The proportion of workers 55 years of age and older will grow to nearly a quarter of the United States (US) labor force by 2018, a 43% increase from 2008 [Toossi, 2009]. As aging workers remain on the job longer, understanding the health and safety needs of an aging workforce will be critical. This will be especially true for physically demanding jobs, such as those in the construction trades, where older workers may be at higher risk of injury and illness. The injuries and illnesses sustained by workers in the construction industry often result in a significant financial burden for the worker, industry and society as a whole [Dong et al., 2007; Waehrer et al., 2007a].

Construction workers have higher rates of injuries than workers in other industries [U.S. Bureau of Labor Statistics, 2009] and the average total cost of their injuries is significantly greater. In the US, the estimated average total cost of construction-related injury was $27,000, with a greater proportion of the total costs related to indirect costs (e.g., wage loss) rather than direct costs (e.g., medical costs), compared to $15,000 across all industries in 2002 [Waehrer et al., 2007a]. Waehrer et al. [2007b] esti-mated that $13 billion is spent annually in the US on workers’ compensation (WC) costs in the construction industry, making it one of the most expensive industries to insure. According to the National Compensation Survey in 2011, the construction industry spent on average, $1.32 per hour worked on WC. This is triple the cost spent 2 Schwatka et al. across all industries ($0.44 per hour worked) [U.S. Bureau of Labor Statistics, 2011].

Given the high cost of work-related injuries and illnesses in the construction industry, it is important to understand what factors contribute to these costs. Older construction workers (>55 years of age) are responsible for a disproportionate risk of work-related health issues [NORA Construction Sector Council, 2008; Dong et al., 2011; Schwatka et al., 2012], but their contribution to the total cost of injuries and illnesses has yet to be quantified. Previous research has either limited their discussion to specific construction trades, types of injuries or WC claims [e.g., Lipscomb et al., 2003; Friedman and Forst, 2009; Lipscomb et al., 2009]. As construction workers age, they will likely experience physical limitations and co-morbidities [Welch et al., 2008; Dong et al., 2011]. These vulnerabilities will adversely affect their ability to perform physically demanding work in the construction industry, leaving them more susceptible to injuries such as musculoskeletal disorders (MSDs), fractures and contusions [Maertens et al., 2012; Schwatka et al., 2012]. Despite these findings, older worker age does not appear to be associated with higher injury rates [Rogers and Wiatrowski, 2005; Schoenfisch et al., 2010; Restrepo and Shuford, 2011].

Older construction workers, however, may be more likely to experience severe-type injuries compared with younger construction workers. This trend may result in higher indemnity, rather than medical costs, as older construction worker’s injuries may require more days away from work and may result in disabilities and physical limitations [Choi, 2009; Schwatka et al., 2012]. Although injuries among older construction workers may result in greater indemnity costs, it is not clear whether they may not reflect a greater total cost per claim as compared to their younger coworkers. In the present study, we performed a comprehensive investigation of the association between worker age, injury type and WC costs (overall and by cost type) among claimants employed in the construction industry. Over 100,000 construction WC claims filed in the state of Colorado between 1998 and 2008 were analyzed in order to test the hypothesis that the positive association between age and WC cost would differ by cost type (e.g., total, medical, and indemnity costs), and that the relationship between injury type and cost would vary by age.

In our dataset of 107,065 WC claims among construction workers in Colorado, the mean claimant age was 34 years (SD = 11) and the median was 33 years (IQR = 26–43). The majority of injured workers who filed a claim were male (97%). Workers under the age of 45 filed approximately 80% of the WC claims. After adjusting all costs to 2010 dollars, the total cost of the 107,064 claims was $931,234,994. The total medical cost and total indemnity cost were $408,613,710 and $461,084,685, respectively.

Of all claims filed, 23% (n = 24,846) were WC claims with medical plus indemnity costs and 77% (n = 82,219) were WC claims with medical only costs. Claimants over the age of 65 filed more medical plus indemnity-type claims (34%) than claimants between the age of 18 and 24 years of age (18%) (χ² = 91.68, P < 0.0001). When the costs ($) of the claim types were compared, claimants over the age of 65 had a higher per-centage of indemnity costs (e.g., 63% of the total costs were due to indemnity costs), compared to claimants aged 18–24 years (e.g., 51% of the total costs were due to indemnity costs) (Table I).

The majority of claims were related to strains (27%), contusions (21%), and lacerations (17%). Other injuries included: foreign body (7.5%), sprain (6.6%), puncture (6.3%), fracture (3.6%), crushing (1.5%) burn (1.47%), and all other injuries (8.6%). The other category included injuries that represented <1% of the claims and “all other” injuries, as defined by the insurer who provided the database of claims. Strains were the most common type of injury among all age groups except for the oldest age group, ≥65, where strains and contusions occurred at similar frequencies, 26% and 27% of all claims, respectively. Lacerations occurred more frequently among younger age groups (18–24 years) compared with older age groups (≥65 years), accounting for 21% and 12% of all injuries, respectively. There were no meaningful differences between age groups for the other types of injuries (data not shown). The mean cost of a claim related to each type of injury (strains, contusions, lacerations, other) generally increased with increasing age group (see Table II) with the greatest proportion of total costs attributed to indemnity expenses. For example, the proportion of the total WC costs attributable to the indemnity costs for a strain type of injury were 59% for claimants ≥65, compared to 52% for claimants 18–24 years. The proportion of the total WC costs attributable to the medical cost of a strain type of injury was and 35% for claimants ≥65 years and 39% for those 18–24 years, respectively (data not shown).

Claimant age (years) and WC costs had a small, but statistically significant correlation with total costs (r = 0.07, P < 0.0001), medical costs (r = 0.05, P < 0.0001), and indemnity costs (r = 0.10, P < 0.0001). Mean costs (total, medical, and indemnity) of a claim increased with increasing age group, with one exception (see Table I). Mean medical cost per claim increased up to the 55- to 64-year age group then slightly decreased for the ≥65-year age group. The differences in mean cost by age group were statistically significant: total cost (F_5,107059 = 123.99, P < 0.0001), medical costs (F_5,107059 = 56.43, P < 0.0001), and indemnity costs (F_5,107059 = 236.86, P < 0.0001).

A priori multiple pairwise comparisons using a Bonferroni-adjusted alpha level of 0.003 per test (0.05/15) was used to evaluate mean costs between claimant age groups (Table III).

There were statistically non-significant differences in mean total cost between claimants 35–44 years of age and those ≥45 years of age [e.g., 45–54 (P = 0.02), 55–64 (P = 0.02), ≥65 years (P = 0.33)]. In other words, total costs increased with increasing age category until ≥35 years of age when total costs plateaued. A similar pattern was observed for medical costs. In terms of indemnity costs, there were statistically non-significant differences in mean indemnity costs between claimants 45–54 years of age and those ≥55 years of age [e.g., 55–64 (P = 1.00) and 65+ (P = 0.014)]. In other words, indemnity costs increased with increasing age category until ≥45 years of age when indemnity costs plateaued.

Simple linear regression analyses were used to further evaluate the relationship between claimant age, type of injury, and WC costs. The first step in assessing this relationship was to evaluate age of claimant and cost in univariate models (see Table IV). When age of claimant was included in the model as a continuous variable, the strongest association was observed between claimant age and indemnity costs. There was a 3.51% increase in the indemnity cost of a claim for each 1-year increase in the age of a claimant. In contrast, there was a smaller 1.11% increase in the medical cost of a claim for each 1-year increase in the age of a claimant. We also included age in the model as a categorical variable, because the ANOVA results indicated that age might not be a linear function of cost. Compared to claimant’s aged 18–24, all other age groups exhibited a greater increase in cost, especially for indemnity costs. For example, a claimant ≥65 years had a 46% higher medical cost than a claimant aged 18–24 but a 372% higher indemnity cost. In summary, we observed the similar relationships between age and cost by type, regardless of whether age was included as a continuous or categorical variable in the linear models.

To further explore the relationship between age and indemnity costs, we conducted linear regression analyses with type of injury as one of the predictors (see Table V). In the univariate model, a strain type of injury was more costly than the other three types of injuries. For example, a contusion type of injury was 46% less costly than a strain type of injury.

In multiple linear regression analyses by claimant age, we observed stronger associations for strains, contusions, and other types of injuries with indemnity cost as age increased (all P for trends <0.05), with the strongest associations observed among claimants ≥65 years. We did not observe evidence for modification by age on the association between laceration type of injury and indemnity cost. The final interaction model for injury type by age group explained 3.1% of the variance in indemnity cost (Table V).

Using a large WC database that was representative of approximately 80% of Colorado construction industry WC policyholders, we evaluated the relationship between age, injury type and costs. The mean total cost of a claim filed by workers 65 years and older was about three times the cost of a claim filed by workers aged 18–24. Yet, workers under the age of 45 filed 80% of the claims. Linear regression analyses revealed that the increase in costs among older workers was driven by increases in indemnity costs, rather than medical costs. We also reported that the indemnity costs associated with specific injuries (e.g., strains and contusions) increased along with age of the claimant. These results suggest a major financial burden, particularly due to indemnity costs (i.e., lost days at work, disabilities, and physical limitations) among the companies that insure workers and the WC insurance agency that will incur among the aging construction workforce.

Our findings indicated that overall there were statistically significant differences by mean WC costs regardless of cost type by increasing age group (18–24, 25– 34, …, ≥65 years). For individual age group comparisons, statistically significant differences in mean total costs were observed among increasing age groups up to age 35, while mean costs did not differ significantly between older age groups (35–44, 45–54, 55–64, and ≥65 years). Medical costs plateaued at 35–44 years of age but indemnity costs plateaued at 45–54 years of age. Our results suggest that how we define “older age,” in terms of a subset of the workforce most susceptible to injury, may need to be adjusted downward. Our results also high-light the importance of evaluating cost type, rather than just total cost when describing relationships between age and cost.

Our findings support previous research demonstrating that worker age was positively associated with WC costs, although the age at which costs begin to plateau differs across studies. In a study of over 20,000 WC claims among Illinois construction workers’ between 2000 and 2005, the mean total cost of compensation peaked for workers aged 55–64 and then declined slightly for workers over the age of 65 [Friedman and Forst, 2009]. Similar findings were reported by Waehrer et al.’s [2007a] study of construction injuries (N = 162,371 injuries) where mean total cost of injuries and illnesses requiring days away from work peaked at ages 45–54 and declined for workers over the age of 55. Their database of occupational injuries, however, came from the Survey of Occupational Injuries and Illnesses (SOII) (2002 Annual Survey) collected by the Bureau of Labor Statistics (BLS) and only represents construction companies with ten or more employees. Of all construction establishments in the US, 79% have less than ten employees, and these establishments make up 24% of the construction workforce [Center for Construction Research and Training, 2008]. These previous studies were either smaller in size [Friedman and Forst, 2009] or had an occupational injury database that was not representative of the entire construction industry [Waehrer et al., 2007a]. Our data, while only representative of the Colorado construction industry, suggest that that the age at which WC costs plateau may be younger than previously reported. A recent report from the National Council on Compensation Insurers found similar trends [Restrepo and Shuford, 2011].

We determined that strains and contusions were more common among construction workers and that age modified the association between injury type and indemnity cost. Our top three frequently occurring types of injuries (strains, contusions, and lacerations) were also cited as the top construction industry related injuries treated in hospital emergency rooms in a recent study using National Electronic Injury Surveillance-Work (NEISS-Work) database [Schoenfisch et al., 2010]. The results of the present study suggest that the divergence in indemnity cost among older and younger workers becomes greater as injury severity increases.

While the present study determined that older workers filed a small percentage of claims related to minor injuries (i.e., lacerations), it is possible that older workers selectively report the most serious of injuries. Older workers may shy away from the negative attention related to injury reporting as they already face the stigma associated with being part an aging workforce. There may be additional fears related to being singled out for their “carelessness” or “unsafe acts” that could lead to retaliation. Additionally, throughout their careers, older workers may have had unpleasant experiences with the WC system and chose not to report minor injuries in order to avoid further frustrations. Despite these issues, older and younger workers who filed a claim for minor injuries had similar indemnity costs.

We reported that indemnity costs, rather than medical costs were driving the higher WC costs among older construction workers. Our findings supports previous research findings that indicate more lost workdays [Lowery et al., 1998; Horwitz and McCall, 2004; Kucera et al., 2009] and increased disability [Courtney et al., 2002; Arndt et al., 2005; Welch et al., 2010], and thus, higher indemnity costs among aging construction workers. Lowery et al. [1998] determined that rate of lost work time among construction workers at a large Owner Controlled Insurance Program site was greatest among workers over the age of 50. Kucera et al. [2009] found that among workers 45 years and older were 60% more likely to have a claim with delayed return to work (>90 days away after injury), compared to workers less than 30 years of age. A study that utilized WC claims from the Oregon construction industry found that the temporary total disability compensated days was greatest among workers 46 and 55 years of age [Horwitz and McCall, 2004]. Our findings not only support previous findings, but also provide novel quantitative data on the increased financial costs associated with lost work time and disability among older construction workers.

Our study supports previous research that indicates injuries to the musculoskeletal system (e.g., strains) are of particular concern for aging construction workers [de Zwart et al., 1999; Welch et al., 2008; Hoonakker and van Duivenbooden, 2010; Welch et al., 2010]. In our database, approximately 50% of all strain injuries were to the spine/back/lower trunk among all age groups. Overexertions of the back are a major source of pain and injury among older construction workers [Welch et al., 2008; Hoonakker and van Duivenbooden, 2010]. A significant amount of lost work time, delayed return to work, and disability from back injuries among aging construction workers has been reported among carpenters in Washington state [Lipscomb et al., 2008]. Lipscomb et al. [2009] reported that payment rates increased with age but the source of payments were not reported (i.e., medical vs. indemnity payments). Our results are consistent with these previous results, but also contribute new information about the actual dollar amount associated with strain-type injuries and how the association with indemnity costs increase with in-creasing age of the claimant.

By the year 2018, the participation rate of workers over the age of 55 will have increased while the participation rate of workers between the ages of 16 and 54 will have decreased [Toossi, 2009]. Maintaining the employability of older workers will be critical in order to compensate for the decreasing labor force participation rates of those in their prime working years. In physically demanding industries like construction, the impact of the aging population can be significant. The physical limitations that older construction workers experience [Dong et al., 2011] may limit older workers ability to perform physically demanding tasks in the construction industry without becoming injured [Schwatka et al., 2012]. Older construction workers may be more likely to hold supervisorial positions due to experience and tenure, and thus may not have the same exposures to illness- and injury-related risk factors than younger workers. However, as the number of skilled construction workers in the labor force decrease, there may be an increased demand for older workers to remain in more laborious positions [U.S. Bureau of Labor Statistics, 2012]. Thus, subsequent efforts to return to work after injury or illness may be hindered by the difficultly to make accommodations in the construction industry [Berecki-Gisolf et al., 2012].

While this study indicated that older workers filed a small percentage of the total WC claims, the WC costs incurred by them were more costly on a per claim basis than their younger counterparts for indemnity rather than medical costs. This study illustrates the economic significance of injuries and illnesses among older construction workers. Additional research is needed to determine if older construction workers are selectively reporting inju-ries, which would likely have an effect on the medical and indemnity costs. The utilization of WC cost data is a useful but lagging indicator of the state of occupational health and safety among construction workers. New research should be aimed at leading indicators (e.g., safety climate/culture) of health and safety that promote the development of proactive injury prevention strategies. Leading indicators have the potential to identify the risk of occupational injuries prior to their occurrence among construction workers of all ages.