer s report of injury/disease (Form 7) and was used to determine whether an individual was working in construction. The industry was determined from the rate group number according to standard guidelines or was inferred from the qualitative description of the occupation at the time of injury. The group of other industries that were not classified as construction included forest products, mining and related, other primary, manufacturing, transportation and storage, retail and wholesale trades, government and related services, and other services Statistical analysis The data abstracted from the medical charts were entered into a custom designed database. The proportions of individuals employed in the construction industry, and not employed in the construction industry (other industries), were examined. The proportion of individuals who had incomplete information for PEO variables was examined, with cell sizes less than five not shown and indicated with an asterisk. Variables were selected for analysis according to the PEO model. Key person variables were age, gender, marital status, highest level of education, first language, and immigration status, type of injury, medical emergency response, and previous medical history. The type of injury was classified into two categories of whether the injury was head exclusive or multiple traumas including a head injury. Emergency response was classified as medical if it involved hospital/emergency department admittance or a doctor s attention and non-medical if it included ambulance/paramedicattention,first aid, on-site attention by a medical professional, or fire and police departments. A previous medical history was classified into five categories including chronic physical conditions, psychiatric conditions, neurological disorders, other conditions, and no past medical history. Key environment variables were agent, contributing factors, and the time of the injury event, including the mechanism of injury. Key occupation variables were occupation type, employment duration, whether an individual was represented by a trade union, had weekly work schedule changes, and classification of employment, including whether an individual was working at the time of admission or attempted a return to work. Chi-square tests for independence without continuity correction were used to examine the differences in proportions for PEO variables between the construction industry and other industries. Differences were considered significant at p<0.01 to address multiple testing. All analyses were performed in SAS v. 9.1 (SAS Institute, NC). 3. Results Out of 435 individuals, 83 individuals (19.1%) were classified as working in the construction industry at the time of injury, whereas 352 individuals (80.9%) were classified as working in other industries. Individuals with incomplete information were excluded for specific variables, except for age. Two individuals with missing information for age were reassigned to the median value from among the total eligible population (n = 435). Table 1 shows the person characteristics. In the construction industry,there was a greater proportion of men employed (data not shown due to low cell counts, p< 0.001), and a greater proportion likely not to have completed secondary school compared to those in other industries. Individuals in the construction industry were also more likely to have experienced multiple traumas. The environment characteristics are shown in Table 2. The most common contributing factor for the construction industry was working in an elevated area, whereas the least likely factor contributing to injuries in the construction industry was slippery terrain/nonterrain, compared to other industries. Individuals in the construction industry were more likely to have sustained their injury due to a fall from an elevation and less likely from a fall from the same level, compared to those in other industries. The occupation characteristics are shown in Table 3. Injuries in the construction industry were more likely

4 286 M. Liu et al. / Traumatic brain injury and the construction industry Table 1 Person characteristics for construction workers vs. other industries Person Characteristics Construction, n = 83 Other, n = 352 p-value n 1 (%) n 1 (%) Age < 45 years 28 (33.7) 108 (30.7) years 55 (66.3) 244 (69.3) Marital status Never married 17 (21) 43 (12.4) Married 55 (67.9) 245 (70.6) Divorced/separated/widowed 9 (11.1) 59 (17) Education Did not complete secondary school 51 (63.8) 146 (42.7) Completed secondary school 16 (20) 87 (25.4) Post-secondary education 13 (16.2) 109 (31.9) Language English/French 54 (65.1) 245 (69.8) Other 29 (34.9) 106 (30.2) Immigrant No 39 (53.4) 159 (50.5) Yes 34 (46.6) 156 (49.5) Injury classification Head exclusive 26 (31.7) 175 (50) Multiple trauma 56 (68.3) 175 (50) Medical emergency response No 9 (10.8) 75 (21.3) Yes 74 (89.2) 277 (78.7) Previous medical history Psychiatric conditions No 68 (81.9) 318 (90.3) Yes 15 (18.1) 34 (9.7) Neurological disorders No 66 (79.5) 298 (84.7) Yes 17 (20.5) 54 (15.3) Chronic physical conditions No 73 (88) 270 (76.7) Yes 10 (12) 82 (23.3) Other conditions No 48 (57.8) 199 (56.5) Yes 35 (42.2) 153 (43.5) Denied past medical history No 55 (66.3) 228 (64.8) Yes 28 (33.7) 124 (35.2) 1 Some variables may have missing values. to have occurred in occupations such as trades, transport and equipment operators, or general labourers, and among those who were employed for a short duration. 4. Discussion This study provides information about the epidemiology of WrTBIs in the construction industry for those who have lasting symptoms from a brain injury. Injuries among individuals working in the construction industry were more detrimental than those sustained in other industries shown by a larger proportion of multiple traumas. This study provides additional data to support the importance of addressing falls as a mechanism of injury to prevent brain injuries, particularly those from elevation. The largely male population that also had a lower level of education suggest safety training and education to minimize WrTBIs in the construction industry. This is the only study to show that construction workers were more likely to have multiple traumas, that is, a head injury combined with another injury such as a spine or neck injury, compared to other industries. One previous study examined the type of injury by employment sector and showed that construction workers ranked third in the total number of injuries after workers in public administration and defense, which included some construction and highway work, and mining and quarrying [10]. In a study of construction work-

5 M. Liu et al. / Traumatic brain injury and the construction industry 287 Table 2 Environment characteristics for construction workers vs. other industries Environment Characteristics Construction, n = 83 Other, n = 352 p-value n 1 (%) n 1 (%) Agent Materials/working area 43 (55.8) 175 (52.1) Human contact/vehicles 8 (10.4) 42 (12.5) Tools/machinery/equipment 23 (6.8) Other unboxed materials 24 (31.2) 96 (28.6) Contributing factors Slippery terrain/non-terrain No 78 (94) 286 (81.2) Yes 5 (6) 66 (18.8) Placement of objects/materials in working area No 77 (92.8) 296 (84.1) Yes 6 (7.2) 56 (15.9) Working in an elevated work area No 45 (54.2) 325 (92.3) < Yes 38 (45.8) 27 (7.7) Action of other persons No 78 (94) 299 (84.9) Yes 5 (6) 53 (15.1) Mechanism of injury Struck by inanimate object No 53 (63.9) 226 (64.2) Yes 30 (36.1) 126 (35.8) Struck against an inanimate object No 74 (89.2) 305 (86.6) Yes 9 (10.8) 47 (13.4) Struck against building structures No 76 (91.6) 309 (87.8) Yes 7 (8.4) 43 (12.2) Fall from the same level No 77 (92.8) 243 (69) < Yes 6 (7.2) 109 (31) Fall from an elevated level No 34 (41) 300 (85.2) < Yes 49 (59) 52 (14.8) Motor vehicle crash No 78 (94) 307 (87.2) Yes 5 (6) 45 (12.8) Time of injury 12:00 am 3:59 am 10 (3.4) :00 am 7:59 am 27 (9.3) 8:00 am 11:59 am 30 (50) 101 (34.7) 12:00 pm 12:59 pm 22 (7.6) 1:00 pm 3:59 pm 19 (31.6) 63 (21.7) 4:00 pm 7:59 pm 42 (14.4) 8:00 pm 11:59 pm 26 (8.9) 1 Some variables may have missing values. ers only, multiple injuries more often occurred among individuals who experienced a fall from a height, followed by a fall from the same level, and a slip without a fall [1]. In both studies, it was not known if those who sustained an additional injury also had a head injury. Consistent with our results, was the elevated rate of hospitalizations shown for construction workers, compared to workers in transportation, manufacturing, and sales and service occupations in a recent study [6]. Only workers in primary industry occupations had rates above that of construction workers [6], as when those workers get injured, they tend to have more serious outcomes associated with WrTBIs [8]. Person factors significantly associated with WrTBIs in the construction industry include gender and education. A larger proportion of males are associated with lost-time claims compared to females [17], and if a larger proportion of males work in construction than females, then they would continually represent the majority of those injured in this sector. A significantly

6 288 M. Liu et al. / Traumatic brain injury and the construction industry Table 3 Occupation characteristics of construction workers vs. other industries Occupation Characteristics Construction, n = 83 Other, n = 352 p-value n 1 (%) n 1 (%) Occupation Trades/transport/equipment operators 44 (55) 119 (34.9) < General labourers 33 (41.2) 17 (5) Other 205 (60.1) Employment duration 0 4 weeks 11 (15.3) 9 (2.7) < month 1 year 18 (25) 61 (18.5) > 1 year 43 (59.7) 260 (78.8) Represented by a trade union No 26 (48.1) 183 (63.3) Yes 28 (51.9) 106 (36.7) Weekly schedule changes No 25 (56.8) 152 (64.7) Yes 19 (43.2) 83 (35.3) Classification of employment hours Full-time 57 (90.5) 240 (88.9) Part-time 6 (9.5) 30 (11.1) Working at time of admission No 65 (82.3) 268 (77.7) Yes 14 (17.7) 77 (22.3) Attempted a return to work No 52 (70.3) 185 (56.1) Yes 22 (29.7) 145 (43.9) 1 Some variables may have missing values. greater proportion of the workers in construction did not complete secondary school and a lower proportion attained post-secondary education, compared to those in other industries. This may be associated with the nature of occupations in the construction industry, in that workers are not required to have completed secondary education for a job in construction. Environment factors significantly associated with WrTBIs in the construction industry include the mechanism of injury and contributing factors. Construction workers were more likely to have sustained their injury due to a fall from an elevation compared to those in other industries, consistent with previous studies [6,12, 18]. Differences between falls from an elevation and falls from the same level have not been studied previously, and in our study, construction workers were less likely to have experienced a fall from the same level. Studying the type of construction work by type of fall has shown crafts and labourers [9], and site development, roadway, airfield, and utility workers [1] to more often experience WrTBIs due to falls from an elevation compared to falls from the same level. One study showed that fatal WrTBIs in the construction industry were most often due to jobs involving miscellaneous equipment, followed by truck and tractor jobs [12]. Studies with increasing detail of both the job type within the construction industry and injury mechanisms may reveal more specific areas for workplace safety and injury prevention such as targeting the occupations within the construction industry for preventive measures or recommending a need for personal protective equipment. Working in an elevated work area was the number one factor contributing to WrTBIs in the construction industry compared to other industries and this is consistent with our results for falls from an elevation. Slippery terrain/non-terrain as a contributing factor was reported less often among workers in the construction industry compared to other industries. However it appears that a major preventive effort with respect to wrtbi should be directed toward falls from roofs, ladders or suspended equipment. Additional information on task-specific details would be needed to help clarify our results. There was also some suggestion that injuries in the construction industry were most likely to occur in the early to late mornings (8:00 am 11:59 am), and early afternoons (1:00 pm 3:59 pm), and this is consistent with previous work that showed TBIs in the construction industry occurred during the daytime working hours [19]. Further analysis should explore the types of construction jobs that are associated with these injury times to target prevention efforts. Occupation factors significantly associated with WrTBIs in the construction industry include occupation and employment duration. Consistent with previ-

7 M. Liu et al. / Traumatic brain injury and the construction industry 289 ous studies [9,12], occupations more likely involved in WrTBIs in the construction industry were trades, transport and equipment operators, and general labourers. Construction workers were more likely to sustain WrT- BIs during the first four weeks of employment with their employer than workers in other industries. One previous study cross-classified jobs within the construction industry by an increasing level of skill with the duration of employment and showed that, those who sustained an accident were working in unskilled operative jobs and employed for a shorter amount of time [13]. Our study results indicate that during the first few weeks of work, prevention efforts and improved safety training should be emphasized. Further analysis in our study is needed to determine the type of occupations within the construction industry that should be targeted. Strengths of this study include primary data collection of variables that have been consistently reported in medical charts, with applied clinical criteria, the population-based design, and information on a number of different factors for an under-studied population. Although retrospective chart review is a method that provides data that are conveniently accessible, it is also limited in that the original data were not collected for the purpose of research. This study was constrained by what was available in the charts; for example, some injury-related information (e.g. presence of concussion, duration of loss of consciousness, posttraumatic amnesia, distance of fall, whether the worker was alone at time of injury etc.) was not always provided for all individuals and the WSIB Form 7 from which some of the occupation information was abstracted was not always complete. For these variables, the proportion of missing data from among the total population was 20 40%, and except for the occupation variables, these variables were not used for the analysis. Another drawback is that relevant variables were not collected with the abstraction tool or were absent from the charts (e.g. use of personal protective equipment, provision of safety training prior to the injury occurring etc.). It should also be noted that the data were not collected with the goal of addressing injury prevention but rather for clinical treatment however the goal of this research was with respect to injury prevention. The study was limited to workers who did not return to work after six weeks, therefore the generalizability of the results is limited to injured workers with mild to moderate traumatic brain injuries who had lingering post-injury sequelae. Our population did not include persons who sustained a head injury and who returned to work and did not have lingering symptoms and also did not include seriously/fatally injured persons. Cases however included those from all parts of the province. Near complete ascertainment of individuals in a geographically defined region was achieved and our population may be considered representative of WrTBIs in Ontario. Since there were only 83 workers in the construction industry, some levels of the PEO variables had small cell counts and therefore our p-value may have been affected. Descriptive statistics indicated that person, environment and occupation factors contributed to the occurrence of WrTBIs in Ontario s construction industry. The knowledge and hypotheses generated from this study can be used to guide more in-depth analysis of our data and contribute to the growing body of literature on WrTBIs in the construction industry. The purpose of this study was to understand the characteristics of workers and work-related factors that contribute to WrTBIs. In applying the PEO model in this study, the following are recommendations for future research in WrTBIs and the construction industry: (1) examine the use of personal protective equipment, (2) investigate whether workers received any safety training, and if so, what type(s) of safety training and education, (3) develop and improve safety training and education for the first few months of employment, the period during which WrTBIs are more likely to occur. A key point drawn from this study is that safety training and education should focus on preventing falls from an elevated level among construction workers during the first few months of work with an employer. In summary, the PEO model was useful for categorizing factors related to WrTBIs in an organized fashion. This model was helpful for framing WrTBI variables to gain an understanding of factors associated with injuries in the construction industry. Occupational therapists can develop WrTBI prevention strategies that target person, environment and occupation components. Occupational therapists can work with employers and workers in the construction industry to decrease the likelihood of WrTBIs and increase occupational performance. Acknowledgements We would like to thank Vartouhi Jazmaji at the Dalla Lana School of Public Health, University of Toronto for creating our database, Dave Mroczek and Jigisha Patel for entering the data, the many clinical professionals and staff at Toronto Rehab Neurology Services,

8 290 M. Liu et al. / Traumatic brain injury and the construction industry and the abstractors: Lisa Engel, Courtney Henderson, Michael Hutchinson, Angela Lalla, Sylvia Magrys, Cristine Provvidenza, and Catherine Wiseman- Hakes. This study was funded by the Ontario Neurotrama Foundation and by the Toronto Rehabilitation Institute with support from the Ontario Ministry of Health and Long-Term Care under the Provincial Rehabilitation Research Program. We thank Sandra Sokoloff for proofreading. References [1] H.J. Lipscomb, J.E. Glazner, J. Bondy, K. Guarini and D. Lezotte, Injuries from slips and trips in construction, Appl Ergon 37 (2006), [2] R.L. Neitzel, N.S. Seixas and K.K. Ren, A review of crane safety in the construction industry, Appl Occup Environ Hyg 16 (2001), [3] M. McCann, Deaths in construction related to personnel lifts, , J Safety Res 34 (2003), [4] H. Hsiao and P. Simeonov, Preventing falls from roofs: A critical review, Ergonomics 44 (2001), [5] B.J. Schaller, A. Kleindienst, T. Kruschat, H. Schliephake, M. Buchfelder and H.A. Merten, Industrial nail gun injury to the anterior skull base: A case report and review of the literature, J Trauma 64 (2008), E [6] H. Kim, A. Colantonio and M. Chipman, Traumatic brain injury occurring at work, NeuroRehabilitation 21 (2006), [7] C.A. Janicak, An examination of occupational fatalities involving impact-related head injuries in the construction industry, J Occup Environ Med 40 (1998), [8] A.C. Tricco, A. Colantonio, M. Chipman, G. Liss and B. McLellan, Work-related deaths and traumatic brain injury, Brain Inj 20 (2006), [9] S.M. Kisner and D.E. Fosbroke, Injury hazards in the construction industry, J Occup Med 36 (1994), [10] V.P. Coleman, Occupational head injury accidents in Great Britain, J Occup Accid 8 (1986), [11] N.J. Heyer and G.M. Franklin, Work-related traumatic brain injury in Washington State, 1988 through 1990, Am J Public Health 84 (1994), [12] C.A. Janicak, An examination of occupational fatalities involving impact-related head injuries in the construction industry, J Occup Environ Med 40 (1998), [13] R.A. Haslam, S.A. Hide, A.G. Gibb, D.E. Gyi, T. Pavitt, S. Atkinson and A.R. Duff, Contributing factors in construction accidents, Appl Ergon 36 (2005), [14] M. Law, B. Cooper, S. Strong, D. Stewart, P. Rigby and L. Letts, The person-environment-occupation model: A transactive approach to occupational performance, CanJOccupTher 63 (1996), [15] R.M. Ruff, G.L. Iverson, J.T. Barth, S.S. Bush, D.K. Broshek and the NAN Policy and Planning Committee, Recommendation for diagnosing a mild traumatic brain injury: A National Academy of Neuropsychology Education Paper, Arch Clin Neuropsychol 24 (2009), [16] L. Engel, C. Henderson, J. Fergenbaum and A. Colantonio, Medical record review conduction model for improving interrater reliability of abstracting medical-related information, Eval Health Prof 32 (2009), [17] Workplace Safety and Insurance Board, Statistical Supplement to the 2007 Annual Report. Retrieved July 4, 2009, from [18] N.J. Heyer and C.M. Franklin, Work-related traumatic brain injury in Washington State, 1988 through 1990, Am J Public Health 84 (1994), [19] A. Colantonio, D. McVittie, J. Lewko and J. Yin, Traumatic brain injuries in the construction industry, Brain Inj 23 (2009),