Abstract
Background:
Workers’ Compensation differs from standard insurance, and it is unclear how or if Workers’ Compensation insurance influences the diagnosis and treatment of hand and wrist disorders. The aim of this study was to compare the diagnosis and course of treatment of hand disorders between patients with Workers’ Compensation insurance and patients with standard insurance.
Methods:
The complete medical records of patients who visited an academic orthopaedic hand clinic between January 2005 and January 2007 were reviewed, and information on patient history, utilization of diagnostic tests, diagnosis, surgery, and wait-time to surgery was collected. Patients with Workers’ Compensation insurance and those with other, third-party coverage were analyzed and compared. Patients without insurance were excluded from this study.
Results:
1413 patients (representing 2121 diagnoses) were included in the study. One hundred and sixteen patients (8%) had Workers’ Compensation insurance and 1297 patients (92%) had standard insurance. Patients with Workers’ Compensation insurance were younger than patients with standard insurance (mean age, forty-three years compared with fifty years, respectively; p < 0.05) and were also more likely to be male (50% compared with 40%, respectively; p = 0.04). Generally, Workers’ Compensation patients more often had neurological conditions (p < 0.01), but there was no significant difference in the most common diagnoses between the two groups. Patients with Workers’ Compensation underwent surgery slightly more often than did patients with standard insurance (44% compared with 35%, respectively; p = 0.07) and had a higher average number of visits before undergoing surgery (2.3 visits compared with 1.2 visits, respectively; p < 0.05). Twenty-three (45%) of the fifty-one patients with Workers’ Compensation insurance who received a diagnosis indicating the need for surgery underwent surgery after the first visit, compared with 316 (69%) of 458 patients with standard insurance (p < 0.05). Patients with Workers’ Compensation insurance were more likely than patients with standard insurance to undergo electrodiagnostic testing (26% compared with 15%, respectively; p < 0.01) or magnetic resonance imaging (16% compared with 9%, respectively; p = 0.02).
Conclusions:
Compared with patients receiving standard insurance, patients receiving Workers’ Compensation insurance have a greater number of clinic visits before undergoing surgery and receive more diagnostic testing. More research is needed to explore these differences and their potential clinical and economic consequences.
Level of Evidence:
Prognostic Level II. See Instructions to Authors for a complete description of levels of evidence.
Workers’ Compensation accounts for approximately 12% of the total payments for orthopaedic services in the United States, which makes Workers’ Compensation insurance the third largest payer in orthopaedics after Medicare and Medicaid1. The treatment of musculoskeletal injuries, and upper-extremity musculoskeletal injuries in particular, is of key economic importance because musculoskeletal injuries account for a major proportion of missed days from work among U.S. workers and upper-extremity injuries are the most common type of all workplace injuries2. Patients who receive Workers’ Compensation also contribute to increased health-care costs because they receive a greater average number of postoperative and therapy visits and spend longer periods of time out of work than do patients who are not receiving Workers’ Compensation3.
Workers’ Compensation offers unique health-care cost and employee salary incentives that typically do not exist for patients who have standard health-care insurance. Private-sector Workers’ Compensation benefits in the U.S. are largely the responsibility of states; this tradition dates back to 1902 when Maryland became the first state to offer compensation to curb litigation by injured workers4. In Massachusetts, where our study was conducted, the Department of Industrial Accidents guarantees to deliver "prompt and fair compensation to victims of occupational injuries and illness, and to see that medical treatment to injured workers is provided in a timely manner while balancing the needs of employers to contain Workers’ Compensation insurance costs."5 The state's Workers’ Compensation program consists of fully reimbursed "reasonable" medical care, including prescription benefits, which can last as long as medical and hospital services are required due to the injury or illness experienced in the workplace. Specifically, the patient does not have any co-payments for the medical services received under Workers’ Compensation.
In addition to health-care costs, patients who receive Workers’ Compensation benefits in Massachusetts receive other temporary and permanent benefits. To qualify for benefits, the patient must miss more than five full or partial workdays. For the first 156 weeks that the patient misses work as the result of an injury, he or she is entitled to "temporary total incapacity benefits" consisting of 60% of the patient's gross average weekly wage (but not to exceed the average weekly wage set by the state). On returning to work, the patient is entitled to earn a percentage of compensating wages that is based on current productivity as compared with previous productivity. For example, if the patient's current productivity is 50% of what it was before the injury, the patient earns 50% of his or her original salary; however, above this amount, the patient is additionally entitled to 45% of the original salary. Thus, the patient who is 50% productive after the injury can still earn 95% of his or her original salary. This benefit can last for up to 260 weeks. Finally, patients with a permanent disability who are unable to work are entitled to two-thirds of their average weekly pay (up to the average weekly wage as set by the state) plus annual cost-of-living adjustments in perpetuity6. According to federal and state tax codes alike, the compensated wage is exempt from all taxes.
The purpose of this study was to examine whether there is a difference in the diagnosis and treatment of hand disorders between patients with Workers’ Compensation insurance and patients with standard health insurance. Following economic incentives, we hypothesized that the availability of Workers’ Compensation may induce differences in two aspects of patient care. First, there may be a difference in the diagnostic tools or tests ordered, such as electrodiagnostic testing, magnetic resonance imaging, and computed tomography scans. Second, patients receiving Workers’ Compensation may experience differences in their treatment, including a greater number of clinic visits between the time that surgery is recommended and the time of the actual surgery.
Patient Population
Complete medical records of all consecutive patients seen between January 2005 and January 2007 in an orthopaedic hand clinic at an academic health-care center were reviewed. Two hand surgeons (C.S.D. and T.D.R.) participated in this study. Information on patient history, symptoms, diagnostic tests, diagnosis, and surgery was collected. Records were reviewed for all visits to the clinic, including those during, prior to, and after the 2005 calendar year. Patients were divided into two groups: those receiving benefits from Workers’ Compensation insurance, and those receiving benefits from standard insurance. Patients without health insurance were excluded from the study.
The patient data with regard to demographics and health behavior that were included in this study were age, sex, occupation, smoking status, and diagnosis. We recorded whether patients received electrodiagnostic testing, radiographs, magnetic resonance imaging, or computed tomography scans during the process of diagnosis. To investigate waiting time until surgery, we counted the number of visits between the time when a patient was offered surgical treatment and the time when the patient underwent surgery. The number of visits was included in this study in order to assess the level of health-care utilization. Only the patients who eventually received surgery were included in the comparison of surgery waiting time between the two groups of patients.
Statistical Methods
Available demographic, socioeconomic, and health-behavior data were summarized in frequencies and percentages for the two groups of patients in the study. The first set of primary outcome variables included an indicator of whether or not patients had undergone electrodiagnostic testing, radiography, magnetic resonance imaging, or computed tomography to aid in establishing a diagnosis. The second outcome variable was an indicator of whether or not a patient had undergone surgery. The third outcome variable was the number of clinic visits that the patient made between the time that surgery was offered and the time of actual surgery. The frequency and percentages of all outcome variables were summarized by patient group (Workers’ Compensation insurance compared with other insurance), and the Fisher exact test for proportions was used to test the hypothesis that there was a difference between the Workers’ Compensation patient group and the standard insurance patient group with respect to all outcomes. Statistical significance was assessed at p < 0.05, and a statistical trend was assessed at p < 0.10.
In addition, we stratified our analysis of surgery rates by occupation type. We compared the rate at which patients receiving Workers’ Compensation benefits and patients receiving standard insurance benefits underwent surgery by stratifying the patients according to four major occupation types: clerical workers, laborers, medical workers, and service workers. The Fisher exact test for proportions was used to assess whether there was a significant difference among occupation types.
To assess the robustness of findings with regard to potential confounding, appropriate multivariate regression models were estimated, controlling for age, sex, occupation, and smoking status. A logistic regression model was applied to estimate the between-group difference in the probability of having a diagnostic test or surgery. Whether or not a diagnostic test was conducted and whether or not surgery was done were dependent variables in two different models. The presence of Workers’ Compensation insurance was a predictor variable and was assigned a value of zero or one, allowing for the estimation of the probability of the dependent variables while controlling for the potential confounders. A Poisson model, the simple and most common count model in the literature, was used to estimate whether (of those patients receiving surgery) patients receiving Workers’ Compensation benefits were likely to have a shorter or longer wait time before surgery than the wait time of patients receiving standard insurance benefits.
Source of Funding
There were no external sources for funding.
Patient Characteristics
A total of 1413 eligible patients, comprising 2121 unique diagnoses, were identified and included in the analysis. Of these 1413 patients, 116 (8%) had Workers’ Compensation insurance, and the remaining 1297 (92%) had private insurance, Medicare, Medicaid, or other standard insurance.
A comparison of demographic variables revealed that patients with Workers’ Compensation differed significantly from those with standard insurance with respect to age (p < 0.05), sex (p = 0.04), and smoking status (p < 0.001). Patients with Workers’ Compensation were, on the average, seven years younger, more likely to be male (50% of patients with Workers’ Compensation were male compared with 40% of patients with standard insurance), and more likely to be smokers (31% of patients with Workers’ Compensation were smokers compared with 14% of patients with standard insurance; Table I).
With respect to overall diagnostic categories, neurological disorders were more common among patients receiving Workers’ Compensation (p = 0.04), while arthritis and cysts were less common among these patients (p = 0.04 and p < 0.001, respectively). There was no significant difference in the prevalence of tendinitis conditions or fracture and/or strain between the two groups (Table II). With respect to specific diagnoses, there was no significant difference between the hand disorders diagnosed in patients with Workers’ Compensation insurance and those diagnosed in patients with standard insurance. The three most common hand disorders in patients with Workers’ Compensation were carpal tunnel syndrome (15%), distal radial fractures (9%), and cubital tunnel syndrome (7%). Among the patients with standard insurance, the most common hand disorders were carpal tunnel syndrome (15%), trigger finger (9%), and distal radial fractures (7%). Table III lists the ten most common diagnoses for the two groups of patients.
When the total patient population (1413 patients) was stratified by occupation, 291 (21%) were in clerical work, 158 (11%) were laborers, 146 (10%) held jobs in medical fields, 176 (12%) were in service occupations, 325 (23%) were in other occupations, 236 (17%) were retired, and eighty-one (6%) were unemployed. When patients receiving Workers’ Compensation were analyzed separately, among the 116 patients, forty-four (38%) were laborers, twenty-one (18%) were in service occupations, nineteen (16%) held clerical jobs, sixteen (14%) worked in medical fields, nine (8%) were in other occupations, five (4%) were retired, and two (2%) were unemployed. In comparison, among the 1297 patients receiving standard insurance, only 9% (114 patients) were laborers, 155 (12%) worked in service occupations, 272 (21%) were clerical workers, 130 (10%) were medical workers, 316 (24%) were in other occupations, 231 (18%) were retired, and seventy-nine (6%) were unemployed. The composition of the occupations of the two patient groups was significantly different with respect to laborers (p < 0.001), others (p < 0.001), and retired persons (p < 0.001).
Use of Diagnostic Tests
Patients with Workers’ Compensation insurance were neither more nor less likely to have undergone radiography or computed tomography for the purposes of formulating a diagnosis. However, patients receiving Workers’ Compensation were statistically more likely to have received electrodiagnostic testing or to have undergone magnetic resonance imaging. Among patients with Workers’ Compensation insurance, thirty (26%) had an electrodiagnostic test performed and nineteen (16%) had a magnetic resonance imaging scan performed, whereas, among patients with standard insurance, a smaller proportion of patients had electrodiagnostic testing (15% or 190 patients; p < 0.01) or magnetic resonance imaging (9% or 118 patients; p = 0.02) performed.
Surgery Rates
Fifty-one patients (44%) receiving Workers’ Compensation underwent surgical treatment for at least one of their diagnoses, while 458 patients (35%) receiving standard insurance underwent surgical treatment (p = 0.07).
When the data were analyzed according to occupation type, twenty-two (50%) of forty-four laborers with Workers’ Compensation insurance underwent surgery and forty-five (39%) of 114 laborers with standard insurance underwent surgery (p = 0.30). Among patients with clerical occupations, ten (53%) of nineteen patients with Workers’ Compensation insurance underwent surgery and eighty-six (32%) of 272 patients with standard insurance underwent surgery (p = 0.08). Among patients working in medical fields, five (31%) of sixteen patients with Workers’ Compensation underwent surgery while thirty-five (27%) of 130 patients with standard insurance underwent surgery (p = 0.30). Patients with Workers’ Compensation tended to undergo surgery at higher rates overall than patients with standard insurance; however, this difference was not significant within any specific occupation subgroup.
Number of Visits Prior to Surgery
Following the recommendation for surgery, the average number of clinic visits that took place before surgery was higher in the Workers’ Compensation group than it was in the standard insurance group (mean, 2.3 visits compared with 1.2 visits, respectively; p < 0.05). In addition, patients with standard insurance were statistically more likely than patients with Workers’ Compensation insurance to receive surgery after the first visit (p < 0.05). Surgery was performed after the first visit in 316 (69%) of 458 patients receiving standard insurance and in twenty-three (45%) of fifty-one patients receiving Workers’ Compensation.
A review of the orthopaedic literature reveals that the majority of Workers’ Compensation studies focus on poor clinical outcomes following specific surgical interventions7-11. In their review, Ghori and Chung addressed the issue of patients receiving Workers’ Compensation who exaggerate their incapacity, a behavior known as malingering12. Those authors identified feigned hand weakness as a prevalent form of malingering. This condition represents a potential drain on resources and therefore requires more clinical attention at the diagnostic stage. Although our study did not examine the issue of malingering in patients receiving Workers’ Compensation, our results do address the issue of health-care resource utilization since patients with hand and wrist disorders who receive Workers’ Compensation benefits undergo electrodiagnostic testing more often, have moderately higher rates of surgery, and have a greater number of clinic visits prior to surgery than other patients do13-15. Thus, our results support our first hypothesis that there may be a difference in the diagnostic tools or tests ordered for patients receiving Workers’ Compensation benefits as compared with patients receiving standard insurance benefits. However, the greater utilization of medical care by patients with Workers’ Compensation insurance must be considered within the broader framework of individual Workers’ Compensation policies, government regulations, and the financial incentives of different Workers’ Compensation systems across the United States.
The evidence in our study is consistent with that in the broader literature on the economics of Workers’ Compensation insurance. According to the literature, injured workers typically respond to the economic incentives provided by Workers’ Compensation insurance; as Workers’ Compensation benefits become more generous, the frequency and severity of claims increase16-19. There is evidence that Workers’ Compensation drives up health-care utilization; according to one estimate, a 10% increase in Workers’ Compensation benefits can lead to as much as an 11% increase in the number of Workers’ Compensation claims and up to the same percentage increase in the duration of a claim20. However, this evidence is based on aggregate data or insurance claims and not on individual clinical data such as those used in our study. This evidence is also based on a broad set of medical conditions and treatments, while our study reports exclusively on upper-extremity disorders, which represent the most common work-related injuries. In addition, the underlying causes for greater health-care utilization are unclear21-25.
The financial incentives of a patient receiving Workers’ Compensation may include reimbursement of medical expenses and paid leave from work in association with more expensive treatment. Health economists have summarized the problems raised by these kinds of incentives under the rubrics of moral hazard and asymmetric information26,27. Moral hazard describes a change in a person's behavior when that individual no longer bears the costs of his or her actions. In the case of Workers’ Compensation, the patient no longer bears the costs of medical care. Thus, it may be plausible that, in patients receiving Workers’ Compensation, this situation contributes to the patient undergoing more diagnostic studies and slightly higher rates of surgery. The concept of asymmetric information describes the fact that only the patient is fully aware of the true symptoms that he or she is experiencing; neither the orthopaedic surgeon nor the insurer has this information and cannot with certainty know the diagnosis or the best course of treatment. In orthopaedics, this problem may be especially relevant in difficult-to-diagnose conditions, especially ones that rely heavily on subjective assessment.
Our results are also consistent with other possible explanations. The slight trend that we observed with regard to a higher rate of surgery in patients receiving Workers’ Compensation may be due to the desire to return to work and other physical activities as quickly as possible. The statutes governing Workers’ Compensation reimbursements and the actual mechanism by which surgeons are reimbursed vary from state to state and may account for some of the incentives that drive up the costs of treating patients who have Workers’ Compensation insurance. Additionally, the selection of electrodiagnostic testing during a diagnostic workup is not up to the patient; it is ordered by the physician to establish and confirm a clinical diagnosis. The driving force for testing could also come from the requirements of an attorney for a patient receiving Workers’ Compensation or may be a requirement of the state's laws regarding Workers’ Compensation.
The results of our study also support our second hypothesis, which is that patients receiving Workers’ Compensation insurance may have a greater number of clinic visits before undergoing surgery. This finding may be due in part to the nature of Workers’ Compensation insurance proceedings. For example, patients receiving Workers’ Compensation in Massachusetts who would be willing to undergo surgery the day after their first appointment could potentially experience prolonged wait times due to the requirement of a letter of approval from their employer's risk-management firm or a second opinion as required by their insurance carrier. Thus, patients receiving Workers’ Compensation cannot necessarily be faulted for more clinic visits (and therefore more health-care resource utilization) before undergoing surgery.
From a policy perspective, our results offer some justification for a growing concern that patients with hand and wrist disorders who receive Workers’ Compensation present a challenge both for providers and policymakers working to contain cost. According to the Survey of Occupational Injuries and Illnesses conducted by the U.S. Department of Labor, the upper extremity is the most common site of work-related injuries2. In 2007, from a total of 1,158,870 diagnoses of work-related injuries and illnesses that were reported in the private sector, 269,240 diagnoses were classified in the survey as upper-extremity disorders (including arm, wrist, hand without finger, and finger), with an additional 75,580 diagnoses classified as shoulder injuries. Together, upper-extremity disorders and shoulder injuries account for 30% of all diagnoses, followed in frequency by disorders of the lower extremity (23%), back (20%), and multiple body parts (10%)2. Patients with upper-extremity diagnoses were most likely to be in the trade, transportation, and utilities industry; the manufacturing industry; or the construction industry (as classified by the U.S. Department of Labor). By focusing on patients with hand and wrist disorders who were from a wide cross-section of occupations, our study suggests areas in which both provision of care and cost can be further addressed: the use of diagnostic tools, surgical treatment, and the waiting time until surgery. Because upper-extremity injuries are the most common ones among patients who receive Workers’ Compensation, an increasing focus on these areas could potentially have a large economic impact on the Workers’ Compensation insurance system as a whole.
Due to its observational nature, our study has the standard limitations that characterize all retrospective reviews of patient medical records. We aimed to minimize any potential selection bias by systematically including all patients who visited the clinic in the specified time period, and we conducted a detailed review of all clinic notes, which enabled us to compare outcomes between the two groups. An advantage of our study is that we were able, with most diagnoses, to recover a full medical history and obtain information about the patient's repeated visits to the clinic. We used the number of visits in our study partly as a proxy for surgery waiting times; our data for this study do not include the exact waiting time in days. However, from an economic perspective, a focus on visits rather than on the number of days is useful for the purpose of assessing whether there are different health-care utilization rates for patients with Workers’ Compensation insurance than for patients with standard insurance, and, consequently, whether the care of these patients requires greater resource allocation. Additionally, the acuity or chronicity of diagnoses was not analyzed in our study. Thus, we cannot directly take into account whether or not patients may have been candidates for electrodiagnostic testing or whether testing was needed to substantiate a diagnosis.
Watkins-Castillo
S. Orthopaedic practice in the U.S. 2005-2006. Department of Research and Scientific Affairs. Rosemont, IL: American Academy of Orthopaedic Surgeons; .
Survey of Occupational Injuries and Illnesses. Counts, rates, and characteristics, 2007. Bureau of Labor Statistics, US Department of Labor; 2008.
Bednar
JM;
Baesher-Griffith
P;
Osterman
AL. Workers compensation. Effect of state law on treatment cost and work status. Clin Orthop Relat Res.
1998;351:74-7.[PubMed][CrossRef]
Maryland Employer and Employee Cooperative Fund. .
Department of Industrial Accidents, Executive Office of Labor and Workforce Development, Commonwealth of Massachusetts. Mission statement. .
Massachusetts General Law, Worker's Compensation, §21 M.G.C. 152. .
Harris
I;
Mulford
J;
Solomon
M;
Van Gelder
JM;
Young
J. Association between compensation status and outcome after surgery: a meta-analysis. JAMA.
2005;293:1644-52.[PubMed]
Misamore
GW;
Ziegler
DW;
Rushton
JL
2nd. Repair of the rotator cuff. A comparison of results in two populations of patients. J Bone Joint Surg Am.
1995;77:1335-9.[PubMed]
Frieman
BG;
Fenlin
JM
Jr. Anterior acromioplasty: effect of litigation and Workers’ Compensation. J Shoulder Elbow Surg.
1995;4:175-81.[PubMed]
Greenough
CG;
Taylor
LJ;
Fraser
RD. Anterior lumbar fusion. A comparison of noncompensation patients with compensation patients. Clin Orthop Relat Res.
1994;300:30-7.[PubMed]
Bosacco
SJ;
Berman
AT;
Bosacco
DN;
Levenberg
RJ. Results of lumbar disk surgery in a city compensation population. Orthopedics.
1995;18:351-5.[PubMed]
Ghori
AK;
Chung
KC. A decision-analysis model to diagnose feigned hand weakness. J Hand Surg Am.
2007;32:1638-43.[PubMed]
Sears
JM;
Wickizer
TM;
Franklin
GM. Overstating the prevalence of symptom exaggeration in Workers’ Compensation cases. J Hand Surg Am.
2008;33:1014-5.[PubMed]
Ghori
AK;
Chung
KC. Overstating the prevalence of symptom exaggeration in Workers’ Compensation diagnoses [In reply to Sears et al.]. J Hand Surg Am.
2008;33:1015-7.
Bellamy
R. Compensation neurosis: financial reward for illness as nocebo. Clin Orthop Relat Res.
1997;336:94-106.[PubMed]
Worrall
JD;
Appel
D. The wage replacement rate and benefit utilization in Workers’ Compensation insurance. J Risk Ins.
1982;49:361-71.
Butler
RJ. Economic determinants of Workers’ Compensation trends. J Risk Ins.
1994;61:383-401.
Butler
RJ;
Worrall
JD. Claims reporting and risk bearing moral hazard in Workers’ Compensation. J Risk Ins.
1991;58:191-204.
Johnson
WG;
Ondrich
J. The duration of post-injury absences from work. Rev Econ Statist.
1990;72:578-86.
Loeser
JD;
Henderlite
SE;
Conrad
DA. Incentive effects of Workers’ Compensation benefits: a literature synthesis. Med Care Res Rev.
1995;52:34-59.[PubMed]
Butler
RJ;
Hartwig
RP;
Gardner
H. HMOs, moral hazard and cost shifting in Workers’ Compensation. J Health Econ.
1997;16:191-206.[PubMed]
Leigh
JP;
Ward
MM. Medical costs in Workers’ Compensation insurance: comment. J Health Econ.
1997;16:619-22.[PubMed]
Royalty
AB. Estimating workers’ marginal valuation of employer health benefits: would insured workers prefer more health insurance or higher wages?J Health Econ.
2008;27:89-105.[PubMed]
Skatun
JD. Take some days off, why don't you? Endogenous sick leave and pay. J Health Econ.
2003;22:379-402.[PubMed]
Baker
LC;
Krueger
AB. Medical costs in Workers’ Compensation insurance: reply. J Health Econ.
1997;16:619-24.[PubMed]
Arrow
K. Uncertainty and the welfare economics of medical care. Am Econ Review.
1963;53:941-73.
Newhouse
J. Pricing the priceless: a health care conundrum. Cambridge: MIT Press; 2004.