An in-depth research study of work-related musculoskeletal disorders in Washington state from 1990-1998 based on worker’s compensation claims.
|Differences from Previous SHARP Reports|
|Magnitude and Costs|
|Comparisons with Other Studies|
|High Risk Industries|
|Limitations and Strengths of Using Worker’s Compensation Data|
In the current study we looked at claims for general and selected specific hand/wrist, elbow, shoulder and back disorders. The focus was on non-traumatic soft tissue musculoskeletal disorders (NTST-MSDs). In previous reports, we referred to these as gradual onset MSDs vii . These NTST-MSDs, when caused or aggravated by work activities (for example exposures to frequent or heavy manual handling, awkward postures, forceful or repetitive exertions) are referred to as Work-related MSDs.
Differences from Previous SHARP Reports
For the first time we have combined the forearm region with elbow, the fingers with hand/wrist, the upper arm with shoulder. This is the most likely reason for increases in the average number of claims per year for these body regions.
Although we used the designation “Non-Traumatic Soft Tissue, ” in the May 1999 Report viii, we made some additional changes in the current report including:
- Exclusion of lower extremity disorders because after looking at coding for nature and type for a random sample of claims, we could not identify a workable coding scheme that would separate actual sudden onset injuries from more gradual non-traumatic disorders. This eliminated approximately 59,000 claims that would have been in the analysis.
- Exclusion of contusions (nature=160) even if they were associated with overexertion because we thought there must be a more acute traumatic component to them. This eliminated about 5,000 claims from the back, neck and upper extremity analyses.
- Exclusion of claims that might have had a body part other than those included in the upper extremity, neck, or back or combination of these in the current report. For example, the primary body part may be leg but the claimant also had epicondylitis. This eliminated about 23,000 claims
- Exclusion of upper extremity claims that had Nature=190 (Dislocation, herniation or pinched nerve) because we reasoned that those were more likely to be dislocation than herniation or pinched nerve. This eliminated approximately 2260 claims.
- Using ANSI z16.2 codes for non-traumatic with the specific diagnoses (sciatica, rotator cuff, epicondylitis and carpal tunnel syndrome).
Therefore, we eliminated approximately 89,000 claims from the analysis that we would have included if we had performed the same analysis as in the May 1999 report.
Magnitude and Cost
4.2.1 State Fund
Neck, back and upper extremity musculoskeletal disorders represent a significant cause of morbidity in the working population. Claims for these disorders represent 46% of all accepted claims in the State Fund. Non-traumatic Soft Tissue MSDs account for 26.4%. While 24% of all State Fund claims during this period resulted in 4 or more days of lost time (compensable), 36.6% of NTST-MSD claims (49.2% of Neck, 40.2% of back, 33.0% of neck-back combination and 35.7% of upper extremity NTST-MSD claims) resulted in compensable claims in the State Fund. Overall, these NTST-MSD claims for neck, back and upper extremity musculoskeletal disorders in the State Fund had direct costs of $2.6 billion in 1998 dollars and 20.5 million lost workdays over the 9-year period. This is equivalent to loosing 82,000 full-time workers from the workforce over the nine-year period. While the claims incidence rates are decreasing for most but not all NTST-MSDs, the rate of decrease is less than for all claims excluding NTST-MSDs particularly for the upper extremity. There is no significant decrease for shoulder or elbow NTST-MSDs or the specific diagnoses of rotator cuff, or sciatica. Epicondylitis has increased. Carpal tunnel syndrome has been decreasing significantly but not as rapidly as all claims excluding NTST-MSDs (Figure 6). The percent of CTS cases resulting in surgery has decreased significantly. Severity (number of lost workdays per 10,000 FTEs) decreased for all claims approximately 17.3% per year, and compensable claims severity rate decreased 14.8% per year. Compensable NTST-MSD claims severity rates decreased about 10.4% per year (Table 35). Severity rates for sciatica decreased about 8.8% per year, for rotator cuff syndrome 5.4% per year, for epicondylitis about 6.5% per year, and for CTS about 7.4% per year.
For the Self-Insured, there were 80,230 compensable neck, back and upper extremity NTST musculoskeletal disorder new claims. It is unfortunate that we have no diagnoses, and the data on costs and lost time are incomplete.
4.3. Comparisons with Other Studies
The 1997 combined State Fund and Self-Insured compensable claims incidence rate for neck, back and upper extremity NTST-MSDs was 122.0 per 10,000 FTEs. This is slightly higher than the Washington 1997 BLS rate for overexertion and repetitive motion disorders (113.4 per 10,000 FTEs) which is limited to private industry only.
Surgical incidence of CTS in the Montreal adult populationix was 9 per 10,000, 19 per 10,000 for male manual workers and 18 per 10,000 for female manual workers, with 75% and 55% of all surgical CTS attributable to work. Food and beverage processors and material handlers were among those occupations at increased risk. Tanaka et alx estimated the overall prevalence of self-reported CTS among the general U.S. adult recently working population at 1.47 %. Using the same national survey data, Blanc et al xi reported a higher proportion of females reporting CTS (61%) than males. Work disability (loss or change in work) increased with an odds ratio of 1.5 per 120 minutes of hand bending (95%CI 1.2-1.7). Atroshi et al xii estimated a prevalence of 2.7% of a general adult population in Sweden had CTS symptoms with clinical or electrodiagnostically positive findings (2.1% for men and 3.0% for women). Using excessive hand force or wrist flexion/extension or using vibratory tools for more than one hour per day was associated with significantly increased prevalence of CTS. Using the same data source as Tanaka et al, Guo et al xiii estimated 22.4 million back pain cases among recent workers (17.9% prevalence) resulting in 149.1 million lost work days in 1988. Approximately 65% of back pain cases were attributable to occupational activities. Similar to our findings, the risks were greatest in construction and nursing occupations.
Although Webster and Snook xiv were unable to estimate incidence rates for upper extremity cumulative trauma disorders based on Liberty Mutual data, they identified 6,067 claims in 1989 for policy holders in 45 states with an average cost of $8,070 and median cost of $824. They then estimated the national cost to be $563 million. In a previous report (40-2-1999) for the same period, we identified 8,791 shoulder, elbow and hand/wrist lost time and medical-only claims with gradual onset in the Washington State Fund. Costs were comparable to the Liberty Mutual claims. The average cost of our 1989 claims ranged from $7,093-$8,250 and median cost ranged between $269- $332.
Using the average
costs and rates for the 1990-1998 State Fund data, and assuming there
are 96 million workers in the US (number covered by OSHAct) with an overall
incidence rate of 116 per 10,000 FTEs for all upper extremity NTST-MSDs
with average claim costs approximately $5,837, the total direct cost would
be approximately $6.5 billion for all upper extremity NTST-MSDs per year.
Webster and Snookxv noted that low back pain workers compensation cases represented 16% of all claims and 33% of all costs. The mean cost for low back pain was $8,321; median cost per case was $396. Again, this is similar to our 1989 data of $6,347 average and $412 median costs. Using our claims incidence rate of 191.6 per 10,000 FTEs for back NTST-MSDs, and applying that to the approximate 96 million US workers, the estimated direct costs would be $10.8 billion. Combined back and upper extremity NTST-MSD costs are remarkably similar to the estimates reported by OSHA and the National Safety Council. There are reasons to believe that these estimates of the magnitude and costs are underestimates. For example, Morse et al xvi , in a survey of the Connecticut working population, found that only 10.6% of those with work-related upper extremity disorders went through the workers compensation system and only 21% of those who had medical visits or procedures reported having them paid for by workers compensation. None of these figures take into account the indirect costs to the employer in lost productivity, quality, training replacement workers, recruitment and other administrative costs. Nor do these figures take into account the quantitative and qualitative costs to the claimant and family (e.g., loss in home production), as well as those workers who never file a workers compensation claim but suffer from work-related back or upper extremity disorders. Morse et al (1998) reported the cases in their survey had much higher difficulties with daily tasks (bathing, child care), having lost their homes, had divorces, etc. than non-cases.
The 1997 Washington BLS rate for CTS resulting in lost time was 4.4 per 10,000 FTEs for private industry, while the 1997 compensable CTS rate for the State Fund was 16.6 per 10,000 FTEs xvii . CTS workers compensation data are not available for the Self-Insured employers and the BLS data does not include public sector employees. It is unlikely that these two differences in populations explain the 4- fold difference in estimated rates. Although there had been some increase in CTS and other NTST-MSDs in Washington State between the late 1980s and mid 1990s, the increase had been quite modest compared to the rapid increase in rates reported by the BLS, suggesting potential underreporting especially in earlier BLS data. Rates have been decreasing for both systems since 1994, as have all injury and illness rates. The incidence of work-related CTS found in this study was 24.5 per 10,000 FTE over the years 1990-1998, while Franklin et al xviii who also used Washington State workers' compensation data reported a claim rate of 17.4 per 10,000 FTEs for carpal tunnel syndrome over the period 1984- 1988. The increase in claim rate over the years is consistent with national data. This probably represents both increases in true incidence as well as in reporting.
The focus of this study has been identifying high-risk industries for both research and prevention purposes. It should be noted that there are high-risk jobs in low-risk industries. However, we did not have adequate denominator data to determine incidence rates by occupation in this study.
We used three different approaches to identifying “high-risk” industries: Frequency count, relative risk or rate ratio, and prevention index. Each has advantages and disadvantages depending on the goals. If we were interested in, for example, reducing the overall number of claims by 10%, we would look for the industries where the most claims are occurring and perhaps focus educational campaigns in those industries. It is likely that the reason why there are so many claims is because these are large industries with many employees, even if the relative risk is low. An example of this can be seen with Eating Places in Table 8. This suggests that any specific restaurant may not have a high concentration of risk factors present. The relative risk or rate ratio is used to identify those industries at highest risk. We would expect to find a higher concentration of risk factors present in most workplaces in these industries. This might be important in focusing inspection activities or research where contrasts in claims incidence rates may be important. However, if the highest risk industry has few employees, the overall industry impact of intervention activities might be small unless control measures have widespread utility. The Prevention Index was developed as a way to obtain the most impact in high-risk industries. It treats frequency and relative risk as equally important. Depending on the type of intervention, education or research focus contemplated, weighting relative risk or incidence rate more heavily than frequency should be considered.
Based on State Fund compensable claims data, the Construction Sector is at highest risk but based on Self-Insured data, Transportation is at the greatest risk (Figures 8 and 9). It is likely that the larger Self-Insured construction initiated prevention activities whereas the small residential contractor in the State Fund has not used the same strategy. It may be that the joint apprenticeship programs, (e.g., carpenter’s apprenticeship program) have focused more on using ergonomics principles and workers from these apprenticeships tend to work for the larger Self-Insured employers. Although tasks in the construction industries are quite varied, they are characterized by manual handling of heavy materials, high peak hand force with periodic repetitive motions (sometimes with segmental vibration as in sawing and drilling), combined with awkward postures. Construction industries are not identified in the top industries for repetitive motion disorders in the BLS data. This discrepancy may reflect our inclusion of "overexertion in lifting" as a non-traumatic onset disorder (whereas the BLS separates overexertion from repetitive motion
It is unclear why the rates are increasing for Transportation among the Self-Insured (twice the State Fund rate in 1997) whereas they are decreasing for the State Fund employers. Additionally, the Self-Insured Public Administration employers have dramatically higher rates than for the State Fund employers. The reasons for this are not readily apparent.
Based on the State Fund Prevention Index and 3 digit SIC code for compensable back, neck and upper extremity NTST-MSDs, the 3-digit SIC code industries involving heavy manual handling are of note: Nursing & Personal Care Facilities, Masonry, Trucking, Residential General Building Contractors, Roofing, Carpentry, Landscaping, Residential Care, Concrete work and Sawmills. These industries are characterized by heavy manual handling tasks and should be the focus of prevention activities. Within the industries at highest risk (RR>2.5), occupations such as drywall installers, roofers, nursing aides, bus drivers accounted for the majority of claims (Table 20). For upper extremity NTST-MSDs, additional industries to include are Miscellaneous Food Preparation, Meat Products, Groceries and Related Products. The Temporary help-assembly WIC was also identified as being at high risk for NTST-MSDs in the State Fund data.
Based on the Self-Insured Prevention Index, the 3-digit SIC code industries requiring the most prevention attention include Trucking and Courier Services , Air Transportation, Grocery Stores, Rolling Mills, Groceries and Related Products, Variety Stores, Nursing & Personal Care Facilities, Personnel Supply Services, Local & Suburban Passenger Transportation and Services to Dwellings and Other Buildings. It is likely that the Variety Store claims are most related to manual handling activities. As shown in Table 29, SIC codes with rate ratios greater than 2.5 had the largest percentage of claims among occupations such as freight/stock handlers and bus or truck drivers. Other industries to include because they are in the top 10 for upper extremity NTST-MSDs are Shipping and Boat Building & Repair, Miscellaneous Durable Goods, Dairy Products and Paperboard Containers & Boxes.
In general, the State Fund WIC codes provide more information on industries requiring attention (Tables 36-47). While there is substantial overlap between industries at high risk for back and upper extremity NTST-MSDs, there are also some important differences. Shoulder (and rotator cuff) disorders tend to be distributed in the same way as back disorders (Nursing homes, Wood frame Building Construction, Wallboard Installation, Wood Products Manufacturing, and Garbage Collection. For hand/wrist NTST-MSDs, Meat Dealers, Wood Products Manufacturing, Temporary Help Assembly, Plastic Products Manufacturing, Sawmills, Fruit and Vegetable Packing, Meat Products Manufacturing, Supermarkets are particularly important for prevention activities.
For the Self-Insured employers, WIC code industries with the highest PI include Parcel Package Delivery, Bus Companies, Airline Ground Crews, Trucking, Warehouses, Wholesale Stores, Schools and Cities with all other employees (grounds, maintenance, etc.) Parcel package delivery, supermarkets, temporary help-administrative, aluminum products manufacturing, meat products, telephone company other employees, sawmills, bakeries were associated with hand/wrist NTST-MSDs. These industries should be the focus of research and prevention activities.
The percentage of female claimants increased somewhat over the years 1990-1998 possibly reflecting the steady increase in the proportion of women in the work force since 1960. In Washington State, the female portion of the workforce increased from 45.3% in 1987 to 46.9% in 1995xix. The average percentage of women claimants with hand/wrist NTST-MSD claims from 1990-1998 was 51% among the State Fund employers and 60.6% among the Self-Insured employers. This can be contrasted to 29.4% and 41.5% respectively for Back NTST-MSDs. Women may differentially select or be selected into highly repetitive work that puts them at increased riskxx whereas men are selected into heavy manual handling work (with the exception of health care).
Although there are a variety of temporary work situations (temporary agencies, leasing, independent contractor, etc.), we were only able to identify claims associated with temporary service agencies. Because of the rapid increase in temporary service agency employment over the study period, summary tables may actually underestimate the current rank order of frequency and incidence of musculoskeletal disorders in segments of this industry. For example, based on the 9-year average incidence rate, Temporary services-assembly ranked 6 th in the State Fund WIC industries for all NTST-MSDs, 9th in neck and 5th in upper extremity NTST-MSDs, 7 th for elbow, 8 th for shoulder, and 3rd for hand/wrist NTST-MSDs. Among the Self-Insured, Temporary Help-Administrative Staff is at high risk for both upper extremity and back disorders based on the prevention index. This WIC actually includes claimants who are temporary help workers in assembly, machine operators, etc. This most likely explains the extremely high rate ratio identified. Because L&I does not manage the self-insured claims like the State Fund, there may be less oversight on the quality of the Self Insured data provided to L&I by the Self-Insured employers. Given the likely continued increase in temporary service employment due to many companies shifting high-risk low skill work away from permanent employees, we can expect to see these temporary service WICs at the top of the Prevention Index in the future. This change in the labor market presents a major challenge for developing effective prevention strategies.
Limitations and Strengths of Using Worker’s Compensation Data.
There were a number of limitations in this study. Potential misclassification of the outcome measures is an important consideration. Underestimation of non-traumatic onset was identified in the records review. This was particularly evident in the Self-Insured data. The use of broad industrial categories as surrogates for exposure may mask high-risk jobs in heterogeneously exposed industries. However, misclassification of exposure is believed to be less for WIC than for SIC because WICs are more closely related to exposure (e.g., clerical workers in a factory have a different WIC than the factory workers) whereas SIC is related to commerce. The only major exception to this is Self-Insured Temporary Help employers who may be misclassifying the claimants using the central office WIC rather than their actual employment.
The use of ANSI z16.2 codes of nature and type are cumbersome. This coding scheme should be replaced with the Occupational Injury and Illness Classification System (OIICS) used by BLS.
Secondly, this study includes very limited data for the largest employers in the state who employ one-third of the workforce (including large aerospace, health care and forest products which are known to involve jobs with work-related risk factors for these musculoskeletal disorders). In some respects, because smaller employers are more represented in the State Fund, costs may be overestimated due to the greater capacity of big employers to return employees to work, even in light duty jobs, thereby reducing lost time days and costs. Differences may also involve higher caseloads for workers compensation adjudicators for the State Fund compared to the Self-Insured, thus delaying the attention needed to address claims once they have been opened.
A third limitation is the inherent bias in reporting. The traumatic onset disorders tend to gain more ready acceptance in the workers compensation system than the more gradual or non-traumatic onset disorders. The medical records review indicated that the State Fund databases were useful for correctly identifying carpal tunnel syndrome, epicondylitis, rotator cuff syndrome and that our coding scheme for determining non-traumatic soft tissue disorders was good for the neck, back and upper extremity but poor for the lower extremity. For example, a random records review of knee and ankle disorders indicated that the vast majority is of traumatic origin.
In this study we
were able to define specific disorders with non-traumatic onset using
State Fund data. Among CTS cases, 86% were NTST claims, for epicondylitis,
this was true for 78% of claims, for sciatica for 73% of claims and for
rotator cuff syndrome, 70% of claims.
We were also able to identify some emerging trends among workers in temporary service agencies. Hopefully, this finding will generate a closer look at health and safety issues affecting contingent workers.
Although much has been written about carpal tunnel syndrome, the magnitude and distribution of sciatica, rotator cuff syndrome and epicondylitis have not been previously described in US working populations. The incidence of rotator cuff syndrome is two-thirds that of CTS (15.3 versus 24.3) but more costly (Table 6). Epicondylitis has less than half of the claims incidence and about half of the cost of CTS per claim but still presents a major lost time problem. Sciatica appears to be a relatively rare diagnosis (incidence rate of 4.9 per 10,000 FTEs) compared to NTST back disorders (incidence rate of 191.6 per 10,000 FTEs), however, the costs are extremely expensive (approximately $40,000 on average). Research and prevention activities focused in industries with high demands for manual handling and repetitive work should contribute to the reduction of these work-related disorders.