Pre-Implementation Strategies to Adapt and ... - Wiley Online Library

4 downloads 8826 Views 432KB Size Report
2 Department of Communication Studies, San Francisco State University, San ... 10 Department of Psychiatry, University of California-San Francisco, San ...
ORIGINAL ARTICLE

Pre-Implementation Strategies to Adapt and Implement a Veteran Peer Coaching Intervention to Improve Mental Health Treatment Engagement Among Rural Veterans Christopher J. Koenig, PhD;1,2 Traci Abraham, PhD;3,4 Kara A. Zamora, MA;1 Coleen Hill, BA;1 P. Adam Kelly, PhD, MBA;5,6,7 Madeline Uddo, PhD;5,6,8 Michelle Hamilton, PhD;5,8 Jeffrey M. Pyne, MD;3,4 & Karen H. Seal, MD, MPH1,9,10 1 San Francisco Veterans Affairs Health Care System, San Francisco, California 2 Department of Communication Studies, San Francisco State University, San Francisco, California 3 Center for Mental Healthcare & Outcomes Research, Health Services Research and Development, Little Rock, Arkansas 4 Department of Psychiatry, University of Arkansas for Medical Sciences, Little Rock, Arkansas 5 Southeast Louisiana Veterans Health Care System, New Orleans, Louisiana 6 South Central Mental Illness Research Education Clinical Center (MIRECC), Little Rock, Arkansas 7 General Internal Medicine & Geriatrics, Tulane University School of Medicine, New Orleans, Louisiana 8 Department of Psychiatry and Behavioral Sciences, Tulane University School of Medicine, New Orleans, Louisiana 9 Department of General Internal Medicine, University of California-San Francisco, San Francisco, California 10 Department of Psychiatry, University of California-San Francisco, San Francisco, California

Abstract Funding: This research was supported by CREATE award #12-083 from the United States Department of Veterans Affairs Health Services Research and Development. Acknowledgments: We would like to thank the veterans and VA clinic staff who took time out of their busy schedules to meet with us in order to discuss the study and how it might fit into their busy daily lives. This article does not reflect the official opinion of the Department of Veterans Affairs, and responsibility for the views expressed lies entirely with the listed authors. For further information, contact: Christopher J. Koenig, PhD, Department of Communication Studies, 1600 Holloway Avenue, San Francisco State University, San Francisco, CA 94132; e-mail: [email protected]. doi: 10.1111/jrh.12201

Purpose: Telephone motivational coaching has been shown to increase urban veteran mental health treatment initiation. However, no studies have tested telephone motivational coaching delivered by veteran peers to facilitate mental health treatment initiation and engagement. This study describes pre-implementation strategies with 8 Veterans Affairs (VA) community-based outpatient clinics in the West and Mid-South United States to adapt and implement a multisite pragmatic randomized controlled trial of telephone peer motivational coaching for rural veterans. Methods: We used 2 pre-implementation strategies, Formative Evaluation (FE) research and Evidence-Based Quality Improvement (EBQI) meetings to adapt the intervention to stakeholders’ needs and cultural contexts. FE data were qualitative, semi-structured interviews with rural veterans and VA clinic staff. Results were rapidly analyzed and presented to stakeholders during EBQI meetings to optimize the intervention implementation. Findings: FE research results showed that VA clinic providers felt overwhelmed by veterans’ mental health needs and acknowledged limited mental health services at VA clinics. Rural veteran interviews indicated geographical, logistical, and cultural barriers to VA mental health treatment initiation and a preference for self-care to cope with mental health symptoms. EBQI meetings resulted in several intervention adaptations, including veteran study recruitment, peer veteran coach training, and an expanded definition of mental health care outcomes. Conclusions: As the VA moves to cultivate community partnerships in order to personalize and expand access to care for rural veterans, preimplementation processes with engaged stakeholders, such as those described here, can help guide other researchers and clinicians to achieve proactive and veteran-centered health care services. Key words health services research, mental health initiation and engagement, pre-implementation research, qualitative research methods, rural veterans.

c 2016 National Rural Health Association The Journal of Rural Health 00 (2016) 1–11 

1

Pre-implementation of a Mental Health Intervention

Koenig et al.

Military veterans who live in rural areas bear a substantial and disproportionate mental health burden relative to veterans who live in urban areas. Studies show that veterans residing in rural areas experience significantly greater mental health severity and poorer outcomes compared to their urban counterparts.1-3 Moreover, veterans living in rural areas have difficulty remaining engaged in mental health care. Prior research has shown that veterans with mental health problems do not receive an adequate course of mental health treatment,4-6 which the VA Mental Health Handbook7 defines as receiving 8 or more mental health treatment sessions or receiving at least 2 months of psychiatric medication and more than 4 visits within 1 year.8 Failing to remain engaged in mental health treatment may result in chronic mental illness and associated social and occupational dysfunction with high costs to individuals, families, and society.9,10 Further, research has demonstrated strong links between mental illness, negative health behaviors and physical health problems, such as smoking, hypertension, and drug and alcohol abuse, resulting in disproportionately high medical services utilization.11-14 Rural veterans have also been shown to have poor retention in mental health services.4,5 One study reported that living more than 25 miles from a VA facility was a strong predictor of failing to receive adequate mental health treatment.4 While geographical distance is a significant logistical barrier, rurality may also be associated with other access and engagement barriers including financial, temporal, digital, and cultural barriers, including heightened levels of stigma, stoicism, self-reliance, lack of perceived need, and negative beliefs about mental health treatment.15-17 Motivational Interviewing (MI) is an evidence-based therapeutic technique that can promote mental health treatment initiation and engagement in veterans. A previous randomized controlled trial with urban Iraq and Afghanistan veterans demonstrated that 3 telephone MI sessions delivered by clinically trained research staff significantly improved mental health treatment initiation and decreased mental health-related stigma, but the sessions did not significantly improve engagement in mental health care.18 Similarly, in an older population of 113 depressed veterans, telephone-based referral care management consisting of 1 to 2 telephone MI sessions plus pre-scheduled appointments significantly improved mental health treatment initiation, but it did not translate into improved clinical outcomes.19,20 While these results show MI improved initiation of mental health treatment, none of these studies were conducted with rural veterans. MI has been effectively used in culturally diverse populations, including rural populations. MI techniques enable patients to explore ambivalence, negative beliefs,

and stigma regarding mental health treatment. In addition, MI can draw on culturally appropriate strengths and coping mechanisms, such as religion and peer social support that are especially important for rural veterans.21 MI has been successfully used over the telephone, and peer counselors have been successfully trained to conduct MI.22,23 In fact, research has shown that veteran peer counselors who shared cultural values and experiences may be highly effective in engaging rural veterans in care.24-27 Using evidence from these previous studies, we designed a study entitled “Motivational Coaching to Enhance Mental Health Engagement in Rural Veterans” (hereafter: COACH). The COACH study extends and adapts prior MI interventions to address the mental health needs of rural veterans in the following ways.18 First, the COACH study uses veteran peers to conduct the telephone motivational coaching intervention because rural veterans may be more receptive to peers when discussing mental health concerns.26 Second, this study adapts the telephone MI intervention to a telephone motivational coaching intervention that uses MI elements in a manner appropriate for veteran peer coaches with limited clinical training. Third, the COACH study targets rural veterans who may particularly benefit from telephone-based peer motivational coaching to facilitate mental health initiation. Finally, this study explores the potential effect of peer motivational coaching on mental health engagement to extend prior research focus on treatment initiation.18 In sum, the current study implements a pragmatic randomized controlled trial of an MI-based telephone motivational coaching intervention to partner rural veterans who use VA community-based outpatient clinics (CBOCs) with veteran peer coaches to facilitate initiation in mental health care. This article has 2 aims: first, to describe the pre-implementation research we conducted with rural veterans and VA clinic staff, and second, to describe how we adapted the overall study design and its implementation in response to the input we received from each of our study sites.

2

c 2016 National Rural Health Association The Journal of Rural Health 00 (2016) 1–11 

Methods The COACH study is designed as a Hybrid Type II implementation-pragmatic effectiveness trial that compares telephone-based veteran peer motivational coaching versus enhanced usual care to encourage rural veteran mental health treatment initiation and explore engagement in mental health treatment.28,29 We defined mental health treatment engagement according to the VA Mental Health Handbook standard of adequate treatment.8 Table 1 describes key components for

Koenig et al.

Pre-implementation of a Mental Health Intervention

Table 1 COACH Study Intervention Components Study Arm

Intervention Component

Intervention

Baseline assessment Tailored VA and community mental health resources Peer veteran motivational coaching calls Ongoing tailored VA and community mental health resources Follow-up assessments a

√ √ √ √ √

Enhanced Control √ √



Time Frame Week 0 Week 0 Week 0-22: Up to 6 motivational coaching sessionsa Occurs during motivational coaching sessions Week 8, 16, 32

Coaching session timing is personalized according to if/when participants initiate mental health treatment.

intervention and enhanced usual care participants. All participants receive a baseline assessment and tailored mental health referrals, but only the intervention group received veteran peer motivational coaching to encourage initiation and to explore engagement in mental health services. The VA Central Institutional Review Board approved this study. From March 2014 to December 2015, we conducted pre-implementation research using Formative Evaluation (FE) and Evidence-Based Quality Improvement (EBQI) methods to prepare for implementing the veteran peer coaching mental health initiation and engagement intervention at each of the 8 sites. The goal of the preimplementation research was to adapt the intervention such that it was culturally appropriate and responsive to the needs of both rural dwelling veterans and the VA clinic staff. Figure 1 illustrates the process we used to elicit stakeholder input to adapt the intervention implementation to the needs of each site. The Promoting Action on Research Implementation in Health Systems (PARiHS) framework guided the implementation strategy.30 According to PARiHS, successful implementation is most likely to occur when scientific evidence supporting an intervention fits with providers’ experience and local cultural norms, the health care organizational context is supportive of implementation, and there are culturally appropriate mechanisms to facilitate implementation.31 PARiHS fits well with the theoretical underpinnings of MI because both require active collaboration among the research team, VA clinic staff, veterans, and local community to promote rural veterans’ initiation and engagement with acceptable forms of mental health care. The following describes the preimplementation steps taken in this study.

visits were 60-90 minutes in length and were typically held over the lunch period to maximize the potential that clinic leadership, primary care and mental health providers and other interested staff might attend. The main goals of the site visits were to introduce the study and study team, meet VA clinic staff, present the evidence for the intervention, and to gauge staff organizational climate, capacity, and interest in study participation.

Formative Evaluation Research Formative Evaluation (FE) is defined as a “rigorous assessment process designed to identify potential and actual influences on the progress and effectiveness of implementation efforts.”32,33 We used FE research to optimize implementation of the study at each study site. The main FE data were qualitative semi-structured interviews with veterans who use community-based outpatient clinics as well as VA clinic leadership and staff at each site. Interviews were conducted with stakeholders in the VA clinic communities with the goal of adapting the intervention implementation to each site and region. FE data were rapidly analyzed, summarized, and presented to VA stakeholders during EBQI meetings described below.28,29,34-36 Semi-Structured Interview Recruitment

First, the extended study team visited each prospective VA clinic site for an in-person “Meet and Greet.” Site

VA staff participants. We used criterion sampling, a common purposeful sampling technique that involves selecting participants meeting pre-determined criteria.37,38 We sampled VA clinic primary care providers who refer veterans to mental health services and VA clinic mental health providers and ancillary VA staff such as front-line staff, social workers, nurse managers, or administrative clinic leadership. Eligibility criteria for VA staff were minimal to reflect the diversity of staff involved in rural veteran health care. The senior author (KHS) asked permission from VA clinic directors to contact each clinic’s primary care, mental health, and other staff involved in

c 2016 National Rural Health Association The Journal of Rural Health 00 (2016) 1–11 

3

VA Community-Based Outpatient Clinic “Meet and Greet” Visits

Pre-implementation of a Mental Health Intervention

Koenig et al.

Figure 1 Pre-Implementation Resesarch Process to Adapt the COACH Study Intervention to Each VA Community-Based Outpatient Clinic.

referring, supporting, or directly providing mental health care to rural veterans. COACH study staff then contacted prospective clinic staff by e-mail to explain the study and request participation in a 1-time interview and several EBQI meetings throughout the study. All participants received a study information sheet and a copy of the informed consent form. Interested participants were scheduled for either an in-person or telephone interview.

Veteran participants. We used national VA administrative data to identify rural veterans who were at least 18 years old and received care at one of the VA clinic study sites. Exclusion criteria included having a hearing impairment and no access to a working telephone. To ensure that a range of veterans were included, we constructed a stratified purposeful sample37,38 of mental health services utilization following a positive mental health screen or diagnosis over a 1-year period that included veterans who: (1) failed to attend any mental health visits; (2) attended 1-2 mental health visits; or (3) attended 6 or more mental health visits. Once identified, veterans were mailed a recruitment letter, an information sheet, and an opt-out postcard. Veterans who did not opt out 14 days after the mailing were called by phone to inquire about participation in a 1-time interview. Interested veterans completed a brief survey to confirm eligibility and, if eligible, were scheduled for an interview. All participants provided informed consent prior to interviews.

piloted the interview guides with 4 VA providers and 5 veterans to ensure that the interview guides captured rich data with minimal participant burden. We used the interview as conversation model to maximize the potential of encountering unexpected data.39 All interviews were digitally recorded for subsequent analysis. The majority of provider interviews were conducted in person, but some were conducted by phone at providers’ request. All veteran interviews were conducted by phone.

Semi-Structured Interview Data Analysis We used rapid qualitative analysis techniques,40-43 including structured templates and matrix displays for VA staff interviews and narrative case summaries for veteran interviews. CBOC provider and staff interviews were rapidly analyzed 1 site at a time. Average time for rapid analysis was 3 weeks. Veteran interviews were analyzed progressively. Qualitative rapid analyses were conducted with the purpose of generating research findings for use in EBQI meetings, which were held within 2-3 months of the CBOC provider and staff interviews.

Using prior literature and the research team’s combined clinical and research experience, we developed separate qualitative interview guides for provider and veteran participants (see Appendix S1). Experienced interviewers

Structured template and matrix display. Provider interview recordings were summarized using a multistep procedure. First, qualitative team members used the interview guide while listening to each recording to synthesize each interview’s content, noting particularly rich responses, defined as responses that were detailed and contextually meaningful.44 Second, we developed a structured template to standardize how interview content was captured. Particularly rich responses were summarized with recording timestamps for future analysis. Third, after all interviews were transferred onto structured templates, the data were aggregated to create site-specific matrices

4

c 2016 National Rural Health Association The Journal of Rural Health 00 (2016) 1–11 

Semi-Structured Interview Data Collection

Koenig et al.

that enabled systematic comparison between sites and regions.45 Finally, the first author audited all templates and matrices to standardize categories and vocabulary across sites, constructing a final cross-site matrix that highlighted regional similarities and differences.

Narrative case summaries. Narrative case summaries are a rapid analytic method for transforming audio data to text by capturing interviewers’ interpretation of participants’ responses to main domains of interest.45 Each interviewer composed a narrative for each veteran participant, including his or her mental health initiation and engagement, sources of social support, experiences of health care, and self-care or other health work practices, defined as physical, emotional, mental, or social actions individuals take outside of clinical settings with the goal of maintaining health and promoting well-being.46,47 Average summary length was 2-3 pages. Analysts ensured narrative summary accuracy by verifying details from the recordings and identifying rich segments across interviews for future in-depth analysis. The qualitative team met weekly to discuss the interviewing process and resulting patterns across sites and regions. Summaries were uploaded into Atlas.TI (v7.x)48 to discern broad patterns for preliminary results used during EBQI meetings and veteran peer coach training described below.

Evidence-Based Quality Improvement Once FE research data were collected and rapidly analyzed, the research team engaged VA clinic staff in 1hour EBQI meetings to tailor the COACH study intervention implementation to each site. EBQI methods use techniques derived from quality improvement to ensure that study implementation is consistent with the intervention evidence base and sufficiently acceptable to stakeholders.43,49,50 EBQI also helps foster partnerships among researchers, clinicians, staff, and administration to help align goals and expectations over the course of the study period.

EBQI Recruitment After endorsement by local clinic leadership, the senior author reached out to primary care and mental health clinic staff via e-mail 1 month before the initial EBQI meeting to request their attendance. We included staff members who were and were not interviewed during the FE, as well as mental health and primary care leadership. We also asked clinic staff for names of influential veterans in the local community to serve as peer opinion leaders. We subsequently mailed these veterans a letter of

c 2016 National Rural Health Association The Journal of Rural Health 00 (2016) 1–11 

Pre-implementation of a Mental Health Intervention

introduction, an information sheet, and an opt-out post card. If after 2 weeks veteran leaders did not opt-out, study staff contacted them by phone to explain the study and invite their participation in the upcoming EBQI meeting.

EBQI Procedures Each 1-hour EBQI meeting began with a team introduction, an outline of the current evidence for MI and peer coaching interventions, a brief outline of the COACH study design, and presentation of the FE results for member checking,51 a form of stakeholder feedback that establishes the observational data credibility. EBQI meeting structure drew on the PARiHS framework30 by presenting and discussing evidence to foster facilitation between our research team and the VA clinic staff in order to adapt the implementation of the COACH study intervention to each site. During meetings, the research team emphasized flexible aspects of the implementation strategy, including rural veteran participant recruitment, communication between the COACH research team and VA clinic staff, and feasible mental health care outcomes for rural veterans. We also elicited VA clinic staff and veteran opinion leaders to discuss perceived barriers and facilitators for study implementation as well as preferences for implementation of the intervention at each site.

Results During the FE research, we partnered with 8 CBOCs in the West and Mid-South US regions shown in Table 2. Seven clinics were mid-size, meaning they served between 1,500 and 5,000 unique veterans per year, and 2 were large, meaning they served between 5,000 and 10,000 unique veterans per year. The location of each community also varied in terms of distance to the nearest urban center. While 2 CBOCs in the Mid-South were located in an urban center, these clinics serve many veterans who live in outlying rural areas. From these 8 CBOCs, we conducted individual interviews with 52 providers and 37 veterans (Table 3). In this section we present the rapid analysis of these interviews, which were summarized and presented to VA clinic leadership and staff during the first EBQI meeting.

Formative Evaluation Results VA Clinic Interviews In their interviews, community-based outpatient clinic leadership and staff described various barriers and facilitators for veteran initiation and engagement in

5

Pre-implementation of a Mental Health Intervention

Koenig et al.

Table 2 Regional Community-Based Outpatient Clinic (CBOC) Characteristics

Region West

Mid-South

CBOC

CBOC Sizea

Unique Patients/Year

Northern Clinic Western Clinic Eastern Clinic Southern Clinic Eastside Clinic Westside Clinic Metro Clinic City Clinic

Mid-size Mid-size Mid-size Mid-size Mid-size Mid-size Large Very Large

4,855 2,841 2,080 2,189 3,633 2,252 5,630 10,722

Local Community Population

Nearest Urban Center

26,913 15,871 15,250 41,114 90,000 43,761 378,715 378,715

270 miles 114 miles 108 miles 15 miles 34 miles 30 miles 0 miles 0 miles

a

VA calculates CBOC sizes based upon total unique patients per year served as follows: Very Large (10,000+); Large (5,000-10,000); Mid-size (1,500-5,000); Small (