| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
The Eugene McDermott Center for Pain Management, The University of Texas Southwestern Medical Center at Dallas, Dallas, Texas
Address correspondence and reprint requests to Dr. Robert J. Gatchel, Department of Psychiatry, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd., Dallas, TX 75390-9044. Address e-mail to robert.gatchel{at}utsouthwestern.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
IMPLICATIONS: Interdisciplinary pain management is effective and cost-effective in treating the major health problem of chronic pain. The present study demonstrated its efficacy using a prospective, 1-yr posttreatment evaluation methodology. Moreover, physical therapy "carve out" practices by insurance carriers had a negative impact on the outcomes, raising important medico-legal and ethical issues.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In 1999, Okifuji et al. (7) also reviewed this literature and found similar improvements. They concluded that multidisciplinary pain centers are effective in helping patients to attain treatment goals. Other reviewers have come to the same conclusion (8). Although some studies have suggested that the overall median effect size may be small (9), and that their helpfulness may diminish over time (10), there were methodological problems associated with these studies, such as the use of only cognitive-behavioral treatment in the programs evaluated by Morley et al. (9), and significant demographic and follow-up rate differences in the three centers assessed by Chapman et al. (10). Indeed, most recently, Turk (11), in a comprehensive review of the clinical and cost-effectiveness of treatments for chronic pain, again concludes that interdisciplinary pain management programs yield significantly better outcomes than other pain treatment approaches for the following factors: return to work, functional activities, closure of disability claims, health care utilization, and substantially fewer iatrogenic consequences and adverse events.
Thus, interdisciplinary pain management is an evidence-based therapeutic modality that has been systematically evaluated and demonstrated to be effective and cost-effective. It is widely embraced by the medical community at large, but a paradoxical problem has developed. Despite the empirical support, many third party payers "carve out" specific services for cost-saving purposes, thereby only covering some of the program’s component services for patients. As a result, these patients cannot take part in a true interdisciplinary program for their pain. They are left with "disconnected" health care professionals operating in different facilities, with no possibility of coordinated care. In an initial study evaluating only short-term effects, we found that such physical therapy "carve out" policies had a significantly negative impact on chronic patients’ pain relief when evaluated from pretreatment to treatment discharge (12). The major purpose of the present study was to more comprehensively evaluate such negative effects of physical therapy "carve outs," prospectively, over a 1-yr follow-up period.
The first goal of this study was to examine the efficacy of treatment for patients in an interdisciplinary pain management center. Patients were compared at pre- and posttreatment, as well as 1 yr, on a variety of outcome measures. In addition, comparisons were made between treatment program completers and treatment dropouts. The second goal was to examine treatment response in physical therapy "carve out" (PT out), relative to the no "carve out" group (PT in). It was hypothesized that the PT in group would show greater improvements, relative to those of the PT out group.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Procedure
Each patient was referred to the pain center by an outside treating health care professional or consultant. An initial intake was performed by one of the staff pain physicians. It included a detailed history, review of the medical records, and physical examination. Patients who had been referred for "medication only" or for interventional pain blocks would typically not undergo interdisciplinary evaluation. However, those patients designated for the full interdisciplinary program (and who are the topic of the present article) then underwent a physical therapy evaluation and an interview and testing by one of the psychological staff. A standard pretreatment assessment battery (reviewed below) was completed by patients during the initial evaluation. Each patient was then discussed in staffing by the entire treatment team, and assigned a treatment plan, which consisted of the following: 1) a specified number of physical therapy sessions; 2) individual cognitive-behavioral sessions with available options of biofeedback, relaxation training, sensation focus exercises, and/or self-hypnosis training; 3) a specified number of group educational and family sessions; and 4) patients with complicated psychological distress issues were referred for one or more appointments with the psychiatrist on the team, primarily for psychotropic medications stabilization.
Assignment of a treatment plan was on an individual tailored basis, determined by severity of pain complaints, psychological profile (including motivation and benefit potential), and level of physical conditioning. On average, there were 10 individual cognitive-behavioral sessions, 10 group educational sessions, 1 family session, 5–10 physical therapy sessions, 1–2 sessions for psychotropic medication monitoring, and 2–4 physician visit sessions. The interdisciplinary team also met at least three times for each patient: at initial evaluation to assign a treatment program, at mid-point to define progress or lack thereof, and at discharge to assess gains and arrange disposition. At these three points, patients completed self-report measures to gauge treatment progress, as listed in the next section.
If patients left the vicinity or were not compliant with their treatment recommendations, a team conference was held to discuss possible modifications in treatment, or to strategize ways to increase compliance. If this was not successful, the patient was discharged from the program and was considered a treatment dropout. Finally, 1 yr after discharge from the program, subjects were contacted by telephone and administered a structured series of follow-up questionnaires. A 100% contact rate was accomplished for this follow-up, although there were some missing data for some patients on various measures.
Instruments and Outcome Measures
The Beck Depression Inventory (BDI) (14) is a self-report measure of depression. A total score of 0–10 is considered normal; 11–14 mild depression; 15–18 moderate depression; 19–30 severe depression; and >30 very severe depression.
The Dallas Pain Questionnaire (DPQ) (15) is a 15-item analog, self-report scale measuring perceived pain and disability. Scores of 0–39 indicate mildly disabling pain, 40–84 moderately disabling pain, and ≥85 severely disabling pain.
The Medical Outcomes Short Form-36 Health-Status Survey (16) is a self-report measure of mental and physical functioning.
The Oswestry Pain Disability Questionnaire (17) measures perceived functional disabilities caused by pain.
In the Pain Drawing Visual Analog (VAS), patients mark their level of pain along a 10-cm visual analog scale.
The Multidimensional Pain Inventory (MPI) (18) yields three coping styles: adaptive, dysfunctional, and interpersonally distressed. There are also three nonprotypic profiles: hybrid, anomalous, and unanalyzable.
The Treatment Helpfulness Questionnaire (19) measures patients’ satisfaction with their assessment-treatment care.
More detailed reviews of these measures can be found in Gatchel (20).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
2 tests with a normal approximate to the binomial were used in analyzing the categorical data.
Program Completer Versus Dropout Comparisons
To examine differences between the completer and dropout groups at pretreatment, an independent samples t-test was conducted for each continuous measure of physical or psychosocial disability. No statistically significant differences were found between groups for any of the measures. In addition, a 2 analysis was conducted to examine differences in the percentage of each type of coping style as classified by the MPI between the two groups. No significant differences were found in the percentage of the coping styles between the two groups. It should be noted that the baseline scores on the DPQ place both of these groups in the "severely disabling pain" category; their scores on the BDI place them in the mild-moderate depression category. Such scores reflect a substantial degree of emotional distress associated with pain, which is reflective of most pain populations of this type.
To examine the effect of dropout on treatment outcome at 1-yr follow-up, one-between and one-within repeated-measures analyses of variance were conducted (Table 1). Significant group tests were found for the Oswestry, F(1,94) = 8.02, P = 0.006; BDI, F(1,95) = 7.30, P = 0.008; VAS, F(1,100) = 9.60, P = 0.003; DPQ, F(1,93) = 9.87, P = 0.002; and the Medical Outcomes Short Form-36 Health-Status Survey Mental Component Score, F(1,88) = 6.98, P = 0.010, with treatment completers showing significantly better outcomes at 1 yr after treatment. Significant Group X Time interaction effects were also found for the VAS, F(1,100) = 11.64, P = 0.001, and DPQ, F(1,93) = 4.82, P = 0.031.
|
2 analysis, we examined whether any significant change took place in MPI coping style in the dropout group between pretreatment and 1-yr follow-up. No significant changes were revealed. In contrast, 2 analysis of the changes in coping style for the treatment completers revealed significant changes, 2(4) = 13.81, P = 0.004. At pretreatment, 28.9% of the completers were adaptive copers, 22.2% were interpersonally distressed copers, and 33.3% were dysfunctional copers; the others were either anomalous or hybrid nonprototypic profiles. At 1-yr follow-up, 42.2% were adaptive copers, 13.3% were interpersonally distressed copers, and 8.9% were dysfunctional copers; the others were nonprototypic.
To measure changes in vocational status after the treatment program, patients were classified into one of the following categories at pretreatment and 1-yr follow-up: currently working; not working because of the original injury/pain; or not working for reasons not related to the original injury/pain. A 2 analysis, conducted at pretreatment, revealed no significant differences between completers and dropouts in work status. At 1-yr follow-up, however, a significant group effect was revealed: 2(2) = 7.59, P = 0.011, with 50.0% of treatment completers working at this time, 15.6% not working because of the original pain/injury, and 34.4% not working for other reasons. In contrast, for the treatment dropout group, only 38.0% were working, 36.6% had not returned to work because of the original pain/injury, and 25.4% were not working for other reasons.
Table 2 presents medication use at pretreatment and 1-yr follow-up for the two groups. 2 analyses revealed a significant difference at pretreatment in the amount of opioid usage between treatment completers and dropouts, 2(1) = 6.14, P = 0.002, reflecting the fact that a smaller percentage of treatment completers were taking opioids. No significant differences were found for benzodiazepine or antidepressant usage at pretreatment. At 1-yr follow-up, treatment dropouts were again almost twice as likely to be taking opioid medication relative to treatment completers, 2(1) = 2.80, odds ratio = 1.95 (confidence interval: 0.889, 4.297), P < 0.05. In addition, treatment completers were almost 3 times more likely than treatment dropouts to be taking antidepressant medication, 2(1) = 6.75, odds ratio = .298 (confidence interval: 0.116, 0.761), P = 0.005. No significant effects for benzodiazepine usage were found for either group at 1-yr follow-up. However, a significant decrease in antidepressant use was noted in the treatment dropout group between pretreatment and 1-yr follow-up, 2(1) = 11.77, P = 0.0005.
|
Physical Therapy "Carve Out" Subgroup Treatment Outcomes
Repeated-measures analyses of covariance were conducted for all measures across the posttreatment and 1-yr follow-up time intervals, with the pretreatment interval level serving as the covariate (Table 3). Significant group effects were found for the Physical Health Composite Score, F(1,39) = 5.37, P = 0.03; Oswestry, F(1,39) = 4.05, P = 0.05; and BDI, F(1,42) = 6.88, P = 0.01, with the PT out group exhibiting significantly worse functioning. A significant interaction effect was also found for the Mental Component Score, F(1,39) = 4.12, P = 0.05, indicating a continuing improvement in mental health status for the PT in group, but a decrease for the PT out group. 2 analyses were also conducted for vocational status differences between groups from pretreatment to posttreatment, as well as from posttreatment to 1-yr follow-up. However, the PT in group did show a significant change in vocational status from pre- to posttreatment, with the percentage of patients who were not working because of their original injury decreasing from 39.4% to 9.1%, 2 = 8.26, P = 0.003. This group also did not show a significant change in vocational status from posttreatment to 1-yr follow-up, indicating that their gains were maintained (Table 4).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
These effects were also found for medication use. The dropout group had a significantly larger amount of opioid usage at pretreatment relative to completers. This is consistent with the expectation that treatment dropouts report higher levels of pain disability at the beginning of a treatment program (12). At one-year follow-up, treatment dropouts were almost twice as likely as completers to be taking opioid medication, and were significantly less likely to be taking antidepressant medication. These findings suggest that treatment dropouts relied more on pain medication than completers to manage their pain, and were also less likely to be appropriately treated for comorbid depression. This notion is further supported by the finding that dropouts showed a significant decrease in antidepressant usage from pretreatment to one-year follow-up. Conversely, treatment completers demonstrated a decrease in opioid usage from pretreatment to posttreatment, as well as an increase in antidepressant usage. These findings are significant, because depression is a common comorbid condition associated with chronic pain, occurring in at least 50% of chronic pain patients. Also, when depression occurs in the context of a medical condition, including chronic pain, this can markedly worsen the prognosis of that condition because of symptoms such as sleep disturbance, decreased energy, chronic fatigue, and worrying. Thus, optimal treatment of depression may improve the outcome of both the depression and chronic pain (22). In addition, research suggests that many chronic pain patients experience substance abuse and dependence, most often with opioids and benzodiazepines (23). In light of these findings, a shift from using opioids to antidepressants in the completer group reflects a positive change, whereas the significant change in the dropout group from antidepressant to opioid use indicates a negative shift. Furthermore, the findings suggest that the treatment program had a lasting positive effect on decreasing opioid usage, and even decreased levels of depression in the completer group one year after treatment.
Relatedly, recent news reports about OxyContin drew renewed attention to what is known about opioids and chronic pain (24,25). A significant percentage of patients who are referred to pain clinics have a problem with medication usage (26). Furthermore, opioid therapy is associated with loss of effectiveness over time, and a lack of functional improvement (27). Even when opioid therapy is indicated for chronic pain, patients are likely to do better with opioids in an interdisciplinary program. The present study displayed how controlled use could be decreased, whereas simultaneously increasing psychosocial status and functioning. Completing this overall display of positive treatment completer outcomes, the dropout patients were associated with more health care utilization in terms of number of health care visits, relative to completers. These findings indicate that completed interdisciplinary treatment significantly reduces the need for patients to seek further medical attention. In fact, treatment dropouts had more than twice as many health care visits as completers in the year after discontinuing the program.
It was found that the PT out group exhibited significantly worse functioning, relative to the PT in group. In terms of vocational status, the PT in group showed a decrease in the percentage of subjects who were not working because of their original injury, and these gains were maintained at one-year follow-up. Conversely, the PT out group did not show significant changes in vocational status at either immediate posttreatment or one-year follow-up. These findings suggest that the PT out group did not experience the same benefits of interdisciplinary pain management as the subjects who received all of their treatment in the same clinic. Of course, it should be noted there were relatively few subjects in the PT out group, thus possibly creating a statistical power issue with respect to obtaining significant results. Thus, these results should be viewed as only preliminary. However, they do extend the results reported previously by Gatchel et al. (12) who found poor outcomes in the PT out group immediately posttreatment evaluation. The present study results extend these findings in demonstrating that such poor outcomes persist even at one-year posttreatment completion.
In conclusion, this one-year follow-up study demonstrated that treatment completers were associated with better physical and psychosocial functioning relative to dropouts, were better able to work, and used less opioid medication. In addition, they required medical attention less often than dropouts in the year after treatment. Treatment completers also demonstrated significant gains from pretreatment to posttreatment for all variables evaluated, and these gains were maintained at one-year follow-up. These findings were consistent with previous studies examining the efficacy of interdisciplinary pain management. In addition, the negative impact of physical therapy "carve out" on treatment response at posttreatment and one-year follow-up was found for physical and psychosocial functioning, with the PT in group showing better functioning. Also, the PT in group showed significant improvements in vocational status, whereas the PT out group did not. These results are especially noteworthy in this age of managed care because they suggest that "carving out" an important component of interdisciplinary care may compromise the therapeutic impact of all of the patients’ medical care and their satisfaction with treatment overall, as originally noted by Gatchel et al. (12). Other clinical researchers have also demonstrated the negative impact on outcomes by not providing full integrated interdisciplinary care for pain patients (28).
With the growing number of insurance carriers contracting treatment "carve outs," these data are especially important to consider because of their significant medico-legal and ethical issues, as well as vocational implications, for patients’ long-term improvement and independent financial security. This attempt to contain costs in the short term is also shortsighted, because chronic pain will continue to be a medical problem requiring additional future long-term treatment costs. With the new "Pain Care Bill of Rights" issued by the American Pain Foundation, chronic pain patients are now in a position to begin demanding the best standard of care for their chronic pain (i.e., interdisciplinary pain management).
Finally, a comment should be made concerning the cost-effectiveness of interdisciplinary pain management programs. Although such cost-effectiveness has been reported in previous studies (5), we did not evaluate this in the present study. Interdisciplinary pain management is, by definition, individualized but billable by specific services or by an hourly rate. Cost-effectiveness is defined by gains and productivity, decreased health care use, and decreased disability burden to a society as a result of interdisciplinary pain management. We would have to rely on previous studies to help define these costs for our present investigation. A future study might more comprehensively document individual costs for such a program, and then define cost benefits for each treated patient on the basis of criteria previously mentioned. For an acceptable cost utility analysis of interdisciplinary pain management, a measure of individual preference with community standardization should be used. However, such an analysis was beyond the scope of this study.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. K. Nicholas Managed Care Policies and Pain Management Programs Anesth. Analg., July 1, 2003; 97(1): 1 - 1. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
