Adapting and Examining Collaborative Decision Skills Training Among Veterans With Serious Mental Illness (CDST)

Adapting and Examining Collaborative Decision Skills Training Among Veterans With Serious Mental Illness

Recovery-oriented care is an imperative for the VA, particularly in mental health programming for Veterans with serious mental illness (SMI). Collaborative decision-making (CDM) is a recovery-oriented approach to treatment decision-making that assigns equal participation and obligation to patients and providers across all aspects of decision-making, thereby empowering patients and facilitating better decision-making based on patient values and preferences. CDM is associated with several important outcomes including improved treatment engagement, treatment satisfaction, and social functioning. However, current levels of CDM among Veterans with SMI are low, and there is not yet an evidence-based method to improve CDM. Improving Veteran skill sets associated with engaging in CDM is a potential intervention strategy. Collaborative Decision Skills Training (CDST) is a promising new intervention that was previously developed by the applicant for use in adult civilians with SMI and found to improve relevant skills and improve sense of personal recovery. The proposed study has two primary stages. First, a small, one-armed, open label trial will establish CDST's feasibility will evaluate CDST among 12 Veterans with SMI receiving services at the VA San Diego Psychosocial Rehabilitation and Recovery Center (PRRC) and identify and complete any needed adaptations to CDST. Stakeholder feedback from Veterans, VA clinicians, and VA administrators will be collected to assess Veteran needs and service context to identify any needed adaptations to the CDST manual or the delivery of CDST to maximize its impact and feasibility. The developers of CDST will review all feedback and make final decisions about adaptations to ensure that CDST retains its essential components to protect against loss of efficacy. For example, a recommendation to adjust role-play topics to better reflect the needs of Veterans would be accepted because it would increase CDST's relevance without impairing its integrity, but a recommendation to remove all role-plays would not be accepted because it would cause loss of a key component. Second, CDST will be compared to active control (AC) using a randomized clinical trial of 72 Veterans. The primary outcome measure will be functioning within the rehabilitation context, operationalized as frequency of Veteran CDM behaviors during Veteran-provider interactions. Secondary outcomes are treatment attendance, engagement, satisfaction, and motivation, along with treatment outcomes (i.e., rehabilitation goal attainment, sense of personal recovery, symptom severity, and social functioning). Three exploratory outcomes will be assessed: Veteran-initiated collaborative behaviors, acute service use and provider attitudes and behavior. Veterans will be randomly assigned to CDST or AC conditions. Veterans in the both groups will attend eight hour-long group sessions held over eight weeks. All Veterans will complete an assessment battery at baseline, post-intervention, and at three-month post-intervention follow-up. Following the trial and adaptation phase, the findings will be used to develop a CDST service delivery manual and design a logical subsequent study. The results of the proposed study will inform the potential for larger trials of CDST and the utility of providing CDST broadly to Veterans with SMI. The results of this study will expand current understanding of CDM among Veterans with SMI by providing data that will: 1) identify adaptations needed to optimize CDST for Veterans receiving services in PRRCs; 2) identify possible benefits of CDST; 3) inform development of alternate interventions or methods to improve CDM; and 4) further elucidate CDM and associated treatment processes among Veterans with SMI receiving VA rehabilitation services.

The proposed study has three phases, each associated with an Aim. First, a small, one-arm, open-label trial will establish feasibility of Collaborative Decision Skills Training (CDST) for Veterans receiving care in Psychosocial Rehabilitation and Recovery Centers (PRRCs), identify and complete adaptations, and pilot the assessment strategy. Stakeholder feedback will be collected via the Dynamic Adaptation Process in order to identify whether aspects of CDST or service delivery need to be adapted before conducting the RCT. Second, an RCT will compare CDST with an active control (AC) among 72 Veterans with serious mental illness (SMI). These two phases will address two key aspects of evaluating promising interventions: 1) whether CDST could feasibly be integrated into usual care practices in the VA San Diego Healthcare System (VASHDS) PRRC and other PRRCs nationwide, and 2) whether CDST leads to significant improvements in important domains for the target population. Third, findings from the first two phases will inform the design of a subsequent merit study and the completion of two primary products a CDST Clinician Manual and CDST Service Delivery Manual. Aim 1: Conduct a small, one-arm, open-label trial to identify and complete any adaptations needed to successfully integrate CDST into PRRC settings. The Dynamic Adaptation Process will be used to produce a CDST Clinician Manual and a CDST Service Delivery Manual. The existing CDST Clinical Manual from the pilot study will be modified as needed during Aim 1 based on findings from this project. A new CDST Service Delivery Manual will be developed during Aim 1 and honed during Aim 1 and Aim 2 study activities. An Adaptation Resource Team (ART) will be formed including 2 Veterans with SMI currently receiving services in the VASDHS PRRC, 2 clinicians who provide treatment in the VASDHS PRRC, and 2 administrators whose duties include administration of the VASDHS PRRC Veterans will be enrolled via: 1) referrals from clinicians, and 2) print advertisements in the PRRC waiting room. Clinicians and administrators will be solicited directly. Written informed consent will be obtained from each participant. Each ART member will receive a copy of the Clinician Manual and a description of the purpose of the Service Delivery Manual. They will be asked to review these materials and make notes of their perceptions, including components that are not relevant, may not resonate with the Veteran population or would not be feasible within the PRRC context. Approximately 2 weeks later, each participant will complete an individual interview, accompanied by the research assistant, using a semi-structured interview approach. These interviews will systematically work through the CDST content and service delivery aspects of the intervention. These interviews will be analyzed, compiled, and presented back to the ART for discussion. The ART will come to consensus about adaptations with supervision from the CDST developers to ensure fidelity to CDST's active ingredients. The open trial will occur following initial adaptation. Twelve Veterans will participate. There will not be a comparison group for the open trial. During the open trial and RCT, there will be three assessment time points: baseline, post-intervention, and follow-up at 3 months postintervention completion. One measure will also be completed at the midpoint but this will be done by recovery coaches and does not require Veteran time. Veterans will receive compensation for assessment completion but not for intervention attendance to minimize the potential impact of compensation on treatment engagement. Veteran, clinician, and administrator ART participants will all be encouraged to continue in the ART during RCT. RCT Veterans will be enrolled on a cohort bases in tandem with the trial. Specifically, after each RCT cohort has completed data collection, two Veterans who were assigned to the CDST group will be recruited to join the adaptation process. Therefore this will equal 10 Veterans total (2 non-trial + 2 open trial + 2from RCT cohort 1 + 2 from RCT cohort 2 + 2 from RCT cohort 3). If more than 2 Veterans from a given cohort wish to join, all Veterans will be admitted (4 Veterans per cohort have been budgeted for to account for this). If Aim 1 ART members choose not to continue, efforts will be made to replace them by a person with a similar role (e.g., another provider, another non-trial Veteran). The ART will meet twice before each new RCT cohort to review and approve additions and changes to the Service Delivery Manual, and will meet once following completion of data collection for each cohort to discuss lessons learned from that cohort to inform the Service Delivery Manual. Following the end of all RCT data collection in Year 5, ART clinicians and administrators will complete one semi-structured interview to assess implementation and sustainability feasibility of CDST for the VASDHS PRRC. In order to have a final sample of 58 Veterans combined across conditions in the RCT, the investigators will recruit 72 Veterans (estimating 20% attrition). There will be three cohorts, with 24 Veteran per cohort. In years 2-4, there will be one cohort recruited annually. Veterans will be randomized by cohort with 12 Veterans randomized to CDST and 12 randomized to AC. There will be two CDST groups per cohort and two AC groups per cohort, and therefore, there will be six total groups of each type. Veterans will be randomly assigned to either the CDST condition or to AC after completing baseline assessments. To maximize power for between-group comparisons, randomization will be 1:1 resulting in 36 participants per group. Participants will be randomly assigned by cohort, so that for each cohort, there are 12 participants each in CDST and AC. Randomization procedure may be adjusted if necessary to fix uneven group sizes or demographics (e.g., due to attrition), for example by introducing a stratification variable. The research assistant who completes the assessments will be blind to condition, and will not attend any clinical team meetings or otherwise participate in any non-study related PRRC activities to maintain blindness. The final aim of this study is to 1) design a follow-up study based on the conclusions of the proposed study, and 2) develop a CDST Service Delivery Manual that will facilitate delivery of CDST in VA PRRCs. The specific study design and aims will be determined based on the outcomes of the proposed study, and will be responsive to possible changes in the VASDHS PRRC, PRRCs nationwide, and the broader VA context. There are three general possibilities for this follow-up study based on three potential outcomes. First, if this study supports CDST's effectiveness among Veterans with SMI and implementation viability for the VASDHS PRRC the follow-up study is likely to be a larger, multisite clinical trial that will allow for confirmation of CDST's effectiveness and work towards broad application of CDST in PRRCs throughout the VA system. In this case, the investigators plan to collaborate with other Desert Pacific PRRCs to conduct a larger trial. This trial may also include maintenance of effects after a 6- or 12- month follow-up period, and ability to personalize CDST service delivery for individual PRRCs. Second, if this study supports CDST's effectiveness among Veterans with SMI but not its implementation viability for VASDHS PRRCs or vice versa, the follow-up study will use data from this study to inform potential ways to resolve this issue so that CDST can become an effective and viable intervention. For example, if effectiveness was supported but viability was not, and results indicated that CDST was too many sessions, then a follow-up study could develop a shorter version of CDST and test it in the VASDHS PRRC. Third, if this study does not support CDST's effectiveness among Veterans with SMI nor its implementation viability in PRRCs, then the follow-up study will pursue other mechanisms to increase collaborative decision-making among Veterans with SMI. The Service Delivery Manual will be developed and refined throughout Aims 1 and 2. This manual is intended to be an accompaniment to the Clinician Manual to help clinicians deliver CDST in their own service contexts. The Service Delivery Manual will include discussion of approved adaptations to the manual that will not dilute or remove CDST's active ingredients. The manual will also include tips from CDST providers and participants that could assist future providers in effective delivery. These approved adaptations and tips could include recommendations for managing dynamics in the group, deciding upon role plays to suit Veteran priorities, and approved service approach variations that will not impact fidelity (e.g., number of meetings per week). In the event that data from this study refutes CDST's viability for Veterans with SMI, a CDST-specific service delivery manual is unlikely to be a useful product. In that case, a more general service delivery product will be developed; for example, a set of collaborative decision-making guidelines for PRRC administrators and providers based on data collected. Products from this study will be disseminated in the VA and outside the VA.

Schizophrenia, Schizoaffective Disorder, Delusional Disorder, Major Depressive Disorder With Psychotic Features

psychosis, serious mental illness, collaborative decision making, shared decision making, veterans, recovery, personal recovery

The experimental design for this study is a 2-group randomized effectiveness trial, with assessments at baseline (month 0), post-intervention (month 2), and 3-month post-intervention follow-up (month 5). Veterans will be randomly assigned to either the CDST condition or to AC after completing baseline assessments. To maximize power for between-group comparisons, randomization will be 1:1 resulting in 36 participants per group. Participants will be randomly assigned by cohort, so that for each cohort, there are 12 participants each in CDST and AC. Stratification may be introduced if necessary due to uneven or statistically different groups in key demographic domains (e.g., age, gender). Quantitative analyses will compare improvements in primary, secondary, and exploratory outcomes between the two groups, as described elsewhere. There are also qualitative outcomes associated with this study.

The research assistant who completes the assessments will be blind to condition, and will not attend any clinical team meetings or otherwise participate in any non-study related PRRC activities to maintain blindness. No study participants will be blinded. No other staff will be blinded.

An ten-session group-based skills training intervention that focuses on treatment-related decision-making in order to facilitate improved engagement in decision-making processes. CDST teaches assertiveness skills, problem solving, goal planning, and conflict negotiation, within the context of treatment planning and decision-making. Each session is 60 minutes long.

An ten-session group intervention that focuses on psychoeducation around befriending. Each session is 60 minutes long.

The Shared Decision-Making Coding System (SDM-CS) is a validated method of coding collaborative behaviors during treatment decision-making among patients with SMI and their providers.The SDM-CDS codes 9 elements of the decision process including goal setting, exploration of patient preference, and treatment alternatives. Coding for this measure is completed by recording frequency of specific interactions (e.g., the patient states a preference). Scores range from 0-18, where higher scores are better (indicating more collaborative behaviors).

Change in collaborative decision-making behaviors from baseline (month 0) to post-intervention (month 2)

The Shared Decision-Making Coding System (SDM-CS) is a validated method of coding collaborative behaviors during treatment decision-making among patients with SMI and their providers.The SDM-CDS codes 9 elements of the decision process including goal setting, exploration of patient preference, and treatment alternatives. Coding for this measure is completed by recording frequency of specific interactions (e.g., the patient states a preference). Scores range from 0-18, where higher scores are better (indicating more collaborative behaviors).

Change in collaborative decision-making behaviors from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

A 14-item clinician-rated measure using a 4-point Likert scale. Scores range from 0-42, where higher scores indicate more service engagement.

16-item self-report measure, with scores ranging from 16-112, with higher scores indicating higher levels of motivation.

Brief self-report survey assessing interest in rehabilitative approaches not previously engaged in. This scale will vary based on what rehabilitative approaches are currently being offered by the VASDHS PRRC and so the total score will also vary. Interest in each approach will be measured on a 0-4 Likert scale, and up to 10 approaches will be measured. Therefore the lowest possible score is 0 and highest possible score is 40. Higher scores will indicate more interest.

Change in interest in adjunctive rehabilitative approaches from baseline (month 0) to post-intervention (month 2)

An 8-item self-report measure of client satisfaction, with scores ranging from 8-32. Higher scores indicate higher levels of satisfaction.

A short interview-based measure that assesses goal attainment in a range of functional areas. In this study, the investigators will use the performance subscale. The performance subscale is rated between 1-10, where higher scores indicate better performance.

An algorithm that identifies goals and goal attainment. Scores range from -2 to +2, with 0 meaning that the expected outcome was achieved, while lower scores mean that less favorable outcomes occurred, and higher scores mean that better than expected outcomes occurred.

A 25-item self-report measure that assesses personal alignment with six primary domains of recovery: self-direction or empowerment, holistic, nonlinear, strengths based, responsibility, and hope. Scores range from 25-125, with higher scores indicating higher levels of personal recovery.

A semi-structured interview that assesses 26 domains of symptoms. Scores range from 26-182, with higher scores indicating more severe symptoms.

A measure of social functioning ability developed for adults with schizophrenia using role-plays of everyday social interactions. Scores range from 17 to 85, with higher scores indicating better social skill.

A clinician-reported scale of patient ability in four areas: socially useful activities, personal and social relationships, self-care, and disturbing and aggressive behavior. Scores range from 1-100, with higher scores indicating better social functioning.

A 10-item quiz of skills targeted by CDST. Scores range from 0-10, with higher scores indicating more correct answers.

3 item measure of involvement in treatment decisions. Scores range from 0-12, with higher scores indicating higher levels of involvement.

9 item measure of involvement in treatment decisions. Scores range from 0-45, with higher scores indicating high levels of involvement in treatment decisions.

Self-report about preferences for involvement in treatment decision-making based on three brief vignettes. There are 18 items and a total score range from 18-90, with lower scores indicating preference for paternalistic (provider-determined) decision-making, higher scores indicating preference for self-determination in treatment decisions, and middle scores indicating preference for provider-self collaboration.

Change in preferences about involvement in treatment decisions from baseline (month 0) to post-intervention (month 2)

Clinician-rated 14-item checklist of targeted skills. Scores range from 0-28, with higher scores indicating better performance on targeted skills.

Veteran semi-structured interviews focused on constructs aligned with quantitative outcomes including involvement in decision-making, treatment engagement, motivation, and satisfaction, and attainment of recovery goals.

A 14-item clinician-rated measure using a 4-point Likert scale. Scores range from 0-42, where higher scores indicate more service engagement.

Change in service engagement from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

Treatment attendance will be measured by attendance at all non-study-related PRRC appointments. These data will be collected from CPRS and confirmed with PRRC providers.

Treatment attendance will be measured by attendance at all non-study-related PRRC appointments. These data will be collected from CPRS and confirmed with PRRC providers.

Change in treatment attendance from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

16-item self-report measure, with scores ranging from 16-112, with higher scores indicating higher levels of motivation.

Change in treatment motivation from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

Brief self-report survey assessing interest in rehabilitative approaches not previously engaged in. This scale will vary based on what rehabilitative approaches are currently being offered by the VASDHS PRRC and so the total score will also vary. Interest in each approach will be measured on a 0-4 Likert scale, and up to 10 approaches will be measured. Therefore the lowest possible score is 0 and highest possible score is 40.

Change in interest in adjunctive rehabilitative approaches from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

An 8-item self-report measure of client satisfaction, with scores ranging from 8-32. Higher scores indicate higher levels of satisfaction.

Change in client satisfaction from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

A short interview-based measure that assesses goal attainment in a range of functional areas. In this study, the investigators will use the performance subscale. The performance subscale is rated between 1-10, where higher scores indicate better performance.

Change in goal attainment from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

An algorithm that identifies goals and goal attainment. Scores range from -2 to +2, with 0 meaning that the expected outcome was achieved, while lower scores mean that less favorable outcomes occurred, and higher scores mean that better than expected outcomes occurred.

Change in goal attainment from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

A 25-item self-report measure that assesses personal alignment with six primary domains of recovery: self-direction or empowerment, holistic, nonlinear, strengths based, responsibility, and hope. Scores range from 25-125, with higher scores indicating higher levels of personal recovery.

Change in personal recovery from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

A semi-structured interview that assesses 26 domains of symptoms. Scores range from 26-182, with higher scores indicating more severe symptoms.

Change in symptom severity from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

A measure of social functioning ability developed for adults with schizophrenia using role-plays of everyday social interactions. Scores range from 17 to 85, with higher scores indicating better social skill.

Change in symptom severity from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

A clinician-reported scale of patient ability in four areas: socially useful activities, personal and social relationships, self-care, and disturbing and aggressive behavior. Scores range from 1-100, with higher scores indicating better social functioning.

Change in social functioning from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

A 10-item quiz of skills targeted by CDST. Scores range from 0-10, with higher scores indicating more correct answers.

Change in targeted skills from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

3 item measure of involvement in treatment decisions. Scores range from 0-12, with higher scores indicating higher levels of involvement.

Change in involvement in treatment decisions from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

9 item measure of involvement in treatment decisions. Scores range from 0-45, with higher scores indicating high levels of involvement in treatment decisions.

Change in involvement in treatment decisions from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

Self-report about preferences for involvement in treatment decision-making based on three brief vignettes. There are 18 items and a total score range from 18-90, with lower scores indicating preference for paternalistic (provider-determined) decision-making, higher scores indicating preference for self-determination in treatment decisions, and middle scores indicating preference for provider-self collaboration.

Change in preferences about involvement in treatment decisions from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

Clinician-rated 14-item checklist of targeted skills. Scores range from 0-28, with higher scores indicating better performance on targeted skills.

Change in targeted skills from from baseline (month 0) to follow-up at 3-months post-intervention (month 5).

Veteran semi-structured interviews focused on constructs aligned with quantitative outcomes including involvement in decision-making, treatment engagement, motivation, and satisfaction, and attainment of recovery goals.

Veteran semi-structured interviews focused on constructs aligned with quantitative outcomes including involvement in decision-making, treatment engagement, motivation, and satisfaction, and attainment of recovery goals.

A validated measure for adults with SMI that codes presence of consumer driven collaborative behaviors, including sharing opinions about treatment effectiveness, making requests, and reflecting on the therapeutic relationship. This measure codes behaviors, so the minimum score is 0, and it has no maximum. Higher scores indicate more collaborative behaviors initiated by the consumer.

Change in consumer driven collaborative behaviors from baseline (month 0) to post-intervention (month 2)

Change in acute service use on a per month basis from 12 months prior to study start through 3 month post-intervention follow-up

Veterans and staff will complete semi-structured interviews focused on providing feedback on CDST to inform adaptations and implementation.

Self-report questionnaire for the intervention studied. Score is a percentage, 0-100%, where higher scores indicate higher levels of satisfaction with the studied intervention.

A validated measure for adults with SMI that codes presence of consumer driven collaborative behaviors, including sharing opinions about treatment effectiveness, making requests, and reflecting on the therapeutic relationship. This measure codes behaviors, so the minimum score is 0, and it has no maximum. Higher scores indicate more collaborative behaviors initiated by the consumer.

Change in consumer driven collaborative behaviors from baseline (month 0) to follow-up at 3-months post-intervention (month 5)

Inclusion Criteria: currently receive services in the VASDHS PRRC i.e., seen in the clinic in the past month and/or completed a PRRC group during the past trimester have an SMI diagnosis per the electronic medical record e.g., schizophrenia, schizoaffective disorder, delusional disorder, and major depressive disorder with psychotic features agree to have a subset of treatment appointments audiotaped Exclusion Criteria: having primary substance use or organic neurological disorder diagnosis are determined by PRRC and/or study staff to be at significant risk of exacerbation of symptoms, suicidal ideation, or other risk due to study participation have a history and/or current risk of violence that PRRC and/or study staff determine to be too high risk to manage effectively at the PRRC's outpatient clinic location (which has less police presence than the primary VASDHS hospital location)

IPD underlying results in a publication will be shared. Deidentified data sets will be used for analysis and only these de-identified sets will be shared. All HIPAA identifiers, combinations of variables that might be used for re-identification, and proprietary information will be excluded. Audio data will not be shared; de-identified transcriptions and coding based on audio data will be. Only data needed to fulfill a requester's stated purpose will be shared. Therefore this plan will protect personal privacy of research subjects, prevent loss of confidentiality, and ensure the secure maintenance of propriety data and information.

Data will be shared with researchers and consumer advocates who provide a rationale and methodological sound plan for their use of the data. Acceptable uses include meta-analysis and mega-analysis, among other uses. Data sets provided to requesters will include data necessary for the requester's stated purpose, as long as the requester's purpose does not violate participant privacy, research ethics, or similar procedures. Requests for access must be made in writing including an explicit assurance that the recipient will not attempt to identify or re-identify any individual. Requests must be signed by a requester currently residing in the United States. The request must reference the publication or other source motivating the request along with the detailed purpose for the request. Requests should be made to the corresponding author or the principal investigator, but can be made to the Associate Chief of Staff for Research.

Treichler EBH, Rabin BA, Spaulding WD, Thomas ML, Salyers MP, Granholm EL, Cohen AN, Light GA. Skills-based intervention to enhance collaborative decision-making: systematic adaptation and open trial protocol for veterans with psychosis. Pilot Feasibility Stud. 2021 Mar 29;7(1):89. doi: 10.1186/s40814-021-00820-4.