Page 301

8


Interventions in the Workplace

This chapter examines the hypothesis that prevention strategies in the workplace can reduce the incidence and impact of musculoskeletal injuries, illnesses, and disorders. Primary prevention occurs when the intervention is undertaken before members of the population at risk have acquired a condition of concern, for example, educational programs to reduce the number of new cases (incidence) of low back pain. Secondary prevention occurs when the intervention is undertaken after individuals have experienced the condition of concern, for example, introduction of job redesign for workers with symptoms of carpal tunnel syndrome. Tertiary prevention strategies are designed for individuals with chronically disabling musculoskeletal disorders; the goal is to achieve maximal functional capacity within the limitations of the individual's impairments.

In this chapter we review the scientific data about interventions in the workplace particularly related to primary and secondary prevention. We have not addressed specific medical interventions, such as drugs, manipulative therapy, and surgery. The scientific bases for medical treatment of musculoskeletal conditions, particularly back pain, have been reviewed in detail by a number of panels of experts, such as the Quebec Task Force and the Agency for Health Care Policy and Research (AHCPR). Although there is significant overlap between secondary and tertiary prevention strategies, we have not addressed the extremely complex issues regarding the medical management of or workplace accommodations for those with chronic musculoskeletal disabilities.

In tertiary prevention the accommodations are usually made on a case by case basis. There often are complex psychosocial issues that may involve a wide array of treatment options, and the medical treatment



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 301
Page 301 8 Interventions in the Workplace This chapter examines the hypothesis that prevention strategies in the workplace can reduce the incidence and impact of musculoskeletal injuries, illnesses, and disorders. Primary prevention occurs when the intervention is undertaken before members of the population at risk have acquired a condition of concern, for example, educational programs to reduce the number of new cases (incidence) of low back pain. Secondary prevention occurs when the intervention is undertaken after individuals have experienced the condition of concern, for example, introduction of job redesign for workers with symptoms of carpal tunnel syndrome. Tertiary prevention strategies are designed for individuals with chronically disabling musculoskeletal disorders; the goal is to achieve maximal functional capacity within the limitations of the individual's impairments. In this chapter we review the scientific data about interventions in the workplace particularly related to primary and secondary prevention. We have not addressed specific medical interventions, such as drugs, manipulative therapy, and surgery. The scientific bases for medical treatment of musculoskeletal conditions, particularly back pain, have been reviewed in detail by a number of panels of experts, such as the Quebec Task Force and the Agency for Health Care Policy and Research (AHCPR). Although there is significant overlap between secondary and tertiary prevention strategies, we have not addressed the extremely complex issues regarding the medical management of or workplace accommodations for those with chronic musculoskeletal disabilities. In tertiary prevention the accommodations are usually made on a case by case basis. There often are complex psychosocial issues that may involve a wide array of treatment options, and the medical treatment

OCR for page 301
Page 302often involves the management of failed surgical procedures, the utilization of pain clinics, functional restoration programs, and controversial drug regimens and surgical procedures. The panel examined the broader systems approach, dealing with primary and secondary prevention, and thought these more applicable to the interface between job demands and the population of workers. This chapter is organized as follows: First, we revisit the conceptual model ( Figure 1.2) that forms the framework for workplace interventions and discuss the related principles of ergonomics. We then summarize the intervention literature specific to the back and upper extremities. Practical considerations, including assessment of cost-outcome effectiveness, are included, particularly as a framework for future research. Finally, we draw conclusions based on the weight of current evidence and suggest efforts to coordinate the gathering and sharing of further information to assess and prevent musculoskeletal disorders in the workplace. CONTEXT The conceptual model described in Figure 1.2 illustrates how the complex relationship among external loads, organizational factors, individual factors, and the social environment of work can lead to adverse outcomes that include pain, impairment, and disability. This conceptual model is also useful to understand how workplace interventions are used to control potentially adverse conditions. Primary and secondary interventions may prevent adverse outcomes by reducing or eliminating external loads, changing organizational factors, altering the social environment, improving individual stress-coping skills, or matching the physical demands of the job with the employee's physical capacities. The literature suggests that some of these approaches are more successful than others. Some interventions have not yet been adequately assessed. External loads in the work environment act on the body to create internal loads on tissues and other anatomical structures. Interventions that focus on the reduction or elimination of exposure to external loads must first identify and quantify the motions and forces acting on the individual, including vibration and thermal exposures. Often a systematic study of the work is required to evaluate these physical exposures and their characteristic properties. When specific physical stress factors are identified, the sources of these loads are ascertained. Workplace redesign may include alterations in tools, equipment, workstations, materials handled, tasks, work methods, work processes, and work environment, based on their contributions to the identified stresses. The majority of the intervention literature is based on this approach to injury reduction.

OCR for page 301
Page 303 Work organizational factors broadly consist of job content and organization characteristics, as well as temporal and economic aspects of the work and task. Job content is the array of tasks and procedures that make up an employee's workload. These requirements have a direct effect on the exposure to external loads associated with the use of equipment and tools. Other aspects of job content may include specifications to handle certain objects, operate machinery, or work under potentially adverse environmental conditions. Interventions in job content may reduce or eliminate exposure to physical stresses by directly altering the job requirements. Examples include assigning a worker to different tasks or eliminating a specific operation through automation. Organizational characteristics describe the management structure of the organization and the level of autonomy an employee has when performing a job. These factors may affect employees' attitudes about their work or influence physical stress exposure. Interpersonal relationships among employees and supervisors may also influence physical stress exposure. For example, physical stresses in a cooperative work environment often are reduced when employees informally assist one other. Temporal aspects include job scheduling such as shift work, number of hours on deadline, job rotation, or the frequency and duration of specific tasks. For example, the continuous performance of a monotonous, repetitive task has been associated with physical and psychological stress. Finally, the economic and compensation policies of a company can affect physical exposures. For example, overtime and extended work can increase the daily duration of exposure to musculoskeletal stressors. Compensation incentives also may affect intensity, frequency, and duration of work and may discourage rest breaks. PRINCIPLES OF ERGONOMICS The application of ergonomic principles forms the basis for much of the intervention literature. Ergonomics is the study of work, including workplace interventions to establish compatibility among the worker, the job, and the work environment. Ergonomics professionals, both researchers and practitioners, reflect the variety of factors that affect safety and productivity in the workplace; the disciplines involved include researchers and practitioners in, for example, medicine, epidemiology, psychology, and industrial engineering and other health-related, technical/ engineering, and behavioral disciplines. The process whereby ergonomics is applied to intervention adheres to the scientific method: available data are gathered and analyzed (e.g., through broad or local surveillance and through analysis of jobs); hypotheses are developed (e.g., specific engineering controls are proposed to address specific factors or condi-

OCR for page 301
Page 304tions of the physical workplace and job characteristics; administrative controls are proposed to address characteristics of the organization and management thought to be relevant; and such individual modifiers as age, gender, body mass index, smoking habits, comorbidity, and reaction to the psychosocial environment at home and at work are addressed); the hypotheses are tested (i.e., the effectiveness of the intervention is assessed); and the hypotheses are maintained (for effective interventions) or refined (an intervention is improved or replaced) in response to the outcomes. Emerging evidence indicates that work-related musculoskeletal disorders and associated disability are the consequence of a complex interplay among medical, social, work organizational, biomechanical, and workplace and individual psychosocial factors. As with most areas of scientific endeavor, the ergonomics of intervention in the workplace is providing an expanding knowledge base and an evolving set of interventions, and continued development of innovative models is needed to allow for the effective management and prevention of these complex disorders. The practice of ergonomics relies on a process that tailors interventions to specific circumstances currently found effective, continues to assess the effectiveness of these interventions in the face of changing workplace and worker factors, and evaluates new interventions. It is therefore neither feasible nor desirable to propose generic interventions expected to apply to every industry, job, and worker, nor “once for all” interventions whose effectiveness need not be regularly assessed. It is, rather, both feasible and desirable to encourage application—with continued assessment—of interventions found useful and promising to date, and to encourage cooperative endeavor and information exchange among researchers, practitioners, and workers/managers in industry/labor, government, and academia. Surveillance and job analysis are critical components of intervention research and practice. Surveillance Surveillance is the ongoing, systematic collection, analysis, and interpretation of health and exposure information in the process of describing and monitoring work-related musculoskeletal disorders and evaluating the effectiveness of the program. Surveillance may include the analysis of existing medical records, worker compensation records (passive surveillance), or the surveying of workers with questionnaire and physical examination (active surveillance). The analysis of existing records is generally less costly, but the reliability of such data is difficult to assess. Standardized questionnaires and physical examinations can be as sensitive as the use of unusually thor-

OCR for page 301
Page 305ough existing occupational medical records; however, it is unclear whether the additional cost of an active surveillance system will deter the routine use of such systems (Fine et al., 1986). Symptoms questionnaires and checklist-based hazard surveillance are feasible within the context of joint labor-management ergonomics programs and are more sensitive indicators of ergonomic problems than preexisting data sources (Silverstein et al., 1997). Surveillance systems are key methods of collecting information that describes the characteristics of workers (individual factors) that may interact with workplace factors to cause or mitigate musculoskeletal disorders or to affect the success of interventions. Surveillance systems may also be used to evaluate the effectiveness of interventions. In addition, surveillance systems (e.g., employee surveys, absentee and turnover analysis) can be used to gather information that complements that of other analytical procedures (e.g., job analysis, work flow analysis, productivity study) in defining relevant features of tasks, jobs, worker satisfaction, and organizational culture. Surveillance is most important in times of rapid change in the economy and when resources for prevention may be limited (Fine, 1999). Surveillance data may be collected and analyzed at different levels of task, job, occupation, and industry (see Chapter 2), and they may elucidate features of the physical work environment, work tasks, work organization, psychosocial environment, and worker characteristics relevant to intervention (see Figure 1.2). Intervention can be indicated by case, data, and risk factors identified through surveillance. Case-initiated interventions are triggered by employee reports of symptoms, concerns, or recommendations. An intervention is data-initiated when it is the outcome of an analysis of medical records and injury reports. An intervention is risk-factor initiated when factors known to be significantly related to musculoskeletal disorders are shown to be present. Surveillance systems at the company level serve as an invaluable resource for evaluating the impact of efforts to reduce risk and to find new risks early, before they cause irreparable harm. For example, in an electronics company in Norway, new cases of sick leave attributed to musculoskeletal disorders were tracked over a number of years. Regular examination of the trends led the company to recognize an increase in the sick leave rates and prompted it to introduce a number of improvements in the work setting. The surveillance of sick leave also served to document the impact of the interventions, showing reductions in the sick leave following the changes. Figure 8.1 (from Westgaard and Aaräs, 1985) shows the percentage of lost work days, due to musculoskeletal disorders, of total work force in the electronics company. Record analyses may also reveal trends across jobs, operations, or other common conditions. Inter-

OCR for page 301
Page 306 ~ enlarge ~ FIGURE 8.1 Percentage lost work days due to musculoskeletal disorders of total work force in an electronics company (based on data from Westgaard and Aaräs, 1985). ventions may be initiated when risk factors are identified from proactive job surveys or other risk factor reviews. Job Analysis When it is determined, based on the clinical evaluation and exposure information, that an employee has a work-related disorder, the employer performs an analysis of the employee's job or of a sample of representative jobs. Job analysis is a procedure used to identify potential exposure to risk factors and to evaluate their characteristic properties. When risk factors are identified, job design or redesign, including engineering or administrative controls, is used for eliminating or reducing work-related risk factors or exposure to those factors. Job analysis consists of a work methods analysis, based on traditional techniques of time-and-motion study to determine the work content of the job, and a systematic analysis of risk factors (Armstrong et al., 1986).

OCR for page 301
Page 307Work methods analysis involves describing the work content of each task as a series of steps or elements and identifying risk factors for each work element. After the major physical stresses have been identified, ergonomic interventions may be applied to minimize exposure to risk factors. Such interventions are often categorized as engineering controls, administrative or production controls, or personal modifiers. Although risk factors are frequently cited in the literature, acceptable exposure levels have not yet been determined for all individual and combined factors. The presence of a risk factor does not necessarily mean that the person doing the job is at excessive risk of an injury or that ergonomic changes are worth their cost. When the presence of risk factors is combined with a history of musculoskeletal disorders among persons doing that job, the risk may be considered excessively high and ergonomic interventions may be worth their cost. In many cases, it is desirable to make changes regardless of the history, because job changes are often inexpensive, and injuries are often very expensive. This is particularly the case when jobs are being initially designed and set up. Examples of Process-Based Ergonomics Intervention Programs The Occupational Safety and Health Administration has issued a standard of practice for assessing and addressing musculoskeletal disorders in the workplace. The American National Standards Institute has accredited a national committee (Accredited Standards Committee, 2000) charged with developing a standard process or program for reducing musculoskeletal disorders associated with work. Within industry, Ford Motor Company, for example, uses a team-based approach at each facility, establishing ergonomic committees consisting of labor and management representatives who collect and analyze local data (reviewing surveillance records and medical cases), propose interventions, and assess their effectiveness (Joseph, 2000). Though not targeting specifically the issue of musculoskeletal disorders in the workplace, the Department of Defense has demonstrated for decades that the elements of an effective ergonomics program can be formally defined in broad scope, and that these elements can be successfully tailored to the needs and features of specific workplaces and workers (see U.S. Army Missile Command, no date). We note here that each of these approaches to ergonomics interventions is process-based (i.e., applies a process of analysis, implementation, and assessment of effectiveness), and that any ergonomics program benefits when the scientific method is applied by researchers and practitioners in those programs who collect and analyze data, propose and implement interventions, and assess their effectiveness.

OCR for page 301
Page 308 SPECIFIC INTERVENTIONS: BACK Few high-quality scientific intervention studies related to the primary and secondary prevention of low back pain are available in the literature. The vast majority of the literature consists of retrospective analyses and case reports of intervention effectiveness, better assessed through prospective studies and, if possible, randomized controlled trials. In order to provide insight as to the effectiveness of interventions in controlling the risk of back pain, the reviewed literature was categorized in three ways. First, high-quality reviews of the literature were evaluated for common indications and trends. Second, since most of these efforts consisted of reviews dating back to the mid-1990s, a review of the literature since that time was performed using similar quality criteria. Finally, a review of reported “best practices” was performed to assess how reported experiences in industry relate to the more rigorous controlled studies of interventions. Collectively, this approach has provided insight into the effectiveness of low back interventions. Results of High-Quality Reviews Through the Mid-1990s Six high-quality reviews of the literature were identified that assess the state of knowledge associated with the ability of ergonomic interventions to influence low back disorders. Based on its criteria for quality, this assessment selected reviews that accepted only studies that met rigorous methodological constraints and reviews that comprehensively examined the basic science behind intervention logic. Six reports met these criteria. The first review, performed by Westgaard and Winkel (1997), assessed 91 studies that met seven inclusion criteria. By these criteria each reviewed study included: (1) field intervention; (2) involvement of a primary or secondary intervention; (3) an exposure-effect analysis; (4) sufficient documentation for repeatability; (5) a unit of analysis that was the entire job (not a task); (6) outcome analyses that included internal exposure, acute responses, and/or health effects; and (7) documentation in English. In addition to these inclusion criteria, this review selected interventions based on seven quality criteria. These quality criteria included: (1) proper statistical analyses, (2) reasonable study group size, (3) variables that were generalizable to other settings, (4) consideration of reliability and sensitivity of variables, (5) inclusion of a control group, (6) an adequate observation period with follow-up measurements, and (7) proper documentation of the intervention process. The inclusion criteria of the second review (van Poppel et al., 1997) accepted studies that included: (1) a controlled clinical trial that is prospective in nature in which the intervention group is randomly assigned

OCR for page 301
Page 309and compared with a randomly assigned control group derived from the same setting as the intervention group; (2) intervention intended to prevent back pain and consisting of education, exercise, or lumbar support; (3) no restrictions on subjects at the start of the study, and (4) intervention occurring in an industrial setting. A total of 14 studies met these inclusion criteria. The third review (Volinn, 1999) closely examined six studies that based their interventions on various principles. These principles included promotion of back safety, back belt use, introduction of ergonomic devices, and back strengthening exercises. By Volinn's inclusion criteria, these studies contained: (1) use of a primary prevention, (2) a dependent measure consisting of a recordable back injury, (3) only workplace-related injuries, and (4) successful outcomes. In the fourth review, Kaplansky and colleagues (Kaplansky, Wei, and Reecer, 1998) evaluated 133 studies of occupational low back pain prevention. Their review categorized the interventions by type and collectively assessed the literature within each intervention type. The inclusion criteria considered the state of the body of knowledge associated with each intervention type, focused on the intervention logic, and used selection criteria that varied across the intervention types. The literature reviewed did not consist only of field studies. The assessment was generally of a critical nature. A fifth review (Scheer, Watanabe, and Radack, 1997) searched 4,000 articles but accepted for review only randomized, controlled trials of interventions: 12 studies met this criterion. The sixth review (Hsiang, Brogmus, and Courtney, 1997) summarized the results of 70 biomechanical and physiological publications. Although the review did not focus on field interventions, it examined the logic behind the advantages and disadvantages of lifting techniques. This represents an alternative assessment of a low back pain intervention approach. Collectively, the data in these six reviews indicate that certain engineering controls (e.g., ergonomic workplace redesign), administrative controls (specifically, adjusting organizational culture), addressing modifying individual factors (specifically, employee exercise), and the inclusion of a combination of interventions are the only strategies that have been shown to be positively associated with the reduction of work-related low back pain. An important finding in these studies is the observation that multiple interventions that actively involve workers in medical management of workers at risk, physical training, and technique education, in combination, can also be effective in controlling risk. Isolated approaches such as control of the production process, use of back belts, training, relaxation, physiotherapy, health education, and worker selection either have not been proven effective or have proven ineffective.

OCR for page 301
Page 310 Results of Recent High-Quality Studies The high-quality reviews discussed above evaluated the back intervention literature up until the mid-1990s. A search of the back intervention literature published since that time identified seven additional high-quality studies. All but one of these studies involved randomized controlled designs; the remaining study was of a case-control design. These studies are summarized in Table 8.1. Three of the studies were not necessarily related to work. Only two of the studies addressed primary interventions, evaluating the effectiveness of back belts, health education, and back schools: none of these interventions was found to be effective. These findings are consistent with the literature reviews discussed in the previous section. The findings of the six secondary intervention studies were also in agreement with the literature reviews described above in reporting that engineering controls, exercise (two of three studies), and multiple interventions are effective. The other positive findings for secondary interventions consisted of functional restoration (one study), physiotherapy (one study), cognitive therapy (one study), medical management (two studies), and light duty (one study). Conflicting results were found for back schools (positive findings in two of four studies) and health education (positive results in one of two studies). Back belts and pain management approaches to interventions were found to be ineffective. Best Practices Since practical experience with interventions can also be of use in interpreting the effectiveness of ergonomic interventions, we reviewed the literature of best practices interventions to see how it compared with the more rigorous scientific studies. To this end, we reviewed the transcripts from the Effective Workplace Practices and Programs Conference held in Chicago in 1997. A total of 13 interventions relating to the control of low back pain were found; these are reported in Table 8.2. These best practices illustrate the type of intervention approaches that are often applied by industry. As the table shows, many of these approaches are the same as those identified in the high-quality reviews and high-quality studies previously discussed. Table 8.2 shows that engineering controls are one of the most commonly used interventions in industry. Many of the industries that employ engineering controls as an intervention also report a positive effect on low back pain control. The second most common control is training, which is typically applied along with engineering controls. Many of the industries reported positive control of low back pain through training. These data

OCR for page 301
Page 311 TABLE 8.1 Review of High-Quality Studies of Back Interventions Since 1995a ~ enlarge ~ a Plus (+) sign indicates positive finding of effectiveness for the given intervention; null (0) sign indicates that the intervention was not found to be effective; empty cell indicates that the intervention was not assessed.

OCR for page 301
Page 319 Duration Control co-variates Outcome measures Positive Effect? 1 year Yes Hand/arm symptom Yes Neck/shoulder/back symptoms Yes 2 year No Neck/shoulder pain Yes Forearm/hand pain No 3 months Yes Hand pain Yes Physical examination Yes Nerve conduction No 6 months Yes Hand/arm pain Yes Physical examination No 6 months Yes Shoulder/forearm/wrist pain Yes Sick leave Yes 18 months No Sick leave Yes Turnover Yes Production costs Yes 6 months Yes Neck/upper extremity pain Yesa Physical examination Yesa 6 months Yes Neck pain No Neck range of motion No Sick leave No 1 year Yes Musculoskeletal complaints Yes Physical examination Yes Lost time No 1 year Yes Neck shoulder discomfort No Studies of the effects of stretching exercises provide only suggestive results. Moore and Garg (1998) found, in a pre-/post-test study of 60 workers, that participants in a stretching exercise program achieved, after 36 sessions over 2 months, greater physical flexibility. Conditioning exercise programs may make participants feel better about themselves (Silverstein, 1988; Moore and Garg, 1998) and may encourage workers to attend to physical activity.

OCR for page 301
Page 320 Taken together, these studies generally demonstrate a benefit of changes in workstation design, tool design, or training on upper extremity symptoms. The effect on related physical examination findings is less consistent. The value of exercise in reducing musculoskeletal symptoms has found mixed support. Best Practices Table 8.4 summarizes the best practices reported by industry representatives at the 1997 conference on effective workplace practices and programs relating to the management and control of upper extremity work-related musculoskeletal disorders. The table reveals a trend similar to that found for the interventions previously described for low back pain. Interventions for upper extremity and neck work-related musculoskeletal disorders involving engineering controls, administrative processes and personal modifiers resulted in positive outcomes for 27 companies. In particular, engineering controls and administrative controls, such as periodic rest breaks and job rotation, were most frequently accompanied by employee involvement, management support, and training. All programs included either engineering or administrative controls in addition to personal modifiers (such as exercise, training, light duty, or medical management). None of the reports demonstrated that personal modifiers were successful by themselves. Companies that instituted exercise or stretching programs also employed engineering or administrative controls. These findings are consistent with the approaches discussed in the high-quality reviews of back interventions in Table 8.1. CHALLENGES DURING THE PROCESS OF ERGONOMIC INTERVENTION Those who develop and implement interventions that address the complexity of factors potentially affecting musculoskeletal disorders in the workplace face challenges during the phases of gathering information, applying it to implement interventions and assess their effectiveness, and disseminating both research and practical knowledge. Given the magnitude and the complexity of the problem of musculoskeletal disorders in the workplace, in response to these challenges it would be beneficial to coordinate the efforts of researchers and practitioners in industry, labor, government, and academia. Collection of Information: Surveillance Challenges to the collection of information exist in the areas of methods and measures for surveillance. Surveillance of information pertinent

OCR for page 301
Page 321to interventions for musculoskeletal disorders in the workplace can be improved by the development of more reliable, valid, sensitive, and specific medical diagnostic tests for relevant musculoskeletal disorders, as well as development of specific criteria that can be used for surveillance for determining cases of musculoskeletal disorders. Criteria should include, separately or in combination as appropriate, systematic methods for the collection of symptom information, minimum physical examination test findings, or physiological measurements. Criteria should also be relevant to specific purposes (for example, surveillance for early identification versus diagnosis for treatment). In addition, effective surveillance would be facilitated by further development of self-report instruments (e.g., by the application of computer technologies). Collection of Information: Research Pertaining to Developing Interventions Challenges to the collection of information also exist in the areas of methods and measures for research on ways to improve interventions and more effectively tailor them to the needs of specific workers, jobs, and industries. The development of interventions would benefit from more research to better quantify exposure limits, including cumulative risk for exposure, to apply them to job design to prevent injury, and to use such data as part of the effort to improve biomechanical models for predicting risk of musculoskeletal disorders in designing jobs. Furthermore, useful research would also define the most effective methods for applying multifactorial interventions (e.g., engineering controls, administrative controls, individual modifiers) and for reducing exposure for targeted industries (e.g., better ways of turning and lifting patients, better hand tools that reduce vibration). A continuing program of cooperative research would provide further understanding of the features of specific interventions that best account for their effectiveness (e.g., elucidating the characteristics of relevant work organizational factors, assessing the effect of job rotation as an intervention, and clarifying the mechanisms whereby psychosocial factors, individual worker characteristics, and organizational culture interact with interventions. Assessing Interventions The development of the repertoire of effective interventions would be facilitated by an infrastructure that supports standardization of test instruments, standardized reporting, and agreed-on outcome measures that would assist researchers and practitioners to assess the effectiveness of interventions and to define the contexts within which specific inter-

OCR for page 301
Page 322 TABLE 8.4 Case Study Best Practices Observations—Upper Extremities and Neck Focus Personal Modifiers Environment/Industry Engineering Controls Administrative/ Production Exercise/ Stretching Large Manufacturing 3M • • Process PPG • • Process Hay & Forge • • Process Ford • • Process Samsonite • • Process Frito-Lay • • Process/Rotation Construction Bechtel • Periodic breaks • TDC • • Apparel Red Wing Shoe • • Rotation • Sequins International • Fieldcrest-Cannon • • Process Sara Lee Knit Products • • Process Small Manufacturing Rocco Enterprises • • Process/Rotation • Charleston Forge • • Process/Rotation Lunt Silversmiths • Perdue, Inc. • • Process/Rotation Woodpro Cabinetry • Rotation Farmland Foods Maritime Bath Iron Works • • Process/Rotation • Newport News Shipbuilding • • Process Utilities PG&E • • Montana Power • • Rest breaks • Warehousing Murphy Warehousing Co. • • Process Office 3M • Process American Express Financial Advisors • • Process L.L. Bean • • Process US West • • Process

OCR for page 301
Page 323 Light Duty Multiple Interventions Employee Involvement Management Support Training Medical Management Positive Impact Reported • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

OCR for page 301
Page 324 ventions are found to be effective. These methods and measures would also be beneficially applied to assessment of the adjuncts to interventions, including training programs and aids (for workers, managers, researchers, and practitioners), work standards on musculoskeletal disorders and prevention, and manuals for interventions and for ergonomics programs. There is a need, addressed below, for the development of standardized methods for cost-effectiveness analysis of interventions. Cost-Outcome Effectiveness Analysis Public and private policy makers, managers, and other leaders and decision makers are faced with the practical challenges of allocating limited resources. They must therefore be informed of the ratio between a program's marginal monetary costs and the marginal effectiveness, expressed both in terms of monetary savings and marginal improvements in health outcomes. Under such rubrics as prevention effectiveness (Haddix et al., 1996) and economic evaluation (Goossens et al., 1999), a set of analytical techniques is available to address questions of cost outcome. These techniques include: cost-benefit analysis, which expresses both costs and benefits in monetary terms; cost-effectiveness analysis, which expresses outcomes in terms that may not be monetary (e.g., lost work days, reported pain); and cost-utility analysis, which expresses outcome in the common form “quality adjusted life years.” Despite their availability, these techniques have not been widely applied to the study of interventions to relieve or prevent musculoskeletal disorders. Therefore, it is not possible, given existing studies, to provide a systematic summary, or comparison across programs, of the cost-outcome effectiveness of such interventions. Few studies of the effects of interventions for musculoskeletal disorders describe the costs associated with the intervention. Studies of interventions for musculoskeletal disorders employ a wide variety of outcome measures, and methods for determining cost also vary across studies that consider costs. Cost and outcome factors can differ widely, depending on the level of intervention (primary, secondary). In fact, most reports of cost-outcome effectiveness for such interventions provide only anecdotal or case data. Goossens et al. (1999) noted that three recently performed reviews of the economic evaluation of interventions for musculoskeletal pain pointed to the impossibility of reaching confident conclusions, because of the lack of clarity and low quality of the costing methodologies employed. There are several key methodological issues associated with cost/outcome analysis in this area:

OCR for page 301
Page 325 1. Different interventions and the circumstances to which they are applied may result in different time profiles. Some costs may be spread over a long period of time, while others are incurred at the beginning of the intervention; similarly, some outcomes appear quickly, while others require long periods of time. 2. Reduction in cost related to a disorder can be deemed a benefit of the intervention (e.g., when an intervention results in a decrease in the prevalence of a disorder at a workplace). This is sometimes clearly demonstrable with a methodology that controls for the effects of competing—or no—interventions. Moreover, predicted reduction in cost is sometimes taken as a benefit (e.g., when an intervention is predicted to prevent cases of a disorder that will not be observed, because they have, by prediction, been prevented). Similarly, assumptions are sometimes made about the interrelationships of benefits to one another (e.g., the assumption that an outcome of improved job satisfaction will imply also an outcome of improved productivity). In all cases, care must be exercised to validate the relationships and predictions. 3. Costs reported under the same label can be determined by different accounting methods (e.g., overhead costs); similarly, outcomes described in similar terms can mask operationally different variables (e.g., job satisfaction, reported pain). The methodology whereby cost and outcome are determined should be precisely explained in any study that assesses cost-outcome effectiveness—a principle violated often in case reports and best practice description. 4. The results of cost-outcome effectiveness studies must be considered only one part of policy and decision analysis, which must also consider such factors as who pays what costs, who reaps what benefits, and what cost/outcome ratio is acceptable. One method of quantifying values is to measure the evaluator's willingness to pay for units of risk reduction. Despite the current dearth of rigorous, comprehensive cost-outcome effectiveness studies of interventions for musculoskeletal disorders, direction can be found in, for example, Haddix et al. (1996), who provide a general framework applicable to public health prevention initiatives, and Goossens et al. (1999), who propose a framework specific to musculoskeletal disorders. The combined framework (see Box 8.1) is augmented by factors described in existing cost-outcome studies for musculoskeletal disorders; this framework identifies types of costs and benefits that can be considered as part of future efforts toward the development of an accepted methodology that will permit comparison across alternative interventions.

OCR for page 301
Page 326 BOX 8.1 Framework for Cost-Outcome Effectiveness Research Examples of Cost Factors for Intervention/Prevention Programs Direct Health Care Costs Institutional inpatient care Institutional outpatient services Home health care Physician services Ancillary services (other professionals and volunteers) Overhead (facility, equipment, etc.) Medications Devices and appliances Diagnostic tests Rehabilitation Direct Nonhealth Care Costs Program development (analysis, program design) Training and education (personnel, materials, facilities) Program management Transportation Modification to workplace (time and cost to design, fabricate/purchase, implement) Indirect Costs: Wages/Time Lost productivity during implementation Psychosocial costs (adjustment, stress due to change) Examples of Beneficial Outcomes of Intervention/Prevention Programs Number of reported injuries Number of Workers' Compensation claims Amount of Workers' Compensation indemnity paid Reduction in lost days Reduction in absenteeism Reduction in recurrences Reduced number of worker replacements and reduced training of new workers Improved physical functioning Motivation Morale Job satisfaction Productivity Improved perceived quality of life Reduction in risky behaviors Reduced fatigue Improved knowledge of workers about safety

OCR for page 301
Page 327 Return to Work Return to work is one desirable outcome of interventions. Research that elucidates the factors that operate on behalf of return to work would further the understanding of the systematic description of factors presented in Figure 1.2. Such research would usefully address development of prognostic indicators of clinical outcomes; determination of how return to work is affected by functional capacities of the worker and by workplace accommodations; identification of barriers to timely access to health care and to filing workers' compensation claims; and identification of disincentives to compliance with suggested programs. Dissemination of Information There is also a challenge of disseminating—to workers, managers, and researchers and practitioners in government, industry, and academia— the surveillance, research, assessment, and effectiveness information discussed above. Features of an improved infrastructure to foster this dissemination would include regular meetings that bring together scientific and best practices experience; an intervention trials network; development and sharing of common tools, training programs, and standards—tailored to users—for implementing ergonomic programs; and mechanisms for disseminating validated best practices. Particularly helpful would be the expansion of education and training programs to assist workers and employers (especially small employers) in understanding and utilizing the range of possible workplace interventions designed to reduce musculoskeletal disorders. Expanding continuing education for a broad range of professionals concerning risk factors for musculoskeletal disorders—in and out of the workplace—and concerning interventions would also be useful. SUMMARY Musculoskeletal injuries in the workplace result from a complex interaction of mechanical stresses, individual host factors, and organizational policy and structure. Furthermore, due to the rapidly changing nature of work and the intense focus on production goals, it is extremely difficult to design and carry out intervention studies in the workplace, particularly those that are prospective, randomized, and controlled. When interpreting the results of existing intervention studies, there are important methodological considerations that must be taken into account to determine the validity of each study.

OCR for page 301
Page 328 There is a very large body of published literature about workplace interventions for the primary and secondary prevention of back and upper extremity work-related disorders. Despite the large number of studies published, few meet the strictest criteria for scientific validity. At the same time, there is a another body of information derived from quality improvement studies and best practices that reflect the practical experience within industry. To the degree that there is agreement between the more scientific literature and best practices, this congruence is important in establishing a weight or pattern of evidence. In applying this information, there are several ethical and cost-benefit considerations, as well as practical barriers particularly in optimizing secondary prevention strategies. The complexity of musculoskeletal disorders in the workplace requires a variety of strategies that may involve the worker, the workforce, and management. The literature shows that no single strategy is or will be effective for all types of industry; interventions are best tailored to the individual situation. Although we have no measure of their relative contributions, there are, however, some program elements that consistently recur within successful programs: management commitment, employee involvement, and directly addressing workplace physical and work organizational factors. The weight and pattern of the evidence supports the conclusion that primary and secondary prevention interventions to reduce the incidence, severity, and consequences of musculoskeletal injuries in the workplace are effective when properly implemented. The evidence suggests that the most effective strategies involve a combined approach that takes into account the complex interplay between physical stressors and the policies and procedures of industries. When the scientific information is combined with the very practical quality improvement data, the panel is persuaded that continued focus on primary and secondary prevention can reduce the incidence and severity of these widespread musculoskeletal conditions affecting the worker in the workplace. Specifically, the panel concludes that: 1. Interventions must mediate physical stressors, largely through the application of principles of ergonomics. 2. Employee involvement is essential to successful implementation. 3. Employer commitment, demonstrated by an integrated program and supported by best practices review, is important for success. 4. Because of limitations in the scientific literature, a comprehensive and systematic research program, supported by an infrastructure linking industry, labor, government, and academic efforts, is needed to further

OCR for page 301
Page 329 clarify and distinguish the features that make interventions effective for specific musculoskeletal disorders. 5. Although generic guidelines have been developed and successfully applied in intervention programs, no single specific design, restriction, or practice for universal application is supported by the existing scientific literature. 6. Study of the relationship between exposure to physical and psychosocial risk factors and musculoskeletal disorders has been constrained by inadequate techniques for quantitative measurement of “dose,” analogous to available measures for noise or chemicals. Existing measures are often based on self-report or qualitative metrics. New tools and research instruments are necessary to provide more reliable and valid exposure estimates; these are most important for the study of the effects, if any, at lower-level exposures and for evaluation of the possibility of interaction among different factors when more than one is present. Better measures of exposures and outcomes would permit more effective evaluation of interventions.