Each year in the United States, approximately 1.7 million people sustain traumatic brain injuries (TBI) from falls, motor vehicle crashes, sports, work, and other circumstances (Faul, Xu, Wald, & Coronado, 2010). These injuries result in sizable direct (e.g., medical) and indirect (e.g., lost productivity) costs; recent estimates suggest that the cost burden of TBI in the United States in 2000 was over $110 billion (Corrigan, Selassie, & Orman, 2010). Beyond the financial ramifications of TBI, injured persons may experience long-term cognitive, physical, psychosocial, and emotional health consequences (Langlois, Rutland-Brown, & Wald, 2006). The impacts of TBI on the U.S. population are profound: 43.3% of hospitalized TBI cases experience residual disability 1 year after hospitalization (Selassie et al., 2008), and over 3 million civilian residents in the U.S. are estimated to experience disability associated with previous TBI (Zaloshnja et al., 2008).

Many of these injuries are sustained at the workplace or during work-related activities, and the National Institute for Occupational Safety and Health (NIOSH) has identified work-related traumatic injury research as one of their key priorities (Stout et al., 1998). A recent analysis of work-related TBI (WRTBI) fatalities in the United States found that TBI accounted for 22% of all work-related injury fatalities between 2003 and 2008, and as much as 46% of work-related fall deaths (Tiesman, Konda, & Bell, 2011). The authors suggest that the magnitude of both fatal and nonfatal TBI in the workplace is substantial enough to warrant future research on the topic (Tiesman et al., 2011). Few U.S.-based studies have focused on WRTBI and we were unable to identify any national estimates of the incidence of nonfatal WRTBI. However, a study of occupational injuries using the Washington State Trauma Registry (WTR) found that more than 20% of fatal or nonfatal work-related trauma reported to the registry involved TBI (Sears, Bowman, Adams, & Silverstein, 2011). Related healthcare and workers’ compensation costs are substantial and may be underestimated, given the long-term physical and psychological consequences of TBI (Langlois et al., 2006). For example, researchers using Washington State workers’ compensation data estimated the median medical, compensation, and pension costs of WRTBI as $61,000 per injured worker (all ages, median; Wrona, 2006).

The long-term repercussions of TBI are substantial among adolescents and young adults, whose brains continue to mature and develop well into their mid-twenties (Pujol, Vendrell, Junque, Marti-Vilalta, & Capdevila, 1993). In 2010, approximately 293,000 adolescents and young adults (ages 15–24) sought emergency department treatment for TBI in the United States (Faul et al., 2010). In this age group, an estimated 3.5% of TBIs were sustained while working (Kim, Colantonio, & Chipman, 2006). Compared to other age groups, young workers have the highest rate of work-related injuries (Jackson, 2001). Several studies that reported WRTBI by age found that 13–23% of all WRTBI occur among young workers (24 years and younger) (Colantonio, et al., 2010; Kim et al., 2006; Wei, Roesler, & Kinde, 2012). For those who sustain TBI before or during adolescence, the economic and physical hardships of treatment and rehabilitation may be compounded by limited return to work or school and diminished attainment of education and vocational milestones (Anderson, Brown, Newitt, & Hoile, 2011; Doctor et al., 2005; Todis, Glang, Bullis, Ettel, & Hood, 2011).

Researchers often rely on workers’ compensation (WC) administrative claims data to study occupational injuries and illnesses. While WC sources often provide extensive data on direct and indirect costs associated with occupational injuries, information on the nature and circumstances around injuries tend to be limited (Reville, Bhattacharya, & Sager Weinstein, 2001). Combining WC data with other data sources thus may provide a more accurate and comprehensive picture of the burden and scope of injuries (Reville et al., 2001). Young workers in particular may be less likely to be included in WC data, because they may be often engaged in informal and unpaid work arrangements, for which injuries are not covered by WC programs or systematically reported to surveillance surveys (Runyan & Zakocs, 2000). For these reasons, exploring data sources beyond WC claims may provide valuable information regarding work-related injuries sustained by adolescents and young adults.

Hospital-based injury surveillance systems, such as trauma registries, provide an alternative source of data to examine occupational traumatic injuries including TBI. Many state trauma registries include an indicator of work-relatedness for injuries (Guice, Cassidy, & Mann, 2007), allowing researchers to identify severe, traumatic occupational injuries that may or may not be associated with a compensable WC claim. Trauma registries contain rich clinical data, including detailed information about the injury type and severity that are not typically available in WC claims data. Because of these advantages, researchers have used trauma registry data from several states for occupational injury surveillance (Friedman & Forst, 2007; Husberg, Conway, Moore, & Johnson, 1998; Sears et al., 2011). Linking trauma registry data with WC records allows researchers to leverage data from both sources and generate a richer, more comprehensive dataset that combines detailed injury information with WC outcomes (e.g. total medical and time-loss compensation costs; Sears, Bowman, Silverstein, & Adams, 2012). The combined data may also provide insight into the characteristics of injuries identified as work-related in the trauma registry but not linked to a WC claim.

The purpose of the research is to obtain a broader understanding of the scope and costs of WRTBI among adolescent and young adult workers in Washington State from 1998 through 2008 by examining traumatic occupational injuries reported to the Washington State Trauma Registry (WTR) and subsequently linked to Washington State WC claims. Specifically, this study focuses on WRTBI among adolescent and young adult workers (age 16–24 years) and aims to: (a) compare patient demographics and injury characteristics for injuries that linked to a WC claim with those that did not; and (b) describe the total amount paid for medical costs and time-loss compensation among WRTBI cases that linked to WC claims, specifically comparing isolated TBI to TBI with other trauma.

We identified 273 WRTBI cases that occurred between 1998 and 2008 among Washington State workers aged 16–24. Of the total, 80 (29.3%) were identified as work-related in the WTR but did not link to a compensable WC claim. One hundred ninety-three individuals had TBIs that were linked to a compensable WC claim, of which 159 (82.4%) were identified as work-related in the WTR and 34 (17.6%) were not (Figure 1). The majority of the 34 linked injuries that were not identified as work-related in the WTR were motor vehicle/traffic related injuries. Most demographic and injury characteristics did not differ significantly by linkage status (Table 1). Of the WRTBIs with L&I listed as an expected payer (N=188), 14.4% did not link to a compensable WC claim (p<0.01). Unlinked TBI was more likely to occur in a street/highway and less likely to occur in an industrial site than linked TBI (p=0.04). The distribution of TBI type did not differ substantially across linkage groups. Overall, 38.5% of identified WRTBI cases involved isolated TBI and 61.5% involved TBI with other trauma.

Among all WRTBIs linked to a WC claim (N=193), 151 (78.2%) were closed, State Fund claims that provided complete cost information from L&I and thus were included in comparative cost analyses by TBI type. Isolated TBI constituted 35.8% of this group, and TBI with other trauma made up 64.2% of linked, closed claims involving WRTBI. Age, sex, race, and other demographic characteristics did not significantly vary across TBI type (Table 2). Injury characteristics also did not differ substantially across groups, except for ISS score: individuals with isolated TBI had significantly lower ISS scores than those with TBI with other trauma (p<0.01). As of December 2010, four workers were found by L&I to have total permanent disability and received a pension due to their injuries (1 with isolated TBI and 3 with TBI and other trauma).

Mean unadjusted medical costs among isolated TBI cases did not differ significantly from TBI with other trauma (Table 3). Medical costs averaged $88,307 (median $16,426) for isolated TBI cases, compared to $73,669 (median $41,167) for TBI with other trauma. Mean unadjusted time-loss compensation costs were significantly higher for TBI with other trauma, compared to isolated TBI (p≤0.01). TBI with other trauma cases received an average of $13,734 (median $2,312) in time-loss compensation, nearly 3 times more than isolated TBI (mean $5,834, median $860).

GLM models showed that medical costs for workers with isolated TBI were 56% lower for workers with isolated TBI compared to TBI with other trauma (Cost Ratio, CR 0.44, 95% CI: 0.29–0.68), after adjusting for age, sex, injury year, and maximum head/neck AIS (Table 4). Compared to TBI with other trauma, time-loss compensation costs were 71% lower for isolated TBI cases (CR 0.29, 95% CI: 0.17–0.51), after adjusting for the same covariates.

In order to provide a minimal picture of the magnitude of costs involved in WRTBI among adolescents and young adults, we also calculated total observed costs, including paid-to-date costs for claims still open as of December 2010. Though incomplete because not all costs have yet been paid for open claims, the observed totals provide a lower bound estimate for the costs of WRTBI for workers covered by the State Fund who had injuries reported to the WTR during the 11 years covered by this study. Combining closed and open WC claims for State Fund workers with WRTBI (N=180), medical costs were nearly $19 million and time loss costs were more than $3.6 million (as of December 2010, when data were extracted from L&I). The median combined medical and time-loss compensation cost was approximately $45,000 per injured worker. The median combined per-worker cost was $19,000 for isolated TBI and $68,000 for TBI with other trauma. Total costs associated with isolated TBI were approximately $6.8 million for medical and $700,000 for time loss, compared to TBI with other trauma, for which medical costs were approximately $12.2 million and time loss costs were almost $3.0 million.

This study identified 273 WRTBI cases among adolescent and young adult workers aged 16–24 in Washington State from 1998 to 2008 using the WTR. Most were WRTBI with other trauma (61.5%) rather than isolated TBI (38.5%). For injuries linked to a closed WC claim, multivariate analyses showed significantly lower medical and time loss compensation costs among isolated TBI compared to TBI with other trauma. The latter group includes injuries to multiple body regions and likely more complex cases, so it is unsurprising that costs are higher in this group.

From 1998 to 2008, observed total costs among adolescent and young adult workers with a compensable State Fund WC claim for a WRTBI reported to the WTR exceeded $22 million. This represents just a portion of the potential costs for WRTBI among young workers in Washington State. It excludes claims paid by self-insured employers (approximately 30%) and injuries that did not link to a WC claim, was based only on those injuries eligible for and reported to the WTR, and underestimates total costs for the subset of injuries included because some claims were still open and potentially accumulating further costs (presumably the most severe). Nevertheless, it demonstrates the significant cost burden imposed by WRTBI among young workers in Washington State. Cost estimates described in this study apply to young workers and readers should use caution when comparing them to older workers, whose time-loss compensation costs may be higher due to increased earning potential with age.

To our knowledge, previous research on the cost burden of WRTBI is limited to regional studies (Colantonio et al., 2010; Wrona, 2006). An analysis of Washington WC claims from 1994 to 2001 estimated that the median cost of medical, pension, and future costs associated with WRTBI was $61,000 per case for workers of all ages (Wrona, 2006), compared to a median total cost in our study of $45,000 per worker for paid-to-date costs among workers aged 16–24. The two cost estimates are not directly comparable because the previous study included pension and future costs and did not restrict the age range, whereas our study’s estimate was based on paid-to-date medical and time-loss compensation costs for closed claims among young workers. Estimated WRTBI costs in Ontario, Canada appeared much lower than those found in our study (e.g., mean healthcare costs C$2,235), however, this estimate did not include time-loss compensation costs, and their definition of WRTBI included concussions and other mild injuries, which may not have been captured in the WTR data used in our study (Colantonio et al., 2010).

Evaluating outcomes for linked WRTBI cases (those with WTR records and compensable WC claims) does not fully capture outcomes for the population of young workers with WRTBI reported to the WTR. Approximately one-third of WRTBI identified as work-related in the WTR did not link to a compensable WC claim. The work-related indicator in the WTR serves as a convenient, independent identifier of work-related traumatic injuries that do not link to WC claims for various reasons, including employment that is exempt or excluded from WC coverage (State of Washington), employment covered by federal WC programs, injuries that may have been covered by L&I but were not reported, or claims for work-related injuries that were rejected for other reasons (Sears, et al., 2012). For young workers who commonly hold temporary, family, or informal jobs, it is possible that an injury that was classified as work-related in the WTR did not result in a WC claim because WC does not cover employment such as domestic work, newspaper delivery, and work conducted by minor children on family farms (State of Washington). Unsurprisingly, most linked WRTBI had L&I listed as the expected payer; it should be noted, however, that 28% of the cases that did not link to a compensable WC claim also listed L&I as an expected payer. This gap may represent valid work-related injuries that were not linked to a WC claim for the reasons described above.

This study utilized data from two sources, the WTR and WC claims, to identify WRTBI cases among young workers and describe the costs associated with these injuries. The different types of information provided by each data source served as an advantage – it allowed us to identify and describe WRTBI using detailed injury data from the WTR and describe and compare medical and time-loss compensation costs across types of TBI using WC claims data. Linking WC claims and WTR data overcomes weaknesses in each data source: WC claims lack detailed clinical information and WTR data patient-level variables relevant to work-related injuries, such as occupation and industry. The utility of trauma registries as data sources for occupational injury surveillance could be greatly improved by including these variables, as has been suggested by others (Forst, Hryhorczuk, & Jaros, 1999; Lowry, et al., 2010; Sears, et al., 2012).

This study has some limitations. First, misclassification of work-related injuries in the WTR may have occurred. An injury identified as work-related in the WTR may not have linked to a WC claim because of the reasons described above or because the injury was not actually work-related. Because this study relied on administrative data alone, we were not able to interview patients for clarification and are therefore unable to discern between these two possibilities. Second, this study captures only the relatively severe WRTBI that qualified for inclusion in the WTR. Injuries that do not meet the criteria for inclusion in the WTR (described in Methods), including those treated in clinics or treated at home, were not included in this study. Thus, the number and cost of WRTBI in Washington State during the time period studied should be considered an underestimate of actual numbers and costs. Third, not all eligible traumas are captured by the WTR. In a nationwide survey conducted in 2004 by Mann and colleagues, the WTR trauma manager estimated that the WTR captured about 85% of trauma victims with injuries satisfying registry inclusion criteria (Mann, Guice, Cassidy, Wright, & Koury, 2006). Finally, many costs associated with WRTBI, such as lost productivity and patient out-of-pocket costs were not measured. Therefore, the costs described in this study are an underestimate of societal costs associated with WRTBI among adolescent and young workers.

The costs of WRTBI reported in this study highlight the impact of WRTBI among young workers in Washington State. Although we were able to measure only a portion of the medical and time-loss costs related to these injuries, they were clearly substantial, well over $22 million for the 11 years covered by this study. The costs described here are also an underestimation of actual societal costs of WRTBI. Injury prevention efforts to reduce the likelihood and severity of WRTBI are warranted, especially in the young worker population, for whom TBI may result in particularly deleterious, long-term consequences. We have demonstrated that the tandem use of multiple data sources (in this case, linking trauma registry records with WC claims) can provide a more comprehensive picture of WRTBI among young workers than a single data source alone. The application of this linked-data source approach extends beyond WRTBI and young workers and holds great potential for future traumatic occupational injury research.

TBI among young workers involves long-term health and psychosocial impacts. In designing prevention strategies, a better understanding of injury characteristics and costs is valuable. Combining trauma registry and workers’ compensation (WC) records forms a unique, detailed dataset to investigate work-related injuries, especially among groups not fully captured by WC data.