B
r ie f
R
epo r ts
MILITARY MEDICINE, 180, 3:263, 2015
D e v e lo p m e n t and E valu atio n o f a B eh a vio ra l Pain M a n a g e m e n t T re a tm e n t P ro g ra m in a V e te ra n s A ffa irs M e d ic a l C e n te r KelceyJ. Stratton, P h D *ft; Mark C. Bender, PhD§; Jennifer J. Cameron, PhD*; Treven C. Pickett, PsyD, ABPP-Rp*fHfj**
ABSTRACT Chronic pain complaints are highly prevalent among Veterans seeking Veterans Affairs health care, and the implementation of effective behavioral health interventions is vital to meet patient needs. Research supports the use of cognitive behavioral therapy for the treatment of chronic pain; however, varying guidelines regarding length of treatment and modality (i.e., group vs. individual) complicate clinical planning and program development. This study aimed to evaluate treatment outcomes and equivalence of 3 variations (12, 10, and 6 weeks of group treatment) of cognitive behavioral therapy for chronic pain using clinical program data collected from Veterans enrolled in Veterans Affairs health services in a large tertiary care setting. Across groups, Veterans showed improvements in negative pain-related thinking and decreases in painrelated disability and distress. In general, patient outcomes regarding pain-related distress and disability for the 6-week group were equivalent or better than the 12- and 10-week groups. Preliminary results support the effectiveness of brief behavioral interventions for chronic pain. The findings have important practical implications, as briefer treatments may offer comparable therapeutic impact as longer, more time-intensive treatment protocols. This study offers a unique examina tion of treatment development and evaluation processes informed by real-world clinical needs and patient feedback.
INTRODUCTION Chronic pain is highly prevalent among U.S. military Veterans treated in the Veterans Affairs (VA) system of care, with upward of 50% of VA medical patients reporting chronic pain.1-3 Chronic pain refers to pain lasting longer than 3 to 6 months and persisting beyond the healing of the initial injury or disease process.4 In contrast to acute pain, which signals injury or disease and disappears once the underlying cause of pain has been treated or healed, chronic pain may *Hunter Holmes McGuire VA Medical Center, 1201 Broad Rock Boulevard (116-B), Richmond, VA 23249. fDepartment of Psychology, Virginia Commonwealth University, 806 West Franklin Street, P.O. Box 842018, Richmond, VA 23284. ^Virginia Institute for Psychiatric and Behavioral Genetics, Virginia Commonwealth University, 800 East Leigh Street, Biotech 1, Suite 101, Richmond, VA 23219. §Tree of Life Services, Inc., 3721 Westerre Parkway, Suite B, Richmond, VA 23233. || Department of Physical Medicine and Rehabilitation, Virginia Commonwealth University, 1223 East Marshall Street, 4th Floor, PO Box 980677, Richmond, VA 23298. ^Department of Psychiatry, Virginia Commonwealth University, 1200 East Broad Street, Richmond, VA 23298. **Defense and Veterans Brain Injury Center, Hunter Holmes McGuire VA Medical Center, 1201 Broad Rock Boulevard (116-B), Richmond, VA 23249. The views expressed are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs or any of the institutions with which the authors are affiliated. doi: 10.7205/MILMED-D-14-00281
M ILITARY MEDICINE, Vol. 180, March 2015
persist for months or years, and the underlying cause of pain might be unknown or poorly understood. Chronic pain is com plex and multidimensional, and a wide array of biomedical, psychosocial, and behavioral factors impact on the nature, severity, and persistence of pain and pain-related disability. Chronic pain can result in numerous negative effects on an individual’s life. Pain intensity is associated with the individ ual’s perception of diminished control in his or her life, partic ularly as related to difficulties in social situations and changes in the activities of daily life, sleep, and appetite.5 Additionally, the suffering associated with chronic pain takes a significant emotional toll, and a growing literature supports a relationship between chronic pain conditions and psychiatric disorders,6 including depression7 and post-traumatic stress disorder (PTSD).8 Indeed, the presence of multiple comorbidities in pain may further complicate assessment and course of treat ment, and several studies demonstrate the increased health care burden, utilization, and costs for the treatment of pain and related disorders.9-12 Furthermore, a broader treatment approach to chronic pain becomes increasingly important as quality of life and mortality concerns regarding the use of opioids for chronic pain management rise.13' 14 Considering the physical and emotional distress and disability associated with chronic pain disorders, the development and dissemina tion of effective behavioral treatments for chronic pain, whether to serve as frontline interventions or to complement medical interventions, is a high priority for VA health care.
263
Clinical Program Evaluation of a Behavioral Pain Treatment for Veterans Cognitive behavioral therapy (CBT) is a leading modality in the behavioral treatment of chronic pain, and evidence sug gests that CBT has small to moderate effects on several aspects of pain, pain-related disability, mood, and pain-related think ing.15 CBT interventions aim to help patients evaluate the impact of pain on their lives, and encourage them to develop means of coping with pain chronicity, thereby developing more adaptive responses to pain in daily life.13 CBT treatments for pain attempt to address not only pain experience but also various factors contributing to overall functioning and quality of life. However, despite evidence that CBT provides some benefit to individuals coping with pain, there is a lack of consensus as to a unifying set of techniques or the mechanisms underlying the treatment.15 Further, CBT protocols for pain management vary widely in terms of length of treatment, from 4 16 to 10 or more sessions17 that may last 60 to 120 minutes each18-20 in either group or individual formats. Although meta-analyses support the effectiveness of CBT compared to other behavioral treatments,5’21 there is limited evidence regarding the comparative effectiveness of CBT variations. The issues of efficiency and efficacy are highly salient for treatment implementation in health care settings, particularly given the high patient needs for adequate pain management treatment and the often limited resources to administer these treatments. Practical considerations regarding time constraints, provider training and availability, and patient needs must be balanced with evidence-based theory and intervention in order to develop successful treatment programs. Following, this study presents and discusses the develop ment and evaluation of clinical program data collected from Veterans enrolled in a VA Medical Center behavioral pain management clinic. Specifically, we compare treatment out comes for three variations of a CBT for chronic pain protocol called emPower Ourselves W ith Every Resource (POWER); the variations of treatment regard differing lengths of a pain management treatment group (12, 10, and 6 weeks). This study examines the equivalence of CBT pain management protocols using clinical program data, and offers a unique examination of a treatment development and evaluation pro cess that was informed by real-world clinical needs and patient feedback. Given the vast number of Veterans seeking pain treatment in VA clinics and the limited resources for providing comprehensive, quality behavioral therapies, the development of high-impact, cost-effective treatments is a critical clinical concern.
METHODS Study Overview Participants were Veterans receiving outpatient treatment for chronic pain at a Veterans Affairs Medical Center (VAMC) between 2010 and 2012. Veterans were referred for treatment by their VA medical or mental health providers, and all participants had an established diagnosis of chronic pain. Pain-related diagnoses varied widely and many Veterans pre
264
sented with multiple sources of pain. Common complaints included musculoskeletal pain in the knee, back, or neck, and neuropathic pain. Less common were headache or arthritis, and there were a few Veterans with phantom limb pain. Onset of pain was often noted to have resulted from injury, overuse, or the etiology was unknown. Veterans underwent a brief orientation and screening in which they were educated about the purpose of the treatment and asked to complete a brief questionnaire to assess functional difficulties because of pain. These screening questionnaires were reviewed by clinical staff, and those Veterans who expressed significant disability because of pain, wished to leam pain self-management strat egies, and were available to attend group elected to enroll in the treatment. At the first treatment session, Veterans completed self-report measures assessing severity of pain symptoms, painrelated disability, general psychological distress, and quality of life. The current investigation is a post hoc, nonrandomized treatment outcome study using archival data from patients treated in a behavioral pain management clinic. Assessment measures completed by patients as part of evidence-based treatments were entered into a clinical database by clinic staff. Outcome data included in this article were taken from all patients completing the pain management treatment over the time period noted above. All Veterans received the POWER group treatment as part of their routine clinical care. Veterans were not subject to any protocol-driven exclusion criteria or given incentives to participate in treatment. This archival study was reviewed and approved by the relevant institu tional review boards.
Participants The patients consisted of 1 Korean W ar Veteran, 28 Vietnam W ar Veterans, 15 post-Vietnam W ar Veterans, and 30 Gulf War/Operation Enduring Freedom/Operation Iraqi Freedom/ Operation New Dawn (OEF/OIF/OND) Veterans. The majority of the sample was male (n = 65; 86.7%) and African-American (« = 47; 62.7%), with a mean age of 53 years (SD = 10.2). All patients were treated by one of two licensed clinical psychologists (M.C.B., J.J.C.) with specialized training and experience providing behavioral pain treatments to Veterans in primary care, mental health, and physical rehabilitation milieus. Treatment sessions were not coded for treatment fidelity; however, therapists participated in regular peer consul tation as part of routine clinical practice to guard against thera pist drift in skills and to further develop the treatment program. On occasion, psychology interns or postdoctoral residents who expressed an interest in pain management cofacilitated the groups with the licensed clinical psychologists.
Intervention The POWER group was developed in response to a grow ing need for pain management treatment and programming as new Veterans entered the VA m edical system, and in particular, the local VAMC postdeploym ent prim ary care
MILITARY MEDICINE, Vol. 180, March 2015
Clinical Program Evaluation o f a Behavioral Pain Treatment for Veterans
clinics for OEF/OIF/OND Veterans. The authors (M.C.B. and J.J.C.) developed a chronic pain group (i.e., POWER) and treatment manual incorporating the extant literature on psychological treatment of chronic pain, which is con sistent with the Veterans Health Administration Directive on Pain Management.-2 This Directive requires interdisciplinary, multimodal approaches to pain management including a focus on quality of life as a treatment outcome (e.g., level of physical and psychosocial functioning), and it emphasizes behavioral therapies including patient participation in care and instruction in self-management. The content of the POWER manual draws from much of the literature on the cognitive behavioral treatment of pain.23-25 The treatment is a CBT group designed to address cognitive, affective, and behavioral aspects of chronic pain, and incorporates specific skills and techniques to increase daily activity, improve pain coping skills, and enhance overall quality of life. Following the last session of each group cohort, a satisfaction ques tionnaire was administered and these results, in addition to open discussions with group participants and between the groups’ facilitators, were used to make modifications to the group content and length. In the initial development of the group, the intervention was designed to consist of 12 onehour sessions delivered weekly. Following patient feed back, significant attrition in longer formats, and clinic needs (including increased referrals) over the course of 2 years, the treatment was further refined to 10 one-hour weekly sessions, and then to six 1-hour weekly sessions. Group facilitators initially coled the group, and then offered two separate groups to better accommodate increased referrals. M e a s u re s
Pain Catastrophizing Scale26
The Pain Catastrophizing Scale (PCS) is a 13-item scale designed to assess catastrophizing in clinical and nonclinical populations. Each PCS item is rated on a 5-point scale: 0 (not at all) to 4 (all the time). Higher scores indicate a greater degree of negative pain-related thinking. The total PCS score represents a single construct of general catastrophizing. The PCS has good reliability and validity in both clinical and nonclinical samples,26'27 and excellent internal reliability was found in the present sample (a = 0.91). World Health Organization Quality of Life—Brief Assessment6
The World Health Organization Quality of Life—Brief Assess ment (WHOQOL-BREF) is an abbreviated version of the 100-item WHOQOL quality of life assessment. The scale produces scores for four domains related to quality of life: physical health, psychological, social relationships, and envi ronment. Domain scores produced by the WHOQOL-BREF demonstrate good discriminant validity, content validity, inter nal consistency, and test-retest reliability in normative sam ples. Internal reliabilities for the domain scores were good in the current sample (as = 0.74-0.78).
MILITARY MEDICINE, Vol. 180, March 2015
Brief Symptom Inventory-18 Item29
General distress was assessed using the Brief Symptom Inventory-18 Item (BSI-18). The BSI-18 yields three primary dimension scores: somatization, depression, and anxiety, and a Global Severity Index (GSI), which summarizes the overall level of distress. The BSI-18 is adapted from the 53-item BSI,311 which is a shortened form of 90-item Symptom Checklist-90-Revised (SCL-90).31 Consistent with the longer forms of the scale, T scores > 63 on the GSI are considered to indicate clinically significant distress. Excellent internal reliability for the GSI was found in the present sample (a = 0.93). Pain Outcomes Questionnaire32
The Pain Outcomes Questionnaire (POQ) is a self-report mea sure of pain treatment outcomes. The scale yields a total score representing the following domains: pain, mobility, activities of daily living, vitality, negative affect, and fear. The POQ was developed on samples of Veterans undergoing inpatient or outpatient pain treatment at VA facilities. In normative outpa tient VA samples, a total score of 120 is considered to be at the 75th percentile. Excellent internal reliability for the total score was found in the present sample (a = 0.89). D a ta A n a ly s is
The program evaluation data were first analyzed using descriptive statistics to determine mean scores and standard deviations for the assessment variables across the three treat ment groups. Significant between-group differences on the demographic variables were analyzed using j 2 and ANOVA, as appropriate. Paired sample /-tests determined the degree of statistically significant change on each treatment outcome variable for each group. The treatment outcome data for each of the three group formats (i.e., 12, 10, and 6 sessions) were then compared in order to assess the equivalence of these variations on the POWER group protocol. Data were col lected as part of standard clinical care and patients were not randomized to groups. Accordingly, the ANCOVA statistical procedure was selected as an appropriate statistical test to evaluate treatment outcomes and conduct between-group comparisons. Because of potential differences in the nonrandomized treatment groups, ANCOVA controls for the dif ferences between groups and adjusts post-treatment scores based on variations in pre-treatment scores. In our analysis, treatment group was defined as the fixed factor, and post treatment scores on the pain-related negative thinking, general distress, pain-related disability and functioning, and quality of life indicators were the dependent variables. Pretreatment scores were entered as covariates. All analyses were conducted using SPSS version 20. RESULTS
We first analyzed differences in the demographic variables between the three treatment groups using %2 analyses for the
265
Clinical Program Evaluation of a Behavioral Pain Treatment for Veterans TABLE I.
Means, Standard Deviations, and Paired Samples i-Tests for Pre-Treatment and Post-Treatment Scores for the 12-, 10-, and 6Week Pain Management Treatment Groups Treatment Group 12-Week Group
Variables
10-Week Group
Pretreatment Post-Treatment M (SD) M (SD)
PCS Total Score 35.9 WHOQOL-BREF Domain 1 10.3 (Physical) Domain 2 10.9 (Psychological) Domain 3 8.6 (Social) Domain 4 11.9 (Environment) BSI-18 Total Score 50.8 POQ Total Score 109.7
t
df
-3.01** 35
6-Week Group
Pretreatment Post-Treatment M (SD) M (SD)
t
df
Pretreatment Post-Treatment M (SD) M (SD)
32.9(12.4)
28.2(13.5)
-2.38* 12
36.2 (8.7)
t
df
31.7 (9.6)
-2.31*
21
(9.5)
32.4(10.8)
(1.7)
10.5 (1.6)
0.55
16
11.4(1.5)
11.0(1.8)
-1.02
9
11.1 (1.2)
11.3 (1.4)
0.61
20
(2.6)
10.9 (2.1)
0.07
16
11.9(1.4)
10.9 (2.2)
-2.67*
9
11.5 (2.2)
12.3 (2.1)
1.74
19
(3.6)
9.0 (3.5)
0.73
17
9.3 (3.3)
9.7 (3.4)
0.71
10
10.9 (3.1)
11.8 (3.8)
1.65
20
(3.0)
12.5(1.9)
1.40
16
12.4 (2.6)
12.4 (2.9)
0.02
9
12.7 (2.5)
13.0 (2.2)
0.70
20
(10.3) (29.0)
50.0 (10.4) 108.0 (30.2)
49.7 (12.2) 96.4 (35.5)
-0.40 -1.36
-0.66 -0.46
36 50.5 (11.1) 38 101.9 (30.2)
10 50.6 (8.3) 13 113.7 (30.9)
50.1 (8.7) 91.3 (40.4)
-0.29 21 -3.40** 21
PCS, Pain Catastrophizing Scale; WHOQOL-BREF, World Health Organization Quality of Life Scale-Short Version; BSI-18, Brief Symptom Inventory-18 Item; POQ, Pain Outcomes Questionnaire. *p< 0.05; **p< 0.01.
categorical variables and ANOVA for the continuous vari able (i.e., age). The three treatment groups did not differ with respect to sex, ethnicity, or service era. There were statistically significant differences with regard to age, with the 12-week group having a lower mean age (Mage = 49.0, SD = 10.6) than both the 10-week group (Mage = 56.9, SD = 6.5) and the 6-week group (Mage = 57.7, SD = 8.7). We then explored whether patients in the three treatment groups reported improvements in pain-related negative think ing, general distress, pain-related disability and functioning, and quality of life following the intervention. Paired samples f-tests were used to highlight statistically significant changes in pre- and post-treatment scores for each of the three groups. Results are presented in Table I. Negative pain-related think ing decreased across all treatment groups. Next, we examined the covariates to test the assumption that the covariates should be independent, that is, they should be roughly the same in all treatment groups. We used an ANOVA test to determine whether the covariates (i.e., pre treatment pain catastrophizing, quality of life, distress, and
general pain disability scores) differed by the three treatment groups (Table II). The assumption of independence was sup ported for all covariates. Because of significant between-group differences for age, Veteran age was added as an additional covariate in the ANCOVA analyses. Results are presented in Table III. The groups were generally equivalent in terms of post-treatment scores after controlling for pre-treatment scores and patient age. Length of treatment only had a significant effect on post treatment POQ and WHOQOL-BREF psychological quality of life domain scores after controlling for pre-treatment scores, F(2, 70) = 4.21, p < 0.05, and F(2, 42) = 3.39, p < 0.05, respectively. Planned contrasts revealed that the 6-week group had significantly lower POQ total scores than the 12-week group, p < 0.01, 95% Cl (-30.8, -4 .3 ), indicating greater treatment gains with regard to decreased pain disability and distress in the shorter treatment group. For the WHOQOLBREF psychological domain, the 10-week group had signifi cantly lower scores than the 6-week group, p < 0.05, 95%
Analysis of Covariance Test of Differences in Pain Management Post-Treatment Scores, After Controlling for the Effects of Pretreatment Scores and Patient Age
TABLE III.
Analysis of Variance Test of Pretreatment Differences in the 3 Pain Management Treatment Groups TABLE II.
Variable
F
df
PCS Total Score WHOQOL-BREF Domain 1 (Physical) Domain 2 (Psychological) Domain 3 (Social) Domain 4 (Environment) BSI-18 Total Score POQ Total Score
0.55
2,69
P 0.58
1.69 0.89 1.86 0.10 0.75 0.68
2, 54 2,53 2, 54 2, 52 2,71 2, 72
0.20 0.42 0.17 0.90 0.48 0.51
PCS, Pain Catastrophizing Scale; WHOQOL-BREF, World Health Organiza tion Quality of Life Scale-Short Version; BSI-18, Brief Symptom Inventory-18 Item; POQ, Pain Outcomes Questionnaire.
266
Variable
F
df
P
PCS Total Score WHOQOL-BREF Domain 1 (Physical) Domain 2 (Psychological) Domain 3 (Social) Domain 4 (Environment) BSI-18 Total Score POQ Total Score
0.10
2, 66
0.91
0.42 3.39 0.57 0.31 0.01 4.21
2, 43 2, 42 2, 45 2, 43 2, 65 2, 70
0.66
r ie f
R
epo r ts
MILITARY MEDICINE, 180, 3:263, 2015
D e v e lo p m e n t and E valu atio n o f a B eh a vio ra l Pain M a n a g e m e n t T re a tm e n t P ro g ra m in a V e te ra n s A ffa irs M e d ic a l C e n te r KelceyJ. Stratton, P h D *ft; Mark C. Bender, PhD§; Jennifer J. Cameron, PhD*; Treven C. Pickett, PsyD, ABPP-Rp*fHfj**
ABSTRACT Chronic pain complaints are highly prevalent among Veterans seeking Veterans Affairs health care, and the implementation of effective behavioral health interventions is vital to meet patient needs. Research supports the use of cognitive behavioral therapy for the treatment of chronic pain; however, varying guidelines regarding length of treatment and modality (i.e., group vs. individual) complicate clinical planning and program development. This study aimed to evaluate treatment outcomes and equivalence of 3 variations (12, 10, and 6 weeks of group treatment) of cognitive behavioral therapy for chronic pain using clinical program data collected from Veterans enrolled in Veterans Affairs health services in a large tertiary care setting. Across groups, Veterans showed improvements in negative pain-related thinking and decreases in painrelated disability and distress. In general, patient outcomes regarding pain-related distress and disability for the 6-week group were equivalent or better than the 12- and 10-week groups. Preliminary results support the effectiveness of brief behavioral interventions for chronic pain. The findings have important practical implications, as briefer treatments may offer comparable therapeutic impact as longer, more time-intensive treatment protocols. This study offers a unique examina tion of treatment development and evaluation processes informed by real-world clinical needs and patient feedback.
INTRODUCTION Chronic pain is highly prevalent among U.S. military Veterans treated in the Veterans Affairs (VA) system of care, with upward of 50% of VA medical patients reporting chronic pain.1-3 Chronic pain refers to pain lasting longer than 3 to 6 months and persisting beyond the healing of the initial injury or disease process.4 In contrast to acute pain, which signals injury or disease and disappears once the underlying cause of pain has been treated or healed, chronic pain may *Hunter Holmes McGuire VA Medical Center, 1201 Broad Rock Boulevard (116-B), Richmond, VA 23249. fDepartment of Psychology, Virginia Commonwealth University, 806 West Franklin Street, P.O. Box 842018, Richmond, VA 23284. ^Virginia Institute for Psychiatric and Behavioral Genetics, Virginia Commonwealth University, 800 East Leigh Street, Biotech 1, Suite 101, Richmond, VA 23219. §Tree of Life Services, Inc., 3721 Westerre Parkway, Suite B, Richmond, VA 23233. || Department of Physical Medicine and Rehabilitation, Virginia Commonwealth University, 1223 East Marshall Street, 4th Floor, PO Box 980677, Richmond, VA 23298. ^Department of Psychiatry, Virginia Commonwealth University, 1200 East Broad Street, Richmond, VA 23298. **Defense and Veterans Brain Injury Center, Hunter Holmes McGuire VA Medical Center, 1201 Broad Rock Boulevard (116-B), Richmond, VA 23249. The views expressed are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs or any of the institutions with which the authors are affiliated. doi: 10.7205/MILMED-D-14-00281
M ILITARY MEDICINE, Vol. 180, March 2015
persist for months or years, and the underlying cause of pain might be unknown or poorly understood. Chronic pain is com plex and multidimensional, and a wide array of biomedical, psychosocial, and behavioral factors impact on the nature, severity, and persistence of pain and pain-related disability. Chronic pain can result in numerous negative effects on an individual’s life. Pain intensity is associated with the individ ual’s perception of diminished control in his or her life, partic ularly as related to difficulties in social situations and changes in the activities of daily life, sleep, and appetite.5 Additionally, the suffering associated with chronic pain takes a significant emotional toll, and a growing literature supports a relationship between chronic pain conditions and psychiatric disorders,6 including depression7 and post-traumatic stress disorder (PTSD).8 Indeed, the presence of multiple comorbidities in pain may further complicate assessment and course of treat ment, and several studies demonstrate the increased health care burden, utilization, and costs for the treatment of pain and related disorders.9-12 Furthermore, a broader treatment approach to chronic pain becomes increasingly important as quality of life and mortality concerns regarding the use of opioids for chronic pain management rise.13' 14 Considering the physical and emotional distress and disability associated with chronic pain disorders, the development and dissemina tion of effective behavioral treatments for chronic pain, whether to serve as frontline interventions or to complement medical interventions, is a high priority for VA health care.
263
Clinical Program Evaluation of a Behavioral Pain Treatment for Veterans Cognitive behavioral therapy (CBT) is a leading modality in the behavioral treatment of chronic pain, and evidence sug gests that CBT has small to moderate effects on several aspects of pain, pain-related disability, mood, and pain-related think ing.15 CBT interventions aim to help patients evaluate the impact of pain on their lives, and encourage them to develop means of coping with pain chronicity, thereby developing more adaptive responses to pain in daily life.13 CBT treatments for pain attempt to address not only pain experience but also various factors contributing to overall functioning and quality of life. However, despite evidence that CBT provides some benefit to individuals coping with pain, there is a lack of consensus as to a unifying set of techniques or the mechanisms underlying the treatment.15 Further, CBT protocols for pain management vary widely in terms of length of treatment, from 4 16 to 10 or more sessions17 that may last 60 to 120 minutes each18-20 in either group or individual formats. Although meta-analyses support the effectiveness of CBT compared to other behavioral treatments,5’21 there is limited evidence regarding the comparative effectiveness of CBT variations. The issues of efficiency and efficacy are highly salient for treatment implementation in health care settings, particularly given the high patient needs for adequate pain management treatment and the often limited resources to administer these treatments. Practical considerations regarding time constraints, provider training and availability, and patient needs must be balanced with evidence-based theory and intervention in order to develop successful treatment programs. Following, this study presents and discusses the develop ment and evaluation of clinical program data collected from Veterans enrolled in a VA Medical Center behavioral pain management clinic. Specifically, we compare treatment out comes for three variations of a CBT for chronic pain protocol called emPower Ourselves W ith Every Resource (POWER); the variations of treatment regard differing lengths of a pain management treatment group (12, 10, and 6 weeks). This study examines the equivalence of CBT pain management protocols using clinical program data, and offers a unique examination of a treatment development and evaluation pro cess that was informed by real-world clinical needs and patient feedback. Given the vast number of Veterans seeking pain treatment in VA clinics and the limited resources for providing comprehensive, quality behavioral therapies, the development of high-impact, cost-effective treatments is a critical clinical concern.
METHODS Study Overview Participants were Veterans receiving outpatient treatment for chronic pain at a Veterans Affairs Medical Center (VAMC) between 2010 and 2012. Veterans were referred for treatment by their VA medical or mental health providers, and all participants had an established diagnosis of chronic pain. Pain-related diagnoses varied widely and many Veterans pre
264
sented with multiple sources of pain. Common complaints included musculoskeletal pain in the knee, back, or neck, and neuropathic pain. Less common were headache or arthritis, and there were a few Veterans with phantom limb pain. Onset of pain was often noted to have resulted from injury, overuse, or the etiology was unknown. Veterans underwent a brief orientation and screening in which they were educated about the purpose of the treatment and asked to complete a brief questionnaire to assess functional difficulties because of pain. These screening questionnaires were reviewed by clinical staff, and those Veterans who expressed significant disability because of pain, wished to leam pain self-management strat egies, and were available to attend group elected to enroll in the treatment. At the first treatment session, Veterans completed self-report measures assessing severity of pain symptoms, painrelated disability, general psychological distress, and quality of life. The current investigation is a post hoc, nonrandomized treatment outcome study using archival data from patients treated in a behavioral pain management clinic. Assessment measures completed by patients as part of evidence-based treatments were entered into a clinical database by clinic staff. Outcome data included in this article were taken from all patients completing the pain management treatment over the time period noted above. All Veterans received the POWER group treatment as part of their routine clinical care. Veterans were not subject to any protocol-driven exclusion criteria or given incentives to participate in treatment. This archival study was reviewed and approved by the relevant institu tional review boards.
Participants The patients consisted of 1 Korean W ar Veteran, 28 Vietnam W ar Veterans, 15 post-Vietnam W ar Veterans, and 30 Gulf War/Operation Enduring Freedom/Operation Iraqi Freedom/ Operation New Dawn (OEF/OIF/OND) Veterans. The majority of the sample was male (n = 65; 86.7%) and African-American (« = 47; 62.7%), with a mean age of 53 years (SD = 10.2). All patients were treated by one of two licensed clinical psychologists (M.C.B., J.J.C.) with specialized training and experience providing behavioral pain treatments to Veterans in primary care, mental health, and physical rehabilitation milieus. Treatment sessions were not coded for treatment fidelity; however, therapists participated in regular peer consul tation as part of routine clinical practice to guard against thera pist drift in skills and to further develop the treatment program. On occasion, psychology interns or postdoctoral residents who expressed an interest in pain management cofacilitated the groups with the licensed clinical psychologists.
Intervention The POWER group was developed in response to a grow ing need for pain management treatment and programming as new Veterans entered the VA m edical system, and in particular, the local VAMC postdeploym ent prim ary care
MILITARY MEDICINE, Vol. 180, March 2015
Clinical Program Evaluation o f a Behavioral Pain Treatment for Veterans
clinics for OEF/OIF/OND Veterans. The authors (M.C.B. and J.J.C.) developed a chronic pain group (i.e., POWER) and treatment manual incorporating the extant literature on psychological treatment of chronic pain, which is con sistent with the Veterans Health Administration Directive on Pain Management.-2 This Directive requires interdisciplinary, multimodal approaches to pain management including a focus on quality of life as a treatment outcome (e.g., level of physical and psychosocial functioning), and it emphasizes behavioral therapies including patient participation in care and instruction in self-management. The content of the POWER manual draws from much of the literature on the cognitive behavioral treatment of pain.23-25 The treatment is a CBT group designed to address cognitive, affective, and behavioral aspects of chronic pain, and incorporates specific skills and techniques to increase daily activity, improve pain coping skills, and enhance overall quality of life. Following the last session of each group cohort, a satisfaction ques tionnaire was administered and these results, in addition to open discussions with group participants and between the groups’ facilitators, were used to make modifications to the group content and length. In the initial development of the group, the intervention was designed to consist of 12 onehour sessions delivered weekly. Following patient feed back, significant attrition in longer formats, and clinic needs (including increased referrals) over the course of 2 years, the treatment was further refined to 10 one-hour weekly sessions, and then to six 1-hour weekly sessions. Group facilitators initially coled the group, and then offered two separate groups to better accommodate increased referrals. M e a s u re s
Pain Catastrophizing Scale26
The Pain Catastrophizing Scale (PCS) is a 13-item scale designed to assess catastrophizing in clinical and nonclinical populations. Each PCS item is rated on a 5-point scale: 0 (not at all) to 4 (all the time). Higher scores indicate a greater degree of negative pain-related thinking. The total PCS score represents a single construct of general catastrophizing. The PCS has good reliability and validity in both clinical and nonclinical samples,26'27 and excellent internal reliability was found in the present sample (a = 0.91). World Health Organization Quality of Life—Brief Assessment6
The World Health Organization Quality of Life—Brief Assess ment (WHOQOL-BREF) is an abbreviated version of the 100-item WHOQOL quality of life assessment. The scale produces scores for four domains related to quality of life: physical health, psychological, social relationships, and envi ronment. Domain scores produced by the WHOQOL-BREF demonstrate good discriminant validity, content validity, inter nal consistency, and test-retest reliability in normative sam ples. Internal reliabilities for the domain scores were good in the current sample (as = 0.74-0.78).
MILITARY MEDICINE, Vol. 180, March 2015
Brief Symptom Inventory-18 Item29
General distress was assessed using the Brief Symptom Inventory-18 Item (BSI-18). The BSI-18 yields three primary dimension scores: somatization, depression, and anxiety, and a Global Severity Index (GSI), which summarizes the overall level of distress. The BSI-18 is adapted from the 53-item BSI,311 which is a shortened form of 90-item Symptom Checklist-90-Revised (SCL-90).31 Consistent with the longer forms of the scale, T scores > 63 on the GSI are considered to indicate clinically significant distress. Excellent internal reliability for the GSI was found in the present sample (a = 0.93). Pain Outcomes Questionnaire32
The Pain Outcomes Questionnaire (POQ) is a self-report mea sure of pain treatment outcomes. The scale yields a total score representing the following domains: pain, mobility, activities of daily living, vitality, negative affect, and fear. The POQ was developed on samples of Veterans undergoing inpatient or outpatient pain treatment at VA facilities. In normative outpa tient VA samples, a total score of 120 is considered to be at the 75th percentile. Excellent internal reliability for the total score was found in the present sample (a = 0.89). D a ta A n a ly s is
The program evaluation data were first analyzed using descriptive statistics to determine mean scores and standard deviations for the assessment variables across the three treat ment groups. Significant between-group differences on the demographic variables were analyzed using j 2 and ANOVA, as appropriate. Paired sample /-tests determined the degree of statistically significant change on each treatment outcome variable for each group. The treatment outcome data for each of the three group formats (i.e., 12, 10, and 6 sessions) were then compared in order to assess the equivalence of these variations on the POWER group protocol. Data were col lected as part of standard clinical care and patients were not randomized to groups. Accordingly, the ANCOVA statistical procedure was selected as an appropriate statistical test to evaluate treatment outcomes and conduct between-group comparisons. Because of potential differences in the nonrandomized treatment groups, ANCOVA controls for the dif ferences between groups and adjusts post-treatment scores based on variations in pre-treatment scores. In our analysis, treatment group was defined as the fixed factor, and post treatment scores on the pain-related negative thinking, general distress, pain-related disability and functioning, and quality of life indicators were the dependent variables. Pretreatment scores were entered as covariates. All analyses were conducted using SPSS version 20. RESULTS
We first analyzed differences in the demographic variables between the three treatment groups using %2 analyses for the
265
Clinical Program Evaluation of a Behavioral Pain Treatment for Veterans TABLE I.
Means, Standard Deviations, and Paired Samples i-Tests for Pre-Treatment and Post-Treatment Scores for the 12-, 10-, and 6Week Pain Management Treatment Groups Treatment Group 12-Week Group
Variables
10-Week Group
Pretreatment Post-Treatment M (SD) M (SD)
PCS Total Score 35.9 WHOQOL-BREF Domain 1 10.3 (Physical) Domain 2 10.9 (Psychological) Domain 3 8.6 (Social) Domain 4 11.9 (Environment) BSI-18 Total Score 50.8 POQ Total Score 109.7
t
df
-3.01** 35
6-Week Group
Pretreatment Post-Treatment M (SD) M (SD)
t
df
Pretreatment Post-Treatment M (SD) M (SD)
32.9(12.4)
28.2(13.5)
-2.38* 12
36.2 (8.7)
t
df
31.7 (9.6)
-2.31*
21
(9.5)
32.4(10.8)
(1.7)
10.5 (1.6)
0.55
16
11.4(1.5)
11.0(1.8)
-1.02
9
11.1 (1.2)
11.3 (1.4)
0.61
20
(2.6)
10.9 (2.1)
0.07
16
11.9(1.4)
10.9 (2.2)
-2.67*
9
11.5 (2.2)
12.3 (2.1)
1.74
19
(3.6)
9.0 (3.5)
0.73
17
9.3 (3.3)
9.7 (3.4)
0.71
10
10.9 (3.1)
11.8 (3.8)
1.65
20
(3.0)
12.5(1.9)
1.40
16
12.4 (2.6)
12.4 (2.9)
0.02
9
12.7 (2.5)
13.0 (2.2)
0.70
20
(10.3) (29.0)
50.0 (10.4) 108.0 (30.2)
49.7 (12.2) 96.4 (35.5)
-0.40 -1.36
-0.66 -0.46
36 50.5 (11.1) 38 101.9 (30.2)
10 50.6 (8.3) 13 113.7 (30.9)
50.1 (8.7) 91.3 (40.4)
-0.29 21 -3.40** 21
PCS, Pain Catastrophizing Scale; WHOQOL-BREF, World Health Organization Quality of Life Scale-Short Version; BSI-18, Brief Symptom Inventory-18 Item; POQ, Pain Outcomes Questionnaire. *p< 0.05; **p< 0.01.
categorical variables and ANOVA for the continuous vari able (i.e., age). The three treatment groups did not differ with respect to sex, ethnicity, or service era. There were statistically significant differences with regard to age, with the 12-week group having a lower mean age (Mage = 49.0, SD = 10.6) than both the 10-week group (Mage = 56.9, SD = 6.5) and the 6-week group (Mage = 57.7, SD = 8.7). We then explored whether patients in the three treatment groups reported improvements in pain-related negative think ing, general distress, pain-related disability and functioning, and quality of life following the intervention. Paired samples f-tests were used to highlight statistically significant changes in pre- and post-treatment scores for each of the three groups. Results are presented in Table I. Negative pain-related think ing decreased across all treatment groups. Next, we examined the covariates to test the assumption that the covariates should be independent, that is, they should be roughly the same in all treatment groups. We used an ANOVA test to determine whether the covariates (i.e., pre treatment pain catastrophizing, quality of life, distress, and
general pain disability scores) differed by the three treatment groups (Table II). The assumption of independence was sup ported for all covariates. Because of significant between-group differences for age, Veteran age was added as an additional covariate in the ANCOVA analyses. Results are presented in Table III. The groups were generally equivalent in terms of post-treatment scores after controlling for pre-treatment scores and patient age. Length of treatment only had a significant effect on post treatment POQ and WHOQOL-BREF psychological quality of life domain scores after controlling for pre-treatment scores, F(2, 70) = 4.21, p < 0.05, and F(2, 42) = 3.39, p < 0.05, respectively. Planned contrasts revealed that the 6-week group had significantly lower POQ total scores than the 12-week group, p < 0.01, 95% Cl (-30.8, -4 .3 ), indicating greater treatment gains with regard to decreased pain disability and distress in the shorter treatment group. For the WHOQOLBREF psychological domain, the 10-week group had signifi cantly lower scores than the 6-week group, p < 0.05, 95%
Analysis of Covariance Test of Differences in Pain Management Post-Treatment Scores, After Controlling for the Effects of Pretreatment Scores and Patient Age
TABLE III.
Analysis of Variance Test of Pretreatment Differences in the 3 Pain Management Treatment Groups TABLE II.
Variable
F
df
PCS Total Score WHOQOL-BREF Domain 1 (Physical) Domain 2 (Psychological) Domain 3 (Social) Domain 4 (Environment) BSI-18 Total Score POQ Total Score
0.55
2,69
P 0.58
1.69 0.89 1.86 0.10 0.75 0.68
2, 54 2,53 2, 54 2, 52 2,71 2, 72
0.20 0.42 0.17 0.90 0.48 0.51
PCS, Pain Catastrophizing Scale; WHOQOL-BREF, World Health Organiza tion Quality of Life Scale-Short Version; BSI-18, Brief Symptom Inventory-18 Item; POQ, Pain Outcomes Questionnaire.
266
Variable
F
df
P
PCS Total Score WHOQOL-BREF Domain 1 (Physical) Domain 2 (Psychological) Domain 3 (Social) Domain 4 (Environment) BSI-18 Total Score POQ Total Score
0.10
2, 66
0.91
0.42 3.39 0.57 0.31 0.01 4.21
2, 43 2, 42 2, 45 2, 43 2, 65 2, 70
0.66
