Testing Mediators of Reduced Drinking for Veterans in Alcohol Care Management

Testing Mediators of Reduced Drinking for Veterans in Alcohol Care Management Abstract Introduction Alcohol Care Management (ACM) is a manualized treatment provided by behavioral health providers working in a primary care team aimed at increasing patients’ treatment engagement and decreasing their alcohol use. Research has shown that ACM is effective in reducing alcohol consumption; however, the mechanisms of ACM are unknown. Therefore, the purpose of this study is to examine the mechanisms of change in ACM in the context of a randomized clinical trial evaluating the effectiveness of ACM. Materials and Methods This study performed secondary data analysis of existing data from a larger study that involved a sample of U.S. veterans (N = 163) who met criteria for current alcohol dependence. Upon enrollment into the study, participants were randomized to receive either ACM or standard care. ACM was delivered in-person or by telephone within the primary care clinic and focused on the use of oral naltrexone and manualized psychosocial support. According to theory, we hypothesized several ACM treatment components that would mediate alcohol consumption outcomes: engagement in addiction treatment, reduced craving, and increased readiness to change. Parallel mediation models were performed by the PROCESS macro Model 4 in SPSS to test study hypotheses. The institutional review boards at each of the participating facilities approved all study procedures before data collection. Results As hypothesized, results showed that treatment engagement mediated the relation between treatment and both measures of alcohol consumption outcomes, the percentage of alcohol abstinent days, and the percentage of heavy drinking days. Neither craving nor readiness to change mediated the treatment effect on either alcohol consumption outcome. Conclusions Findings suggest that ACM may be effective in changing drinking patterns partially due to an increase in treatment engagement. Future research may benefit from evaluating the specific factors that underlie increased treatment engagement. The current study provides evidence that alcohol use disorder interventions should aim to increase treatment engagement and reduce barriers to care. Alcohol Care Management (ACM) is a manualized intervention targeting treatment engagement and alcohol consumption.1 A behavioral health provider working in a primary care team conducts the sessions and provides alcohol use monitoring, support, and education. Behavioral health providers also encourage treatment adherence and promote the use of oral naltrexone (50 mg), an evidence-based pharmacotherapy that has been shown to attenuate alcohol cravings,2,3 reduce heavy alcohol consumption,4–7 and increase alcohol abstinence days.8,9 However, not all patients receive naltrexone due to either medication contraindication or lack of interest. Therefore, providers offer two forms of ACM: medical management individual counseling sessions in which patients take naltrexone and medical attention individual counseling sessions where patients receive psychosocial support alone.10 Oslin, Lynch, Maisto et al1 conducted a randomized clinical trial (RCT) to test the effectiveness of the ACM model of care. Patients in the RCT consisted of 163 U.S. military veterans who were randomized to receive either ACM or standard care in a specialty outpatient addiction treatment program. Details regarding components of each study arm are described in Table I. Results of the study showed that ACM was more effective than standard care in reducing the percentage of heavy drinking days (% HDD). However, it remains unclear how ACM works to reduce alcohol consumption. Identifying necessary components (mediators) of effective alcohol use disorder (AUD) treatments is critical to develop or improve interventions. Table I. Components of Each Study Arm Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Note. Complete details are described in the original study article.1 Table I. Components of Each Study Arm Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Note. Complete details are described in the original study article.1 One potential mediator of ACM’s effects on alcohol consumption is engagement in addiction treatment because engagement is a main target in the ACM model of care. Oslin et al1 showed that participants who received ACM were more likely to attend at least two addiction treatment visits per month compared with those who received standard care. Therefore, ACM increased engagement in addiction treatment and this increase may partly account for ACM’s effect on alcohol consumption. Although it is unclear how treatment engagement contributes to treatment outcomes,11 researchers have also examined engagement as a mediator in other treatment models, such as motivational interviewing (MI)12 and cognitive behavior therapy.13 A second potential mediator of ACM’s treatment effect is the attenuation of craving that may have occurred through ACM’s components that promote naltrexone use and introduce skills to cope with craving. One hypothesis is that naltrexone attenuates alcohol craving, which increases one’s ability to limit or abstain from alcohol consumption. This hypothesis is partly derived from neurobiological research, suggesting that alcohol activates the brain’s reward system,14 and conversely that naltrexone causes a μ-opioid receptor blockade, which may reduce the reinforcing properties of alcohol and attenuate cravings.15,16 Another hypothesis is that ACM may have influenced craving through skills training because the manualized behavioral intervention included a component where the provider introduced modules based on the veterans’ needs, one of which focused on coping with craving and urges.10 A third possible mediator of change in ACM is motivation or readiness to change. According to the transtheoretical model of change, individuals progress through several stages to make a behavioral change (e.g., reducing alcohol consumption).17 Further, empirically supported treatments such as MI and motivational enhancement therapy are based on the premise that motivation is central to achieving reductions in alcohol consumption. Indeed, research suggests that motivation predicts post-treatment drinking outcomes18–20 and mediates alcohol use intervention drinking outcomes.21,22 Consequently, increases in readiness to change may mediate reductions in alcohol consumption for patients receiving ACM. The purpose of this study is to examine the mechanisms of change in ACM in the context of the Oslin et al clinical trial.1 To this end, we investigated whether alcohol outcomes (percentage of alcohol abstinent days [% AAD] and % HDD) among veterans with AUD receiving ACM were mediated by (a) increased treatment engagement, (b) reduced alcohol craving, and (c) increased readiness to change. We hypothesized that treatment engagement, readiness to change, and reduced craving mediate drinking outcomes among those who receive ACM. Method Design and Sample This study involved secondary data analyses of a RCT evaluating the effectiveness of ACM, a primary care-based program for alcohol consumption. Complete details describing the original study are published elsewhere.1 Briefly, participants were randomly assigned to a 26-wk intervention of either ACM or standard addiction outpatient care, standard care (see Table I). The sample was 163 U.S. military veterans (96.9% male, Mage = 55.92, SD = 10.82, age range: 21–82 yr) from one of the three participating Veterans Health Administration (VHA) facilities who were referred to the study by their primary care physicians. Eighty-six veterans were randomized to ACM and 77 to standard care. A naltrexone prescription was filled by 56 (66.6%) and 9 (12.3%) veterans in the ACM condition and standard care condition, respectively. This discrepancy was expected because ACM providers were intended to promote naltrexone, whereas VA providers typically provide addiction pharmacotherapy at low rates.23 Rates in the standard care condition are comparable with the percentage of patients in AUD specialty care receiving naltrexone according to a nationally representative longitudinal study,24 but somewhat lower than in other surveys (17.5–24.6%).25 Participants were included in the study based on the following criteria: (a) at least 18 yr of age, (b) met DSM-IV criteria for current alcohol dependence, and (c) drank more than an average of two standard alcoholic drinks per day in the 60 d before randomization. Participants were excluded if they (a) endorsed current illicit drug (except marijuana) or opioid abuse or dependence; (b) exhibited symptoms of dementia, hearing loss, psychosis, or mania; or (c) participated in VHA substance abuse treatment (standard care) in the last 12 mo. Procedure Patients were referred to the Behavioral Health Laboratory,26 an integrated care program, by their primary care providers. This program delivers evidence-based mental health care in primary care and administers a battery of clinical assessments. Following receipt of referral, program staff conducted clinical assessments and offered research participation to eligible patients. Interested veterans were then scheduled for in-person baseline assessments with research staff where eligible participants were randomized to receive ACM or standard care. Follow-up interviews were completed by telephone or in-person at 3 and 6 mo following randomization. The institutional review boards at participating VHA facilities approved all study procedures before data collection. Measures Outcome Measure The outcomes examined in this study were % AAD and % HDD as measured by the Alcohol Timeline Followback (TLFB). Heavy drinking was defined as 5+ standard drinks in 1 day for men and 4+ for women.27 The TLFB uses a calendar format and memory cues to enhance retrospective self-reporting of daily alcohol consumption in standard drink units.28 The TLFB has good psychometric properties, including high reliability (ICC = 0.60–0.99)29 and validity (i.e., accuracy of reports; r = 0.74–0.89).30 At baseline, alcohol use for the previous 90 d was recorded. At each follow-up, the TLFB encompassed the time since the last interview. Mediator Measures The mediators examined were treatment engagement, alcohol craving, and readiness to change. Treatment engagement included the total number of addiction visits from baseline to 3-mo follow-up. Engagement in addiction treatment was tracked using the VHA electronic medical record supplemented by a questionnaire of services received outside the VHA. Specifically, participants reported the number of visits to a medical office outside of the VHA and specified whether the visit was mental health or substance abuse. Alcohol craving was measured using the Penn Alcohol Craving Scale (PACS), a five-item scale measuring past-week intensity, frequency and duration of cravings, and ability to resist acting on cravings. This instrument has been correlated with other measures of craving31 and demonstrated good to excellent internal consistency across time points (α = 0.899–0.936). Item scores range from 0 to 6 with higher scores indicating more craving; all items are summed to create a total score. Lastly, readiness to change was measured using the Contemplation Ladder.32 Participants used a numbered image of a ladder to rate their readiness to quit using alcohol on an 11-point scale (0 = “No thought of quitting” to 10 = “Taking action to quit”) containing five anchor statements representing each stage of change:33 (0–3) precontemplation, (4–6) contemplation, (7–8) preparation, (9) action, (10). These readiness categories are based on the transtheoretical model of change. Higher numbers represent greater readiness to change. Potential Covariates Additional measures were examined to assess for potential group differences. The Behavioral Health Laboratory assessments included demographic data such as age, gender, race, and employment status. The post-traumatic stress disorder (PTSD) Checklist-Situation (PCL-C) for DSM-IV was used to assess for PTSD symptoms related to any traumatic event.34 Symptom severity of depression was measured using the Patient Health Questionnaire (PHQ-9).35 The Shortened Inventory of Problems (SIP-2R) was used to assess consequences of drinking over the past 3 mo.36 These measures demonstrated good to excellent reliability in the current sample (α = 0.933, 0.993, and 0.893 for PCL-C, PHQ-9, and SIP-2R, respectively). Statistical Analysis Univariate statistics were conducted to describe participant characteristics. Chi-square analyses for categorical variables and Mann–Whitney U-tests for continuous variables were conducted to examine any differences between ACM and standard care in baseline characteristics. To examine treatment engagement, alcohol craving, and readiness to change as mediators of the treatment effect, we performed parallel mediation models by the PROCESS macro Model 4 in SPSS. This approach uses an ordinary least squares regression framework and produces a test of total, direct, and indirect effects.37 Parallel mediation models offer several advantages over conducting simple mediation models for each mediator. For example, parallel mediation provides information on whether the mediation is independent of the effect of the other mediator(s) and reduces the potential for parameter bias due to omitted variables.38 The model was specified as depicted in Figure 1. The independent variable was condition (ACM = +1, standard care = −1) and the parallel mediators were treatment engagement, alcohol craving, and readiness to change. Alcohol outcomes (% HDD or % AAD) were the dependent variables and pre-baseline levels of alcohol use were entered as covariates due to their significant correlations (Table II). Separate models were run for each alcohol outcome (i.e., % HDD or % AAD). The % HDD variables were transformed for normality using square root transformations before analyses. Temporal precedence of the mediators was established to increase confidence in interpreting study findings. Specifically, the total number of addiction visits from 0 to 3 mo was calculated for treatment engagement and change scores over the same time were calculated for alcohol craving and readiness to change (3 mo – baseline). Alcohol outcomes were calculated as change in % HDD and % AAD from 4 to 6 mo. Figure 1. View largeDownload slide Parallel mediation model analyzed in this study with craving change, treatment engagement, and readiness to change as parallel mediators. Figure 1. View largeDownload slide Parallel mediation model analyzed in this study with craving change, treatment engagement, and readiness to change as parallel mediators. Table II. Correlations Between Alcohol Use, Craving, Treatment Engagement, and Readiness to Change Variables r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 Note. % AAD, percentage of alcohol abstinent days, % HDD, percentage of heavy drinking days (square root transformed). Pre-baseline is based on 90 d before the baseline. Craving change and readiness to change is a change score (3-mo score − 0-mo score). Treatment engagement is a total score (0–3 mo). *p < 0.05; **p < 0.01; ***p < 0.001. N = 130. View Large Table II. Correlations Between Alcohol Use, Craving, Treatment Engagement, and Readiness to Change Variables r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 Note. % AAD, percentage of alcohol abstinent days, % HDD, percentage of heavy drinking days (square root transformed). Pre-baseline is based on 90 d before the baseline. Craving change and readiness to change is a change score (3-mo score − 0-mo score). Treatment engagement is a total score (0–3 mo). *p < 0.05; **p < 0.01; ***p < 0.001. N = 130. View Large We tested for mediation according to guidelines published by Baron and Kenney.39 Although the guidelines were provided in the context of single mediator models, the causal steps have also been extended to multiple mediator models.38 If mediation is present, paths a, b, and c (shown in Fig. 1) will be significant, and path c’ will decrease or become non-significant. The statistical significance of indirect effects (path ab) were assessed using 10,000 resamples and bias-corrected CI. The indirect effects were considered significant if zero was not within the 95% CI. Primary analyses required complete data for all variables entered into the model. Therefore, missing data were handled in two ways. First, given the distribution of data and empirical precedent, alcohol consumption variables were coded as missing when <85% of the TLFB data were available.40,41 When ≥85% of the TLFB data were available, the number of non-missing days were entered into the denominator to calculate alcohol use percentages. Second, participants who were missing any data included in the proposed models were subsequently removed from the main analyses. Of the 163 individuals randomized, 33 participants had missing data that restricted them from being entered into the parallel mediation analyses. We referenced published guidelines to estimate power; our sample of n = 130 provided approximately 75% (α 0.05) power to detect small to medium effects (0.26) in the mediation analyses.42 Results Patient Characteristics Table III depicts characteristics of randomized participants. The sample was mostly male, but racially diverse with approximately half identifying as Black (52.8%; Caucasian 42.3%; 4.9% other) and almost 5% as Hispanic. Twenty-one percent of the sample screened positive for probable PTSD (PCL-C ≥50).34 No baseline differences were found between individuals assigned to ACM or standard care (Table III). Participants with missing data were more likely to be Caucasian and have lower physical health scores (SF12 PCS score) (p < 0.05), as compared with those without missing data, but did not differ on any other baseline variables (p’s > 0.05). Table III. Descriptive Statistics for Study Variables Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Note. ACM, alcohol care management; SC, specialty care; % HDD, percentage of heavy drinking days; % AAD, percentage of alcohol abstinent days; PHQ-9, Patient Health Questionnaire (0–27 with higher scores indicating greater depression); SIP, Short Inventory of Problems; PCS-SF12, Physical Component Score from the Medical Outcomes scale (SF36); MCS-SF12, Mental Component Score from the Medical Outcomes scale (SF36); PTSD, probable post-traumatic stress disorder, ≥50 on the PTSD Checklist for DSM-IV (PCL-C); treatment engagement, total addiction treatment visits from baseline to 3 mo. aNo significant differences between ACM and SC conditions at baseline (all p’s > 0.063). bn = 130, represents the sample included in main analyses. Table III. Descriptive Statistics for Study Variables Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Note. ACM, alcohol care management; SC, specialty care; % HDD, percentage of heavy drinking days; % AAD, percentage of alcohol abstinent days; PHQ-9, Patient Health Questionnaire (0–27 with higher scores indicating greater depression); SIP, Short Inventory of Problems; PCS-SF12, Physical Component Score from the Medical Outcomes scale (SF36); MCS-SF12, Mental Component Score from the Medical Outcomes scale (SF36); PTSD, probable post-traumatic stress disorder, ≥50 on the PTSD Checklist for DSM-IV (PCL-C); treatment engagement, total addiction treatment visits from baseline to 3 mo. aNo significant differences between ACM and SC conditions at baseline (all p’s > 0.063). bn = 130, represents the sample included in main analyses. Naltrexone Prescription Data Prescription data for six participants (3.7 %) were missing (n = 4 ACM; n = 2 standard care). Of the 157 participants for whom prescription data were available, 65 (41.4%) participants filled at least one naltrexone prescription. Initial fill dates occurred on average 38.5 (SD = 47.54) days following randomization. Providers individualized the naltrexone dosage as indicated (M = 63.26 mg, SD = 24.06 mg); dosages ranged from 50 to 150 mg per day. Specific to the 130 participants included in the parallel mediation analyses, 54 participants (n = 7 standard care, n = 47 ACM) filled a naltrexone prescription. Of note, five ACM participants filled their initial naltrexone prescription after the 3-mo time point. However, among the 130 participants, there was no significant correlation between the amount of craving change and the number of days between randomization and the initial fill date. Evaluation of Research Hypotheses Table II presents the simple correlations among key variables for the participants included in these analyses (i.e., participants with complete data for parallel mediation models). Table IV displays results of the parallel mediation models (i.e., total, direct, indirect, and total indirect effects). The first parallel mediation analysis examined the hypothesized role of treatment engagement, craving change, and readiness to change as mediators between condition (ACM and standard care) and % HDD, controlling for pre-baseline % HDD. Except for path c, which was non-significant (p = 0.06), required mediation criteria were met for treatment engagement as a mediator of condition and % HDD (paths a1 and b1 were significant, p < 0.05). The total effect (path c, b = −0.513; p = 0.06) reduced in significance when the mediators were accounted for (path c’, b = −0.182; p = 0.51). The mediation effect of treatment engagement (indirect effect, path ab1) was significant according to bootstrapping results (b = −0.226 [−0.509, −0.024]). This analysis indicates that participants in the ACM condition were more engaged in treatment than those in the standard care condition (path a1, b = 1.845, p < 0.001), and increases in treatment engagement explained variance in reductions in the % HDD. Required mediation criteria were not met for craving change or readiness to change as mediators of condition and % HDD. Condition did not significantly influence craving change or readiness to change (path a2, b = −0.724; p = 0.39; path a3, b = 0.133; p = 0.60). Table IV. Parallel Mediating Effects of Craving Change, Treatment Engagement, and Readiness to Change in the Relation Between Condition and Alcohol Outcomes Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Note. % HDD, percentage of heavy drinking days square root transformed; % AAD, percentage of alcohol abstinent days; IE, indirect effect. All estimates are unstandardized. Confidence intervals (CIs) resulted from 10,000 bootstrap draws. Alcohol use outcomes are based on use 4–6 mo post-baseline. Change in craving and readiness to change are change scores (3−0 mo). Treatment engagement is total addiction treatment visits 0–3 mo. All models are controlling for pre-baseline value of the alcohol outcome examined in the model. *p < 0.05. †p = 0.06. ‡CI does not include zero. N = 130. Table IV. Parallel Mediating Effects of Craving Change, Treatment Engagement, and Readiness to Change in the Relation Between Condition and Alcohol Outcomes Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Note. % HDD, percentage of heavy drinking days square root transformed; % AAD, percentage of alcohol abstinent days; IE, indirect effect. All estimates are unstandardized. Confidence intervals (CIs) resulted from 10,000 bootstrap draws. Alcohol use outcomes are based on use 4–6 mo post-baseline. Change in craving and readiness to change are change scores (3−0 mo). Treatment engagement is total addiction treatment visits 0–3 mo. All models are controlling for pre-baseline value of the alcohol outcome examined in the model. *p < 0.05. †p = 0.06. ‡CI does not include zero. N = 130. Following similar procedures, a second parallel mediation analysis was conducted for % AAD, when controlling for pre-baseline % AAD. Criteria were met for treatment engagement as a mediator of condition and % AAD (paths a1, b1, and c were significant, p < 0.05). The total effect (path c, b = 5.973; p < 0.05) reduced in significance when the mediators were accounted for; the regression coefficient decreased and was no longer significant (path c’, b = 1.379; p = 0.64). The mediation effect of treatment engagement (indirect effect, path ab1) was significant (b = 3.627 [1.625, 6.675]). Results indicate that participants in the ACM condition were more engaged in treatment than those in the standard care condition (path a1, b = 1.843, p < 0.001), and increases in treatment engagement explained variance in increases in % AAD. Required mediation criteria were not met for craving change or readiness to change as mediators of condition and % AAD. Condition did not significantly influence craving or readiness to change (path a2, b = −0.715; p = 0.40; path a3, b = 0.136; p = 0.60) and readiness to change did not significantly influence % AAD (b3, b = 0.707; p = 0.45). Discussion The results of this study supported the hypothesis that ACM was effective in reducing alcohol consumption partly because participants attended more addictions treatment visits. Data showed that participants who received ACM as compared with standard care were less likely to consume any alcohol (p < 0.05) and less likely to drink heavily (p = 0.06). Notably, in contrast to our non-significant results regarding heavy drinking, the original RCT analyses by Oslin et al,1 which included all randomized participants, showed a significant treatment effect on heavy drinking days, favoring ACM. Treatment effects on drinking outcomes were mediated by number of treatment visits. Therefore, results suggest that treatment engagement is an important target for AUD treatment. The finding that engagement in addiction treatment was a key component in ACM aligns with research suggesting an association between level of involvement in addiction treatment and treatment outcomes.43–45 Various components of ACM likely contributed to improved treatment engagement. First, the ACM model of care encourages providers to discuss attendance with patients, specifically, to explore the reasons for missing appointments and to reinforce attendance.10 As such, providers may have assisted in problem-solving and increased patients’ commitment to engage in treatment. Also, given that theory suggests patients may be more invested in treatment when they set their own goals,46 another important component may have been that patients set their own drinking-reduction goals (as opposed to an abstinence-only goal). Finally, it is possible that by placing addiction treatment in primary care, as is the case with ACM, patients’ access to treatment services improved47 and consequently increased likelihood of engaging in treatment.48 Although data were not available in this study to test specific factors that may have improved treatment engagement (e.g., therapeutic relationship, commitment to treatment, good clinic access, perceived helpfulness of treatment, and decreased stigma), it is clear that ACM was successful in reducing alcohol use, in part by increasing treatment engagement. Future research may benefit from evaluating the specific factors that underlie increased treatment engagement so that these important factors can be targeted in interventions. Contrary to hypotheses, results indicated that alcohol craving and readiness to change did not mediate treatment effects. Therefore, findings suggest that craving and readiness to change are not mechanisms by which ACM has its effect on alcohol consumption. Although results do not support craving as a mediator of ACM’s treatment effect or the theorized biobehavioral mechanism of action of naltrexone,15,16 our findings do align with other research that did not find evidence for craving as a mediator of alcohol treatment outcomes.2,49 One explanation for null craving mediation findings in our study is the presence of co-occurring PTSD and AUD. Indeed, similar to another naltrexone treatment study with null craving mediation findings,18 participants in our study had co-occurring PTSD (20.9% based on the PCL-C). Co-occurring AUD and PTSD are known to bi-directionally affect one another, and research shows that trauma cues influence alcohol craving.50 Therefore, it is possible that naltrexone was alleviating cravings, but trauma cues or PTSD symptoms simultaneously increased craving. We conducted additional analyses on the relation between PTSD symptoms and craving to explore this hypothesized explanation for null mediation findings. Results showed that PTSD symptoms at baseline were not correlated with craving change (baseline to 3 mo; r = 0.609). However, the current study does not contain a measure of PTSD symptoms over time, which would provide a more sensitive test of this hypothesis. A second explanation for null mediation findings for craving is that the categorizations of treatment conditions (ACM and standard care) did not offer a precise test of the hypothesized relation that was based on the presumed effect of naltrexone. Namely, several participants in the ACM condition did not fill a naltrexone prescription, while prescriptions were filled by some participants in the standard care condition. Indeed, research shows a direct relation between naltrexone treatment compliance and improved alcohol outcomes.51 Further, several patients did not begin taking naltrexone at the outset of treatment, thereby limiting the ability to detect subsequent changes in craving from baseline to 3 mo. Therefore, to create a more precise test of craving as a mediator, secondary mediation analyses were conducted on a subgroup of participants who were either (1) in standard care and never filled a naltrexone prescription or (2) in ACM and filled a naltrexone prescription within 3 mo post-randomization. Although results of these secondary analyses may be limited by a reduced sample size (n = 96), results remained non-significant and did not support alcohol craving as a mediator (% HDD: path a2b = −1.458; p = 0.12; % AAD: path a2b = −1.082; p = 0.24). When interpreting the present findings, it is important to note that we are unable to disentangle the effects of naltrexone versus the behavioral treatment effects of ACM. For instance, it may be the combination of both, not one or the other that drives the treatment effect. Contrary to expectation, readiness to change was not a mediator of the effect of ACM on alcohol outcomes. This is surprising given the theory-based assumption that a prerequisite for making a change in drinking patterns is reaching a sufficiently high level of readiness to change.17 However, other studies also have produced null mediation findings for readiness to change,52 even when the intervention presumably targets readiness to change specifically.53–55 Nevertheless, it is worth noting that, on average, participants entering the study reported a relatively high level of readiness to change at baseline (M = 7.19/10, SD = 2.54), representative of being in the preparation stage of change. Therefore, ceiling effects may have limited findings; participants may not have had much room to increase readiness to change further over the treatment course. Notable strengths of this study include the recruitment of a diverse population at multiple VHA facilities, conservative testing of research hypotheses by controlling for pre-baseline alcohol consumption patterns, examining multiple mediators in a single model, and temporally isolating the mediators and outcome variables. However, several caveats should be considered when interpreting the results of this study. First, our results are not based on an intent-to-treat analysis because missing data were not tolerated in our analyses. Although individuals in the randomized conditions did not differ at baseline, those who had missing data were more likely to be Caucasian and score higher on a physical health measure, which may limit generalizability. Second, despite temporally ordering the mediator and outcome variables, it is possible that participants who were more likely to abstain or reduce alcohol consumption were also individuals who were more likely to engage in treatment.56 Third, our interpretation regarding the mediation effect for treatment engagement is just one interpretation. The relation is likely complex; for instance, it is possible that veterans who experience diminished craving may feel more inclined to attend appointments. Fourth, the treatment effects observed in this study are not specific to naltrexone treatment per se because not all ACM participants received naltrexone. Lastly, evidence suggests that treatment-naïve individuals are less ready to change and less likely to achieve abstinence than persons who have attended treatment more frequently.57 Therefore, differences may exist between this sample, which consists of individuals who have not attended addiction treatment in the last year and individuals who have attended treatment more recently. Future research should examine how ACM works across individuals with diverse treatment histories and varying levels of readiness to change. Also, the literature includes several other hypotheses (e.g., altering the subjective effects of alcohol, increasing cognitive control) that may underlie the effects of AUD treatments that use naltrexone.15 Future research should aim to empirically examine additional factors that are hypothesized to mediate the effects of naltrexone treatment on alcohol consumption. Conclusions This study represents the first examination of theoretically supported mediators of ACM. Our data indicate that ACM, a naltrexone-based AUD treatment, was effective in reducing alcohol consumption partly by increasing treatment engagement. Findings also suggest that readiness to change and reduced alcohol craving may not be essential to changing alcohol patterns. These results have important implications for researchers and clinicians in AUD treatment and highlight the importance of engaging participants in treatment. Interventions for AUD should aim to reduce barriers to care and increase treatment engagement. Funding Funding was provided by Health Services Research and Development Program of the Department of Veteran Affairs (IIR); The VISN 4 Mental Illness Research, Education, and Clinical Center at the Cpl. Michael J. Crescenz VA Medical Center; The VISN 2 Center for Integrated Healthcare; Career Development Award (K05 AA16928 [Dr. Maisto]). References 1 Oslin DW , Lynch KG , Maisto SA , et al. : A randomized clinical trial of alcohol care management delivered in Department of Veterans Affairs primary care clinics versus specialty addiction treatment . J Gen Intern Med 2013 ; 29 ( 1 ): 162 – 8 . Google Scholar CrossRef Search ADS PubMed 2 Richardson K , Baillie A , Reid S , et al. : Do acamprosate or naltrexone have an effect on daily drinking by reducing craving for alcohol? Addiction. 2008 ; 103 ( 6 ): 953 – 9 . Google Scholar CrossRef Search ADS PubMed 3 Monti PM , Rohsenow DJ , Hutchison KE , et al. : Naltrexone’s effect on cue-elicited craving among alcoholics in treatment . Alcohol Clin Exp Res 1999 ; 23 ( 8 ): 1386 – 94 . Google Scholar PubMed 4 Volpicelli JR , Alterman AI , Hayashida M , et al. : Naltrexone in the treatment of alcohol dependence . Arch Gen Psychiatry 1992 ; 49 ( 11 ): 876 – 80 . Google Scholar CrossRef Search ADS PubMed 5 O’Malley SS , Jaffe AJ , Chang G , et al. : Naltrexone and coping skills therapy for alcohol dependence. A controlled study . Arch Gen Psychiatry 1992 ; 49 ( 11 ): 881 – 7 . Google Scholar CrossRef Search ADS PubMed 6 Anton RF , Moak DH , Latham P , et al. : Naltrexone combined with either cognitive behavioral or motivational enhancement therapy for alcohol dependence . J Clin Psychopharm 2005 ; 25 ( 4 ): 349 – 57 . Google Scholar CrossRef Search ADS 7 Pettinati HM , O’Brien CP , Rabinowitz AR , et al. : The status of naltrexone in the treatment of alcohol dependence: specific effects on heavy drinking . J Clin Psychopharmacol 2006 ; 26 ( 6 ): 610 – 25 . Google Scholar CrossRef Search ADS PubMed 8 Maisel NC , Blodgett JC , Wilbourne PL , et al. : Meta-analysis of naltrexone and acamprosate for treating alcohol use disorders: when are these medications most helpful? Addiction 2013 ; 108 ( 2 ): 275 – 93 . Google Scholar CrossRef Search ADS PubMed 9 Rosner S , Hackl-Herrwerth A , Leucht S , et al. : Opioid antagonists for alcohol dependence . Cochrane Database Syst Rev 2010 ; ( 12 ): CD001867 . 10 Pettinati HM , Weiss RD , Miller WR , et al. Medical Management treatment manual. A Clinical Research Guide for Medically Trained Clinicians Providing Pharmacotherapy as Part of the Treatment for Alcohol Dependence COMBINE Monograph Series. 2004 ; 2. 11 Dearing RL , Barrick C , Dermen KH , et al. : Indicators of client engagement: influences on alcohol treatment satisfaction and outcomes . Psychol Addict Behav 2005 ; 19 ( 1 ): 71 – 8 . Google Scholar CrossRef Search ADS PubMed 12 Moyers TB , Moyers TB , Miller WR , et al. : How does motivational interviewing work? Therapist interpersonal skill predicts client involvement within motivational interviewing sessions . J Consult Clin Psychol 2005 ; 73 ( 4 ): 590 – 8 . Google Scholar CrossRef Search ADS PubMed 13 Spangler DL , Baldwin SA , Agras WS : An examination of the mechanisms of action in cognitive behavioral therapy for bulimia nervosa . Behav Ther 2004 ; 35 ( 3 ): 537 – 60 . Google Scholar CrossRef Search ADS 14 Koob GF , Roberts AJ , Schulteis G , et al. : Neurocircuitry targets in ethanol reward and dependence . Alcohol Clin Exp Res 1998 ; 22 ( 1 ): 3 – 9 . Google Scholar CrossRef Search ADS PubMed 15 Ray LA , Chin PF , Miotto K : Naltrexone for the treatment of alcoholism: clinical findings, mechanisms of action, and pharmacogenetics . CNS Neurol Disord Drug Targets 2010 ; 9 ( 1 ): 13 – 22 . Google Scholar CrossRef Search ADS PubMed 16 Spanagel R , Zieglgansberger W : Anti-craving compounds for ethanol: new pharmacological tools to study addictive processes . Trends Pharmacol Sci 1997 ; 18 ( 2 ): 54 – 9 . Google Scholar CrossRef Search ADS PubMed 17 Prochaska JO , DiClemente CC : Transtheoretical therapy: toward a more integrative model of change . Psychotherapy (Chic) 1982 ; 19 ( 3 ): 276 – 88 . Google Scholar CrossRef Search ADS 18 Maisto SA , Krenek M , Chung T , et al. : A comparison of the concurrent and predictive validity of three measures of readiness to change alcohol use in a clinical sample of adolescents . Assessment 2011 ; 23 ( 4 ): 983 – 94 . Google Scholar CrossRef Search ADS 19 Heather N , McCambridge J : Post-treatment stage of change predicts 12-month outcome of treatment for alcohol problems . Alcohol Alcohol 2013 ; 48 ( 3 ): 329 – 36 . Google Scholar CrossRef Search ADS PubMed 20 Hogue A , Dauber S , Morgenstern J : Validation of a contemplation ladder in an adult substance use disorder sample . Psychol Addict Behav 2010 ; 24 ( 1 ): 137 – 44 . Google Scholar CrossRef Search ADS PubMed 21 Stein LAR , Minugh PA , Longabaugh R , et al. : Readiness to change as a mediator of the effect of a brief motivational intervention on posttreatment alcohol-related consequences of injured emergency department hazardous drinkers . Psychol Addict Behav 2009 ; 23 ( 2 ): 185 – 95 . Google Scholar CrossRef Search ADS PubMed 22 Kelly JF , Myers MG , Brown SA : A multivariate process model of adolescent 12-step attendance and substance use outcome following inpatient treatment . Psychol Addict Behav 2000 ; 14 ( 4 ): 376 – 89 . Google Scholar CrossRef Search ADS PubMed 23 Harris AH , Kivlahan DR , Bowe T , et al. : Pharmacotherapy of alcohol use disorders in the Veterans Health Administration . Psychiatr Serv 2010 ; 61 ( 4 ): 392 – 8 . Google Scholar CrossRef Search ADS PubMed 24 Roman PM , Johnson JA : Adoption and implementation of new technologies in substance abuse treatment . J Subst Abuse Treat 2002 ; 22 ( 4 ): 211 – 8 . Google Scholar CrossRef Search ADS PubMed 25 Fuller BE , Rieckmann T , McCarty D , et al. : Adoption of naltrexone to treat alcohol dependence . J Subst Abuse Treat 2005 ; 28 ( 3 ): 273 – 80 . Google Scholar CrossRef Search ADS PubMed 26 Tew J , Klaus J , Oslin DW : The Behavioral Health Laboratory: building a stronger foundation for the patient-centered medical home . Fam Syst Health 2010 ; 28 ( 2 ): 130 – 45 . Google Scholar CrossRef Search ADS PubMed 27 Wechsler H , Dowdall GW , Davenport A , et al. : A gender-specific measure of binge drinking among college students . Am J Public Health 1995 ; 85 ( 7 ): 982 – 5 . Google Scholar CrossRef Search ADS PubMed 28 Sobell LC , Sobell MB : Timeline Followback User’s Guide: A Calendar Method for Assessing Alcohol and Drug Use . Toronto, ON , Addiction Research Foundation , 1996 . 29 Sobell LC , Agrawal S , Annis H , et al. : Cross-cultural evaluation of two drinking assessment instruments: alcohol timeline followback and inventory of drinking situations . Subst Use Misuse 2001 ; 36 ( 3 ): 313 – 31 . Google Scholar CrossRef Search ADS PubMed 30 Sobell LC , Sobell MB , Leo GI , et al. : Reliability of a timeline method: assessing normal drinkers’ reports of recent drinking and a comparative evaluation across several populations . Br J Addict 1988 ; 83 ( 4 ): 393 – 402 . Google Scholar CrossRef Search ADS PubMed 31 Flannery BA , Volpicelli JR , Pettinati HM : Psychometric properties of the Penn Alcohol Craving Scale . Alcohol Clin Exp Res 1999 ; 23 ( 8 ): 1289 – 95 . Google Scholar CrossRef Search ADS PubMed 32 Biener L , Abrams DB : The Contemplation Ladder: validation of a measure of readiness to consider smoking cessation . Health Psychol 1991 ; 10 ( 5 ): 360 – 5 . Google Scholar CrossRef Search ADS PubMed 33 Prochaska JO , DiClemente CC , Norcross JC : In search of how people change. Applications to addictive behaviors . Am Psychol 1992 ; 47 ( 9 ): 1102 – 14 . Google Scholar CrossRef Search ADS PubMed 34 Weathers FW , Litz BT , Herman DS , et al. : The PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. Paper presented at: annual convention of the international society for traumatic stress studies, San Antonio, TX 1993 . 35 Kroenke K , Spitzer R , Williams JB : The PHQ-9: validity of a brief depression severity measure . J Gen Intern Med 2001 ; 16 ( 9 ): 606 – 13 . Google Scholar CrossRef Search ADS PubMed 36 Kenna GA , Longabaugh R , Gogineni A , et al. : Can the short index of problems (SIP) be improved? Validity and reliability of the three-month SIP in an emergency department sample . J Stud Alcohol 2005 ; 66 ( 3 ): 433 – 7 . Google Scholar CrossRef Search ADS PubMed 37 Hayes AF : Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach . New York, NY, Guilford Press , 2013 . 38 Preacher KJ , Hayes AF : Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models . Behav Res Methods 2008 ; 40 ( 3 ): 879 – 91 . Google Scholar CrossRef Search ADS PubMed 39 Baron RM , Kenny DA : The moderator–mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations . J Pers Soc Psychol 1986 ; 51 ( 6 ): 1173 – 82 . Google Scholar CrossRef Search ADS PubMed 40 Stanger C , Scherer EA , Babbin SF , et al. : Abstinence based incentives plus parent training for adolescent alcohol and other substance misuse . Psychol Addict Behav 2017 ; 31 ( 4 ): 385 . Google Scholar CrossRef Search ADS PubMed 41 Stanger C , Ryan SR , Scherer EA , et al. : Clinic- and home-based contingency management plus parent training for adolescent cannabis use disorders . J Am Acad Child Adolesc Psychiatry 2015 ; 54 ( 6 ): 445 – 53. e442 . Google Scholar CrossRef Search ADS PubMed 42 Fritz MS , Mackinnon DP : Required sample size to detect the mediated effect . Psychol Sci 2007 ; 18 ( 3 ): 233 – 9 . Google Scholar CrossRef Search ADS PubMed 43 De Weert-Van Oene GH , Schippers GM , De Jong CA , et al. : Retention in substance dependence treatment: the relevance of in-treatment factors . J Subst Abuse Treat 2001 ; 20 ( 4 ): 253 – 61 ; discussion 263-4. Google Scholar CrossRef Search ADS PubMed 44 Harris AS , Humphreys K , Bowe T , et al. : Does meeting the HEDIS substance abuse treatment engagement criterion predict patient outcomes? J Behav Health Serv Res 2010 ; 37 ( 1 ): 25 – 39 . Google Scholar CrossRef Search ADS PubMed 45 Simpson DD , Joe GW , Rowan-Szal GA : Drug abuse treatment retention and process effects on follow-up outcomes . Drug Alcohol Depend 1997 ; 47 ( 3 ): 227 – 35 . Google Scholar CrossRef Search ADS PubMed 46 Bandura A : Social Foundations of Thought and Action: A Social Cognitive Theory . Englewood Cliffs, NJ , Prentice-Hall, Inc , 1986 . 47 Samet JH , Friedmann P , Saitz R : Benefits of linking primary medical care and substance abuse services: patient, provider, and societal perspectives . Arch Intern Med 2001 ; 161 ( 1 ): 85 – 91 . Google Scholar CrossRef Search ADS PubMed 48 Bartels SJ , Coakley EH , Zubritsky C , et al. : Improving access to geriatric mental health services: a randomized trial comparing treatment engagement with integrated versus enhanced referral care for depression, anxiety, and at-risk alcohol use . Am J Psychiat 2004 ; 161 ( 8 ): 1455 – 62 . Google Scholar CrossRef Search ADS PubMed 49 McLean CP , Su YJ , Foa EB : Mechanisms of symptom reduction in a combined treatment for comorbid posttraumatic stress disorder and alcohol dependence . J Consult Clin Psychol 2015 ; 83 ( 3 ): 655 – 61 . Google Scholar CrossRef Search ADS PubMed 50 Coffey SF , Schumacher JA , Stasiewicz PR , et al. : Craving and physiological reactivity to trauma and alcohol cues in posttraumatic stress disorder and alcohol dependence . Exp Clin Psychopharm 2010 ; 18 ( 4 ): 340 – 9 . Google Scholar CrossRef Search ADS 51 Volpicelli JR , Rhines KC , Rhines JS , et al. : Naltrexone and alcohol dependence: role of subject compliance . Arch Gen Psychiatry 1997 ; 54 ( 8 ): 737 – 42 . Google Scholar CrossRef Search ADS PubMed 52 Collins SE , Carey KB , Smyth J : Relationships of linguistic and motivation variables with drinking outcomes following two mailed brief interventions . J Stud Alcohol Drugs 2005 ; 66 ( 4 ): 526 – 35 . Google Scholar CrossRef Search ADS 53 Barnett NP , Apodaca TR , Magill M , et al. : Moderators and mediators of two brief interventions for alcohol in the emergency department . Addiction 2010 ; 105 ( 3 ): 452 – 65 . Google Scholar CrossRef Search ADS PubMed 54 Carey KB , Henson JM , Carey MP , et al. : Perceived norms mediate effects of a brief motivational intervention for sanctioned college drinkers . Clin Psychol-Sci Pr 2010 ; 17 ( 1 ): 58 – 71 . Google Scholar CrossRef Search ADS 55 LaChance H , Feldstein Ewing SW , Bryan AD , et al. : What makes group MET work? A randomized controlled trial of college student drinkers in mandated alcohol diversion . Psychol Addict Behav 2009 ; 23 ( 4 ): 598 – 612 . Google Scholar CrossRef Search ADS PubMed 56 Judd CM , Kenny DA : Data analysis in social psychology: recent and recurring Issues. In: Handbook of Social Psychology . Hoboken, NJ, John Wiley & Sons, Inc , 2010 . 57 LoCastro JS , Potter JS , Donovan DM , et al. : Characteristics of first-time alcohol treatment seekers: the COMBINE study . J Stud Alcohol Drugs 2008 ; 69 ( 6 ): 885 – 895 . Google Scholar CrossRef Search ADS PubMed Author notes The contents do not represent the views of the U.S. Department of Veterans Affairs or the United States Government. © Association of Military Surgeons of the United States 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Military Medicine Oxford University Press

Testing Mediators of Reduced Drinking for Veterans in Alcohol Care Management

Military Medicine , Volume 183 (9) – Sep 1, 2018

Loading next page...
 
/lp/ou_press/testing-mediators-of-reduced-drinking-for-veterans-in-alcohol-care-jcmgmzCaBg
Publisher
Oxford University Press
Copyright
© Association of Military Surgeons of the United States 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
0026-4075
eISSN
1930-613X
D.O.I.
10.1093/milmed/usy024
Publisher site
See Article on Publisher Site

Abstract

Abstract Introduction Alcohol Care Management (ACM) is a manualized treatment provided by behavioral health providers working in a primary care team aimed at increasing patients’ treatment engagement and decreasing their alcohol use. Research has shown that ACM is effective in reducing alcohol consumption; however, the mechanisms of ACM are unknown. Therefore, the purpose of this study is to examine the mechanisms of change in ACM in the context of a randomized clinical trial evaluating the effectiveness of ACM. Materials and Methods This study performed secondary data analysis of existing data from a larger study that involved a sample of U.S. veterans (N = 163) who met criteria for current alcohol dependence. Upon enrollment into the study, participants were randomized to receive either ACM or standard care. ACM was delivered in-person or by telephone within the primary care clinic and focused on the use of oral naltrexone and manualized psychosocial support. According to theory, we hypothesized several ACM treatment components that would mediate alcohol consumption outcomes: engagement in addiction treatment, reduced craving, and increased readiness to change. Parallel mediation models were performed by the PROCESS macro Model 4 in SPSS to test study hypotheses. The institutional review boards at each of the participating facilities approved all study procedures before data collection. Results As hypothesized, results showed that treatment engagement mediated the relation between treatment and both measures of alcohol consumption outcomes, the percentage of alcohol abstinent days, and the percentage of heavy drinking days. Neither craving nor readiness to change mediated the treatment effect on either alcohol consumption outcome. Conclusions Findings suggest that ACM may be effective in changing drinking patterns partially due to an increase in treatment engagement. Future research may benefit from evaluating the specific factors that underlie increased treatment engagement. The current study provides evidence that alcohol use disorder interventions should aim to increase treatment engagement and reduce barriers to care. Alcohol Care Management (ACM) is a manualized intervention targeting treatment engagement and alcohol consumption.1 A behavioral health provider working in a primary care team conducts the sessions and provides alcohol use monitoring, support, and education. Behavioral health providers also encourage treatment adherence and promote the use of oral naltrexone (50 mg), an evidence-based pharmacotherapy that has been shown to attenuate alcohol cravings,2,3 reduce heavy alcohol consumption,4–7 and increase alcohol abstinence days.8,9 However, not all patients receive naltrexone due to either medication contraindication or lack of interest. Therefore, providers offer two forms of ACM: medical management individual counseling sessions in which patients take naltrexone and medical attention individual counseling sessions where patients receive psychosocial support alone.10 Oslin, Lynch, Maisto et al1 conducted a randomized clinical trial (RCT) to test the effectiveness of the ACM model of care. Patients in the RCT consisted of 163 U.S. military veterans who were randomized to receive either ACM or standard care in a specialty outpatient addiction treatment program. Details regarding components of each study arm are described in Table I. Results of the study showed that ACM was more effective than standard care in reducing the percentage of heavy drinking days (% HDD). However, it remains unclear how ACM works to reduce alcohol consumption. Identifying necessary components (mediators) of effective alcohol use disorder (AUD) treatments is critical to develop or improve interventions. Table I. Components of Each Study Arm Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Note. Complete details are described in the original study article.1 Table I. Components of Each Study Arm Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Components Study Arm Alcohol Care Management Standard Care Setting Primary care VHA specialty outpatient addiction program Treatment/drinking goal Promotes abstinence, but allows participants to set their own drinking-reduction goals Abstinence; based on a 12-step facilitation model Session format Weekly 30-min meetings (either in-person or via telephone, based on veteran preference), decreasing to twice per month after the first 3 mo as participants improved 2–4 half-day treatment sessions for up to 6 wk (intensive outpatient program) then group therapy 1–2 times per week with ancillary services as needed Interventions Alcohol use assessment Individualized education about alcohol use disorder as a treatable problem Education about pharmacotherapy Encouragement for treatment adherence Skills to cope with craving and urges Support and education Monitor new or worsening medical problems Action plans focused on achievable goals (drinking or other health/life goals) for each week Encouragement to attend mutual-support groups Assessments and evaluations Outpatient detoxification Counseling (individual, family, and group) Pharmacotherapy (naltrexone, acamprosate, and disulfiram) Psychotherapy Psycho-educational groups Outreach and referral Acupuncture Expected to attend Alcoholics Anonymous, but attendance was not tracked or mandatory Note. Complete details are described in the original study article.1 One potential mediator of ACM’s effects on alcohol consumption is engagement in addiction treatment because engagement is a main target in the ACM model of care. Oslin et al1 showed that participants who received ACM were more likely to attend at least two addiction treatment visits per month compared with those who received standard care. Therefore, ACM increased engagement in addiction treatment and this increase may partly account for ACM’s effect on alcohol consumption. Although it is unclear how treatment engagement contributes to treatment outcomes,11 researchers have also examined engagement as a mediator in other treatment models, such as motivational interviewing (MI)12 and cognitive behavior therapy.13 A second potential mediator of ACM’s treatment effect is the attenuation of craving that may have occurred through ACM’s components that promote naltrexone use and introduce skills to cope with craving. One hypothesis is that naltrexone attenuates alcohol craving, which increases one’s ability to limit or abstain from alcohol consumption. This hypothesis is partly derived from neurobiological research, suggesting that alcohol activates the brain’s reward system,14 and conversely that naltrexone causes a μ-opioid receptor blockade, which may reduce the reinforcing properties of alcohol and attenuate cravings.15,16 Another hypothesis is that ACM may have influenced craving through skills training because the manualized behavioral intervention included a component where the provider introduced modules based on the veterans’ needs, one of which focused on coping with craving and urges.10 A third possible mediator of change in ACM is motivation or readiness to change. According to the transtheoretical model of change, individuals progress through several stages to make a behavioral change (e.g., reducing alcohol consumption).17 Further, empirically supported treatments such as MI and motivational enhancement therapy are based on the premise that motivation is central to achieving reductions in alcohol consumption. Indeed, research suggests that motivation predicts post-treatment drinking outcomes18–20 and mediates alcohol use intervention drinking outcomes.21,22 Consequently, increases in readiness to change may mediate reductions in alcohol consumption for patients receiving ACM. The purpose of this study is to examine the mechanisms of change in ACM in the context of the Oslin et al clinical trial.1 To this end, we investigated whether alcohol outcomes (percentage of alcohol abstinent days [% AAD] and % HDD) among veterans with AUD receiving ACM were mediated by (a) increased treatment engagement, (b) reduced alcohol craving, and (c) increased readiness to change. We hypothesized that treatment engagement, readiness to change, and reduced craving mediate drinking outcomes among those who receive ACM. Method Design and Sample This study involved secondary data analyses of a RCT evaluating the effectiveness of ACM, a primary care-based program for alcohol consumption. Complete details describing the original study are published elsewhere.1 Briefly, participants were randomly assigned to a 26-wk intervention of either ACM or standard addiction outpatient care, standard care (see Table I). The sample was 163 U.S. military veterans (96.9% male, Mage = 55.92, SD = 10.82, age range: 21–82 yr) from one of the three participating Veterans Health Administration (VHA) facilities who were referred to the study by their primary care physicians. Eighty-six veterans were randomized to ACM and 77 to standard care. A naltrexone prescription was filled by 56 (66.6%) and 9 (12.3%) veterans in the ACM condition and standard care condition, respectively. This discrepancy was expected because ACM providers were intended to promote naltrexone, whereas VA providers typically provide addiction pharmacotherapy at low rates.23 Rates in the standard care condition are comparable with the percentage of patients in AUD specialty care receiving naltrexone according to a nationally representative longitudinal study,24 but somewhat lower than in other surveys (17.5–24.6%).25 Participants were included in the study based on the following criteria: (a) at least 18 yr of age, (b) met DSM-IV criteria for current alcohol dependence, and (c) drank more than an average of two standard alcoholic drinks per day in the 60 d before randomization. Participants were excluded if they (a) endorsed current illicit drug (except marijuana) or opioid abuse or dependence; (b) exhibited symptoms of dementia, hearing loss, psychosis, or mania; or (c) participated in VHA substance abuse treatment (standard care) in the last 12 mo. Procedure Patients were referred to the Behavioral Health Laboratory,26 an integrated care program, by their primary care providers. This program delivers evidence-based mental health care in primary care and administers a battery of clinical assessments. Following receipt of referral, program staff conducted clinical assessments and offered research participation to eligible patients. Interested veterans were then scheduled for in-person baseline assessments with research staff where eligible participants were randomized to receive ACM or standard care. Follow-up interviews were completed by telephone or in-person at 3 and 6 mo following randomization. The institutional review boards at participating VHA facilities approved all study procedures before data collection. Measures Outcome Measure The outcomes examined in this study were % AAD and % HDD as measured by the Alcohol Timeline Followback (TLFB). Heavy drinking was defined as 5+ standard drinks in 1 day for men and 4+ for women.27 The TLFB uses a calendar format and memory cues to enhance retrospective self-reporting of daily alcohol consumption in standard drink units.28 The TLFB has good psychometric properties, including high reliability (ICC = 0.60–0.99)29 and validity (i.e., accuracy of reports; r = 0.74–0.89).30 At baseline, alcohol use for the previous 90 d was recorded. At each follow-up, the TLFB encompassed the time since the last interview. Mediator Measures The mediators examined were treatment engagement, alcohol craving, and readiness to change. Treatment engagement included the total number of addiction visits from baseline to 3-mo follow-up. Engagement in addiction treatment was tracked using the VHA electronic medical record supplemented by a questionnaire of services received outside the VHA. Specifically, participants reported the number of visits to a medical office outside of the VHA and specified whether the visit was mental health or substance abuse. Alcohol craving was measured using the Penn Alcohol Craving Scale (PACS), a five-item scale measuring past-week intensity, frequency and duration of cravings, and ability to resist acting on cravings. This instrument has been correlated with other measures of craving31 and demonstrated good to excellent internal consistency across time points (α = 0.899–0.936). Item scores range from 0 to 6 with higher scores indicating more craving; all items are summed to create a total score. Lastly, readiness to change was measured using the Contemplation Ladder.32 Participants used a numbered image of a ladder to rate their readiness to quit using alcohol on an 11-point scale (0 = “No thought of quitting” to 10 = “Taking action to quit”) containing five anchor statements representing each stage of change:33 (0–3) precontemplation, (4–6) contemplation, (7–8) preparation, (9) action, (10). These readiness categories are based on the transtheoretical model of change. Higher numbers represent greater readiness to change. Potential Covariates Additional measures were examined to assess for potential group differences. The Behavioral Health Laboratory assessments included demographic data such as age, gender, race, and employment status. The post-traumatic stress disorder (PTSD) Checklist-Situation (PCL-C) for DSM-IV was used to assess for PTSD symptoms related to any traumatic event.34 Symptom severity of depression was measured using the Patient Health Questionnaire (PHQ-9).35 The Shortened Inventory of Problems (SIP-2R) was used to assess consequences of drinking over the past 3 mo.36 These measures demonstrated good to excellent reliability in the current sample (α = 0.933, 0.993, and 0.893 for PCL-C, PHQ-9, and SIP-2R, respectively). Statistical Analysis Univariate statistics were conducted to describe participant characteristics. Chi-square analyses for categorical variables and Mann–Whitney U-tests for continuous variables were conducted to examine any differences between ACM and standard care in baseline characteristics. To examine treatment engagement, alcohol craving, and readiness to change as mediators of the treatment effect, we performed parallel mediation models by the PROCESS macro Model 4 in SPSS. This approach uses an ordinary least squares regression framework and produces a test of total, direct, and indirect effects.37 Parallel mediation models offer several advantages over conducting simple mediation models for each mediator. For example, parallel mediation provides information on whether the mediation is independent of the effect of the other mediator(s) and reduces the potential for parameter bias due to omitted variables.38 The model was specified as depicted in Figure 1. The independent variable was condition (ACM = +1, standard care = −1) and the parallel mediators were treatment engagement, alcohol craving, and readiness to change. Alcohol outcomes (% HDD or % AAD) were the dependent variables and pre-baseline levels of alcohol use were entered as covariates due to their significant correlations (Table II). Separate models were run for each alcohol outcome (i.e., % HDD or % AAD). The % HDD variables were transformed for normality using square root transformations before analyses. Temporal precedence of the mediators was established to increase confidence in interpreting study findings. Specifically, the total number of addiction visits from 0 to 3 mo was calculated for treatment engagement and change scores over the same time were calculated for alcohol craving and readiness to change (3 mo – baseline). Alcohol outcomes were calculated as change in % HDD and % AAD from 4 to 6 mo. Figure 1. View largeDownload slide Parallel mediation model analyzed in this study with craving change, treatment engagement, and readiness to change as parallel mediators. Figure 1. View largeDownload slide Parallel mediation model analyzed in this study with craving change, treatment engagement, and readiness to change as parallel mediators. Table II. Correlations Between Alcohol Use, Craving, Treatment Engagement, and Readiness to Change Variables r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 Note. % AAD, percentage of alcohol abstinent days, % HDD, percentage of heavy drinking days (square root transformed). Pre-baseline is based on 90 d before the baseline. Craving change and readiness to change is a change score (3-mo score − 0-mo score). Treatment engagement is a total score (0–3 mo). *p < 0.05; **p < 0.01; ***p < 0.001. N = 130. View Large Table II. Correlations Between Alcohol Use, Craving, Treatment Engagement, and Readiness to Change Variables r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 r 1 2 3 4 5 6 7 1. % AAD pre-baseline 1 2. % HDD pre-baseline −0.641*** 1 3. % AAD 4–6 mo 0.320*** −1.67 1 4. % HDD 4–6 mo −0.233** 0.266** −0.735*** 1 5. Craving change 0–3 mo 0.021 −0.106 −0.379*** 0.382*** 1 6. Treatment engagement 0–3 mo 0.118 −0.084 0.353*** −0.283** −0.163 1 7. Readiness to change 0–3 mo 0.028 −0.117 0.194* −0.308*** −0.284** 0.139 1 Note. % AAD, percentage of alcohol abstinent days, % HDD, percentage of heavy drinking days (square root transformed). Pre-baseline is based on 90 d before the baseline. Craving change and readiness to change is a change score (3-mo score − 0-mo score). Treatment engagement is a total score (0–3 mo). *p < 0.05; **p < 0.01; ***p < 0.001. N = 130. View Large We tested for mediation according to guidelines published by Baron and Kenney.39 Although the guidelines were provided in the context of single mediator models, the causal steps have also been extended to multiple mediator models.38 If mediation is present, paths a, b, and c (shown in Fig. 1) will be significant, and path c’ will decrease or become non-significant. The statistical significance of indirect effects (path ab) were assessed using 10,000 resamples and bias-corrected CI. The indirect effects were considered significant if zero was not within the 95% CI. Primary analyses required complete data for all variables entered into the model. Therefore, missing data were handled in two ways. First, given the distribution of data and empirical precedent, alcohol consumption variables were coded as missing when <85% of the TLFB data were available.40,41 When ≥85% of the TLFB data were available, the number of non-missing days were entered into the denominator to calculate alcohol use percentages. Second, participants who were missing any data included in the proposed models were subsequently removed from the main analyses. Of the 163 individuals randomized, 33 participants had missing data that restricted them from being entered into the parallel mediation analyses. We referenced published guidelines to estimate power; our sample of n = 130 provided approximately 75% (α 0.05) power to detect small to medium effects (0.26) in the mediation analyses.42 Results Patient Characteristics Table III depicts characteristics of randomized participants. The sample was mostly male, but racially diverse with approximately half identifying as Black (52.8%; Caucasian 42.3%; 4.9% other) and almost 5% as Hispanic. Twenty-one percent of the sample screened positive for probable PTSD (PCL-C ≥50).34 No baseline differences were found between individuals assigned to ACM or standard care (Table III). Participants with missing data were more likely to be Caucasian and have lower physical health scores (SF12 PCS score) (p < 0.05), as compared with those without missing data, but did not differ on any other baseline variables (p’s > 0.05). Table III. Descriptive Statistics for Study Variables Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Note. ACM, alcohol care management; SC, specialty care; % HDD, percentage of heavy drinking days; % AAD, percentage of alcohol abstinent days; PHQ-9, Patient Health Questionnaire (0–27 with higher scores indicating greater depression); SIP, Short Inventory of Problems; PCS-SF12, Physical Component Score from the Medical Outcomes scale (SF36); MCS-SF12, Mental Component Score from the Medical Outcomes scale (SF36); PTSD, probable post-traumatic stress disorder, ≥50 on the PTSD Checklist for DSM-IV (PCL-C); treatment engagement, total addiction treatment visits from baseline to 3 mo. aNo significant differences between ACM and SC conditions at baseline (all p’s > 0.063). bn = 130, represents the sample included in main analyses. Table III. Descriptive Statistics for Study Variables Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Characteristic ACM (n = 86) SC (n = 77) Overall (n = 163)a Baseline Baseline Baseline Age 54.97 ± 11.41 56.97 ± 10.10 55.92 ± 10.83 Male, N (%) 85 (98.8%) 73 (94.8%) 158 (96.9%) Caucasian, N (%) 36 (41.9%) 33 (42.9%) 69 (42.3%) Employed, N (%) 31 (36%) 19 (24.7%) 50 (30.7%) Current smokers, N (%) 41 (47.7%) 42 (54.5%) 83 (50.9%) PTSD, N (%) 20 (23.3%) 14 (18.2%) 34 (20.9%) PHQ-9 score 9.58 ± 6.17 10.36 ± 7.58 9.95 ± 6.86 SIP score 10.51 ± 8.69 8.92 ± 7.13 9.76 ± 8.01 Baselineb 3-mob 6-mob Baselineb 3-mob 6-mob Baselineb % HDD 53.46 ± 32.42 18.39 ± 26.96 16.96 ± 28.13 53.18 ± 30.78 25.98 ± 28.60 24.44 ± 29.63 54.24 ± 31.15 % AAD 30.32 ± 26.07 65.75 ± 32.88 69.31 ± 34.32 30.08 ± 28.99 55.50 ± 35.62 57.26 ± 37.17 28.98 ± 26.80 PACS score 14.43 ± 8.79 9.96 ± 8.34 6.61 ± 3.68 14.43 ± 7.96 11.38 ± 8.57 2.92 ± 4.35 14.4 ± 8.29 Readiness to change 7.35 ± 2.71 8.14 ± 2.43 — 7.08 ± 2.25 7.61 ± 2.20 — 7.19 ± 2.54 Treatment engagement — 7.16 ± 3.89 — — 3.95 ± 5.54 — — Note. ACM, alcohol care management; SC, specialty care; % HDD, percentage of heavy drinking days; % AAD, percentage of alcohol abstinent days; PHQ-9, Patient Health Questionnaire (0–27 with higher scores indicating greater depression); SIP, Short Inventory of Problems; PCS-SF12, Physical Component Score from the Medical Outcomes scale (SF36); MCS-SF12, Mental Component Score from the Medical Outcomes scale (SF36); PTSD, probable post-traumatic stress disorder, ≥50 on the PTSD Checklist for DSM-IV (PCL-C); treatment engagement, total addiction treatment visits from baseline to 3 mo. aNo significant differences between ACM and SC conditions at baseline (all p’s > 0.063). bn = 130, represents the sample included in main analyses. Naltrexone Prescription Data Prescription data for six participants (3.7 %) were missing (n = 4 ACM; n = 2 standard care). Of the 157 participants for whom prescription data were available, 65 (41.4%) participants filled at least one naltrexone prescription. Initial fill dates occurred on average 38.5 (SD = 47.54) days following randomization. Providers individualized the naltrexone dosage as indicated (M = 63.26 mg, SD = 24.06 mg); dosages ranged from 50 to 150 mg per day. Specific to the 130 participants included in the parallel mediation analyses, 54 participants (n = 7 standard care, n = 47 ACM) filled a naltrexone prescription. Of note, five ACM participants filled their initial naltrexone prescription after the 3-mo time point. However, among the 130 participants, there was no significant correlation between the amount of craving change and the number of days between randomization and the initial fill date. Evaluation of Research Hypotheses Table II presents the simple correlations among key variables for the participants included in these analyses (i.e., participants with complete data for parallel mediation models). Table IV displays results of the parallel mediation models (i.e., total, direct, indirect, and total indirect effects). The first parallel mediation analysis examined the hypothesized role of treatment engagement, craving change, and readiness to change as mediators between condition (ACM and standard care) and % HDD, controlling for pre-baseline % HDD. Except for path c, which was non-significant (p = 0.06), required mediation criteria were met for treatment engagement as a mediator of condition and % HDD (paths a1 and b1 were significant, p < 0.05). The total effect (path c, b = −0.513; p = 0.06) reduced in significance when the mediators were accounted for (path c’, b = −0.182; p = 0.51). The mediation effect of treatment engagement (indirect effect, path ab1) was significant according to bootstrapping results (b = −0.226 [−0.509, −0.024]). This analysis indicates that participants in the ACM condition were more engaged in treatment than those in the standard care condition (path a1, b = 1.845, p < 0.001), and increases in treatment engagement explained variance in reductions in the % HDD. Required mediation criteria were not met for craving change or readiness to change as mediators of condition and % HDD. Condition did not significantly influence craving change or readiness to change (path a2, b = −0.724; p = 0.39; path a3, b = 0.133; p = 0.60). Table IV. Parallel Mediating Effects of Craving Change, Treatment Engagement, and Readiness to Change in the Relation Between Condition and Alcohol Outcomes Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Note. % HDD, percentage of heavy drinking days square root transformed; % AAD, percentage of alcohol abstinent days; IE, indirect effect. All estimates are unstandardized. Confidence intervals (CIs) resulted from 10,000 bootstrap draws. Alcohol use outcomes are based on use 4–6 mo post-baseline. Change in craving and readiness to change are change scores (3−0 mo). Treatment engagement is total addiction treatment visits 0–3 mo. All models are controlling for pre-baseline value of the alcohol outcome examined in the model. *p < 0.05. †p = 0.06. ‡CI does not include zero. N = 130. Table IV. Parallel Mediating Effects of Craving Change, Treatment Engagement, and Readiness to Change in the Relation Between Condition and Alcohol Outcomes Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Outcomes Mediators Indirect Effect Total Indirect Effect Direct Effect Total Effect Percent Mediation (%) % HDD Treatment engagement −0.226‡ –0.331‡ −0.182 −0.513† 44.1% Change in craving −0.081 — Change in readiness −0.024 — % AAD Treatment engagement 3.627‡ 4.593‡ 1.379 5.973*‡ 60.7% Change in craving 0.869 — Change in readiness 0.096 — Note. % HDD, percentage of heavy drinking days square root transformed; % AAD, percentage of alcohol abstinent days; IE, indirect effect. All estimates are unstandardized. Confidence intervals (CIs) resulted from 10,000 bootstrap draws. Alcohol use outcomes are based on use 4–6 mo post-baseline. Change in craving and readiness to change are change scores (3−0 mo). Treatment engagement is total addiction treatment visits 0–3 mo. All models are controlling for pre-baseline value of the alcohol outcome examined in the model. *p < 0.05. †p = 0.06. ‡CI does not include zero. N = 130. Following similar procedures, a second parallel mediation analysis was conducted for % AAD, when controlling for pre-baseline % AAD. Criteria were met for treatment engagement as a mediator of condition and % AAD (paths a1, b1, and c were significant, p < 0.05). The total effect (path c, b = 5.973; p < 0.05) reduced in significance when the mediators were accounted for; the regression coefficient decreased and was no longer significant (path c’, b = 1.379; p = 0.64). The mediation effect of treatment engagement (indirect effect, path ab1) was significant (b = 3.627 [1.625, 6.675]). Results indicate that participants in the ACM condition were more engaged in treatment than those in the standard care condition (path a1, b = 1.843, p < 0.001), and increases in treatment engagement explained variance in increases in % AAD. Required mediation criteria were not met for craving change or readiness to change as mediators of condition and % AAD. Condition did not significantly influence craving or readiness to change (path a2, b = −0.715; p = 0.40; path a3, b = 0.136; p = 0.60) and readiness to change did not significantly influence % AAD (b3, b = 0.707; p = 0.45). Discussion The results of this study supported the hypothesis that ACM was effective in reducing alcohol consumption partly because participants attended more addictions treatment visits. Data showed that participants who received ACM as compared with standard care were less likely to consume any alcohol (p < 0.05) and less likely to drink heavily (p = 0.06). Notably, in contrast to our non-significant results regarding heavy drinking, the original RCT analyses by Oslin et al,1 which included all randomized participants, showed a significant treatment effect on heavy drinking days, favoring ACM. Treatment effects on drinking outcomes were mediated by number of treatment visits. Therefore, results suggest that treatment engagement is an important target for AUD treatment. The finding that engagement in addiction treatment was a key component in ACM aligns with research suggesting an association between level of involvement in addiction treatment and treatment outcomes.43–45 Various components of ACM likely contributed to improved treatment engagement. First, the ACM model of care encourages providers to discuss attendance with patients, specifically, to explore the reasons for missing appointments and to reinforce attendance.10 As such, providers may have assisted in problem-solving and increased patients’ commitment to engage in treatment. Also, given that theory suggests patients may be more invested in treatment when they set their own goals,46 another important component may have been that patients set their own drinking-reduction goals (as opposed to an abstinence-only goal). Finally, it is possible that by placing addiction treatment in primary care, as is the case with ACM, patients’ access to treatment services improved47 and consequently increased likelihood of engaging in treatment.48 Although data were not available in this study to test specific factors that may have improved treatment engagement (e.g., therapeutic relationship, commitment to treatment, good clinic access, perceived helpfulness of treatment, and decreased stigma), it is clear that ACM was successful in reducing alcohol use, in part by increasing treatment engagement. Future research may benefit from evaluating the specific factors that underlie increased treatment engagement so that these important factors can be targeted in interventions. Contrary to hypotheses, results indicated that alcohol craving and readiness to change did not mediate treatment effects. Therefore, findings suggest that craving and readiness to change are not mechanisms by which ACM has its effect on alcohol consumption. Although results do not support craving as a mediator of ACM’s treatment effect or the theorized biobehavioral mechanism of action of naltrexone,15,16 our findings do align with other research that did not find evidence for craving as a mediator of alcohol treatment outcomes.2,49 One explanation for null craving mediation findings in our study is the presence of co-occurring PTSD and AUD. Indeed, similar to another naltrexone treatment study with null craving mediation findings,18 participants in our study had co-occurring PTSD (20.9% based on the PCL-C). Co-occurring AUD and PTSD are known to bi-directionally affect one another, and research shows that trauma cues influence alcohol craving.50 Therefore, it is possible that naltrexone was alleviating cravings, but trauma cues or PTSD symptoms simultaneously increased craving. We conducted additional analyses on the relation between PTSD symptoms and craving to explore this hypothesized explanation for null mediation findings. Results showed that PTSD symptoms at baseline were not correlated with craving change (baseline to 3 mo; r = 0.609). However, the current study does not contain a measure of PTSD symptoms over time, which would provide a more sensitive test of this hypothesis. A second explanation for null mediation findings for craving is that the categorizations of treatment conditions (ACM and standard care) did not offer a precise test of the hypothesized relation that was based on the presumed effect of naltrexone. Namely, several participants in the ACM condition did not fill a naltrexone prescription, while prescriptions were filled by some participants in the standard care condition. Indeed, research shows a direct relation between naltrexone treatment compliance and improved alcohol outcomes.51 Further, several patients did not begin taking naltrexone at the outset of treatment, thereby limiting the ability to detect subsequent changes in craving from baseline to 3 mo. Therefore, to create a more precise test of craving as a mediator, secondary mediation analyses were conducted on a subgroup of participants who were either (1) in standard care and never filled a naltrexone prescription or (2) in ACM and filled a naltrexone prescription within 3 mo post-randomization. Although results of these secondary analyses may be limited by a reduced sample size (n = 96), results remained non-significant and did not support alcohol craving as a mediator (% HDD: path a2b = −1.458; p = 0.12; % AAD: path a2b = −1.082; p = 0.24). When interpreting the present findings, it is important to note that we are unable to disentangle the effects of naltrexone versus the behavioral treatment effects of ACM. For instance, it may be the combination of both, not one or the other that drives the treatment effect. Contrary to expectation, readiness to change was not a mediator of the effect of ACM on alcohol outcomes. This is surprising given the theory-based assumption that a prerequisite for making a change in drinking patterns is reaching a sufficiently high level of readiness to change.17 However, other studies also have produced null mediation findings for readiness to change,52 even when the intervention presumably targets readiness to change specifically.53–55 Nevertheless, it is worth noting that, on average, participants entering the study reported a relatively high level of readiness to change at baseline (M = 7.19/10, SD = 2.54), representative of being in the preparation stage of change. Therefore, ceiling effects may have limited findings; participants may not have had much room to increase readiness to change further over the treatment course. Notable strengths of this study include the recruitment of a diverse population at multiple VHA facilities, conservative testing of research hypotheses by controlling for pre-baseline alcohol consumption patterns, examining multiple mediators in a single model, and temporally isolating the mediators and outcome variables. However, several caveats should be considered when interpreting the results of this study. First, our results are not based on an intent-to-treat analysis because missing data were not tolerated in our analyses. Although individuals in the randomized conditions did not differ at baseline, those who had missing data were more likely to be Caucasian and score higher on a physical health measure, which may limit generalizability. Second, despite temporally ordering the mediator and outcome variables, it is possible that participants who were more likely to abstain or reduce alcohol consumption were also individuals who were more likely to engage in treatment.56 Third, our interpretation regarding the mediation effect for treatment engagement is just one interpretation. The relation is likely complex; for instance, it is possible that veterans who experience diminished craving may feel more inclined to attend appointments. Fourth, the treatment effects observed in this study are not specific to naltrexone treatment per se because not all ACM participants received naltrexone. Lastly, evidence suggests that treatment-naïve individuals are less ready to change and less likely to achieve abstinence than persons who have attended treatment more frequently.57 Therefore, differences may exist between this sample, which consists of individuals who have not attended addiction treatment in the last year and individuals who have attended treatment more recently. Future research should examine how ACM works across individuals with diverse treatment histories and varying levels of readiness to change. Also, the literature includes several other hypotheses (e.g., altering the subjective effects of alcohol, increasing cognitive control) that may underlie the effects of AUD treatments that use naltrexone.15 Future research should aim to empirically examine additional factors that are hypothesized to mediate the effects of naltrexone treatment on alcohol consumption. Conclusions This study represents the first examination of theoretically supported mediators of ACM. Our data indicate that ACM, a naltrexone-based AUD treatment, was effective in reducing alcohol consumption partly by increasing treatment engagement. Findings also suggest that readiness to change and reduced alcohol craving may not be essential to changing alcohol patterns. These results have important implications for researchers and clinicians in AUD treatment and highlight the importance of engaging participants in treatment. Interventions for AUD should aim to reduce barriers to care and increase treatment engagement. Funding Funding was provided by Health Services Research and Development Program of the Department of Veteran Affairs (IIR); The VISN 4 Mental Illness Research, Education, and Clinical Center at the Cpl. Michael J. Crescenz VA Medical Center; The VISN 2 Center for Integrated Healthcare; Career Development Award (K05 AA16928 [Dr. Maisto]). References 1 Oslin DW , Lynch KG , Maisto SA , et al. : A randomized clinical trial of alcohol care management delivered in Department of Veterans Affairs primary care clinics versus specialty addiction treatment . J Gen Intern Med 2013 ; 29 ( 1 ): 162 – 8 . Google Scholar CrossRef Search ADS PubMed 2 Richardson K , Baillie A , Reid S , et al. : Do acamprosate or naltrexone have an effect on daily drinking by reducing craving for alcohol? Addiction. 2008 ; 103 ( 6 ): 953 – 9 . Google Scholar CrossRef Search ADS PubMed 3 Monti PM , Rohsenow DJ , Hutchison KE , et al. : Naltrexone’s effect on cue-elicited craving among alcoholics in treatment . Alcohol Clin Exp Res 1999 ; 23 ( 8 ): 1386 – 94 . Google Scholar PubMed 4 Volpicelli JR , Alterman AI , Hayashida M , et al. : Naltrexone in the treatment of alcohol dependence . Arch Gen Psychiatry 1992 ; 49 ( 11 ): 876 – 80 . Google Scholar CrossRef Search ADS PubMed 5 O’Malley SS , Jaffe AJ , Chang G , et al. : Naltrexone and coping skills therapy for alcohol dependence. A controlled study . Arch Gen Psychiatry 1992 ; 49 ( 11 ): 881 – 7 . Google Scholar CrossRef Search ADS PubMed 6 Anton RF , Moak DH , Latham P , et al. : Naltrexone combined with either cognitive behavioral or motivational enhancement therapy for alcohol dependence . J Clin Psychopharm 2005 ; 25 ( 4 ): 349 – 57 . Google Scholar CrossRef Search ADS 7 Pettinati HM , O’Brien CP , Rabinowitz AR , et al. : The status of naltrexone in the treatment of alcohol dependence: specific effects on heavy drinking . J Clin Psychopharmacol 2006 ; 26 ( 6 ): 610 – 25 . Google Scholar CrossRef Search ADS PubMed 8 Maisel NC , Blodgett JC , Wilbourne PL , et al. : Meta-analysis of naltrexone and acamprosate for treating alcohol use disorders: when are these medications most helpful? Addiction 2013 ; 108 ( 2 ): 275 – 93 . Google Scholar CrossRef Search ADS PubMed 9 Rosner S , Hackl-Herrwerth A , Leucht S , et al. : Opioid antagonists for alcohol dependence . Cochrane Database Syst Rev 2010 ; ( 12 ): CD001867 . 10 Pettinati HM , Weiss RD , Miller WR , et al. Medical Management treatment manual. A Clinical Research Guide for Medically Trained Clinicians Providing Pharmacotherapy as Part of the Treatment for Alcohol Dependence COMBINE Monograph Series. 2004 ; 2. 11 Dearing RL , Barrick C , Dermen KH , et al. : Indicators of client engagement: influences on alcohol treatment satisfaction and outcomes . Psychol Addict Behav 2005 ; 19 ( 1 ): 71 – 8 . Google Scholar CrossRef Search ADS PubMed 12 Moyers TB , Moyers TB , Miller WR , et al. : How does motivational interviewing work? Therapist interpersonal skill predicts client involvement within motivational interviewing sessions . J Consult Clin Psychol 2005 ; 73 ( 4 ): 590 – 8 . Google Scholar CrossRef Search ADS PubMed 13 Spangler DL , Baldwin SA , Agras WS : An examination of the mechanisms of action in cognitive behavioral therapy for bulimia nervosa . Behav Ther 2004 ; 35 ( 3 ): 537 – 60 . Google Scholar CrossRef Search ADS 14 Koob GF , Roberts AJ , Schulteis G , et al. : Neurocircuitry targets in ethanol reward and dependence . Alcohol Clin Exp Res 1998 ; 22 ( 1 ): 3 – 9 . Google Scholar CrossRef Search ADS PubMed 15 Ray LA , Chin PF , Miotto K : Naltrexone for the treatment of alcoholism: clinical findings, mechanisms of action, and pharmacogenetics . CNS Neurol Disord Drug Targets 2010 ; 9 ( 1 ): 13 – 22 . Google Scholar CrossRef Search ADS PubMed 16 Spanagel R , Zieglgansberger W : Anti-craving compounds for ethanol: new pharmacological tools to study addictive processes . Trends Pharmacol Sci 1997 ; 18 ( 2 ): 54 – 9 . Google Scholar CrossRef Search ADS PubMed 17 Prochaska JO , DiClemente CC : Transtheoretical therapy: toward a more integrative model of change . Psychotherapy (Chic) 1982 ; 19 ( 3 ): 276 – 88 . Google Scholar CrossRef Search ADS 18 Maisto SA , Krenek M , Chung T , et al. : A comparison of the concurrent and predictive validity of three measures of readiness to change alcohol use in a clinical sample of adolescents . Assessment 2011 ; 23 ( 4 ): 983 – 94 . Google Scholar CrossRef Search ADS 19 Heather N , McCambridge J : Post-treatment stage of change predicts 12-month outcome of treatment for alcohol problems . Alcohol Alcohol 2013 ; 48 ( 3 ): 329 – 36 . Google Scholar CrossRef Search ADS PubMed 20 Hogue A , Dauber S , Morgenstern J : Validation of a contemplation ladder in an adult substance use disorder sample . Psychol Addict Behav 2010 ; 24 ( 1 ): 137 – 44 . Google Scholar CrossRef Search ADS PubMed 21 Stein LAR , Minugh PA , Longabaugh R , et al. : Readiness to change as a mediator of the effect of a brief motivational intervention on posttreatment alcohol-related consequences of injured emergency department hazardous drinkers . Psychol Addict Behav 2009 ; 23 ( 2 ): 185 – 95 . Google Scholar CrossRef Search ADS PubMed 22 Kelly JF , Myers MG , Brown SA : A multivariate process model of adolescent 12-step attendance and substance use outcome following inpatient treatment . Psychol Addict Behav 2000 ; 14 ( 4 ): 376 – 89 . Google Scholar CrossRef Search ADS PubMed 23 Harris AH , Kivlahan DR , Bowe T , et al. : Pharmacotherapy of alcohol use disorders in the Veterans Health Administration . Psychiatr Serv 2010 ; 61 ( 4 ): 392 – 8 . Google Scholar CrossRef Search ADS PubMed 24 Roman PM , Johnson JA : Adoption and implementation of new technologies in substance abuse treatment . J Subst Abuse Treat 2002 ; 22 ( 4 ): 211 – 8 . Google Scholar CrossRef Search ADS PubMed 25 Fuller BE , Rieckmann T , McCarty D , et al. : Adoption of naltrexone to treat alcohol dependence . J Subst Abuse Treat 2005 ; 28 ( 3 ): 273 – 80 . Google Scholar CrossRef Search ADS PubMed 26 Tew J , Klaus J , Oslin DW : The Behavioral Health Laboratory: building a stronger foundation for the patient-centered medical home . Fam Syst Health 2010 ; 28 ( 2 ): 130 – 45 . Google Scholar CrossRef Search ADS PubMed 27 Wechsler H , Dowdall GW , Davenport A , et al. : A gender-specific measure of binge drinking among college students . Am J Public Health 1995 ; 85 ( 7 ): 982 – 5 . Google Scholar CrossRef Search ADS PubMed 28 Sobell LC , Sobell MB : Timeline Followback User’s Guide: A Calendar Method for Assessing Alcohol and Drug Use . Toronto, ON , Addiction Research Foundation , 1996 . 29 Sobell LC , Agrawal S , Annis H , et al. : Cross-cultural evaluation of two drinking assessment instruments: alcohol timeline followback and inventory of drinking situations . Subst Use Misuse 2001 ; 36 ( 3 ): 313 – 31 . Google Scholar CrossRef Search ADS PubMed 30 Sobell LC , Sobell MB , Leo GI , et al. : Reliability of a timeline method: assessing normal drinkers’ reports of recent drinking and a comparative evaluation across several populations . Br J Addict 1988 ; 83 ( 4 ): 393 – 402 . Google Scholar CrossRef Search ADS PubMed 31 Flannery BA , Volpicelli JR , Pettinati HM : Psychometric properties of the Penn Alcohol Craving Scale . Alcohol Clin Exp Res 1999 ; 23 ( 8 ): 1289 – 95 . Google Scholar CrossRef Search ADS PubMed 32 Biener L , Abrams DB : The Contemplation Ladder: validation of a measure of readiness to consider smoking cessation . Health Psychol 1991 ; 10 ( 5 ): 360 – 5 . Google Scholar CrossRef Search ADS PubMed 33 Prochaska JO , DiClemente CC , Norcross JC : In search of how people change. Applications to addictive behaviors . Am Psychol 1992 ; 47 ( 9 ): 1102 – 14 . Google Scholar CrossRef Search ADS PubMed 34 Weathers FW , Litz BT , Herman DS , et al. : The PTSD Checklist (PCL): Reliability, validity, and diagnostic utility. Paper presented at: annual convention of the international society for traumatic stress studies, San Antonio, TX 1993 . 35 Kroenke K , Spitzer R , Williams JB : The PHQ-9: validity of a brief depression severity measure . J Gen Intern Med 2001 ; 16 ( 9 ): 606 – 13 . Google Scholar CrossRef Search ADS PubMed 36 Kenna GA , Longabaugh R , Gogineni A , et al. : Can the short index of problems (SIP) be improved? Validity and reliability of the three-month SIP in an emergency department sample . J Stud Alcohol 2005 ; 66 ( 3 ): 433 – 7 . Google Scholar CrossRef Search ADS PubMed 37 Hayes AF : Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach . New York, NY, Guilford Press , 2013 . 38 Preacher KJ , Hayes AF : Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models . Behav Res Methods 2008 ; 40 ( 3 ): 879 – 91 . Google Scholar CrossRef Search ADS PubMed 39 Baron RM , Kenny DA : The moderator–mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations . J Pers Soc Psychol 1986 ; 51 ( 6 ): 1173 – 82 . Google Scholar CrossRef Search ADS PubMed 40 Stanger C , Scherer EA , Babbin SF , et al. : Abstinence based incentives plus parent training for adolescent alcohol and other substance misuse . Psychol Addict Behav 2017 ; 31 ( 4 ): 385 . Google Scholar CrossRef Search ADS PubMed 41 Stanger C , Ryan SR , Scherer EA , et al. : Clinic- and home-based contingency management plus parent training for adolescent cannabis use disorders . J Am Acad Child Adolesc Psychiatry 2015 ; 54 ( 6 ): 445 – 53. e442 . Google Scholar CrossRef Search ADS PubMed 42 Fritz MS , Mackinnon DP : Required sample size to detect the mediated effect . Psychol Sci 2007 ; 18 ( 3 ): 233 – 9 . Google Scholar CrossRef Search ADS PubMed 43 De Weert-Van Oene GH , Schippers GM , De Jong CA , et al. : Retention in substance dependence treatment: the relevance of in-treatment factors . J Subst Abuse Treat 2001 ; 20 ( 4 ): 253 – 61 ; discussion 263-4. Google Scholar CrossRef Search ADS PubMed 44 Harris AS , Humphreys K , Bowe T , et al. : Does meeting the HEDIS substance abuse treatment engagement criterion predict patient outcomes? J Behav Health Serv Res 2010 ; 37 ( 1 ): 25 – 39 . Google Scholar CrossRef Search ADS PubMed 45 Simpson DD , Joe GW , Rowan-Szal GA : Drug abuse treatment retention and process effects on follow-up outcomes . Drug Alcohol Depend 1997 ; 47 ( 3 ): 227 – 35 . Google Scholar CrossRef Search ADS PubMed 46 Bandura A : Social Foundations of Thought and Action: A Social Cognitive Theory . Englewood Cliffs, NJ , Prentice-Hall, Inc , 1986 . 47 Samet JH , Friedmann P , Saitz R : Benefits of linking primary medical care and substance abuse services: patient, provider, and societal perspectives . Arch Intern Med 2001 ; 161 ( 1 ): 85 – 91 . Google Scholar CrossRef Search ADS PubMed 48 Bartels SJ , Coakley EH , Zubritsky C , et al. : Improving access to geriatric mental health services: a randomized trial comparing treatment engagement with integrated versus enhanced referral care for depression, anxiety, and at-risk alcohol use . Am J Psychiat 2004 ; 161 ( 8 ): 1455 – 62 . Google Scholar CrossRef Search ADS PubMed 49 McLean CP , Su YJ , Foa EB : Mechanisms of symptom reduction in a combined treatment for comorbid posttraumatic stress disorder and alcohol dependence . J Consult Clin Psychol 2015 ; 83 ( 3 ): 655 – 61 . Google Scholar CrossRef Search ADS PubMed 50 Coffey SF , Schumacher JA , Stasiewicz PR , et al. : Craving and physiological reactivity to trauma and alcohol cues in posttraumatic stress disorder and alcohol dependence . Exp Clin Psychopharm 2010 ; 18 ( 4 ): 340 – 9 . Google Scholar CrossRef Search ADS 51 Volpicelli JR , Rhines KC , Rhines JS , et al. : Naltrexone and alcohol dependence: role of subject compliance . Arch Gen Psychiatry 1997 ; 54 ( 8 ): 737 – 42 . Google Scholar CrossRef Search ADS PubMed 52 Collins SE , Carey KB , Smyth J : Relationships of linguistic and motivation variables with drinking outcomes following two mailed brief interventions . J Stud Alcohol Drugs 2005 ; 66 ( 4 ): 526 – 35 . Google Scholar CrossRef Search ADS 53 Barnett NP , Apodaca TR , Magill M , et al. : Moderators and mediators of two brief interventions for alcohol in the emergency department . Addiction 2010 ; 105 ( 3 ): 452 – 65 . Google Scholar CrossRef Search ADS PubMed 54 Carey KB , Henson JM , Carey MP , et al. : Perceived norms mediate effects of a brief motivational intervention for sanctioned college drinkers . Clin Psychol-Sci Pr 2010 ; 17 ( 1 ): 58 – 71 . Google Scholar CrossRef Search ADS 55 LaChance H , Feldstein Ewing SW , Bryan AD , et al. : What makes group MET work? A randomized controlled trial of college student drinkers in mandated alcohol diversion . Psychol Addict Behav 2009 ; 23 ( 4 ): 598 – 612 . Google Scholar CrossRef Search ADS PubMed 56 Judd CM , Kenny DA : Data analysis in social psychology: recent and recurring Issues. In: Handbook of Social Psychology . Hoboken, NJ, John Wiley & Sons, Inc , 2010 . 57 LoCastro JS , Potter JS , Donovan DM , et al. : Characteristics of first-time alcohol treatment seekers: the COMBINE study . J Stud Alcohol Drugs 2008 ; 69 ( 6 ): 885 – 895 . Google Scholar CrossRef Search ADS PubMed Author notes The contents do not represent the views of the U.S. Department of Veterans Affairs or the United States Government. © Association of Military Surgeons of the United States 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Journal

Military MedicineOxford University Press

Published: Sep 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off