A research paper

ContentServer2.pdf

Home >Social Science homework help >A research paper

MILITARY MEDICINE, 182, 11/12:e1941, 2017

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Margaret-Anne Mackintosh, PhD*; James Niehaus, PhD†; Casey T. Taft, PhD‡§; Brian P. Marx, PhD‡§; Kathleen Grubbs, PhD∥; Leslie A. Morland, PsyD∥¶**

ABSTRACT Objective: Anger is a symptom of post-traumatic stress disorder (PTSD) associated with a range of clinical and functional impairments, and may be especially prevalent among veterans with PTSD. Effective anger man- agement therapies exist but may be undermined by poor engagement or lack of treatment availability. Finding ways to engage veterans in anger management therapy or to improve access can be helpful in improving clinical outcomes. This randomized controlled trial compared anger management treatment (AMT) with AMT augmented by a mobile application (app) system, Remote Exercises for Learning Anger and Excitation Management (RELAX). Methods: Par- ticipants were 58 veterans enrolled in 12 sessions of either AMT alone or AMT with the RELAX system (AMT + RELAX). The RELAX system includes the RELAX app, a wearable heart rate monitor, a remote server, and a web- based therapist interface. RELAX allows the user to practice skills, monitor symptoms, and record physiological data. The server collects data on app use. A web-based interface allows the therapist to access data on between-session prac- tice, and skills use. Measures administered at baseline, post-treatment, and 3-and 6-month follow-up include state and trait anger, dimensions of anger, PTSD, depression, interpersonal functioning, and satisfaction. We used multilevel modeling to account for the nesting of time points within participants and participants within treatment groups. Predictors were Treatment Condition (AMT + RELAX and AMT), Linear Time (baseline, post-treatment, 3-and 6-month follow-up), and Quadratic Time and Treatment Condition × Linear Time interaction. All analyses were conducted using SPSS 21 (Armonk, New York). Approval was obtained from the institutional review board. Results: Across groups, the treatment dropout rate was 13.8%; of those who remained in treatment, 90% received an adequate dose of treatment (10 or more sessions). There were no significant differences between groups on attendance or treatment completion. Participants in both treatments demonstrated statistically significant and clinically meaningful reductions in anger severity and significant post-treatment reductions in PTSD. Veterans did not report significant changes in depression or interper- sonal functioning. Veterans in the AMT + RELAX group reported spending significantly less time on homework assignments, and they rated the AMT + RELAX app as helpful and easy to use, with these ratings improving over time. Conclusion: Findings suggest that AMT + RELAX was beneficial in reducing anger symptoms and promoting efficient use of the between-session practice; however, AMT + RELAX did not outperform AMT. This study is an important contribution as it is one of the first randomized controlled trials to study the efficacy of a technology- enhanced, evidence-based psychotherapy for anger management. Findings are limited because of small sample size and modifications to the technology during the trial. However, the results highlight the possible benefits of mobile app–supported treatment, including increasing the accessibility of treatment, lowering therapist workload, reducing costs of treatment, reducing practice time, and enabling new activities and types of treatments. This study presents pre- liminary evidence that mobile apps can be a valuable addition to treatment for patients with anger difficulties. Future research should evaluate how much therapist involvement is needed to support anger management.

INTRODUCTION Anger is a symptom of post-traumatic stress disorder (PTSD) that is frequently reported by veterans with PTSD across war eras.1–3 Higher anger symptoms among those with PTSD may result in higher overall PTSD symptom

chronicity and severity, and may interfere with PTSD treatment.4,5 Anger problems also increase the risk for aggression, intimate partner violence, legal problems, and interference with social relationships.6–8 Because of the diffuse clinical, interpersonal, and social consequences of

*Dissemination and Training Division, National Center for PTSD, VA Palo Alto Health Care System, 795 Willow Road, Building 334 (NC-PTSD), Room C-130, Menlo Park, CA 94025.

†Charles River Analytics, 625 Mt. Auburn Street, Cambridge, MA 02138. ‡Behavioral Science Division, National Center for PTSD, VA Boston

Healthcare System, U.S. Department of Veterans Affairs, 150 South Huntington Avenue, Boston, MA 02130.

§Boston University School of Medicine, 72 East Concord Street, Boston, MA 02118.

∥Pacific Islands Division, National Center for PTSD, VA Pacific Islands Health Care System, 3375 Koapaka Street, Suite I-560, Honolulu, HI 96819.

¶VA San Diego Healthcare System, U.S. Department of Veterans Affairs, 3350 La Jolla Village Drive (664BU), San Diego, CA 92161.

**Department of Psychiatry, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093.

The investigator(s) adhered to the policies regarding the protection of human subjects as prescribed by Code of Federal Regulations (CFR) Title 45, Volume 1, Part 46; Title 32, Chapter 1, Part 219; and Title 21, Chapter 1, Part 50 (Protection of Human Subjects). The views, opinions and/or findings contained in this report are those of the author(s) and should not be construed as an official Department of the Army position, policy, or decision so designated by other documentation.

doi: 10.7205/MILMED-D-17-00063

MILITARY MEDICINE, Vol. 182, November/December 2017 e1941

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

anger among those with PTSD, it is critical to develop effective anger management interventions.

Most of the existing evidence-based interventions for anger are on the basis of cognitive-behavior therapy (CBT). A meta-analysis of CBT treatment studies for anger, including studies with veterans with PTSD, demonstrated moderate-to- significant reductions in anger (Cohen’s d = 0.59–0.69), aggression (Cohen’s d = 0.99–0.89), and arousal (Cohen’s d = 0.52–0.59), and increases in positive behavior (Cohen’s d = 0.79–0.99) and attitudes (Cohen’s d = 0.83–1.05).9

CBT for anger is supported for combat veterans diagnosed with PTSD.10,11 Most CBT protocols emphasize skill develop- ment (e.g., deep breathing, relaxation, cognitive restructuring), which requires ongoing practice to reinforce new ways of cop- ing with all aspects of anger. A study investigating the mecha- nisms of change in anger management treatment (AMT) for veterans with PTSD identified the development of calming skills as a critical mediator of anger reduction.12

Despite support for CBT-based interventions, there are bar- riers to access, engagement, and retention in care.13,14 Acquir- ing skills through CBT requires repetition and feedback,15

which may be limited by sessions separated by several days. Treatment engagement (e.g., attendance, homework comple- tion) has predicted better outcomes at post-treatment and follow-up in studies of CBT for anxiety,16,17 underscoring the importance of between-session practice. Practice after the con- clusion of psychotherapy can improve maintenance of gains.

Mobile devices and their applications (apps) have the potential to improve treatment engagement. Preliminary data on mental health apps suggest that they are useful, safe, and appealing.18–20 Apps may be especially helpful in supporting anger management interventions because they can promote real-time access to self-management resources, which could promote practice. To explore this potential, we developed the Remote Exercises for Learning Anger and Excitation Management (RELAX) system. RELAX (1) enables skills practice, (2) integrates with evidence-based treatments, (3) provides feedback through physiological sensor integration, and (4) supports communication with the therapist. It is accessible between sessions to promote skill practice and maintenance. It interfaces with a heart rate monitor, which can be used to increase awareness of physiological arousal, prompt utilization of skills, and monitor changes in symp- toms. Real-time feedback encourages the use of adaptive cop- ing (e.g., calming skills) instead of negative coping to reduce distress (e.g., aggression).1,21 Before each therapy session, the therapist can access a log of exercises, physiological data, and view typed exercises. During the session, the therapist can use these data to tailor the therapy to the patient.

In this study, we examined whether veterans who received AMT accompanied by RELAX (AMT + RELAX) experi- enced significantly greater reductions in (1) state/trait anger, (2) dimensions of anger, (3) PTSD symptoms, and (4) symp- toms of depression when compared with AMT alone. We also examined interpersonal functioning, homework, and treatment

satisfaction between conditions. We also explored participant satisfaction with RELAX.

METHOD

Participants Veterans (N = 58) were recruited from the Department of Veterans Affairs (VA) in Honolulu, Hawaii. Figure 1 pro- vides the CONSORT table for study recruitment. Eligible veterans were men, over 18 years, with a State-Trait Anger Expression Inventory (STAXI) Trait Scale score of 20 or higher. Exclusionary criteria were active suicidal or homicidal ideation, alcohol or drug abuse, and current psychiatric hos- pitalization. Blinded clinicians conducted all baseline and follow-up assessments. The mean age of the participants was 53.3 years (SD = 12.5, range = 24–71 years). The sample was ethnically diverse, with 70% of veterans self-identifying as ethnoracial minorities. Roughly a third of veterans served during recent conflicts Operation Iraqi Freedom/Operation Enduring Freedom (OIF/OEF) and one quarter were cur- rently employed. Approval was obtained from the institu- tional review board.

Measures

State-Trait Anger Expression Inventory-2

The State-Trait Anger Expression Inventory-2 (STAXI-2)22 is a 57-item measure assessing state-anger, trait-anger, anger- expression, and anger-regulation. It was administered at all assessment time points. It has demonstrated strong reliability and validity.22,23 This study included only the 10-item Trait Anger (STAXI-Trait) and 32-item Anger Expression Index (STAXI-AEI) subscales. In this study, Cronbach’s alphas for the STAXI-Trait and STAXI-AEI were 0.93–0.96 and 0.70– 0.76, respectively.

Dimensions of Anger Reactions

The Dimensions of Anger Reactions-5 (DAR-5)24 is a brief (5-item) anger screener adapted from the original DAR scale.25 The DAR was administered at all assessment time points. Items assess anger frequency, intensity, duration, aggressiveness, and interpersonal problems. Each item is measured on a 5-point scale, ranging from 1 (not at all) to 5 (very much). Higher scores correspond to more intense anger. A score of 12 is the cutoff for psychological distress and/or functional impairment. The brief format is a reliable and valid screening tool.24 Cronbach’s alphas in this sam- ple were 0.80–0.89.

Brief Inventory of Psychosocial Functioning

The Brief Inventory of Psychosocial Functioning (B-IPF)26 is a 7-item self-report instrument measuring functional impair- ment over the last 30 days. The B-IPF was administered at all assessment time points. Dimensions of functioning include romantic and family relationships, work, friendships,

MILITARY MEDICINE, Vol. 182, November/December 2017e1942

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

socializing, parenting, education, and self-care. This instru- ment is validated for use with veterans.27 Cronbach’s alphas ranged from 0.90 to 0.94.

Patient Health Questionnaire—Depression Subscale

The Patient Health Questionnaire-9 (PHQ-9)28,29 is a 9-item depression screener. The PHQ-9 was administered at all assessment time points. The PHQ-9 measures the severity of depressive symptoms over the previous month. The measure is reliable, valid, and sensitive to change.28,30,31 Cronbach’s alphas ranged from 0.88to 0.92.

PTSD Checklist for DSM-5 (PCL-5)

The PTSD Checklist for DSM-5 (PCL-5)32 is a 20-item mea- sure of PTSD symptoms corresponding to the DSM-5.33 The PCL-5 was administered at all assessment time points. The PCL-5 has adequate psychometric properties.34 Cronbach’s alphas ranged from 0.95 to 0.97.

Technology Feedback Questionnaire

The Technology Feedback Questionnaire asked for feedback on the app and was administered during Sessions 3, 6, and

9, immediately post-treatment, and at the 3- and 6-month follow-ups. Items on the survey focused on ease, frustra- tion, and helpfulness of the app. Participants responded on a 6-point scale, with lower scores representing negative rat- ings and higher scores representing favorable ratings.

Treatment Engagement

We used three variables to investigate the veterans’ treat- ment engagement: (1) dropout (fewer than 12 therapy ses- sions); (2) session attendance (mean number of sessions); and (3) treatment completion (attended at least 10 sessions).

Homework Engagement

For Sessions 2–12, participants were asked to report how many minutes they spent on skill-building activities. Minutes for all sessions were summed to create an aggregated estimate.

Anger Management Treatment AMT is a 12-session, 90-minute group therapy.35,36 It includes psychoeducation, self-monitoring training, skills training (e.g., relaxation, assertiveness training, communica- tion skills, anger plans), and cognitive restructuring.

FIGURE 1. CONSORT diagram.

MILITARY MEDICINE, Vol. 182, November/December 2017 e1943

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

Remote Exercises for Learning Anger and Excitation Management

The RELAX system includes multiple components that sup- port AMT, including (1) the RELAX app, (2) a wearable heart rate monitor, (3) a remote server, and (4) a web-based therapist interface. It allows the user practice skills, monitor symptoms, and record physiological data. The server collects data on app use. A web-based interface allowed the therapist to access data on between-session practice, and use of skills. Using the app, veterans log anger frequency, intensity, and cues; access suggestions for behavioral strategies; and create individually tailored anger management plans to prompt them throughout the day.

Treatment Modifications

Several additions were made to the AMT protocol to accom- modate technical support. Up to 20 minutes were added to several sessions and an interactive television monitor facili- tated training. Checklists and scripts for introducing the technology were used to promote treatment fidelity and a RELAX user manual was created for veterans.

Procedures

Providers from the VA, Vet Center, and Department of Defense referred veterans to the study from March 2014 to June 2015. Screening procedures introduced the study,

TABLE I. Differences in Demographic and Baseline Variables Between Treatment Conditions

Variable

AMT + RELAX(N = 28) AMT(N = 30) Test Statistics

n % n % χ p

Living with Partnera 17 60.7 23 76.7 1.72 0.189 Primary Ethnicity 3.88 0.693 NHOPI 12 42.9 12 40.0 Asian 4 14.3 3 10.0 Black 1 3.6 1 3.3 NA/NA 0 0.0 1 3.3 Hispanic 3 1.9 1 3.3 Mixed Race 2 7.1 1 3.3 White 6 21.4 11 36.7

Education 3.29 0.511 Less than High School 1 3.6 0 0.0 High School Graduate 8 28.6 6 20.0 Some College 14 50.0 16 53.3 College Graduate 3 10.7 7 23.3 Advanced/Professional Degree 2 7.1 1 3.3

Current Students 8 28.6 3 10.0 3.25 0.071 Employment Status 4.07 0.539 Working Full Time 7 25.0 7 23.3 Working Part Time 1 3.6 1 3.3 Unemployed, Seeking Employment 1 3.6 4 13.3 Unemployed, Not Seeking 6 21.4 3 10.0 Unemployed Because of Disability 7 25.0 5 16.7 Retired 6 21.4 10 33.3

Service Connection Income 21 75.0 22 80.0 0.02 0.885 Branch of Serviceb

Army 20 71.4 16 53.4 2.01 0.156 Navy 3 10.7 10 33.3 4.26 0.039 Air Force 3 10.7 3 10.0 0.01 0.929 Marines 3 10.7 1 3.3 1.23 0.268

Military Service Erab

Cold War II (1954–1964) 0 0.0 1 1.7 1.00c

Vietnam (1964–1975) 8 28.6 13 43.3 1.37 0.242 Cold War III (1976–1991) 6 21.4 3 10.0 1.44 0.230 DS/DS (1990–1991) 3 10.7 4 13.3 0.09 0.760 Post–Cold War (1991–2001) 2 7.1 1 3.3 0.605c

OIF/OEF/OND (2001–present) 10 35.7 11 36.7 0.01 0.940 Service Connection Conditionb

PTSD 10 35.7 11 36.7 0.01 0.940 Non-PTSD Mental Health 2 7.1 7 23.3 0.147c

Medical Condition 21 75.0 24 80.0 0.21 0.648 Legal Issues 20 71.4 17 56.7 1.37 0.242

NA/NA, Native American/Native Alaskan; NHOPI, Native Hawaiian/Other Pacific Islander; DS/DS, Desert Storm/Desert Shield. aThese veterans were either married and living with their partner or unmarried and living together. bVeterans may have served in more than 1 branch of service or war era and be service connected for more than one condition. cFisher’s exact test.

MILITARY MEDICINE, Vol. 182, November/December 2017e1944

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

assessed veteran interest and availability, and assessed pre- liminary inclusion (≥20 on the 10-item STAXI-Trait)9,35

and exclusion (alcohol or substance use). Veterans meeting the screening criteria completed an in-person assessment. Participants were randomized using a block randomiza- tion by a blinded member of the study staff. Participants randomized into AMT + RELAX received a brief RELAX orientation.

Among the 74 participants, one-half were randomized to AMT + RELAX (n = 37) and one-half were randomized to AMT (n = 37). In the AMT + RELAX group, 28 participants completed treatment. In the AMT group, 30 participants com- pleted treatment. The intent-to-treat analysis included 28 par- ticipants from AMT + RELAX (per protocol n = 23) and 30 participants from AMT (per protocol n = 22).

AMT sessions were held twice per week for 6 weeks. Participants completed assessment measures before each ses- sion. Data from the app were uploaded to the remote server. One doctoral-level therapist and one therapist with masters- level training were group therapy coleaders. Post-treatment assessments were conducted immediately after the final ses- sion and at 3 and 6 months, in-person and by masters-level clinicians blind to treatment condition.

Data Analysis This study investigated whether veterans who received AMT + RELAX demonstrated significantly larger reductions in anger and PTSD symptoms compared with those who received AMT alone. Outcomes included (1) state and trait anger, (2) dimen- sions of anger, (3) PTSD symptoms, (4) depression symptoms, (5) interpersonal functioning, and (6) satisfaction. We used multilevel modeling to account for the nesting of time points within participants and participants within treatment groups, which were treated as random variables. Predictors were Treat- ment Condition (AMT + RELAX and AMT), Linear Time (baseline, post-treatment, 3- and 6-month follow-up), and Qua- dratic Time. We also included the Treatment Condition × Lin-

ear Time interaction term. We used restricted maximum likelihood and multiple imputation analyses to handle missing data using a Markov chain Monte Carlo model with 10 data sets. All analyses were conducted using SPSS 21.

RESULTS Table I describes the demographic and background charac- teristic. The only baseline demographic difference was in branch of services, with 33.3% of AMT veterans in the Navy veterans in AMT compared with 10.7% in AMT + RELAX. Table II provides descriptive statistics for all outcome measures as well as standardized effect size estimates (i.e., Cohen's d) measuring change from baseline to each of the post-treatment assessments (i.e., immediately post-treatment, 3- and 6-month follow-ups). There were no significant differences between treatment conditions on any measure. For each of the six models, a null model was estimated to determine the amount of variance of each outcome associated with each nested level. Less than 3% of the variance in outcomes was associated with subjects nested within treatment groups. Per recommendations, this level was dropped.37

Changes in Anger and Anger-Related Impairment Table II provides pre- to post-treatment standardized effect size estimates for each measure at each follow-up time point. Table III presents the parameter estimates with 95% confidence intervals (CIs) from the mixed models for the primary anger outcome measures, and Table IV presents similar results for the secondary mental health measures. Across all measures, the Treatment Condition × Linear Time interactions were not significant, with all Student t tests greater than or equal to −0.31 and all p values greater than or equal to 0.76. Anger reductions were similar across time in both groups. Between-group differences in standardized effect sizes (Cohen’s d) across time points ranged from 0.06 to 0.27 for the STAXI Trait, −0.20 to 0.13 for the STAXI-AEI, and −0.02 to 0.08 for the

TABLE II. Descriptive Statistics and Standardized Pre- to Post-Treatment Effect Size Estimates for Outcome Measures

Outcome Measure Group

Baseline Post-Treatment 3 Months Post-Treatment 6 Months Post-Treatment

M SE M SE ES M SE ES M SE ES

STAXI-Trait AMT + RELAX 25.8 0.94 19.0 1.45 1.07 17.5 1.28 1.29 19.1 1.35 1.05 AMT 27.0 0.96 20.5 1.97 1.01 20.5 2.00 1.02 20.8 1.98 0.98

STAXI-AEI AMT + RELAX 46.4 1.71 37.9 2.04 0.79 33.9 2.02 1.17 35.2 1.87 1.05 AMT 48.9 1.47 38.3 2.59 0.99 37.2 2.50 1.10 39.1 2.61 0.91

DAR-5 AMT + RELAX 15.5 0.70a 10.7 1.94 0.80 11.2 2.03 0.73 11.5 1.90 0.66 AMT 17.2 0.75a 12.6 1.58 0.78 12.8 1.64 0.75 13.7 1.40 0.59

PCL-5 AMT + RELAX 43.3 3.39 32.3 9.33 0.40 33.9 9.64 0.34 32.0 9.26 0.41 AMT 50.3 3.60 43.4 6.68 0.25 43.1 7.25 0.26 40.6 6.33 0.35

PHQ-9 AMT + RELAX 12.3 1.02 8.9 4.07 0.32 10.0 4.04 0.22 9.0 4.04 0.31 AMT 14.0 1.34 8.30 2.25 0.54 10.6 2.59 0.32 10.2 2.01 0.36

B-IPF AMT + RELAX 2.8 0.34 2.3 1.08 0.15 2.2 1.06 0.20 2.0 1.04 0.26 AMT 3.3 0.31 2.6 0.78 0.24 2.6 0.76 0.22 2.4 0.63 0.30

ES, pre- to post-treatment standardized effect sizes estimates on the basis of Cohen’s d; SE, standard error. aIndependent sample Student t tests comparing treatment conditions at the specified time points indicate outcome means are marginally different statistically at 0.05 < p < 0.10.

MILITARY MEDICINE, Vol. 182, November/December 2017 e1945

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

DAR-5 (negative values favoring AMT and positive values favoring AMT + RELAX).

Investigation of effects for time indicated significant lin- ear reductions in all anger outcomes across conditions (all Student t tests greater than or equal to −4.36 and p values less than or equal to 0.001). These results suggested large d for pre- to post-treatment changes in anger-related symp- toms across time points (STAXI-Trait mean d = 1.00, STAXI-AEI mean d = 1.07, and DAR-5 mean d = 0.72). There were significant quadratic effects for each anger out- come, indicating scores increased at later time points (see means in Table II; all Student t tests were greater than or equal to 3.28 and p values were less than or equal to 0.002). In spite of large reductions in anger symptoms from pre- treatment to each follow-up assessment (i.e., d ≥ 0.80), there were only small differences between conditions (i.e., d ≤ 0.20).

The total number of minutes for skill-building activities did not predict any anger outcomes. Other model parameters indicated that there was significant variability in baseline scores (intercepts) on all anger outcomes ( p values less than or equal to 0.01). None of the parameter estimates for vari- ability over time were statistically significant.

Changes in Secondary Clinical Outcomes We also used mixed models to investigate the effects of Treat- ment Condition, Linear and Quadratic Time, and Treatment Condition × Linear Time interaction on mental health symp- toms (PCL-5 and PHQ-9) and interpersonal functioning (B-IPF). The right half of Table III presents the parameter esti- mates with 95% CIs from the mixed models for these mea- sures. Table II provides pre- to post-treatment Cohen’s d at each follow-up time point by condition. None of the Treatment Condition × Linear Time interactions were statistically signifi- cant. Between-group differences in standardized effect sizes across the post-treatment assessment points were small and ranged from 0.06 to 0.15 for the PCL,−0.21 to −0.05 for the PHQ-9, and −0.08 to −0.02 for the B-IPF (negative values favoring AMT and positive values favoring AMT + RELAX).

Inspection of baseline scores indicated both statistically and clinically significant levels of PTSD symptoms (t = 2.75, p = 0.01), moderate levels of depressive symptoms (t = 1.79, p = 0.08), and interpersonal difficulties (t = 1.87, p = 0.07). Effects for time indicated significant linear reductions on the PCL-5 (t = −2.51, p = 0.01, mean pre- to post-treatment follow-up d = 0.33). A marginal reduction was found on

TABLE IV. Mixed Modeling Results for Secondary Mental Health Outcomes

Parameter

Secondary Mental Health Outcomes

PCL-5 PHQ-9 B-IPF

Estimate 95 CI% Estimate 95 CI% Estimate 95 CI%

Intercept 57.8 14.76, 100.89*** 13.4 −1.93, 28.71* 3.4 −0.28, 7.07* Linear Time −7.3 −13.04, −1.54** −1.9 −4.08, 0.23* −0.2 −0.75, 0.43 Quadratic Time 1.4 −0.61, 3.33 0.4 −0.36, 1.13 −0.3 −0.22, 0.16 Treatment Condition −10.3 −20.84, 0.24* −1.5 −5.39, 2.39 −0.5 −1.54, 0.54 Homework (Minutes)a −1.1 −8.26, 6.11 0.01 −2.46, 2.49 −0.1 −0.61, 0.56 Linear Time by Condition 0.9 −3.15, 4.92 0.01 −1.43, 1.58 0.1 −0.28, 0.52 Intercept Variance 277.4 146.11, 408.59**** 32.9 15.98, 49.84**** 2.2 1.07, 3.40**** Time Variance 12.3 −7.98, 32.48 0.8 −1.98, 3.54 0.1 −0.07, 0.18 Residual Variance 27.9 66.73, 176.13**** 18.8 11.59, 25.97**** 1.5 0.96, 1.94****

*p < 0.10, **p < 0.05, ***p < 0.01, ****p < 0.001. aBecause of significant skewness, the log of minutes spent on homework was used in all analyses.

TABLE III. Mixed Modeling Results for Primary Anger Outcomes

Parameter

Primary Anger Outcomes

STAXI-Trait STAXI-AEI DAR-5

Estimate 95 CI% Estimate 95 CI% Estimate 95 CI%

Intercept 23.3 14.84, 31.70*** 57.5 39.74, 75.28*** 19.2 12.63, 25.76*** Linear Time −6.9 −8.75, −5.01*** −10.8 −15.76, −5.91*** −5.3 −6.92, −3.69*** Quadratic Time 1.7 1.13, 2.25*** 2.60 1.02, 4.19** 1.4 0.84, 1.90*** Treatment Condition −1.3 −3.80, 1.26 −2.9 −8.60, 2.87 −2.1 −4.11, 0.91* Homework (Minutes)a 0.6 −0.73, 1.97 −1.5 −4.29, 1.35 −0.4 −1.39, 0.69 Linear Time by Condition −0.2 −1.34, 1.00 −0.5 −3.36, 2.44 −0.1 −1.01, 0.91 Intercept Variance 9.2 2.74, 15.62** 43.9 9.88, 77.89*** 5.8 1.80, 9.72** Time Variance 1.5 −0.01, 3.06* 8.8 −0.74, 18.38* 0.6 −0.31, 1.49 Residual Variance 15.4 8.96, 21.84*** 89.3 65.53, 114.03*** 10.2 6.20, 14.12***

*p < 0.10, **p < 0.05, ***p < 0.001. aBecause of significant skewness, the log of minutes spent on homework was used in all analyses.

MILITARY MEDICINE, Vol. 182, November/December 2017e1946

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

the PHQ-9 (t = −1.76, p = 0.08, mean pre- to post-treatment d = 0.35). No change was found on the B-IPF. There were no significant effects for Quadratic Time for any measure.

Treatment Engagement There were no differences between treatment conditions on measures of engagement. Dropout rates were 7% for AMT + RELAX and 20% for AMT (Fisher’s exact test p = 0.26), and average number of sessions attended was 10.1 sessions (SD = 2.5) for AMT + RELAX and 9.3 ses- sions (SD = 3.8) for AMT. Finally, there were no differ- ences in the proportion of non-dropouts who received an adequate dose or therapy 88.5% in AMT + RELAX (88.5%) and in AMT (91.7%).

Homework Engagement Veterans in the AMT group reported spending a signifi- cantly longer time practicing skill-building activities between sessions compared with those in AMT + RELAX group, t(164) = 2.30, p = 0.023. Means were 666.0 (SD = 564.47) minutes for AMT and 365.6 (SD = 303.41) minutes for AMT + RELAX. The between-group effect size at post-treatment was d = 0.58. We included minutes spent on skill-building as a covariate in all analyses; the value was log transformed because of positive skew.

User Ratings of the RELAX App Figures 2–4 depict mean ratings of the RELAX app for (1) ease, (2) helpfulness, and (3) frustration (reverse scored) mea- sured at Sessions 3, 6, 9, post-treatment, and 3- and 6-month follow-up, respectively. Paired sample Student t tests were conducted using Session 3 scores compared with each time point. Overall, veterans rated RELAX positively with ratings improving as treatment progressed.

DISCUSSION Findings suggest a lack of significant differences between conditions. Contrary to expectations, those in AMT + RELAX did not differ from those in AMT with respect to changes in anger, functional impairment, depression, and PTSD. Evidence suggested the possibility of greater engage- ment in AMT + RELAX, with only a 7% treatment dropout

rate relative to a 20% dropout rate for AMT although this difference did not reach statistical significance. It appeared that RELAX may have facilitated homework completion, requiring less time to complete assignments. Finally, partici- pants rated the RELAX app as easy to use and helpful.

AMT has demonstrated efficacy in previous trials so it is unsurprising that the addition of RELAX to AMT did not improve clinical outcomes.35,38 Furthermore, the enhance- ment of the homework experience may not represent a large-enough program change to find statistical significance. It is noteworthy that for almost all outcomes, mean scores for those receiving RELAX, were more favorable than those receiving AMT. A larger controlled trial with greater statisti- cal power would be needed to detect these potential benefits.

Since other research indicates that telehealth-administered anger management is equally as effective as in-person- administered anger management,38 the next logical step in this area of research would be a stand-alone app or with limited clinical support. Evidence from this study—that veterans found RELAX to be helpful, easy to use, and efficient—suggests that a stand-alone intervention could be delivered in an effec- tive and acceptable manner for veterans with anger problems.

This study is an important contribution as it is one of the first randomized controlled trials of a technology-enhanced, evidence-based psychotherapy for anger management. RELAX is novel because it incorporates a portable electronic version of therapy materials and a physiological sensor to enhance the patient experience. Participants were a diverse sample of veterans with varying degrees of experience using technol- ogy. Findings are limited because of small sample size and modifications to the technology during the trial. These

FIGURE 2. Mean ratings for ease of use of the RELAX app. mo, month. *Ratings at specified time point were significantly different from Session 3 ratings using a paired sample Student t test at p < 0.05.

FIGURE 3. Mean ratings for helpfulness of the RELAX app. mo, month. *Ratings at specified time point were significantly different from Session 3 ratings using a paired sample Student t test at p < 0.05.

FIGURE 4. Mean ratings for lack of frustration related to the RELAX app. mo, month. *Ratings at specified time point were significantly different from Session 3 ratings using a paired sample Student t test at p < 0.05.

MILITARY MEDICINE, Vol. 182, November/December 2017 e1947

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

limitations are inherent to clinical trials involving mobile technology and could be difficult to overcome using tradi- tional methods of investigation.39 In spite of these limita- tions, this study is an important contribution because there are currently few published studies investigating the efficacy of mobile devices in clinical practice.

This study may also point toward the limitations of tradi- tional research designs in assessing the benefits of new technology-enhanced interventions. Possible benefits include increasing the accessibility of treatment, lowering therapist workload, reducing costs of treatment, and enabling new activities and types of treatments. For example, although treatment options are offered at many VA Medical Centers and military Medical Treatment Facilities, practical issues often prevent veterans and service members from taking full advantage of these options. Not every veteran and service member has access to locations or clinic offerings. Therapies supported or wholly delivered by mobile technologies enable patients to receive treatment anywhere.

Overall, findings suggest that AMT + RELAX was bene- ficial in reducing anger symptoms and promoting efficient use of the between-session practice; however, AMT + RELAX did not outperform AMT in our sample. This study presents preliminary evidence that mobile apps can be a via- ble option for patients with anger difficulties; future research could evaluate how much therapist involvement is needed to support effective anger management interventions using an app-based intervention.

ACKNOWLEDGMENTS The work is, in part, supported by the U.S. Army Medical Research and Material Command under Contract No. W81XWH-12-C-0067 (Department of Defense SBIR 12.2, Topic No. OSD11-H13, Phase II as a subcontract through Charles River Analytics, ERMS Log No. 11271015). Charles River Analytics has Small Business Innovation Research data rights to the RELAX app system. Support was also provided by the VA National Center for PTSD and with resources and the use of facilities at the Spark M. Matsunaga VA Pacific Islands Health Care System.

REFERENCES 1. Novaco RW, Chemtob CM: Anger and combat-related posttraumatic

stress disorder. J Trauma Stress 2002; 15: 123–32. 2. Sayer NA, Noorbaloochi S, Frazier P, Carlson K, Gravely A, Murdoch

M: Reintegration problems and treatment interests among Iraq and Afghanistan combat veterans receiving VA medical care. Psychiatr Serv 2010; 61: 589–97.

3. Rosen C, Adler E, Tiet Q: Presenting concerns of veterans entering treat- ment for posttraumatic stress disorder. J Trauma Stress 2010; 26: 640–43.

4. Andrews B, Brewin CR, Rose S, Kirk M: Predicting PTSD symptoms in victims of violent crimes: the role of shame, anger and childhood abuse. J Abnorm Psychol 2000; 109: 69–73.

5. Frueh BC, Henning KR, Pellegrin KL, Chobot K: Relationship between scores on anger measures and PTSD symptomatology, employment, and compensation-seeking status in combat veterans. J Clin Psychol 1997; 53: 871–78.

6. Novaco RW, Chemtob CM: Violence associated with combat-related posttraumatic stress disorder: the importance of anger. Psychol Trauma 2015; 7: 485–92.

7. Taft CT, Street AE, Marshall AD, Dowdall DJ, Riggs DS: Posttraumatic stress disorder, anger, and partner abuse among Vietnam combat veterans. J Fam Psychol 2007; 21: 270–77.

8. Wilk JE, Quartana PJ, Clarke-Walper K, Kok BC, Riviere LA: Aggres- sion in US soldiers post deployment: associations with combat exposure and PTSD and the moderating role of trait anger. Aggress Behav 2015; 41(6): 556–65.

9. DiGiuseppe R, Tafrate RC: Anger treatment for adults: a meta-analytic review. Clin Psychol 2003; 10(1): 70–84.

10. Chemtob CM, Novaco RW, Hamada RS, Gross DM: Cognitive- behavioral treatment for severe anger in posttraumatic stress disorder. J Consult Clin Psychol 1997; 65: 184–89.

11. Reilly PM, Clark HW, Shopshire MS, Lewis EW, Sorensen DJ: Anger management, post-traumatic stress disorder, and substance abuse. In: NIDA Research Monograph: Problems of Drug Dependence, Proceedings of the 57th Annual Scientific Meeting, College on Problems of Drug Dependence, p. 322. Rockville, MD, National Institute on Drug Abuse, 1996.

12. Mackintosh MA, Morland LA, Frueh BC, Greene CJ, Rosen CS: Peeking into the black box: mechanisms of action for anger manage- ment treatment. J Anxiety Disord 2014; 28: 687–95.

13. Elliott P, Biddle D, Hawthorne G, Forbes D, Creamer M: Patterns of treatment response in chronic posttraumatic stress disorder: an application of latent growth mixture modeling. J Trauma Stress 2005; 18: 303–11.

14. Morland LA, Love AR, Mackintosh M, Greene CJ, Rosen CS: Treating anger and aggression in military populations: research updates and clinical implications. Clin Psychol 2012; 19: 305–22.

15. Freeman A: The use of homework in cognitive behavior therapy: working with complex anxiety and insomnia. Cogn Behav Pract 2007; 14: 261–67.

16. Glenn D, Golinelli D, Rose RD, et al: Who gets the most out of cognitive behavioral therapy for anxiety disorders? The role of treatment does and patient engagement. J Consult Clin Psychol 2013; 81: 639–49.

17. LeBeau RT, Davies CD, Culver NC, Craske MG: Homework compli- ance counts in cognitive-behavioral therapy. Cogn Behav Ther 2013; 42: 171–79.

18. Kuhn E, Greene C, Hoffman J, et al: Preliminary evaluation of PTSD Coach, a smartphone app for post-traumatic stress symptoms. Mil Med 2014; 179: 12–18.

19. Luxton DD, McCann RA, Bush NE, Mishkind MC, Reger GM: mHealth for mental health: integrating smartphone technology in behav- ioral healthcare. Prof Psychol Res Pract 2011; 42: 505–12.

20. Reger GM, Hoffman J, Riggs D, et al: The “PE Coach” smartphone application: an innovative approach to improving implementation fidelity, and homework adherence during prolonged exposure. Psychol Serv 2013; 10: 342–49.

21. Taft CT, Weatherill RP, Scott JP, Thomas SA, Kang HK, Eckhardt CI: Social information processing in anger expression and partner violence in returning U.S. veterans. J Trauma Stress 2015; 28: 314–21.

22. Spielberger CD: STAXI-2: State-Trait Anger Expression Inventory-2: Pro- fessional Manual. Lutz, FL, Psychological Assessment Resources, 1999.

23. Spielberger CD, Reheiser EC: Measuring anxiety, anger, depression, and curiosity as emotional states and personality traits with the STAI, STAXI, and STPI. In: Comprehensive Handbook of Psychological Assessment, Vol. 2, pp 70–86. Edited by Hilsenroth MJ, Segal DL, Hersen M, Hoboken, NJ, John Wiley & Sons, 2003.

24. Forbes D, Alkemade N, Mitchell D, et al: Utility of the dimensions of anger reactions-5 (DAR-5) scale as a brief anger measure. Depress Anxiety 2014; 31: 166–73.

25. Forbes D, Hawthorne G, Elliott P, et al: A concise measure of anger in combat-related posttraumatic stress disorder. J Trauma Stress 2004; 17: 249–56.

26. Rodriguez P, Holowka DW, Marx BP: Assessment of posttraumatic stress disorder-related functional impairment: a review. J Rehabil Res Dev 2012; 49: 649–66.

MILITARY MEDICINE, Vol. 182, November/December 2017e1948

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

27. Marx B, Schnurr P, Rodriguez P, et al: Development of a functional impairment scale for active duty service members and veterans. Paper presented at the 25th annual meeting of the International Society for Traumatic Stress Studies, Atlanta, GA, 2009.

28. Kroenke K, Spitzer RL, Williams JB: The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001; 16: 606–13.

29. Spitzer RL, Kroenke K, Williams JBW: Validation and utility of a self- report version of PRIME-MD: the PHQ primary care study. JAMA 1999; 282: 737–44.

30. Löwe B, Kroenke K, Herzog W, Gräfe K: Measuring depression out- come with a brief self-report instrument: sensitivity to change of the Patient Health Questionnaire (PHQ-9). J Affect Disord 2004; 81: 61–66.

31. Löwe B, Spitzer RL, Gräfe K, et al: Comparative validity of three screening questionnaires for DSM-IV depressive disorders and physi- cians’ diagnoses. J Affect Disord 2004; 78: 131–40.

32. Weathers FW, Litz BT, Keane TM, Palmieri PA, Marx BP, Schnurr PP: The PTSD Checklist for DSM-5 (PCL-5). Boston, MA, National Center for PTSD, Behavioral Science Division, 2013.

33. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Ed 5, Washington, DC, American Psychiatric Associ- ation, 2013.

34. Bovin MJ, Marx BP, Weathers FW, et al: Psychometric properties of the PTSD Checklist for DSM-5 (PCL-5) in Veterans. Psychol Assess 2015; 28(11): 1379–91.

35. Reilly PM, Shopshire MS: Anger Management for Substance Abuse and Mental Health Clients: A Cognitive Behavioral Therapy Manual, DHHS Pub. No. (SMA) 02-3756. Rockville, MD, Center for Substance Abuse Treatment, Substance Abuse and Mental Health Services Adminis- tration, 2002.

36. Reilly PM, Shopshire MS, Durazzo TC, Campbell TA: Anger Manage- ment for Substance Abuse and Mental Health Clients: Participant Workbook, DHHS Pub. No. (SMA) 02-36622002. Rockville, MD, Center for Substance Abuse Treatment, Substance Abuse and Mental Health Services Administration, 2002.

37. Heck RH, Thomas SL, Tabata LN. Multilevel and Longitudinal Model- ing with IBM SPSS. New York, NY, Routledge, 2010.

38. Morland LA, Greene CJ, Rosen CS, et al: Telemedicine for anger management therapy in a rural population of combat veterans with posttraumatic stress disorder: a randomized noninferiority trial. J Clin Psychiatry 2010; 71: 855–63.

39. Asch DA, Rosin R: Innovation as discipline, not fad. N Engl J Med 2015; 373: 592–94.

MILITARY MEDICINE, Vol. 182, November/December 2017 e1949

Using a Mobile Application in the Treatment of Dysregulated Anger Among Veterans

Downloaded from publications.amsus.org: AMSUS - Association of Military Surgeons of the U.S. IP: 202.177.173.189 on Nov 05, 2017.

Copyright of Military Medicine is the property of AMSUS and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.