Managing Chronic Pain in Veterans With Substance Use Disorders

The proposed project will provide crucial data on the effectiveness of a pain management intervention designed for veterans with co-occurring pain and substance use disorders. The development of an empirically validated psychosocial intervention for managing pain and substance misuse could greatly enhance the current set of options for treating this large and understudied group of veterans.

Background: Chronic non-cancer pain and substance use disorders frequently co-occur in VA primary care and addictions treatment settings. The treatment for pain is complicated in veterans with a history of alcohol or drug misuse because of a potential for abuse and diversion of many prescription pain medications. Thus, treatment providers are left without successful and empirically-supported methods for treating the large number of patients with chronic pain who also have past or current substance use disorders. One potential strategy is to use cognitive-behavioral treatment (CBT) interventions to manage pain and decrease substance use abuse/misuse. Psychosocial interventions such as CBT have demonstrated efficacy for reducing pain and improving functioning in persons with a broad spectrum of pain-related problems. However, this form of treatment has not been explicitly tested in patients with co-occurring substance use disorders. Determining the effectiveness of a psychosocial intervention targeting both pain and substance use disorders would greatly increase the available treatment options for veterans with these conditions. Project Objectives: This project is designed to test the effect of a CBT intervention for patients with chronic pain who are already engaged in VA treatment for substance use disorders compared to an educational supportive condition on measures of pain intensity, pain-related disability and pain-tolerance. A set of secondary analyses will examine if group CBT compared to the educational supportive condition is associated with greater reductions in illicit substance use over the follow-up time period and whether during-treatment reductions in pain mediate the effect of treatment assignment on later substance use. Finally, a series of secondary analyses will explore whether the hypothesized mechanisms of action for the CBT intervention mediate the effect of this intervention on pain intensity and pain tolerance. Methods: The proposed study is a randomized controlled trial of group-based CBT compared to an educational supportive condition in patients currently receiving drug and alcohol treatment. At least 128 Veterans with alcohol or drug abuse or dependence and at least moderate pain currently treated in the Ann Arbor VA Substance Use Disorder treatment clinic will be assessed at baseline and then randomly assigned to one of these two conditions. These patients will be re-assessed at 3-months (i.e. immediately post treatment), 6-months and 12-months. The primary outcomes will be pain intensity, pain-related disability and pain tolerance. Mixed model regression analyses will be used for all primary analyses to estimate between-group differences in changes in individual slopes over time. Finally, a series of secondary analyses will explore (a) whether the intervention influences post-treatment substance use, (b) if change in pain during treatment mediates the effect of the intervention on subsequent substance use and (c) whether the hypothesized mechanisms of action for the CBT intervention mediate the effect of this intervention on pain intensity and pain tolerance.

cognitive-behavioral treatment (CBT) interventions to manage pain and decrease substance use abuse/misuse

cognitive-behavioral treatment (CBT) interventions to manage pain and decrease substance use abuse/misuse

The outcome measure below represents the average pain levels in each Arm/Group, across the entire follow-up period. Average pain in the last week (NRS-1) is a scale ranging from 0 (no pain) to 10 (worst possible pain). The results reported below are based on longitudinal repeated measures modeling, adjusted for baseline NRS-1 average pain level. The variance was modeled to account for time interval (3-mo, 6-mo, 12-mo), and for session block; at the time of randomization, participants were assigned to blocks for a series of 10 sessions, in parallel for the intervention and control. Session blocks were utilized as a blocking variable to address between-block variation.

The WHY MPI General Activity score is a composite construct designed to assess the extent to which physical pain impacts various aspects of daily living. Aspects include ability to perform chores inside the home, activities away from the home, and social functioning. The outcome measure below represents the average pain levels in each Arm/Group, across the entire follow-up period, and was adjusted for baseline WHY MPI General Activity score.

#days used alcohol in the past 30 days, as assessed via timeline follow-back(TLFB) calendars. #days in a controlled environment (e.g. hospitalization, incarceration) also recorded to identify the duration of days at risk. The outcome measure below represents the average #days alcohol use in each Arm/Group, across the entire follow-up period, and was adjusted for baseline TLFB #days used alcohol.

# days used illicit drugs in the past 30 days, as assessed via timeline follow-back (TLFB) calendars. Use of illicit drugs and misuse of prescription drugs (e.g., Rx opioids) were recorded for the 30 days prior to the follow-up visit. #days in a controlled environment (e.g., hospitalization, incarceration) also recorded to identify the duration of days at risk. The outcome measure below represents the average #days drug use in each Arm/Group, across the entire follow-up period, and was adjusted for baseline TLFB #days used illicit drugs.

Pain tolerance was measured by the amount of time the participant could hold their hand immersed in a vessel of cold water in seconds. The maximum allowed duration of the test was two minutes. The outcome measure below represents the average duration of task in each Arm/Group, across the entire follow-up period, and was adjusted for baseline level of pain tolerance.

As measured via CPSS PSE score at each time point. The CPSS is a 22-item questionnaire designed to measure chronic pain patients' perceived self-efficacy to cope with the consequences of chronic pain (Anderson et al. 1995). The CPSS PSE subscale is a measure of self-efficacy for pain management. These scales have high reliability. Scores range from 10-100, with higher scores representing a better outcome. The outcome measure below represents the average CPSS PSE score in each Arm/Group, across the entire follow-up period, and was adjusted for baseline CPSS PSE.

The CPSS is a 22-item questionnaire designed to measure chronic pain patients' perceived self-efficacy to cope with the consequences of chronic pain (Anderson et al. 1995). The CPSS subscale, self-efficacy for physical function (FSE), utilized to assess management of pain-related disability, with respect to several aspects of daily functioning. This scale have high reliability. Scores range from 10-100, with higher scores representing a better outcome. The outcome measure below represents the average CPSS FSE score in each Arm/Group, across the entire follow-up period, and was adjusted for baseline level of CPSS FSE.

Per study protocol, participants were asked to attend 10 sessions, which comprised the intervention for CBT participants and educational support for EUC participants. Some participants in each group did not attend any sessions, yet participated in follow-up assessments. The As-treated analysis was undertaken to evaluate the impact of the intervention, among participants that attended at least 1 session. As in the Primary Analysis, average pain last week (NRS-1) is a scale ranging from 0 (no pain) to 10 (worst possible pain). The results reported below are based on longitudinal repeated measures modeling, adjusted for baseline NRS-1 average pain level. The variance was modeled to account for time interval (3-mo, 6-mo, 12-mo), and for session block; at the time of randomization, participants were assigned to blocks for a series of 10 sessions, in parallel for the intervention and control. Session blocks were utilized as a blocking variable to address between-block variation

As-treated Analysis: The Impact of Pain on Activities and Functioning: General Activity Score (WHY MPI)

The WHY MPI General Activity score is a composite construct designed to assess the extent to which physical pain impacts various aspects of daily living. The outcome measure below represents the average WHY MPI General Activity score in each Arm/Group, across the entire follow-up period, adjusted for baseline WHY MPI General Activity score, but restricted here to study participants that partook in at least one treatment group session.

Inclusion Criteria: A report of pain of at least moderate or greater over the three months prior to baseline assessment as indicated by an average score of four or greater on the Numeric Rating Scale of pain intensity (NRS-I; Farrar et al. 2001); Completion of an intake assessment in the Ann Arbor VA's SUD clinic within the three months prior to baseline assessment. The pain-related inclusion criterion will be used to ensure that participants are experiencing elevated levels of pain and, consequently, are appropriate for pain-related treatment. Exclusion Criteria: Acute suicidality based on responses on the Beck Depression Inventory (BDI) confirmed by in-depth assessment of patient by the research associate; Psychiatric condition that precludes participation in outpatient group treatment based on: a Mini-Mental State Examination (MMSE; Folstein et al. 1975) score less than 21, psychiatric hospitalization within the past month, or endorsement of current psychotic symptoms on the Brief Symptom Inventory (BSI; Derogatis 1993) combined with noticeable bizarre thoughts or behavior during the interview; Inability to speak and understand English; and The inability to give informed, voluntary, written consent. Decisions about exclusion criteria 1 and 2 will be made using a combination of self-reported symptoms and researcher's judgment.