Estimating uninsured costs of work-related accidents, part II: an incidence-based model

Transcription

1 Theoretical Issues in Ergonomics Science ISSN: X (Print) X (Online) Journal homepage: Estimating uninsured costs of work-related accidents, part II: an incidence-based model O. Paez, T. Uahinui, A. Genaidy, W. Karwowski, L. Sun & N. Daraiseh To cite this article: O. Paez, T. Uahinui, A. Genaidy, W. Karwowski, L. Sun & N. Daraiseh (2006) Estimating uninsured costs of work-related accidents, part II: an incidence-based model, Theoretical Issues in Ergonomics Science, 7:3, , DOI: / To link to this article: Published online: 02 Nov Submit your article to this journal Article views: 74 View related articles Citing articles: 4 View citing articles Full Terms & Conditions of access and use can be found at Download by: [University of Arizona] Date: 17 March 2016, At: 14:03

2 Theoretical Issues in Ergonomics Science Vol. 7, No. 3, May June 2006, Estimating uninsured costs of work-related accidents, part II: an incidence-based model O. PAEZy, T. UAHINUIz, A. GENAIDY*y, W. KARWOWSKIx, L. SUNô and N. DARAISEHy yuniversity of Cincinnati, Department of Mechanical, Industrial, and Nuclear Engineering, Cincinnati, OH , USA zengineering Support, Toyota Motor Manufacturing North America, Inc. Kentucky, USA xuniversity of Louisville, Center for Industrial Ergonomics, Louisville, KY 40292, USA ôuniversity of Cincinnati, Department of Civil and Environmental Engineering, Cincinnati, OH , USA (Received 1 February 2005; in final form 15 April 2005) Background: This is the second of two papers focused on insured costs of workplace accidents. In part I, a systematic review of the literature was conducted to identify previous models. Although uninsured costs have been identified, there is no reliable method for employers to estimate their cost. Objective: This paper proposes a quantitative model to represent the behaviour of the cost of accidents. It provides a cost range into which the accident is most likely to fall, based on information known at the time of the accident. Methods: A logarithmic distribution is proposed to represent the insured cost as a function of the severity of the accident and the occupation of the workforce exposed. Point estimates for administrative costs and production losses are obtained based on a generic case in order to quantify the equation in terms of monetary value. Results: For a given workforce composition, it is possible to obtain uninsured cost intervals that range from times the insured costs. The evidence supports a logarithmic relationship between uninsured and insured costs, contrary to the principle of linear ratios. Conclusion: The proposed model requires only basic information about the accident: the severity of the injury and the composition of the workforce exposed. The value added from the model is not only an interval range for the insured costs but also the financial contribution of safety programmes to the workplace that is often ignored by employers. Keywords: Accident model; Uninsured costs; Accident cost estimation 1. Introduction The use of quantitative methods to estimate the costs of workplace accidents has been hindered by the multiple activities and parties involved, as exposed by the *Corresponding author. Ash.Genaidy@ac.edu Theoretical Issues in Ergonomics Science ISSN X print/issn X online # 2006 Taylor & Francis DOI: /

3 248 O. Paez et al. testimony delivered by Marquis (1992) to the US Senate Committee on Labour and Human Resources. Households, employers and ultimately taxpayers share the burden of the costs of workplace accidents. For employers, the costs of accidents go beyond the expenses covered by the worker s compensation programme. This paper proposes a quantitative model that works exclusively with insured costs carried by the employer and is the second of a two-part series of articles in which the first deals with a systematic review and critical appraisal of the published literature on the subject. The division of the cost of accidents into insured and uninsured costs was developed by Simonds for the US National Safety Council (1998a). The insured costs are those handled through an insurance policy (self-insured or third-party insured) and refer to lost wages, medical expenses and property damage. The uninsured costs are those carried by the employer regardless of the insurance policy. That is, operational and administrative resources used to handle the consequences of the accident in the workplace. While there is an extensive effort to estimate the insured costs, there has been little and perhaps disparate efforts to understand and ultimately to predict the insured component of accidents. Most models work with prevalence-based analyses, that is the total costs over a number of years are used to draw conclusions about the average cost of an accident. Prevalence-based analyses (based on historic and aggregate records) do not explain the behaviour of individual accidents and assume that variation in the distribution of the cost of accidents can be neglected. Incidencebased analyses, which work with individual accident occurrences, provide a clearer description of how the costs behave. Accordingly, this article proposes an incidence-based approach to predict the cost of accidents. 2. Methods The proposed model is the result of a four-step strategy. First, a systematic review of the available literature determined the main components and predictors of uninsured costs of accidents. Secondly, the incidence of accidents was further analysed to find a continuous distribution that best represents the relationship between predictors and cost components. Thirdly, the cost function was applied to each uninsured component in order to obtain an estimate of its cost. Finally, the proposed equations were tested with hypothesized examples Critical appraisal The critical appraisal was conducted in part I (Sun et al. 2005) to identify meaningful components and predictors of uninsured costs. Publications were identified through various search engines as well as business administration, safety engineering and medical journals. Five criteria were selected to rate the articles. The results of the critical appraisal are summarized in figure 1. Four main components of the uninsured costs were identified: administrative and legal costs, investigation of the accident, replacement and productivity loss. Three predictors were also identified: the type of injury, type of industry and skill level. The type of injury can work as a predictor of both administrative and operational costs. The other two predictors

4 Estimating uninsured costs of work-related accidents 249 Accident costs Insured costs Uninsured costs Wages Medical expenses Property damage Administrative costs Productivity loss Replacement Investigation Figure 1. Type of injury Occupation (Type of industry and skill level) Conceptual model based on critical appraisal. represent the occupation (industry and skill level) that it is required to estimate the wages associated with the operational losses Incidence-based distribution of the costs of accidents Simonds and Grimaldi (1963) explain the behaviour of accidents with some examples. A first-aid accident such as a small cut or a bruise might not require any treatment beyond a bandage. If the time lost away from the workstation is less than 1 h, it is very likely the worker is not relieved. The administrative work and medical supplies needed by this accident are almost negligible. However, if the injured worker has to be taken to a facility outside the company new costs will occur: someone has to take the worker to the health care provider, someone will have to cover the worker s job and someone has to handle the indemnity of the case. The handling of the case is more complex as it is elevated to a temporary or permanent disability. Now, administrative personnel as well as different management levels may become involved. As the injury becomes more serious, new components appear and ultimately spur the costs incurred in an accident. A cost function, C(S), is used to express the relationship between the severity of the accident and its uninsured costs, where S is the severity level of the accident. Assessment of severity is an integral part of risk assessment methods (Tweeddale 2003). To measure the effect of the severity level on the cost of accident, sample data of compensation costs per hour worked c i was taken from the US Bureau of Labor Statistics (BLS) (2004). The employer costs for employee compensation survey contains a sample of occupations from both public sector and private economy during the first two quarters of After sorting the cost of accidents by cost value, the data was split into 40 consecutive groups all with the same size. Each group represents a different severity level S i. The severity level S i was represented as a function of the average cost value ½c i Š j. Regression analysis was used to test two models S ¼ ½c i Š and S ¼ Lnð½c i ŠÞ that explain a linear and logarithmic relationship respectively.

5 250 O. Paez et al. C( S ) = ln( 1 S ) 0 <= S <1 m = 1 s = 1 Cost C(S) 400% 200% 1 C(1 ) = m e 100% 50% 25% Figure 2. 0% 20% 40% 60% 80% 100% Least severe Most severe Severity (S) Relationship between accident costs and severity. For C(S), the exponential percentage point function best represents this behaviour as shown in figure 2. The function C(S, ) ¼ ln(1 S); 0 S<1, >0 represents the estimated cost of an accident (outcome). The parameter is required to transform the probability distribution into a monetary value. Instead of obtaining an average to estimate, which would require a vast sample with representative instances of deaths and permanent injuries, the proposed approach only requires to estimate the cost for one case. The best case is the one that is closer to the mean cost value, that is S i ¼ 1 1/e 63.21%. As a continuous function, C(S, ) does not provide point estimators but interval ranges according to the level of severity. For instance, the bottom 20% of accidents will result in between 0 22% of the average cost of accidents. The top 20% of accidents will result in between % of the average cost of accidents. In this case, the cost of accidents can grow infinitely; however, a practical boundary was set at 99.99% of the severity level (which results in 921% of the mean cost). As the severity of the injury increases so do the uninsured costs. Uninsured costs increase because of the additional resources involved in the handling of the accident. Leigh et al. (2000) suggest a categorization of levels of severity based on data from the National Council on Compensation Insurance (NCCI) in (1) First-aid. Non-disabling injury that is not Occupational Safety and Health Administration (OSHA) recordable (60.03% of injuries). (2) Medical only, no indemnity. Injury that needs medical treatment and, therefore, is OSHA recordable (19.68% of injuries). (3) Temporary total. Injury that causes lost time from the worker (14.60% of injuries). (4) Permanent partial major. Injury that results in permanent disability; e.g. amputation of limb, loss of senses (5.55% of injuries).

6 Estimating uninsured costs of work-related accidents 251 Table 1. Severity level by type of injury in 1993 (NCCI). Severity (S i ) L Bound (%) U Bound (%) 1. First aid Medical only Temporary Permanent minor Permanent major Permanent total (5) Permanent total. Injury that results in total disability, so the worker is no longer able to perform productive work (0.09% of injuries). (6) Death. Injury that results in fatality (0.05% of injuries). The previous values are represented in the matrix S (6 2), where S i represents the severity level i for each of the six categories as shown in table 1. S represents the frequency intervals for each category. The bounds are set according to the cumulative frequency of injuries for each category Cost components The equation C a ¼ C i þ C u represents the main components of the cost of an accident, where C a is the cost of the accident, C i its insured component and C u its uninsured component. The uninsured portion C u can be expressed as C u ¼ C Adm þ C Oper, where C Adm represents the administrative losses and C Oper the operational losses due to the accident. Administrative losses account for additional expenses triggered by the occurrence of an accident. The effect of administrative losses appears in ledgers as a negative variation in general expenses. On the other hand, operational losses account for productivity losses after the occurrence of an accident. The effect of operational losses appears in ledgers as a negative variation in salaries. In this sense, the uninsured costs do not represent hidden components but rather variances in standard account ledgers. The proposed model estimates C Adm as C Adm ¼ C(S i, Adm ), where S i is the severity level of the injury and the estimated cost value for operational losses Adm is obtained based on an estimate of the total cost at C(S i ) ¼ 100%, that is, S i ¼ 1 1/e63.21% (figure 2). S i falls into the medical only category [63.03%, 79.71%]. Operational losses represent the effect of the accident on other employees. In risk management this is known as exposure assessment (Coleman and Marks 1999). Oxenburgh et al. (2004) define productivity losses as reductions in output below the maximum possible if all elements of the system are working in optimum conditions. The productivity loss associated with the accident depends on the number of people exposed to the injury. Those exposed to the accident are not only the people who saw the accident occur, but people who work on other shifts or in other departments can also be affected by knowledge of the accident. The management structure is also involved in handling the consequences of the accident. Employees cannot be treated as one population since wage differentials result in different costs. According to the critical appraisal, the type of industry and the skill level account for the wage level. The proposed model works with one

7 252 O. Paez et al. W Table 2. Hourly wages by occupation ($). Service occupations Operators/craft Sales/support Professional/technical Managerial/executive matrix; E as containing the composition of the workforce by occupation and two vectors; H representing the hours lost and W the hourly wages for each occupation. To estimate the salary, the National Compensation Survey (US BLS 2004) provides statistics by industry and skill level. The corresponding values for 2003 have been used to establish the values for W, shown in table 2. The time lost is represented by the matrix product E H. E is a [5 3] matrix, each row represents a salary level and three columns represent different populations exposed. The first column contains the injured workers, which in most cases would be only one. The second column represents the crew, that is the group of people that work directly with the injured person. The third population is the overhead, that is the group in charge of follow-up. The proposed model estimates C Oper as C Oper ¼ C(S i, Oper ), where Oper is obtained as Oper ¼ (E H)W. The model requires an estimate of the number of hours lost H i for each population. Following the same previous approach, a point estimate is obtained at the medical only category (S i ¼ 63.21%) Total cost Operational and administrative losses can be combined to produce the following equation, C u (S i, E) ¼ ( Adm þ (E W)H) ln(1 S i ). The input information is the severity category S i and E the exposed workforce by occupation. Values of Adm, W and H are assumed to be constant but they can be changed based on employer-specific information. To obtain a total value of uninsured costs for several accidents C u, one main consideration is whether the aggregation is within the same team or across different teams. The assumption that the population exposed E is the same will inflate the estimated costs. In this case, P it is better to work with different exposed populations E k for each team k: Cu ¼ P all j ð Adm þðe j WÞHÞ P all i lnð1 S ijþ. This P equation can be simplified if there is only one population exposed to: Ci ¼ ð þðewþh Þ P Adm all i lnð1 S ijþ. 3. Results 3.1. Behaviour of the costs of accidents The results from the regression model applied to the sample of the US Bureau of Labor Statistics shows that while r 2 for the linear model was 0.21, r 2 for the logarithmic model was Out of the 40 groups, the first 20 groups have a cost

8 Estimating uninsured costs of work-related accidents Cost per hour work ln(ci) Figure 3. Severity group (Si) Cost per hour worked as a function of severity. Table 3. Estimation of administrative costs. Index Element Estimate at S ¼ Description I C Adm i ^C Adm i 1 First aid $20 Bandage, gloves, wraps, braces, antiseptics 2 Clean up $20 Sanitizer, gloves, eyewear, disposable wipes, biohazard bags 3 Transportation $10 Gasoline (40 miles) 4 Claim handling $30 One hour for an HR specialist (see wages table) 5 Replacement $200 Cost-per-hire temporary worker per hour worked below $1, while the top 10 groups range from $ Figure 3 shows a scatter plot between the severity level and the logarithm of the cost per hour. The logarithmic function might not be accurate at the extremes. It over-estimates the bottom 10% as well as the top 10% of severity levels. However, the assumption that the cost incurred by accidents follows a logarithmic behaviour is a valid approximation explained by the fact that the few most severe accidents carry the largest portion of the cost Cost components Table 3 shows the elements of administrative costs applied to medical only injuries. Five components were identified as: first aid supplies, clean up supplies, transportation, claim handling and replacement. In the case of a medical-only injury: first aid supplies can range from antiseptics to braces; clean up might be required for bloodborne pathogens; an injured worker may need to be taken to the hospital in a vehicle; the claim has to be handled through the insurance company; and the absence could require hiring a temporary employee or someone from a different position. Adm is estimated as the sum of each element for a medical only injury. The estimated value is Adm ¼ c Adm i ¼ $280.

9 254 O. Paez et al. Population exposed (E) Element Table 4. Estimation of operational losses. Estimate at S ¼ Description J H Oper i ^H Oper i Worker Training 8 h Nature of the task, specifications, safety precautions, task trial and error. It could be substituted by overtime costs Crew Downtime 0.5 h First aid reaction, clean-up, waiting for replacement Crew Production pace 0.5 h Schedule disruption, perceived risk and psychological impact Overhead Follow-up 3 h Accident investigation, development of countermeasures, employee awareness efforts Table 4 shows the elements of operational losses applied to medical only injuries. Five components were identified. The following costs can be associated with operational losses: training, downtime, production pace, follow-up. Training corresponds to the time spent introducing the newcomer to the task. Downtime refers to the time lost due to operational breakdown immediately after the accident and before the position is covered and production pace refers to slower pace due to schedule changes, low morale and perceived risk due to the accident. Follow-up time is needed to document the accident, identify the root cause and develop corrective actions that prevent future occurrences. H is estimated as the sum of each element for a medical only injury. The estimated value is H ¼ hours: Testing Using an electronic worksheet, it is possible to implement the equation with the following input: (1) The severity level of the injury (S i ); and (2) The population exposed to the accident by occupation (E). The model works with two main variables: the workforce exposed and the severity of the injury. To illustrate the results of the model, hypothesized examples are presented for different types of injury as well as different workforce configurations. To illustrate the effect of severity in the model, the following population E ¼ was used to estimate the insured cost for different types of injury. Table 5 presents the cost intervals for each type of injury. Insured costs range from $

10 Estimating uninsured costs of work-related accidents 255 Table 5. Uninsured cost intervals by type of injury. Id Injury type Uninsured costs from ($) Uninsured costs to ($) Ratio C u /C i from Ratio C u /C i to 1 First aid (non-disabling) Medical only Temporary total and partial Permanent partial Permanent total Deaths Ratio C u /C i Figure 4. First aid Medical only Temporary total and partial Permanent partial Permanent total Ratio intervals for different types of injury. Deaths The ranges for each category are variable. The broadest range is for permanent partial injuries ($2405) while the shortest range is for medical only ($439). The insured costs increase more than seven times from medical only to permanent partial and they increase less than twice from permanent total to death. Figure 4 presents an estimated ratio using the statistics from the NCCI (Leigh et al. 2000). The ratios are not close to each other; in fact they range from times the insured costs. For non-disabling or medical only injuries the amount of insured costs would not be significant compared to the disruptions caused in the company. As severity increases, the cost of compensation for disabilities will exceed the cost of disruptions for the rest of the workforce. To illustrate the effect of the composition of the workforce in the model, four different populations were tested for the same type of injury. Examples were used from a construction crew E ¼ , 0 0 1

11 256 O. Paez et al. Table 6. Uninsured cost intervals by type of industry. Id Workforce exposed Uninsured costs from ($) Uninsured costs to ($) Ratio C u /C i from Ratio C u /C i to 1 Construction crew Service crew Software team Manufacturing line a service crew a software team and a manufacturing line E ¼ , E ¼ E ¼ to estimate the insured cost for different types of injury. Table 6 presents the cost intervals for each type of injury. Insured costs range from $ for a medical-only injury. The ranges for each category are variable. The broadest range is for the manufacturing line ($575) while the shortest range is the service crew ($369). The uninsured costs are higher for the manufacturing line with respect to the other three cases. Figure 5 presents an estimated ratio using the same statistics from the NCCI (Leigh et al. 2000). The ratios range from times the insured costs. For non-disabling or medical only injuries the amount of insured costs would not be significant compared to the disruptions caused in the company. As severity increases, the cost of compensation for disabilities will exceed the cost of disruptions for the rest of the workforce. 4. Discussion The use of linear ratios introduced by Heinrich (1959) between direct and indirect costs is not justified due to the low correlation between the incidence rate of accidents and the cumulative accident costs from (Mital et al. 1999). The results of both the statistical testing and the hypothesized examples provide further evidence of the variability of the ratio between uninsured and

12 Estimating uninsured costs of work-related accidents Ratio C u /C i Figure 5. Construction crew Service crew Software team Manufacturing line Ratio intervals for different types of injury. insured costs. Uninsured costs are lower than insured costs for the most severe injuries. Insured costs are determined by the cost of medical treatment and the extent of the employee s absence. Uninsured costs are determined by the impact that personnel absences have on the rest of the organization. Logarithmic functions are commonly used in quantitative risk assessment (Schubach 1995, Binkowitz and Wartenberg 2001). The most common relationship is the cumulative probability function of n fatalities or the FN-curve, which is used to set limits for hazardous activities (Jonkman et al. 2003). The equation FðnÞ ¼Fð1Þ=n a used by Hirst (1998) where a is an aversion multiplier, introduces a power relationship between the number of fatalities and the cumulative probability. Similarly, the model proposed here introduces a power relationship between the severity of the injury and the costs of accidents. Uninsured costs consist of administrative and operational losses. Administrative losses correspond to additional general expenses in medical supplies, transportation and employee replacement as a result of the accident. Operational losses correspond to lost labour hours for other employees directly or indirectly involved in the accident. Administrative costs are estimated based on a constant value. Operational losses are estimated based on the workforce exposed, which is represented by a matrix representing different salary levels. In addition, the model involves an estimate of the hours lost for the crew, the administration and the replacement. The equation of the model is used to estimate the cost of a single accident. However, it is possible to estimate an aggregated cost for a number of accidents, knowing the workforce exposed. Aggregation by teams (crew, department or business unit) can use the same equation since they share the same population. Increased health and safety expenditure can be justified not only by the human impact but also by the administrative and operational losses that the absence of an employee triggers. In fact, the US National Safety Council (1998b) stated that contractors are likely to under-estimate the financial advantages of accident prevention. The model does not work with non-quantifiable variables or hidden costs that are not carried by the employer. The costs involved are future cash flows for the

13 258 O. Paez et al. employer associated with the injury. Most employers assume these losses as part of the operational expenses, so they lose track of the event or events that are wasting resources in the company. Because of the constraints for employers to gather accurate data on an operational basis, the proposed model does not require any input beyond information known at the time of the accident; the type of injury and the population exposed. A pre-defined worksheet can provide real time estimates and no expert intervention is required as long as the input provided is reliable. It is important to realize that the generalization introduced by the model has some limitations. The model estimates values closer to zero for non-disabling injuries, which is contrary to the fact that any type of injury will result at least in some sort of operational loss. On the other extreme, fatalities or acute disabilities might result in total interruptions that go beyond the estimations of the model. However, if the company has business interruption insurance which protects against losses resulting from a temporary shutdown because of fire or other insured events the estimations will be closer for fatalities. The model uses three factors that are assumed to remain constant: an estimate of administrative costs, an estimated of production losses and hourly wages by occupation. Those factors should be corrected for future years using up-to-date prices and wages. In fact, those values can be further customized if required by the company. 5. Conclusions The proposed model does not work with specific ratios based on the insured costs of accidents. Rather, it uses a logarithmic distribution and point estimates to calculate the uninsured costs of any type of injury. Even though the cost estimation is a function of severity of the injury, the model does not translate into a fixed ratio with respect to uninsured costs. Organizational losses are not associated with the medical compensation expenses that constitute the insured costs of accidents. The proposed model introduces an incidence-based approach which clearly explains the consolidation of the costs of accidents. An aggregated estimation can still be obtained by adding the incidence-based statistics. The value of the proposed model is to measure the impact of safety initiatives at the workplace, to support additional investment in infrastructure or safety design considerations. Accordingly, the financial impact of safety programmes can be justified if the frequency of the injuries is reduced. Further research should validate the proposed model with a sample across industries to measure statistical significance of severity and occupation as predictors of the overall cost. References Binkowitz, B. and Wartenberg, D., 2001, Disparity in quantitative risk assessment: a review of input distributions. Risk Analysis, 21, pp Coleman, M. and Marks, H., 1999, Qualitative and quantitative risk assessment. Food Control, 10, pp

14 Estimating uninsured costs of work-related accidents 259 Heinrich, H., 1959, Industrial Accident Prevention (New York: McGraw-Hill Book Co.). Hirst, I., 1998, Risk assessment. A note on F-n curves, expected numbers of fatalities, and weighted indicators of risk. Journal of Hazardous Materials, 57, pp Jonkman, S., van Gelder, P. and Vrijlinkg, J., 2003, An overview of quantitative risk measures for loss of life and economic damage. Journal of Hazardous Materials, 99, pp Leigh, J., Markowitz, S., Fahs, M. and Landrigan, P., 2000, Costs of Occupational Injuries and Illnesses (Ann Arbor, MI: The University of Michigan Press). Marquis, S., 1992, Economic consequences of work-related injuries. Statement Prepared for the Senate Committee on Labor and Human Resources, CT-103 (Santa Monica, CA: RAND). Mital, A., Pennathur, A. and Kansal, A., 1999, Nonfatal occupational injuries in the United States Part I overall trends and data summaries. International Journal of Industrial Ergonomics, 25, pp Oxenburgh, M., Marlow, P. and Oxenburgh, A., 2004, Increasing Productivity and Profit Through Health & Safety (Boca Raton, FL: CRC Press). Schubach, S., 1995, Comparison of probit expressions for the prediction of lethality due to toxic exposure. Journal of Loss Prevention in the Process Industries, 8, pp Simonds, R. and Grimaldi, J., 1963, Safety Management; Accident Cost and Control (Homewood, IL: Richard D. Irwin, Inc.). Sun, L., Paez, O., Lee, D., Daraiseh, N. and Salem, O., 2005, Estimating the uninsured costs of work-related accidents, part I: a systematic review. Theoretical Issues in Ergonomics, 7, pp Tweeddale, M., 2003, Managing Risk and Reliability of Process Plants (Burlington, MA: Elsevier Science). US Bureau of Labor Statistics, 2003, National salaries and wages survey. NCS Available online at: visited 1 November US Bureau of Labor Statistics, 2004, Employer costs for employee compensation. CM (Q1-2/2004). Available online at: visited 1 November US National Safety Council, 1998a, Administration and programs. Accident Prevention Manual for Industrial Operations, Vol. 1 (Chicago, IL: R.R. Donnelley & Sons). US National Safety Council, 1998b, Engineering and technology. Accident Prevention Manual for Industrial Operations, Vol. 2 (Chicago, IL: R.R. Donnelley & Sons). About the authors Omar Paez earned his BSc in Xavier University at Bogota, Colombia. He has worked in the implementation of performance measurement systems in Latin America. He has developed quantitative models for inventory allocation, routing and forecasting. Since 2002, he joined Dr Genaidy s research programme in human performance measurement and now is pursuing a PhD in Industrial Engineering. He has developed object-oriented algorithms to describe human performance variables. His research is now focused on the impact of human resources management in organizational performance. Troy K. Uahinui obtained his BS in Mechanical Engineering from Rose-Hulman Institute of Technology. He is currently working as project manager of Safety & Health in the Corporate Strategy Division at Toyota Motor Manufacturing North America in Erlanger, Kentucky. He has 5-years of experience in Safety Management and his current research interest is in the financial assessment of safety initiatives. Ash M. Genaidy holds degrees in engineering (BS, MS, PhD) and epidemiology (PhD). He is an Associate Professor of Industrial & Manufacturing Engineering

15 260 O. Paez et al. and Environmental Health and is the Deputy Director of the Safety & Health Engineering programme in the NIOSH-Sponsored Cincinnati Education and Research Center, all at the University of Cincinnati. Dr Genaidy s teaching and research focus on ergonomics, safety and health engineering, statistics, epidemiology and engineering economy and entrepreneurship. Dr Genaidy is the author or co-author of over 80 peer-reviewed papers published in international journals including Ergonomics, Human Factors, Applied Ergonomics, Theoretical ssues in Ergonomics Sciences, American Journal of Industrial Medicine, Journal of Occupational Medicine, Spine, Occupational and Environmental Medicine and Journal of Safety Research. He is an Associate Editor for Theoretical Issues in Ergonomics Sciences, has edited a book on Computer-Aided Ergonomics and is the recipient of the Outstanding Young Engineer of the Year by the Engineering Society of Detroit. Liaoming Sun has been a Graduate student of the Civil and Environmental Engineering Department, University of Cincinnati, since He got a Bachelor s degree of Civil Engineering from Guangxi Institute of Technology, China, in He worked in Liuzhou Municipal Construction and Landscape Company from His recent research focuses on Construction safety management. He won a 2004 Summer Graduate Student Research Fellowship of the University of Cincinnati and University Graduate Scholarship of University of Cincinnati from Waldemar Karwowski, ScD PhD, PE, CPE is Professor of Industrial Engineering and Director of the Center for Industrial Ergonomics at the University of Louisville, Louisville, Kentucky, USA. He holds an MS (1978) in Production Engineering and Management from the Technical University of Wroclaw, Poland and a PhD (1982) in Industrial Engineering from Texas Tech University. He received an Honorary Doctorate in Science (Doctor Honoris Causa) from the South Ukrainian State KD Ushynsky Odessa Pedagogical University of Ukraine (May 2004) for his outstanding contributions to the fields of human factors engineering and management of work systems. Dr Karwowski serves as Editor of Human Factors and Ergonomics in Manufacturing and Editor-in-Chief of Theoretical Issues in Ergonomics Science (TIES). Nancy M. Daraiseh is a research assistant professor at the University of Cincinnati in the Department of Industrial and Manufacturing Engineering. She received her PhD in Industrial Engineering with a focus on occupational safety/ergonomics from the University of Cincinnati. Dr Daraiseh holds a BS in Electrical Engineering from The Jordan University of Science & Technology and an MS in Industrial Engineering from the University of Cincinnati. Her research interests include: workplace assessment focusing on the identification of risk factors (physical, psychological and organizational) impacting various outcomes (musculoskeletal symptoms/injuries, ageing, stress, performance, job satisfaction/dissatisfaction) as well as intervention studies.