Neck pain is a well-recognized source of disability in the working population.1 Surveys of workers suggest that the annual prevalence of activity limitations related to neck pain varies from 11.0% in the UK to 14.1% in Quebec, Canada.2,3 However, these statistics do not agree with claim rates reported by Workers’ Compensation Boards. According to workers’ compensation statistics, work-related neck pain represents a minor health burden to society.
For example, in 2005 in Saskatchewan, Canada, neck pain claims accounted 1.7% of all claims accepted by the Workers’ Compensation Board.4 During the same period in Ontario, claims related to the cervical and thoracic regions accounted for 4.5% of all lost-time claims.5 Moreover, Silverstein et al6 reported that between 1990 and 1998, 3.0% of annual claims to the Washington State Department of Labor and Industries were for “non-traumatic” “soft tissue disorders of the neck.” In Washington State, the annual incidence of neck pain lost-time claims was 19.1 per 10,000 full-time equivalents (FTE).6
The discrepancy between survey and administrative statistics is partly attributable to differences in neck pain severity (i.e., lost-time claimants likely represent the proportion of workers with worse injuries). However, it is also possible that workers’ compensation statistics do not accurately capture the true burden of disability related to specific disorders. Although worker’s compensation records offer a rich source of data for epidemiologic research, their use may lead to erroneous estimates of the prevalence and incidence of work absenteeism related to a type of disorder.
Reporting (underreporting and inaccurate reporting) and diagnostic uncertainty can lead to measurement error when using workers’ compensation data in research.7–9 However, for injuries such as soft tissue disorders of the neck, an additional source of systematic error may be related to the coding protocols used by workers’ compensation boards. According to these protocols, an injured worker with multiple injuries would only have his/her “most severe injury” coded and recorded in the database. Therefore, sprain and strains of the cervical spine would not be coded because they are given a low coding priority relative to other injuries such as concussions and lacerations. We studied this potential source of bias and found that omitting to count these claimants would lead to a gross underestimation of the true number of lost-time claimants with soft-tissue disorders involving the neck. Specifically, we found that the number of lost-time claimants with soft-tissue disorders to the neck varied from 88% in those coded with a disorder of the neck region, to 69% for claimants coded with disorders of the back and 55% for claimants with a brain injury (concussion).10 The objective of this article is to report on the prevalence and incidence of work absenteeism involving neck pain in a cohort of claimants to the Ontario Workplace Safety & Insurance Board (WSIB).
Ontario Workplace Safety & Insurance Board
The Ontario WSIB is a public insurance system legislated by the Workplace Safety & Insurance Act to provide no-fault insurance for workplace injuries and diseases to Ontario workers and workplaces.11 In 1998, approximately 65% of Ontario workers were covered by the WSIB.
Two main types of coverage exist within the WSIB: “Schedule 1” and “Schedule 2.” Under “Schedule 1” coverage, employers are not individually liable to pay benefits directly to workers or their survivors. These employers are required to submit a claim to the WSIB within 3 days of a worker’s injury if the injury resulted in lost time from work, wage loss, or the worker receiving health care. Employers covered under “Schedule 2” are companies that do not pay premiums to the WSIB but are compulsorily covered through a system of individual liability (they are required to pay an administrative fee with each claim submitted—approximately 29% of the total injury costs). These companies include telephone and telegraph companies, navigation companies, international bridges, provincial government (including boards, commissions, and Crown agencies), airlines with a regularly scheduled international passenger service, municipalities (including municipal boards and commissions, except hospital boards), public library boards, and school boards. “Schedule 2” employers are required to report all workplace injuries to the WSIB and they are responsible to pay the total costs of benefits for their injured employees. In this article, all injuries reported by “Schedule 2” employers are aggregated under the rubric “Schedule 2.” Workers who are self-employed or employed by companies that are not required to have WSIB coverage are not required to report injuries to the WSIB.
Source Population and Design
The source population included all injured workers 18 years of age or older with an accepted lost-time claim to the WSIB in 1998. Our design includes a cross-sectional and a prospective component. First, we measured the prevalence from a cross-sectional analysis of all injured workers with active lost-time claims in 1998. Second, we measured the incidence of lost-time claims involving neck pain by forming a historical cohort of injured workers with a new lost-time claim corresponding to injuries that occurred between January 1 and December 31, 1998. The incidence analysis only included claimants who did not make a claim for neck pain in the year before their index claim. The study received ethics approval from the University of Toronto Ethics Review Board.
Definition of Neck Pain
We aimed to measure neck pain related to soft-tissue disorders of the cervical spine/shoulder area including disc lesions and radiculopathy. The clinical presentation of neck pain may include symptoms in the head, trapezius muscle, shoulder, upper back, upper arm and arm.13,14 We excluded neck pain associated with fractures, tumors, infections, rheumatoid arthritis, mylopathies, lacerations to the neck–shoulder region, shoulder tendinitis as well as dislocations and subluxations of the glenohumeral joint.
Data Sources and Linkage
WSIB Claim Data File. We used the WSIB claim data file to identify lost-time claims involving neck pain. This file includes (1) age; (2) sex; (3) date of accident; (4) injured part of body; (5) nature of injury; and (6) occupation.
Employer File. The employer file provides information about the sector of activity of all firms registered with the WSIB. This data were used to classify firms into 1 of 16 industrial sectors.
We used firm numbers to deterministically link records from the claim data and employer files. The linked data file used for analysis was stripped of all identifiers.
Enumeration and Validity of Neck-Pain Claimants
We used the “part of body” and “nature of injury” codes from the WSIB claim data file to identify injured workers with lost-time claims involving neck pain. The nature of the injury describes the principal physical characteristics of the injury or disease (e.g., sprains bruises, tendonitis). The part of body identifies the worker’s anatomic location that was directly affected by the nature of injury (e.g., cervical region/cervical vertebrae, multiple back regions). The methodology used to identify lost-time claimants is described in detail elsewhere.10
In summary, we reviewed all 610 nature of injury and 188 part of body codes used by WSIB coders to code claims.10 These codes comply with the National Work Injuries Statistics coding standards.15 We initially selected 60 nature of injury and 27 part of body codes that may relate to neck pain and reached consensus on a final set of 173 code combinations. We combined these codes into 10 groups based on body regions (e.g., cervical region, back region). We then randomly selected 434 claims for an in-depth chart review. Based on the available information, the reviewer (D.V.E.) who was blind to the code combination determined whether a claimant had neck pain according to the definition provided earlier.10
Our final set of body regions included 9 groups formed from 60 nature of injury codes and 22 part of body codes. The proportion of neck pain cases in each of the 9 final groups varied from 10% for the shoulder and head and face region to 88% for the cervical region (Table 1). Twelve percent of the cervical region cases did not meet our criteria for soft-tissue injuries of the neck and could be classified as false-positive cases (3 workers had back pain and 1 had a cervical disc herniation with spinal cord compression). The back region group included a significant proportion of neck pain cases (69%).
Estimation of the Population of Workers Covered by the WSIB
We estimated the size of the population covered by the WSIB according to the method developed by Smith et al (2004) using Statistic Canada’s Labor Force Survey (LFS)16,17. The LFS estimates the working-age population in Canada (employed, unemployed, and nonlabor force participants) using a sample of approximately 54,000 Canadian households each month. For our analysis, the LFS was restricted to “non-self-employed” labor force participants who were working in Ontario. Because of data availability, we restricted our analysis to workers who are at least 20 years of age. We adjusted our estimates to account for differences in mandatory coverage and reporting of work-related injuries to the WSIB among particular industry groups in Ontario.5 To account for differences in hours of employment between labor force participants (e.g., age and gender groups), we present all denominators as FTE. One FTE corresponds to 2000 hours of work in 1 year (50 weeks × 40 hours of work).
Prevalence. We computed the cumulative, age-specific, gender-specific, and industrial sector-specific annual prevalence of neck pain in lost-time claimants using 2 sets of denominators. First, we calculated the prevalence of neck pain as a proportion of all lost-time claimants. Second, we measured the prevalence of neck pain as the proportion of all Ontario workers employed by firms with mandatory WSIB coverage using FTE’s (N1998 = 2,819,437 FTE). Our numerator included one claim per year per claimant, that is, claimants with multiple claims in a calendar year were only counted once.
Incidence. We computed the annual cumulative, age-specific, and gender-specific incidence of neck pain in lost-time claimants. We restricted this analysis to workers who made a new lost-time claim related to neck pain between January 1 and December 31, 1998. The incidence rate was calculated using the total number of individuals employed by firms with mandatory WSIB coverage as the denominator.17 This denominator does not represent a true population of workers at risk of making a lost-time claim because it does not exclude workers with prevalent lost-time claims involving neck pain. However, given the low prevalence of lost-time claims involving neck pain in the working population, we expect the resulting underestimation of incidence to be minimal. As for prevalence, we describe the impact of selecting different WSIB code combinations on our estimates. All analyses were done using SAS Version 8.02 on a UNIX system.18
Assessment and Correction for “Hidden” Neck Pain
The selection of “part of body” and “nature of injury” codes used to enumerate cases from WSIB claims data may influence the measurement of prevalence.10 We used 2 numerators to describe the impact of this bias. First, we used the number of WSIB claimants with injuries coded with “part of body” and “nature of injury” codes referring to the cervical region only. As indicated by our validation study, 88.2% of claimants coded as having an injury to the cervical region and neck have a soft-tissue disorder.10 Therefore, we weighted the first numerator by 0.882. Our second numerator was a weighted proportion of lost-time claimants with neck pain sampled from all 9 groups derived from combinations of “part of body” and “nature of injury” codes10 (Table 1). These groups included the cranial region, neck region, shoulder, and upper arm with specific diagnoses, upper arm, back region, brain, trunk and multiple regions, face and head regions and multiple body regions. For example, we found that 42.7% of claimants coded as “multiple body” injuries had neck pain. Therefore, our numerator included the total number of claimants with “multiple body” injury multiplied by 0.427. We adjusted the variance of the weighted prevalence and incidence estimates by using the variance (standard deviation) of the weight in the calculation of their 95% CI.
We estimated that the Ontario WSIB covered 2,819,437 workers in 1998 (Table 2). Most of the Ontario workforce was younger than 49 years and almost 60% were males. In 1998, 114,836 injured workers had active lost-time claims and received lost-time benefits (date of injury could have been in 1998 or earlier, but the claim was active in 1998) (Table 2). The 3 industrial sectors with the highest proportion of lost-time claimants are the service sector, the manufacturing sector, and “Schedule 2” employers (Table 2). These statistics cannot be used to infer that these sectors are the most hazardous because they include the highest proportion of workers and are therefore likely to include the highest proportion of claims.
Table 3 presents the characteristics of claimants coded with disorders of the neck region and those coded with other code combinations from the 9 identified groups. Claimants coded with disorders of the neck region included a higher proportion of claimants in the 30 to 39 years age-group and more women. We found no systematic differences in the distribution of claimants across industrial sectors.
Annual Prevalence of Work Absenteeism Associated Neck Pain in WSIB Lost-Time Claimants
When we relied on neck codes only, we found that 2.8% (95% CI 2.5–3.3) of claimants were off work each year because of neck pain (Table 4). However, when we accounted for neck pain in workers coded with other injuries, the annual prevalence of neck pain increased to 11.3% (95% CI 9.5–13.1). This 403% difference suggests that a significant amount of neck pain is hidden within the classification system used to code claims. Because of this significant bias, we focus our description of the burden of work absenteeism associated neck pain to the figures obtained from all code combinations.
Of all claims with time-loss in 1998, 14.4% of claims by females and 10.1% of claims by males involved neck pain. We did not find that the annual prevalence varied across age groups (range, 10.4% in 20–29 year old group to 12.1% in 60–69 year old group) (Table 4). When stratified by industrial sector, we found that lost-time claims involving neck pain were most common in the following sectors: health care (18.9%), Schedule 2 (14.3%), electrical (13.5%), transportation (13.3%), and education (12.9%) (Table 5).
Annual Prevalence of Work Absenteeism Associated Neck Pain in Ontario Workers Covered by the WSIB
In Ontario in 1998, the annual prevalence of work absenteeism involving neck pain was 30 per 10,000 FTE (95% CI 25–34) (Table 6). Again, we observed that using the neck codes only would lead to a large underestimation of the prevalence. Using the WSIB covered population as a denominator, we found that the annual prevalence of work absenteeism involving neck pain was higher in males (33 per 10,000 FTE) than females (24 per 10,000 FTE). This finding may seem contradictory to our previous estimates. However, it reflects the gender differences in the composition of the denominators used to compute the annual prevalence; i.e., the male to female ratio is larger in the population of WSIB lost-time claimants than it is in the Ontario covered population (Table 2). The annual prevalence of neck pain peaked in the 30 to 39 year old group (33 per 10,000 workers) and was the highest in 30 to 39 year old males (38 per 10,000 FTE) (Table 6).
Annual Incidence of Work Absenteeism Associated Neck Pain in Ontario Workers Covered by the WSIB
In 1998, the annual incidence of work absenteeism associated neck pain in workers covered by the WSIB was 23 per 10,000 FTE (95% CI 20–27) (Table 7). Using claimants with neck codes only would have led to a 383% underestimation in the incidence. The incidence was higher in males (25 per 10,000 FTE; 95% CI 21–30) than in females (19 per 10,000 FTE; 95% CI 16–21). Among male workers the incidence varied from 18 per 10,000 FTE (95% CI 15–21) in the 50 to 59 year old group to 29 per 10,000 FTE (95% CI 24–35) in those between the ages of 20 to 39 years. In female workers, the incidence varied from 15 per 10,000 FTE (95% CI 13–17) in 60 to 69 year old age group to 20 per 10,000 FTE (95% CI 17–23) in those between the ages of 40 and 59 years (Table 7).
We conducted a cohort study to determine the prevalence and incidence of work absenteeism involving neck pain in Ontario workers. Our study has 2 main findings. First, it provides evidence that neck pain is associated with a significant burden of disability in workers. Specifically, we found that 11.3% of Ontario workers who received lost-time benefits in 1998 had neck pain associated with their claim. Moreover, our analysis suggests that each year approximately 23 new cases of work absenteeism per 10,000 FTE’s will involve neck pain. Second, our analysis demonstrates that strictly relying on neck-specific codes such as those used by Canadian Workers’ Compensation Boards can lead to a gross underestimation of the burden of neck pain in workers. Historically, North American Workers’ Compensation Boards have not considered neck pain to be a leading source of disability in workers; their statistics suggested that less than 5% of claims involved soft-tissue disorders of the neck.
The 2000–2010 Bone and Joint Task Force on Neck Pain and its Associated Disorders accepted only one study describing the population-based incidence of lost-time claims in workers. Silverstein et al6 reported that in Washington State, the annual incidence of work absenteeism involving neck pain (off work for more than 4 days) was 19.1 per 10,000 FTE’s. Despite methodological differences, we reported a similar incidence (23 per 10,000 FTE’s) in Ontario. There are 2 main methodological differences between our studies. First, Silverstein et al only included primary injuries to the neck in their numerator. In contrast, we accounted for injured workers who had neck pain associated with other injuries.6 Second, we excluded workers who had a lost-time claim for neck pain in the previous year whereas the Washington estimates did not impose such restriction.
Our study has limitations. First, we likely underenumerated all lost-time claims involving neck pain. It is likely that the certain code combinations that were omitted from our analyses included a small number of injured workers with neck pain. Second, we could not use a “true” denominator of the workers at risk of being off work because of neck pain. Although, this bias led to an underestimation of our incidence estimates, it was likely minimal. The low number of prevalent cases of work absenteeism because of neck pain relative to the size of the population covered by the WSIB would not result in an important bias. Third, we estimated the prevalence and incidence of lost-time claims involving neck pain among Ontario workers covered by the WSIB using FTE’s rather than workers as a denominator. FTE’s are based on the assumption that a full-time employee works 2000 hours per year (50 weeks × 40 hours per week). However, some workers (e.g., part-time workers) work less whereas others work more than 40 hours per week. Therefore, the number of FTE’s may not accurately reflect the number of workers covered by the WSIB. Finally, our results cannot be used to infer that neck disorders are the sole cause of work absenteeism in workers with prevalent or incident neck pain. Rather, our analysis provides us with crude figures of the association between neck pain and lost-time claims. Future studies should focus on measuring the contribution of neck pain to an episode of work absenteeism in a cohort of injured workers with multiple injuries.
Our study highlights that neck pain is a disabling condition in workers. Traditional application of the workers’ compensation statistics provide a gross underestimation of the burden of neck pain in workers due to errors in the ways these claims are described and coded. We believe that the collection of workers’ compensation data could be improved to allow proper surveillance of neck pain in workers by allowing more than one injury per claim to be recorded and by allowing diagnoses to change over time. Valid surveillance of work-related disorders such as neck pain is essential to measure secular trends in compensable injuries and to evaluate the impact of population-based prevention strategies.
- Little is known about the prevalence and incidence of neck pain in workers making lost-time claims to Workers’ Compensation Board.
- We found that the prevalence and incidence of neck pain in workers making lost-time claims varies considerably with the method used to ascertain cases.
- The estimated percentage of lost-time claimants with neck pain in 1998 ranged from 2.8% (95% CI 2.5–3.3) using codes specific for neck pain to 11.3% (95% CI 9.5–13.1) using a weighted estimate of codes capturing neck pain cases. The annual incidence of neck pain among the Ontario working population ranged from 6 per 10,000 FTE (95% CI 5–6) using only codes specific for neck pain to 23 per 10,000 FTE (95% CI 20–27) using a weighted estimate of codes capturing neck pain cases.
- Male workers between the ages of 20 and 39 years were the most likely to experience an episode of work absenteeism involving neck pain.
1.Côté P, van der Velde G, Cassidy JD, et al. The burden and determinants of neck pain
in workers. Results of the Bone and Joint Decade 2000–2010 Task Force on Neck Pain
and its Associated Disorders. Spine
2.Palmer KT, Walker-Bone K, Griffin MJ, et al. Prevalence
and occupational associations of neck pain
in the British population. Scand J Work Environ Health
3.Leroux I, Dionne CE, Bourbonnais R, et al. Prevalence
of musculoskeletal pain and associated factors in the Quebec working population. Int Arch Occup Environ Health
6.Silverstein B, Viikari-Juntura E, Kalat J. Use of a prevention index to identify industries at high risk for work
-related musculoskeletal disorders of the neck, back, and upper extremity in Washington state, 1990–1998. Am J Ind Med
7.Azaroff LS, Levenstein C, Wegman DH. Occupational injury and illness surveillance: conceptual filters explain underreporting. Am J Public Health
8.Beaton DE. Examining the clinical course of work
-related musculoskeletal disorders of the upper extremity using the Ontario Workers’ Compensation
Board administrative database, Thesis, University of Toronto, Toronto, 1995.
9.Zakaria D, Mustard C, Robertson J, et al. Identifying cumulative trauma disorders of the upper extremity in workers’ compensation
databases. Am J Ind Med
10.Van Eerd D, Côté P, Beaton D, et al. Capturing cases in workers’ compensation
databases: the example of neck pain
. Am J Ind Med
11.Workplace Safety and Insurance Act. Government of Ontario; 1997.
13.Bogduk N. Innervation and Pain Patterns of the Cervical Spine. Physical Therapy of the Cervical and Thoracic Spine.
3rd ed. New York: Churchill Livingstone; 2002:1–13.
14.Spitzer WO, Skovron ML, Salmi LR, et al. Scientific monograph of the Quebec Task Force on Whiplash-Associated Disorders: redefining “whiplash” and its management. Spine
15.Association of Worker’s Compensation Boards of Canada. National Work
Injury/Disease and Fatality Statistics Program. Coder Reference Manual: Coding guidelines, issues and cases Version 2. [Association of Worker’s Compensation Boards of Canada.]. 1999;2.
16.Minister of Industry. Statistics Canada Guide to the Labour Force Survey. 71-543-GIE. 2003. Ottawa. Ref Type: Report.
17.Smith PM, Mustard CA, Payne JI. Methods for estimating the labour force insured by the Ontario Workplace Safety and Insurance Board: 1990–2000. Chronic Dis Can
18.SAS/STAT User’s Guide, Version 6.
4th ed. SAS Institute; 1990.