J Occup Rehabil DOI 10.1007/s10926-014-9539-0

Cost-Effectiveness of Early Versus Delayed Functional Restoration for Chronic Disabling Occupational Musculoskeletal Disorders Brian R. Theodore • Tom G. Mayer Robert J. Gatchel



 Springer Science+Business Media New York 2014

Abstract Purpose Despite extensive evidence for the treatment effectiveness of interdisciplinary functional restoration (FR) for chronic disabling occupational musculoskeletal disorders (CDOMD), there is little documentation on the cost-effectiveness of early rehabilitation using FR. Methods A total of 1,119 CDOMD patients were classified according to duration of disability on FR entry, corresponding to early rehabilitation (ER: 4–8 months of disability, N = 373), intermediate duration (ID: 9–18 months, N = 373), and delayed rehabilitation (DR: [18 months, N = 373). Groups were matched on sex, age, ethnicity, and injured musculoskeletal region. One-year post-rehabilitation outcomes included return-to-work, work retention and healthcare utilization. Economic analyses included a cost-effectiveness analysis of the FR program, and estimation of the total cost-ofillness. Results At 1-year post-rehabilitation, all groups were

comparable on return-to-work (overall 88 %), work retention (overall 80 %), and additional healthcare utilization (overall, 2.2 % of patients received re-operations/new surgeries, 2 visits to new healthcare provider). Savings of up to 64 % in medical costs, and up to 80 % in disability benefits and productivity losses was associated with the ER group. The cost of rehabilitation was also up to 56 % lower when administered early. Overall, ER resulted in estimated cost savings of up to 72 % (or almost $170,000 per claim). Conclusions Duration of disability does not negatively impact objective work or healthcare utilization outcomes following interdisciplinary FR. However, early rehabilitation is more likely to be a costeffective solution compared to cases that progress[8 months and receiving FR as a treatment of ‘‘last resort’’. Keywords Chronic pain  Disability  Musculoskeletal disorders  Functional restoration  Socioeconomic outcomes  Cost-effectiveness analysis

B. R. Theodore PRIDE Research Foundation, Dallas, TX, USA B. R. Theodore Department of Anesthesiology and Pain Medicine, University of Washington, Seattle, WA, USA T. G. Mayer (&) Department of Orthopedic Surgery, The University of Texas Southwestern Medical Center at Dallas, 5701 Maple Ave., Suite 100, Dallas, TX 75235, USA e-mail: [email protected] R. J. Gatchel Department of Psychology, College of Science, The University of Texas at Arlington, Arlington, TX, USA R. J. Gatchel Department of Anesthesiology and Pain Management, The University of Texas Southwestern Medical Center, Dallas, TX, USA

Introduction In the United States, most patients with occupational injuries fall under the jurisdiction of some variety of state or federal workers’ compensation (WC) systems. Although the vast majority of injured workers recover and resume productivity, a small but significant percentage of these patients remain chronically disabled [1]. Reviews of WC databases have indicated that as little as 7 % of these chronically disabled patients account for approximately 70 % of the annual costs to the WC system [2–4]. In addition to medical costs, the state and federal WC systems also cover indemnity benefits such as wage replacement, settlements for workplace injuries, and impairment benefits for any resulting permanent disability. Additionally,

123

J Occup Rehabil

patients in occupational disability compensation systems have been noted for having persistently poor outcomes following standard treatment modalities (especially among patients receiving lumbar fusion surgeries), associated with the fact that acute lumbar injuries account for the highest percentage of acute musculoskeletal injuries becoming chronically disabling [5–8]. Because all WC claims must involve pre-injury work, and patients unable to work are paid disability benefits (temporary total disability or TTD), the study of this unique component of the healthcare system allows documentation of objective socioeconomic outcomes, such as work status, healthcare utilization related to the claim, recurrent injuries after work return and financial costs of disability and lost productivity. For this reason, the present study deals exclusively with a cohort of chronic disabling occupational musculoskeletal disorders (CDOMDs), of which 75 % of the cohort have a compensable work-related injury with an active WC claim to the spinal region. The prevailing theoretical models of pain and disability indicate that chronicity is associated with significant psychosocial barriers to recovery, poorer outcomes following standard care, and often require intensive multidisciplinary biopsychosocially-oriented treatment modalities to address the complex interaction of various physiological, psychological, and socioeconomic factors [9–11]. Additionally, these theoretical models clearly delineate stages or phases of pain and disability, and underscore the cost implications and treatment resistance of pain and disability that have progressed to chronic stages. A prior study on the effectiveness of interdisciplinary rehabilitation using the functional restoration approach addressed the implications of the Krause & Ragland phase model of disability [11] that predicted significantly poorer treatment outcomes among patients with extended durations of disability. To test this model’s assertion that patients beyond 18 months of disability would respond poorly to rehabilitation, Jordan et al. [12] compared a cohort of short-term disability patients (4–8 months duration) against two groups of patients with long-term disability (18–23 months and [24 months), all of whom received functional restoration for work-related spinal injuries between years 1990 and 1993. The results indicated statistically significant reductions in 1-year workrelated outcomes for the long-term disability groups. However, the magnitudes of outcome rates for the longterm disability were relatively comparable to the short-term disability group. For example, the long-term disability groups of 18–23 months and [24 months had return-towork rates of 80 % and 75 % respectively, relative to a 93 % return-to-work rate in the short-term disability group. Similar differences in magnitudes were observed for work retention at 1-year post-rehabilitation, with the long-term

123

disability groups having retained work at 72 and 66 %, relative to the short–term disability group work retention rate of 80 %. Therefore, patients with extremely long durations of disability could still be rehabilitated with respectable success rates after completing a course of intensive, interdisciplinary rehabilitation, and these results are consistent with the documented effectiveness of the functional restoration model over the last three decades [13–15]. However, there has been very little research to date on the link between cost and outcomes for early as opposed to delayed rehabilitation for CDOMDs. The present study was designed to investigate the value or cost-effectiveness of early interdisciplinary functional restoration (FR) for CDOMDs, once a normal healing period has passed and a CDOMD is established (4-month threshold). Early FR will be compared with FR received during an intermediate duration of 9–18 months post-injury and delayed FR at more than 18 months post-injury. The study utilized matched groups of patients who received functional restoration for CDOMDs following a WC injury claim. Extensive demographic, economic and post-treatment outcome information were obtained from patients, employers, insurance carriers and state/federal WC system databases. A full economic analysis was conducted to determine the costeffectiveness of early versus delayed rehabilitation.

Methods Patients A consecutive cohort of 2,491 patients was initially identified as being eligible for this study, based on diagnoses for CDOMDs. All patients were admitted to a functional restoration program at the Productive Rehabilitation Institute of Dallas for Ergonomics (PRIDE), between the years of 1998 and 2005. The workers’ compensation jurisdiction for this cohort is predominantly from the Texas workers’ compensation system (90 %), with the remaining patients from other state workers’ compensation jurisdictions, federal disability compensation systems, as well as private insurance and personal injury policies. Program participation criteria included: (1) four or more months elapsed since a work-related injury; (2) surgical and/or nonoperative care failed to improve symptoms sufficient to allow full-duty return to work; (3) surgery had not produced relief, resolution or simply was not an option; (4) severe pain and functional limitations remained; and (5) ability to communicate in English, Spanish or other translatable languages. In addition to the general program inclusion criteria, this initial cohort was subject to several exclusion criteria to provide a working cohort for the study. The following exclusion criteria apply:

J Occup Rehabil







Exclusion of patients covered by the Federal Employees Liability Act (FELA) for railroad workers (N = 24, 1 % of eligible cohort), due to their nature as extreme outliers in the disability compensation system. Patients covered by FELA are a unique sub-sample that are noted for receiving significantly higher amounts of disability compensation or settlements, relative to preinjury wages, and are noted for significantly poorer outcomes relevant to the workers’ compensation setting (such as return-to-work). Furthermore, due to the very small prevalence of FELA cases available in this cohort, analyzing them as a separate sub-group will not be feasible. Exclusion of patients covered by private insurance (N = 13,\1 % of eligible cohort), who did not receive the full course of interdisciplinary functional restoration and were only treated on an out-patient basis approximately 2 or 3 times per week, mainly with physical therapy. Program non-completers (N = 589, 24 % of eligible cohort) were also excluded from the study. This exclusion criterion is necessary as 1-year outcomes on program non-completers are only obtained at a contact rate of approximately 50 %. Furthermore, no details on any other treatment modalities undergone by noncompleters during the usual 1-year program outcome period are available, and would therefore prevent accurate quantification of any post-program costs and their impact on the limited amount of 1-year outcomes available on non-completers.

A matched groups design was utilized for this study. Eligible patients were divided into three groups based on length of disability. The division of the groups was based on the chronic disability phases of the developmental disability model proposed by Krause et al. [11], as well the cohort structure utilized by Jordan et al. [12] to investigate the effect of extended disability on outcomes following functional restoration. The three groups in this study correspond to: early rehabilitation (ER: 4–8 months); intermediate duration (ID: 9–18 months); and delayed rehabilitation (DR: [18 months). The rationale for extending the duration of ER up to 8 months (as opposed to 6 months in the phase model of disability) is to take into account the time needed for recuperation from any surgery, as well as administrative time needed for insurance carrier approval for the patient’s enrollment in the functional restoration program. Matching Criteria The three groups were matched on sex, age, ethnicity, and injured region of the body (based on International

Statistical Classification of Diseases and Related Health Problems, Ninth Edition—ICD-9-codes). These factors are associated with poorer outcomes at post-treatment, such as lower return-to-work rates, lower likelihood of retaining work, and increased healthcare utilization during one- and 2-years follow-up [7, 16–21]. Following the application of the exclusion criteria described above, a cohort of N = 1,865 program completers remained eligible for the study. The patients were approximately equally distributed among the three groups with N = 683 (37 %), N = 608 (33 %), and N = 574 (31 %) corresponding to the ER, ID, and DR groups, respectively. The ER group was used as the index group based upon which the ID and DR groups were matched. The method of matching used was category matching [22, 23], to ensure that all groups had a similar distribution on age, sex, injured musculoskeletal regions, and ethnicity. Computer-generated matching was utilized to match each patient in the index ER group on the selected matching variables with individual patients from the two extended disability duration groups. Figure 1 summarizes the process of matching the three groups. The final matched groups consisted of N = 1,119 patients, with N = 373 patients in each of the three groups. All three groups had exactly similar distributions on the three demographic variables, as well as the injured musculoskeletal regions of the body. It should be noted that this method of group matching, when utilized in cohort studies, reliably eliminates the confounding effects of the matched variables among the groups and on the evaluated outcomes [24]. Procedure All patients were enrolled in a functional restoration program. Functional restoration is an intensive, medicallysupervised interdisciplinary program, combining quantitatively-directed exercise progression with a multimodal disability management approach, incorporating psychological and case management techniques [25]. Developed in 1983, this rehabilitation process is a variant of chronic pain management based on the biopsychosocial model of pain and disability, and is specifically intended for rehabilitation of compensation injuries (i.e., those for which patients receive financial compensation while determined medically disabled). The efficacy of functional restoration for CDOMD, as well as the objective outcomes for treatment monitoring, have been extensively reviewed in the literature [13–15, 26–29]. All patients within this cohort received an initial interdisciplinary evaluation. The first phase of the program included narcotic detoxification, psychotropic medication initiation, and aerobic and mobility training. Treatment progresses to an intensive second phase involving strength

123

J Occup Rehabil Fig. 1 Cohort selection

and endurance training under the direction of physical and occupational therapists. Concurrent psychosocial interventions are targeted towards addressing any fear-avoidance issues, and to achieve program goals of managing pain and returning to work. At discharge, patients were provided with education geared towards maintaining program goals, instruction on home/gym exercise regimens for fitness maintenance, occupational placement services, and any necessary psychotropic or anti-inflammatory medications as part of a long-term care plan. All data used in the present study were part of the patients’ standard medical files, and therefore the study was granted exemption status by the Institutional Review Board of the University of Texas at Arlington. All patients also signed a Health Insurance Portability and Accountability Act (HIPPA) authorization prior to program participation. All information was protected by this HIPPA rule, and was confidentially maintained in the institute’s database as part of ongoing quality assurance procedures. Measures Demographic Data and Medical History Patient medical records were utilized to obtain all demographic data (e.g., age, sex, ethnicity), medical history (e.g., length of disability, injuries, treatment modalities), occupational characteristics (pre-injury wage, pre-FR work status, pre-injury job availability), and WC -related information (e.g., types of disability payments, attorney retention).

123

Psychosocial Measures and Psychiatric Diagnosis Psychosocial measures were collected both at a pre-rehabilitation intake interview, and at the completion of the functional restoration program. Validated clinician-rated and self-report questionnaires used in this study included, the Hamilton Depression Rating Scale (HAM-D) [30], the Million Visual Analog Scale (MVAS) [31] for measuring perceived disability, and the visual analog scale measuring pain intensity [32]. In addition, post-injury Axis I psychiatric diagnoses, based on the diagnostic and statistical manual of mental disorders, fourth edition—text revision (DSM-IV TR) [33], were established at the pre-rehabilitation intake interview using the structured clinical interview for the DSM-IV (SCID-I) [34], and included post-injury major depressive disorder, anxiety disorders, and opioid dependence disorder. One-Year Socioeconomic Outcomes One-year post-rehabilitation socioeconomic outcomes were collected using a structured telephone interview [35]. Data were also gathered from additional sources, such as employers, insurance carriers, family members, and/or attorneys. These outcomes have been consistently reported from this program, and have been shown to be reliable discriminant indicators of patients who complete the program compared to those that refuse treatment or do not complete it, with direct contact rates of 90–93 %, and with partial data obtained on 98 % for all outcomes [13, 14, 36].

J Occup Rehabil

These socioeconomic outcomes used in this study included: •

• •

• • • •

Return-to-work (RTW): Return to work at any job demand level at any time during the post-rehab year, categorized as Yes or No; Work retention (WR): Work retained at 1-year postrehab at any job demand level, categorized as Yes or No; Percentage seeking treatment for original injury from a new healthcare provider: patient declining a long-term care plan (LTCP) with treating or referral doctor, and seeking new provider; Mean visits to a new healthcare provider for the original injury: measure of intensity of healthcare utilization; Number of new surgeries to original site of injury: includes only surgeries relevant to compensable injury; Number of new compensable injuries with or without work absence: new WC claims after work return; Type of persisting post-rehabilitation disability income, if any: covers private, state or federal payments.

Cost Estimation This study takes into account healthcare costs, estimated lost productivity due to disability, indemnity costs through disability insurance systems such as state and federal workers compensation, and return to work status of the patients under study. Therefore, it takes a broad societal perspective in the estimation of costs. All cost data discussed below were adjusted to the base year of 2011 US Dollars using both the general and medical care components of the Consumer Price Index. Pre-rehabilitation Medical Costs Because direct pre-rehabilitation medical costs were unavailable on our patients in this study, estimates for medical costs were obtained from a 10 % random sample (N = 4,726) of unique claims, between years 1998 and 2005, from a comprehensive claims database managed by the Ohio Bureau of workers’ compensation (OBWC). Medical costs included all associated costs for a workrelated injury, including surgeries, diagnostic tests, injection therapeutics, hospital costs, physical and occupational therapy modalities, chiropractic treatment, pain management, psychological services, case management services, pharmaceuticals, and costs associated with physicians’ consultation, evaluation and management services. The acquisition of the dataset was based on a match of ICD-9 diagnosis codes for musculoskeletal disorders that were prevalent in the PRIDE patient population. Further refining of a comparable subset from these data was undertaken to facilitate accurate estimates of medical costs by using the

category matching method to ensure that both the PRIDE and OBWC cohorts were comparable for the purpose of extracting estimates of medical costs (i.e., sex, age, length of disability, injured musculoskeletal regions, and type of surgery). Medical costs estimated in this study represent the costs reimbursed for medical care received. Lost Productivity and Disability Benefits The costs associated with lost productivity due to disability were estimated using the human-capital method of costing productivity [37, 38]. Specifically, this method entails multiplying the average weekly wage of the patient by the duration of time absent from work (i.e., time from injury to program admit–time working in any capacity = temporary total disability). For the costs associated with disability benefits, estimates were obtained by applying jurisdictional rules relevant to the individual WC system that patients were covered by. Specific rules for each state or federal system were obtained from an online directory of WC jurisdictions, maintained by the School of Labor & Industrial Relations at Michigan State University (http://www.lir.msu.edu/wcc/ wcid/wc_state.php). For the present study, quantifiable disability benefits included state- or federally-administered benefits associated with the Workers’ Compensation system such as wage replacement benefits, lump sum settlement payments for resulting disability from the workplace injury, and impairment benefits that are associated with any permanent disability and ratings of maximum medical improvement by healthcare providers. Rehabilitation Costs Cost data directly associated with the functional restoration program included treatment billing hours for the program, based on current reimbursement guidelines from the Texas state WC jurisdiction. Well over 95 % of the claims in this study utilize the identical hourly treatment billing code (including almost all out-of-state claims who agree to use Texas fees). Federal reimbursement guidelines were also hourly in nature, and standardized across the three groups. One-Year Post-rehabilitation Medical Costs Costs incurred during the 1-year post-rehabilitation period included any new surgeries for the originally injured compensable region, and the cost of visits to new healthcare providers. These costs were estimated from the OBWC database for all costs-per-claim associated with the specific type of surgery. The cost estimate for each category of surgery included the ‘‘all-in’’ costs reimbursed by workers’ compensation (that includes ancillary costs such as hospital charges, diagnostic tests, and anesthesia/surgeon’s fees).

123

J Occup Rehabil

Visits to new healthcare providers during the post-rehabilitation year were estimated using a unit cost per office-based physician visit provided by the Agency for Healthcare and Research Quality (AHRQ), adjusted to 2011 dollars [39]. Data Analysis Tests of association were conducted based on the Pearson Chi square (v2) test statistic for analyses of differences among the ER, ID, and DR groups, on all categorical variables. One-way analysis of variance (ANOVA) and analysis of covariance (ANCOVA) were conducted for analyses of differences among the ER, ID, and DR groups on continuous variables. For example, post-rehabilitation measures on the HAM-D, MVAS and pain VAS were analyzed with the ANCOVA method, using the respective baseline measures as the covariate. Both tests also included planned comparisons for linear trends across the disability duration groups. Effect sizes for all significant p values are reported consisting of: partial eta-squared (g2) for all continuous variables (0.01 = small, 0.09 = medium, 0.25 = large); and Cohen’s w for all categorical variables (\0.3 = small, 0.3–0.5 medium, [0.5 = large). Additionally, Dunnett’s post hoc tests were conducted to compare each of the late disability duration groups (ID and DR groups) against the ER group (significance of a = .025 after controlling for Type I error). For categorical variables, a logistic regression analysis was conducted, using disability duration groups as the predictor with ER as the reference group, and a Bonferroni correction that adjusted the critical value for significance to a = .025. The oddsratio and 95 % CI was reported for all significant post hoc comparisons on categorical variables. A full economic analysis is presented utilizing the costeffectiveness analysis (CEA) methodology recommended by the international panel on cost-effectiveness in health and medicine [40]. The analytical timeframe begins at the point of admission to the functional restoration program, and terminates at the point of data collection for the 1-year outcomes. The cost component of the CEA included the rehabilitation costs, as well as the 1-year post-rehabilitation costs. The denominator for the CEA utilized work retention (WR) at 1-year, which is an outcome relevant to the WC setting. The results of the CEA are represented as Average cost-effectiveness ratios and incremental cost-effectiveness ratios, respectively, defined by the following equations: Average Ratiogroup ¼

Average Costgroup WRgroup

ð1Þ

Incremental RatioGroup1 ¼

ðAverage CostGroup1  Average CostGroup2 Þ ðWRGroup1  WRGroup2 Þ

123

ð2Þ

The average ratios indicate the cost per unit outcome of WR for each of the groups. The incremental ratio defines the cost of one additional outcome of the intervention under consideration, relative to a comparator. The group with poorer outcomes serves as the comparator to enable computation of the cost of one additional outcome of the more effective rehabilitation strategy under consideration (i.e., early versus delayed). The non-parametric bootstrap method was utilized to generate a joint distribution of cost and outcome variables, thus providing a probability distribution of the parameter estimate for the cost-effectiveness ratios [41]. The bootstrap method is a technique used to overcome the limitations associated with the uncertainty around a single parameter estimate of the incremental cost-effectiveness ratios, and helps provide a confidence interval around the true estimate of the ratio. This method involved 2,000 samples for each of the disability duration groups, with each sample consisting of 373 observations (the sample size of each group) sampled with replacement. Per sample, the average cost-effectiveness ratio is calculated. Parameter uncertainty surrounding the average cost-effectiveness ratios were expressed using 95 % confidence intervals, and constructed using the percentile method by taking the 2.5th and the 97.5th percentile values of the resulting bootstrapped distribution of ratios. For the incremental costeffectiveness ratios, uncertainty was demonstrated using cost-effectiveness acceptability planes based on the distribution of bootstrapped estimates. Each plane described the incremental cost versus the incremental effectiveness of the ER group, relative to delayed rehabilitation groups. The probabilities associated with cost-savings (Southern Quadrants) and greater effectiveness (Eastern Quadrants) were indicated using the four quadrants of the plane. Data points appearing on the lower right-hand quadrant indicate the lower costs and greater effectiveness.

Results Demographic and Pre-rehabilitation Characteristics Table 1 summarizes the demographic and pre-rehabilitation characteristics of the three groups. Small percentages of each group were on social security disability income (SSDI) (0–6 %) or doing ‘‘light-duty’’ work (11–16 %), with the highest percentages among DR patients. Significant differences among the groups were observed on the percentage of patients receiving pre-rehabilitation surgery for the WC claim (p \ .001). Compared to the ER group, both the DR group (OR 9.9; 95 % CI 7.0, 14.0) and the ID group (OR 4.4; 95 % CI 3.2, 6.2) were more likely to have

J Occup Rehabil Table 1 Demographic and pre-rehabilitation characteristics of the cohort (N = 1,119) Variables

ER N = 373

ID N = 373

DR N = 373

13.0 (2.8)

37.5 (28.0)

p values group diff/ lin. trend

Effect size Partial g2 or Cohen’s w

.822

Length of disability in months (SD)

5.4 (1.5)

Log length of disability (SD)

1.7 (0.3)

2.5 (0.2)*

3.5 (0.5)*

\.001/\.001

44.7 (8.7) 53.1 (198)

45.1 (8.6) 53.1 (198)

45.5 (8.3) 53.1 (198)

.396/.174 1.000/1.000

57.9 (216)

57.9 (216)

57.9 (216)

Age (SD) Sex [% male (n)] Ethnicity [% (n)] Caucasian

1.000/1.000

African-American

21.2 (79)

21.2 (79)

21.2 (79)

Hispanic

20.9 (78)

20.9 (78)

20.9 (78)

Injured musculoskeletal regions [% (n)] Cervical only Lumbar/thoracic only Multiple spinal

1.000 2.1 (8)

2.1 (8)

2.1 (8)

43.4 (162)

43.4 (162)

43.4 (162)

5.6 (21)

5.6 (21)

5.6 (21)

Upper/lower extremities only

24.9 (93)

24.9 (93)

24.9 (93)

Multiple musculoskeletal (at least 1 spinal)

23.9 (89)

23.9 (89)

23.9 (89)

16.8 (62)

47.2 (174)*

66.5 (244)*

\.001/\.001

.413

5.7 (21)

20.9 (77)*

37.1 (136)*

\.001/\.001

.314

Other musculoskeletal surgeries

10.8 (40)

24.7 (91)*

30.5 (112)*

\.001/\.001

.200

Other procedures Attorney retained [% (n)]

1.4 (5) 11.8 (43)

3.3 (12) 24.7 (91)*

6.0 (22)* 25.5 (92)*

.003/.001 \.001/\.001

.103 .154

5.6 (22)*

\.001/\.001

.178

Received any type of surgery [% (n)] Spine surgeries

Receiving SSDI at admission [% (n)] Pre-injury weekly wage (SD)

0.3 (1) $509 ($252)

0.3 (1) $544 ($355)

$539 ($291)

Working at admission [% (n)]

14.9 (55)

11.3 (41)

16.1 (59)

Original job available [% (n)]

66.5 (244)

47.1 (172)*

34.4 (125)*

.237/.177 .150/.650 \.001/\.001

.264

* Significantly different compared to ER group, at p \ .05 after adjusting for inflation of Type I error

had surgeries between injury and rehabilitation with FR. Additionally, there was a significantly decreasing likelihood of having the original job available (p \ .001) as disability duration increased. Both the DR group (OR 3.3; 95 % CI 2.5, 5.0) and the ID group (OR 2.5; 95 % CI 1.7, 3.3) were also less likely to have the original job available, relative to the ER group. Finally, increased likelihood of being on SSDI was observed (p \ .001), with the DR group being more likely to be on SSDI, relative to the ER group at pre-rehabilitation (OR 22.2; 95 % CI 3.0, 165.9).

95 % CI 1.8, 3.9). The HAM-D revealed significant differences in clinician ratings of depressive symptoms among the groups (p \ .001), with small-to-moderate effects for both the DR group (M: 18.0; SD: 5.4; d: 0.4) and the ID group (M: 17.5; SD: 5.7; d: 0.3), relative to the ER group (M: 15.6; SD: 5.8). Additionally, the MVAS indicated overall significant differences among groups (p = .005). Post-hoc tests indicated that only the DR group (M: 96.8; SD: 23.8; d: 0.2) differed significantly from the ER group (M: 90.7; SD: 26.9).

Pre-rehabilitation Psychosocial Measures

Post-rehabilitation Psychosocial Measures

Table 2 summarizes the pre-rehabilitation psychosocial variables for three groups. Both the ID and DR groups evidenced significantly higher rates of major depressive disorder (MDD)—at approximately double the odds, relative to the ER group (p \ .001). Furthermore, significantly higher rates of post-injury opioid dependence disorder was observed for the DR group (p \ .001), relative to the ER group. Patients in the DR group were more likely to be dependent on opioids, relative to the ER group (OR 2.6;

At post-rehabilitation, all groups were comparable on psychosocial measures. However, a significant linear trend (p = .031) was observed for increasing depressive symptoms on the HAM-D, as disability duration increased (Table 2). One-Year Post-rehabilitation Socioeconomic Outcomes All groups were comparable on the objective socioeconomic outcomes, other than a significantly decreasing

123

J Occup Rehabil Table 2 Psychosocial and functional measures at pre- and post-rehabilitation (N = 1,119) Variables

ER N = 373

ID N = 373

DR N = 373

p values group diff/ lin. trend

Effect size Partial g2 or Cohen’s w

Any Axis I DX [% (n)]

51.4 (179)

70.0 (252)*

74.9 (262)*

\.001/\.001

.211

Major depressive disorder [% (n)]

43.4 (151)

61.1 (220)*

62.6 (219)*

\.001/\.001

.175

Anxiety disorder [% (n)] Opioid dependence disorder [% (n)]

12.4 (43) 13.5 (47)

16.9 (61) 18.9 (68)

18.9 (66) 29.1 (102)*

.056/.019 \.001/\.001

.160

15.6 (5.7)

17.5 (5.7)*

18.0 (5.4)*

\.001/\.001

7.9 (4.4)

8.7 (4.2)

9.4 (4.5)

.072/.031

PRE

6.6 (1.8)

6.5 (1.8)

6.8 (1.9)

.050/.103

POST

4.2 (2.2)

4.3 (2.2)

4.4 (2.2)

.656/.479

PRE

90.7 (26.9)

93.5 (23.9)

96.8 (23.8)*

.005/.001

POST

56.5 (29.6)

59.2 (29.3)

61.9 (30.5)

.612/.326

HAM-D (SD) PRE POST

.033

Pain intensity VAS (SD)

Million VAS (SD) .010

* Significantly different compared to ER group, at p \ .05 after adjusting for inflation of Type I error

linear trend (p = .049) for work retention at 1-year (Table 3). Patients in the late rehabilitation groups were more likely to be still receiving disability benefits at 1-year post-rehabilitation. Both the ID and DR groups were significantly more likely to be receiving financial benefits (p = .003), relative to the ER group (OR 2.9; 95 % CI 1.4, 6.0; OR 3.3; 95 % CI 1.6, 6.9 for the ID and DR groups respectively). Additionally, both the ID group (OR 14.0; 95 % CI 1.8, 107.5) and DR group (OR 30.9; 95 % CI 4.2, 228.6) were also more likely to be on federal SSDI at 1-year post-rehabilitation (p \ .001). Unlike ID and DR groups, there were no ER patients entering the federal SSDI system during the post-rehabilitation year. Total Cost of Illness The estimated cost of illness over the entire duration of disability includes the sum of the estimated medical costs, estimated disability benefits, estimated productivity losses, and the costs associated with the analytical timeframe of the functional restoration program (program costs ? 1year post-rehabilitation costs). Table 4 summarizes the estimated total cost of illness associated with this cohort. Relative to the DR group, the average savings in medical costs over the duration of disability was $54,912 (or approximately 64 % in costs saved) when rehabilitation is administered within 4–8 months post-injury. In a comparison of the ID group to the ER group, the average savings over the duration of disability amount to $14,694 or approximately 32 % in costs saved. The cost of accrued disability benefits can be potentially reduced by 80 % ($32,816), when early rehabilitation is administered,

123

relative to allowing the duration of disability to be prolonged beyond 18 months. With respect to the ID group, potential savings in disability benefits are 60 % ($12,346), if rehabilitation is administered early within 4–8 months post-injury. Similarly, the costs of lost productivity can be reduced by 80 % over the duration of disability when comparing the DR group to the alternative of early rehabilitation as demonstrated by the ER group. For the ID group, savings in productivity losses correspond to 55 %, relative to the ER group. In terms of the costs over the duration of the functional restoration program, savings of 43–56 % ($8,257–$14,156) are indicated for early rehabilitation, relative to late rehabilitation. Finally, in terms of the total cost of illness across the disability duration groups, a savings of 72 % ($167,075) was indicated for early rehabilitation, relative to the DR group. For the ID group, potential savings in total costs over the duration of disability correspond to 45 % ($55,304) if rehabilitation is administered early within 4–8 months post-injury. Cost-Effectiveness of the Functional Restoration Program The average costs per unit outcome of work retention can be defined as the costs incurred to return a patient to work and having them retain it at 1-year. These costs were: $13,604 (95 % CI $12,591–$14,731) for the ER group; $23,674 (95 % CI $22,002–$25,597) for the ID group; and $33,305 ($31,084–$35,791) for the DR group (Table 5). The non-overlapping confidence intervals indicate that the average cost per unit outcome of WR significantly differs among the three disability duration groups. Additionally,

J Occup Rehabil Table 3 One-year post-rehabilitation socioeconomic outcomes (N = 1,119) Variables

ER N = 337

ID N = 337

DR N = 337

p values group diff/ lin. trend

Return to work [% (n)]

89.4 (287)

88.3 (288)

84.8 (284)

.170/.072

Work retention [% (n)]

82.5 (259)

82.3 (265)

76.4 (252)

.082/.049

Seeking treatment from new provider [% (n)] Mean visits (SD) to new provider [for original compensable injury]

18.7 (61) 2.0 (6.2)

18.4 (61) 1.7 (5.2)

17.9 (60) 1.9 (6.1)

.960/.776 .776/.791

New surgery for original injury [% (n)]

1.5 (5)

2.4 (8)

2.7 (9)

.570/.310

Recurrent injury with lost time [% (n)]

2.3 (7)

0.9 (3)

0.9 (3)

.244/.147

Effect size Partial g2 or Cohen’s w

Disability income [% (n)] Wage replacement benefits

3.0 (9)

1.6 (5)

Case settlement benefits

3.3 (10)

8.9 (28)*

3.2 (10)

.386/.845

10.2 (32)*

.003/.001

Short-/long term disability (STD/LTD)

1.6 (5)

1.9 (6)

2.9 (9)

.535/.290

Social security disability income (SSDI)

0.3 (1)

4.4 (14)*

9.3 (29)*

\.001/\.001

.113 .172

* Significantly different compared to ER group, at p \ .05 after adjusting for inflation of Type I error

Table 4 Estimated savings in costs per claim over the duration of disability (N = 1,119) Estimated savings over duration of disability* Medical costs

ER N = 373

ID N = 373

DR N = 373

$31,003

$45,697

$85,915

Average savings per claim



$14,694

$54,912

% Saved



32 %

64 %

Disability benefits

$8,162

$20,508

$40,978

Average savings per claim



$12,346

$32,816

% Saved



60 %

80 %

$16,203

$36,211

$81,395

Average savings per claim



$20,008

$65,192

% Saved



55 %

80 %

$11,156

$19,413

$25,312

Functional restoration cost 1-YR surgical costs to original injured area (n = 22)**

$10,486 $47,502

$18,321 $59,301

$24,247 $37,686

1-YR cost of visits to new healthcare providers (n = 182)***

$983

$844

$969

Average savings per claim % Saved

– –

$8,257 43 %

$14,156 56 %

Productivity losses

CEA analytical timeframe (rehabilitation ? 1-year)

Total cost of illness

$66,525

$121,829

$233,600

Average savings per claim



$55,304

$167,075

% Saved



45 %

72 %

* All p values, except where noted below, were significant for differences between ER, ID, and DR at p \ 0.001 ** Non-significant, p value = 0.926 *** Non-significant, p value = 0.917

the negative ICER values for the ID and DR groups, when compared to the ER group, indicate early interdisciplinary functional restoration as the most favorable option. Figure 2 summarizes the probability of the bootstrap estimates within each quadrant of the incremental cost-effectiveness planes accompanying each comparison. These values represent the probability of a given strategy (i.e., early versus delayed rehabilitation) being cost-effective, based on the distribution of the bootstrapped estimates along the four quadrants of the plane. Specifically, the probability of the intervention being cost-saving in comparison with the comparator is given by the distribution of bootstrapped data points located in the South-East Quadrant and the South-West Quadrant, whereas the probability of greater effectiveness (expressed in percentage of work retention at 1-year) is given by the distribution of bootstrapped data points located in the North-East Quadrant and South-East Quadrant. Overall, the probability distribution indicated that either strategy of earlier rehabilitation (4–8 months, and 9–18 months) were likely to be cost-saving and more effective than delayed rehabilitation beyond 18-months of disability.

Discussion Overall, the present study provides preliminary evidence on the societal costs accrued as a result of prolonged disability following a work-related injury, and the potential cost-effectiveness of administering early interdisciplinary

123

J Occup Rehabil Table 5 Average cost-effectiveness ratios and incremental cost-effectiveness ratios (ICER) for the analytical timeframe of cost of rehabilitation (cost of functional restoration ? costs of 1-YR healthcare utilization) based on 1-year work retention (WR) Group

Cost of rehabilitation

% WR

Average ratio

(95 % CI)

DR

$25,312

0.76

$33,305

($31,084; $35,791)

ID

$19,413

0.82

$23,674

($22,002; $25,597)

ER

$11,156

0.82

$13,604

($12,591; $14,731)

D Cost

D WR

Group

Cost of rehabilitation

% WR

ICER

DR

$25,312

0.76

ID

$19,413

0.82

-$5,899

0.06

-$104,380

ER

$11,156

0.82

-$8,257

\ 0.01

-$474,465

Fig. 2 Incremental cost-effectiveness planes comparing disability duration groups on work retention

functional restoration within the 4–8 month duration of disability. Estimated savings that could potentially be obtained through early intervention strategies for CDOMDs was 72 %, or almost $170,000 per claim, when considering patients with [18 months of disability that made up the DR group. Similarly, savings were also observed on medical costs, productivity losses, and disability benefits. Providing a context for these figures is challenging, due to the fact that estimates of total cost of illness stratified by duration of disability are not available in the literature. Furthermore, cost estimates are often quoted as annual expenditure amounts. Some useful

123

benchmarks, however, are available in the literature. For example, the average cost per claim (medical and disability benefits) in Texas for all injured workers’ with lost time claims was $14,323 during fiscal years 2005–2006 [42]. The average cost, nevertheless, masks the substantial amount of costs incurred by the small minority of patients who persist with chronic disability [2, 3]. Within this context, early functional restoration was also associated with significantly lower rates of medical utilization of surgical procedures, both in terms of overall rates as well as in terms of the number of surgical procedures per patient. Readers who are unfamiliar with the WC claim

J Occup Rehabil

setting may assume that patients in the ID and DR groups had more complex injuries or needed surgeries or other non-interdisciplinary interventional procedures, and therefore were ‘‘self-selecting’’ into those groups. However, that assumption would be correct only in a small minority of cases. While many factors may cause patients to move into ID and DR groups as defined in this group of chronic pain patients, the predominant reasons include primarily psychosocial issues or those associated with the indecisiveness created by national treatment guidelines and attendant disputes over treatment [43], rather than physiological findings [44]. Some of the critical factors include: (a) lengthy disputes over diagnostic tests, surgery or other treatments; (b) financial ‘‘secondary gain’’ on patient’s part, causing resistance to work return and case resolution; (c) disputes over disability between insurance carriers and attorneys; (d) patient ‘‘doctor shopping’’ and opioid use in response to perceived treatment failure; (e) ‘‘pain prone patients’’ with high perception of pain/disability, often with fear-avoidance beliefs; (f) terminated employment causing patients not to have work to return to, or pain limiting physical reconditioning leaving patients unable to perform heavy jobs. While there are many other factors involved in developing chronic pain patients, these predominate in this and similar populations, and ultimately are much more about a Biopsychosocial Model, than about severity of injury or complex surgery. In terms of outcomes following treatment, the functional restoration program resulted in comparable 1-year socioeconomic outcome rates independent of disability duration studied here, consistent with a previous study on the impact of disability duration on socioeconomic outcomes [12]. However, early rehabilitation in this patient population may potentially prevent or mitigate the development of complex psychosocial barriers to treatment, including depression, narcotic dependence, and complex sequelae of comorbid psychopathology, as demonstrated by the higher prevalence rates of DSM-IV Axis I disorders such as major depressive disorder and opioid dependence disorder in the late rehabilitation groups. Attention to early predictors and risk factors for disability and further healthcare utilization have received increased attention recently [45, 46], and may potentially play a role in reducing the likelihood of prolonged disability if addressed within an interdisciplinary-oriented early rehabilitation program. Despite comparable 1-year socioeconomic outcome rates among the three groups, a greater proportion of costshifting out of the WC system, and towards federal disability benefits on SSDI (Social Security Disability Income), was observed for both the late rehabilitation groups. Although the vast majority of patients across the three groups had their WC claim closed during the 1-year post-rehabilitation period (97–99 % in this cohort), this

metric masks the incurring of additional indemnity costs that perpetuate ‘‘disability’’ outside the WC system. It should also be noted that federal disability costs are not a stand-alone entity, but are likely accompanied by further productivity losses due to the injured worker not returning to an active occupational status, the entitlement to Medicare provided 2 years after SSDI acquisition, and the distortion of the true under-65 unemployment rate by these unbudgeted federal transfer payments [47]. Finally, the economic evaluation provides preliminary evidence for early interdisciplinary rehabilitation of patients persisting with disability [4 months. For each comparison corresponding to the incremental cost-effectiveness ratios of the disability duration groups, the probability of cost-savings and greater effectiveness was presented in incremental cost-effectiveness planes [48]. As would be expected from the negative ratios reported, the ER group demonstrated greater association with cost-savings compared to both ID and DR groups, and overall more effective relative to the DR group. The present study also departed from the traditional model of cost-effectiveness analysis utilizing self-reported quality of life measures. Considering that the WC system provides an opportunity for measuring objective outcomes directly linked to resource use, greater emphasis was placed on more objective outcomes, such as occupational status and healthcare utilization, rather than commonly used patientreported outcomes. In this regard, early work by Loisel et al. [49] has explored the use of the number of days disabled as the effectiveness component in comparing several treatment options as part of a cost-effectiveness analysis. The various stakeholders and decision makers within the Workers’ Compensation setting emphasize the importance of resuming an active occupational status. Therefore, a re-definition of effectiveness measures unique to the Workers’ Compensation setting should be formulated as part of the theoretical framework addressing the application of economic evaluation methods to issues related to occupational disability. Given the nature of this study, there are some obvious limitations in terms of the availability of data as well as the design of the study. Estimates of disability benefits were not all-inclusive, specifically on the costs of social security disability insurance (SSDI) or any private disability benefits like short-term disability (STD) and long-term disability (LTD). Only the presence or absence of these indemnity payments was noted in patient records. An additional source of cost data not available for inclusion was the costs of any prescription medications utilized during the 1-year post-rehabilitation period. A further limitation involved the design of the study, in that it was not a randomized controlled trial (RCT), thus preventing any inferences of causality. Even at the outset of injury, it

123

J Occup Rehabil

will be impossible to a priori randomly assign patients into disability duration groups. A prospective matched groups design was best suited to explore the research question of interest in this study (i.e., the effect of disability duration on both the treatment- and cost-effectiveness of a functional restoration program, with outcomes collected prospectively at 1-year post-treatment). Despite matching the groups on demographics and type of compensable injury (as defined by the ICD-9 codes), there are of course prognostic factors (e.g., psychosocial barriers such as major depressive disorder and opioid dependence) that differ among the groups, and that could potentially raise concerns about confounding by indication due to inherent differences in the duration of disability among patients in the three groups. It should be noted that it is virtually impossible to match for a priori or statistically adjust post hoc for such factors since prolonged disability, by its nature, is defined by greater psychosocial disorders [50]. While the results cannot ascertain causality on the impact of disability duration on cost and outcomes following functional restoration, it does provide evidence of associations that should be subject to more controlled research. A final limitation of this study is that the cohort consists of only program completers. Such an exclusion criterion may result in two potential confounders in this study, namely, introducing a selection bias among the disability duration groups, and risk of overestimating the true effect of early rehabilitation relative to the delayed rehabilitation groups. However, inclusion of non-completers is not possible for three main reasons. Firstly, 1-year socioeconomic outcomes are available only on a limited number of noncompleters, at a contact rate of approximately 50 %. This will complicate cost-effectiveness calculations utilizing return to work and work retention outcomes as well as 1-year cost estimates based on some of the 1-year outcomes, like visits to healthcare providers and new surgeries. Finally, program non-completers also constitute a separate population of patients that present at pre-rehabilitation with a unique set of psychosocial barriers to recovery that differentiate them from program completers. These different characteristics include elevated rates of Axis II DSM-IV disorders, increased medico-legal barriers to recovery, and significant amounts of secondary gain issues related to relationship with the employer [36, 51, 52]. For these reasons, it should be noted that the true effect sizes on outcomes and cost-effectiveness of early rehabilitation may be more modest if it were possible to include non-completers in the study. In conclusion, the data provide preliminary evidence on the cost-effectiveness of early interdisciplinary rehabilitation between 4 and 8 months post-injury. If this early intervention cannot be offered, providing interdisciplinary functional restoration during the intermediate duration (ID)

123

length of disability is more cost-effective than delaying until patients are more than 18-months disabled. These findings are consistent with the general principle, now widely accepted in the WC claim literature, that one of the ‘‘first principles to avoid delayed recovery’’ is critical in enhancing cost-effectiveness of care for acute and chronic musculoskeletal disorders. Despite several limitations, this study does provide a snapshot of the economic impact due to extended durations of disability. Acknowledgments The Ohio Bureau of Workers Compensation provided access to non-PHI data in their large claims database to facilitate estimates of medical costs reported in this study.

References 1. Mayer T, Polatin P, Smith B, Gatchel R, Fardon D, Herring S, et al. Spine rehabilitation: secondary and tertiary nonoperative care. Spine J. 2003;3:28–36. 2. Hashemi L, Webster BS, Clancy EA, Courtney TK. Length of disability and cost of work-related musculoskeletal disorders of the upper extremity. J Occup Environ Med. 1998;40(3):261–9. 3. Hashemi L, Webster BS, Clancy EA, Volinn E. Length of disability and cost of workers’ compensation low back pain claims. J Occup Environ Med. 1997;39(10):937–45. 4. Hashemi LB, Webster BS, Clancy EA. Trends in disability duration and cost of workers’ compensation low back pain claims (1988–1996). J Occup Environ Med. 1998;40(12):1110–9. 5. Franklin GM, Haug J, Heyer NJ, McKeefrey SP, Picciano JF. Outcome of lumbar fusion in Washington state worker’s compensation. Spine. 1994;19(17):1897–904. 6. Franklin GM, Mai J, Wickizer T, Turner JA, Fulton-Kehoe D, Grant L. Opioid dosing trends and mortality in Washington State workers’ compensation, 1996–2002. Am J Ind Med. 2005;48(2):91–9. 7. Maghout Juratli S, Franklin GM, Mirza SK, Wickizer TM, Fulton-Kehoe D. Lumbar fusion outcomes in Washington State workers’ compensation. Spine. 2006;31(23):2715–23. 8. Taylor VM, Deyo RA, Ciol M, Kreuter W. Surgical treatment of patients with back problems covered by workers compensation versus those with other sources of payment. Spine. 1996;21(19): 2255–9. 9. Chou R, Shekelle P. Will this patient develop persistent disabling low back pain? JAMA. 2010;303(13):1295–302. 10. Freburger JK, Holmes GM, Agans RP, Jackman AM, Darter JD, Wallace AS, et al. The rising prevalence of chronic low back pain. Arch Intern Med. 2009;169(3):251–8. 11. Krause N, Ragland DR. Occupational disability due to low back pain: a new interdisciplinary classification used in a phase model of disability. Spine. 1994;19:1011–20. 12. Jordan KD, Mayer TG, Gatchel RJ. Should extended disability be an exclusion criterion for tertiary rehabilitation? Socioeconomic outcomes of early versus late functional restoration in compensation spinal disorders. Spine. 1998;23(19):2110–6. 13. Mayer TG, Gatchel RJ, Kishino N, Keeley J, Capra P, Mayer H, et al. Objective assessment of spine function following industrial injury: a prospective study with comparison group and one-year follow-up. Spine. 1985;10(6):482–93. 14. Mayer TG, Gatchel RJ, Mayer H, Kishino ND, Keeley J, Mooney V. A prospective two-year study of functional restoration in industrial low back injury. An objective assessment procedure [published erratum appears in JAMA 1988 Jan 8;259(2):220]. JAMA. 1987;258(13):1763–7.

J Occup Rehabil 15. Rainville J, Kim RS, Katz JN. A review of 1985 Volvo Award winner in clinical science: objective assessment of spine function following industrial injury: a prospective study with comparison group and 1-year follow-up. Spine. 2007;32(18):2031–4. 16. DeBerard MS, Masters KS, Colledge AL, Schleusener RL, Schlegel JD. Outcomes of posterolateral lumbar fusion in Utah patients receiving workers’ compensation: a retrospective cohort study. Spine. 2001;26(7):738–46. 17. Dersh J, Mayer TG, Gatchel RJ, Polatin PB, Theodore BR, Mayer EA. Prescription opioid dependence is associated with poorer outcomes in disabling spinal disorders. Spine. 2008;33(20): 2219–27. 18. Gatchel RJ, Mayer TG, Kidner CL, McGeary DD. Are gender, marital status or parenthood risk factors for outcome of treatment for chronic disabling spinal disorders? J Occup Rehabil. 2005;15(2):191–201. 19. McGeary DD, Mayer TG, Gatchel RJ, Anagnostis C, Proctor TJ. Gender-related differences in treatment outcomes for patients with musculoskeletal disorders. Spine J. 2003;3(3):197–203. 20. Mannion AF, Elfering A. Predictors of surgical outcome and their assessment. Eur Spine J. 2006;15(Suppl 1):S93–108. 21. Mayer TG, Gatchel RJ, Evans TH. Effect of age on outcomes of tertiary rehabilitation for chronic disabling spinal disorders. Spine. 2001;26(12):1378–84. 22. Dicker RC. Designing studies in the field. In: Gregg MB, editor. Field epidemiology. New York, NY: Oxford University Press; 2002. 23. Rothman KJ, Greenland S, Lash TL. Modern epidemiology. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2008. 24. Aschengrau A, Seage GR. Essentials of epidemiology in public health. London: Jones & Bartlett Publishers; 2003. 25. Mayer TG, Gatchel RJ. Functional restoration for spinal disorders: the sports medicine approach. Philadelphia: Lea & Febiger; 1998. 26. Bendix AF, Bendix T, Vaegter K, Lund C, Frølund L, Holm L. Multidisciplinary intensive treatment for chronic low back pain: a randomized, prospective study. Cleve Clin J Med. 1996;63(1): 62–9. 27. Hildebrandt J, Pfingsten M, Saur P, Jansen J. Prediction of success from a multidisciplinary treatment program for chronic low back pain. Spine. 1997;22(9):990–1001. 28. Jousset N, Fanello S, Bontoux L, Dubus V, Billabert C, Vielle B, et al. Effects of functional restoration versus 3 hours per week physical therapy: a randomized controlled study. Spine. 2004; 29(5):487–93. 29. Mayer T, McMahon MJ, Gatchel RJ, Sparks B, Wright A, Pegues P. Socioeconomic outcomes of combined spine surgery and functional restoration in workers’ compensation spinal disorders with matched controls. Spine. 1998;23(5):598–605. 30. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960;23:56–62. 31. Anagnostis C, Mayer TG, Gatchel RJ, Proctor TJ. The Million Visual Analog Scale: its utility for predicting tertiary rehabilitation outcomes. Spine. 2003;28:1051–60. 32. McGeary DD, Mayer TG, Gatchel RJ. High pain ratings predict treatment failure in chronic occupational musculoskeletal disorders. J Bone Joint Surg Am. 2006;88(2):317–25. 33. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed, Text Revision. Washington, DC: American Psychiatric Association; 2000. 34. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured clinical interview for DSM-IV Axis I disorders, research version, non-patient edition (SCID-I/NP). New York: Biometrics Research, New York State Psychiatric Institute; 2002. 35. Mayer TG, Prescott M, Gatchel RJ. Objective outcomes evaluation: Methods and evidence. In: Mayer TG, Polatin PB, Gatchel

36.

37. 38. 39.

40.

41. 42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

RJ, editors. Occupational musculoskeletal disorders: function, outcomes and evidence. Philadelphia: Lippincott Williams & Wilkins; 2000. p. 651–67. Proctor TJ, Mayer TG, Theodore B, Gatchel RJ. Failure to complete a functional restoration program for chronic musculoskeletal disorders: a prospective 1-year outcome study. Arch Phys Med Rehabil. 2005;86(8):1509–15. Hodgson TA. Cost of illness in cost-effectiveness analysis: a review of the methodology. Pharmacoeconomics. 1994;6:536–52. Rice D, Hodgson T, Epstein A. The economic costs of illness: a replication and update. Health Care Financ Rev. 1985;7:61–80. Machlin SR, Carper K. Medical expenditure panel survey statistical brief #166: expenses for office-based physician visits by specialty, 2004. In: Medical expenditure panel survey. Washington, DC: Agency for Healthcare Research and Quality; 2007. Gold MR, Siegel JE, Russell LB, Weinstein MC, editors. Costeffectiveness in health and medicine. New York: Oxford University Press; 1996. Chernick MR. Bootstrap methods: a practitioner’s guide. Wiley series in probability and statistics. New York: Wiley; 1999. Workers’ Compensation Research Institute. Average benefit payment per claim with more than 7 days of lost time, 2006/2007 and 2004/2007. 2007. Available from http://www.wcrinet.org/ benchmarks/benchmarks_08/benchmarks_08_fig_1.a.html. Accessed 5 Jan 2009. Gatchel RJ, Kishino ND, Noe C. ‘‘Carving-Out’’ Services from Multidisciplinary Chronic Pain Management Programs: Negative Impact on Therapeutic Efficacy. In: Schatman ME, Campbell A, editors. Chronic pain management: guidelines for multidisciplinary program development. New York, NY: Informa Healthcare; 2007. p. 39–48. McNee P, Shambrook J, Harris EC, Kim M, Sampson M, Palmer KT, Coggon D. Predictors of long-term pain and disability in patients with low back pain investigated by magnetic resonance imaging: a longitudinal study. BMC Musculoskelet Disord. 2011;12:234. Gatchel RJ, Polatin PB, Noe C, Gardea M, Pulliam C, Thompson J. Treatment- and cost-effectiveness of early intervention for acute low-back pain patients: a one-year prospective study. J Occup Rehabil. 2003;13(1):1–9. Turner JA, Franklin G, Fulton-Kehoe D, Sheppard L, Stover B, Wu R, et al. ISSLS prize winner: early predictors of chronic work disability: a prospective, population-based study of workers with back injuries. Spine. 2008;33(25):2809–18. US General Accounting Office, SSA disability: other programs may provide lessons for improving return-to-work efforts: GAO01-153, 2001. Heitjan DF. ICER vs. INB: The statistical issues. Paper presented at the international conference on health policy statistics. Philadelhpia, PA; January 17th 2008. Loisel P, Lemaire J, Poitras S, Durand MJ, Champagne F, Stock S, et al. Cost-benefit and cost-effectiveness analysis of a disability prevention model for back pain management: a six year follow up study. Occup Environ Med. 2002;59(12):807–15. Gatchel RJ. Comorbidity of chronic mental and physical health disorders: the biopsychosocial perspective. Am Psychol. 2004;59:792–805. Howard KJ, Mayer TG, Theodore BR, Gatchel RJ. Patients with chronic disabling occupational musculoskeletal disorders failing to complete functional restoration: analysis of treatment-resistant personality characteristics. Arch Phys Med Rehabil. 2009;90(5):778–85. Dersh J, Mayer TG, Gatchel RJ, Towns B, Theodore BR, Polatin P. Psychiatric comorbidity in chronic disabling occupational spinal disorders has minimal impact on functional restoration socioeconomic outcomes. Spine. 2007;32(17):1917–25.

123

Cost-effectiveness of early versus delayed functional restoration for chronic disabling occupational musculoskeletal disorders.

Despite extensive evidence for the treatment effectiveness of interdisciplinary functional restoration (FR) for chronic disabling occupational musculo...
428KB Sizes 0 Downloads 8 Views