Capacity Coaching: A Focused Ethnographic Evaluation in Clinical Practice

Objective To qualitatively evaluate the implementation of Capacity Coaching, an intervention to address the work patients must undertake to manage their conditions, implemented as a quality improvement pilot in 1 of 2 implementing US Department of Veterans Affairs medical centers. Participants and Methods Two Veterans Affairs medical centers in the Midwest sought to implement Capacity Coaching as a quality improvement pilot in their Patient-Aligned Care Teams for 6 months (April 1, 2017, through October 31, 2017). Following the pilot, we conducted a focused ethnographic evaluation (on-site data collection, January 2-4, 2018), including interviews, a focus group, and observations with staff at one site to assess the implementation of capacity coaching. Data were analyzed inductively and findings were cross-referenced with implementation theory. Results We found that implementation was feasible and achieved changes that were aligned with reducing patient work and increasing capacity. We found that the key facilitators for the implementation of this program were in participants making sense of the intervention (coherence) and working collectively to enact the program (collective action). The main challenges for the program were in planning the work of implementation and enrolling a diverse coalition of staff to expand referrals to the program (cognitive participation) and in evaluating the impact of the program on outcomes that upper leadership was interested in (reflexive monitoring). Conclusion Implementation of Capacity Coaching is feasible in clinical practice and may be a promising intervention for the care of chronic conditions. Further research should focus on testing capacity coaching using these lessons learned.

pidemiological studies reveal that by middle age, one-half of the population already lives with 1 chronic condition and one-third already have two or more chronic conditions. 1 Once a problem of the elderly, multimorbidity is now a reality for many working-age adults. As more treatments and tests are added to the regimen for these patients to manage their conditions, an important problem surfacesdthe competing priorities of life. Patients with multimorbidity are at increased risk for "structural noncompliance," which is the inability to adhere to care plans because the number and complexity of the health care tasks are simply overwhelming. 2 May et al 2 proposed a framework to address this challenge and called it Minimally Disruptive Medicine (MDM). 2 MDM is a model of care that is oriented by an overarching conceptual model: the cumulative complexity model (CuCoM). 3 The CuCoM illustrates that there is patient work necessary to implement health interventions (eg, attending appointments, taking medications, self-monitoring). [3][4][5] Second, patients invest capacity to implement patient work, which is the same capacity that they must use for life activities (eg, care for families, employment). 3,6 When patient work exceeds capacity, it impairs patients' abilities to access and use health care and enact self-care and may result in worsening health outcomes. 3 Two middle-range theories describe the concepts of workload and capacity used in the CuCoM in greater detail: the Normalization Process Theory (NPT) and the Theory of Patient Capacity (TPC), respectfully. 4,5,7,8 Briefly, NPT focuses on 4 key domains: coherencedhow participants make sense of the work required to take up the intervention; cognitive participation to enroll others and plan the work; collective action to enact the work of implementing the intervention; and reflexive monitoring, to appraise if the intervention is worth the effort. 4,5,8 Normalization process theory can be used at the individual level to understand patient behavior and at the team level to understand health care teams' implementation of interventions. The TPC describes patient capacity as a dynamic interaction between patients and their Biographies, Resources, Environment, patient Work, and Social network (BREWS). 7 The Figure demonstrates the relationships between MDM's conceptual models and theories.
Capacity Coaching is a patient-centered intervention that has been designed in alignment with MDM. It is novel in its application of MDM but also in its combination of 2 types of interventions that have been used separately for populations with multimorbiditydintensive team-based management and coaching. 9 Capacity Coaching begins the first coach-patient conversation with the ICAN (Instrument for Patient Capacity Assessment) Discussion Aid (Supplemental Appendix 1, available online at http://www.mcpiqojournal.org), which was designed to support MDM in practice to understand patients' health care workload and capacity. 10 Capacity Coaching builds on traditional Health and Wellness Coaching (HWC), but is specifically designed to work within the context of the health care team to reduce treatment burden and increase capacity for overwhelmed patients living with chronic illness. 9 To date, capacity coaching remains untested. It is unknown whether it can be implemented into primary care or if implementation is successful, how it would impact patient outcomes. The purpose of this study was to understand its feasibility for implementation in practice. Understanding if this intervention is feasible to implement clinically, as well as key barriers and facilitators to its implementation, can enable the testing of capacity coaching in a larger clinical trial.

PARTICIPANTS AND METHODS
Two US Department of Veterans Affairs (VA) medical centers in the Midwest sought to implement capacity coaching as a quality improvement pilot initiative in their Patient-Aligned Care Teams (PACTs). In one location, 4 PACT teams in the Women's Health Clinic implemented the program (site 1); at the other location, the mental health care PACT in the Women's Health Clinic implemented the program (site 2). PACT teams include physicians, nurses, social workers, pharmacists, nutritionists, and peer support specialists (PSSs). The teams were trained during a 1-day workshop delivered by one of the authors (K.R.B.); approximately 40 team members across the 5 PACTs were trained. Those trained included all members from site 1 and 2 team members from site 2. The workshop content was delivered over a single-day 8-hour session. It included training on MDM and its conceptual and theoretical models, the ICAN Discussion Aid, Capacity Coaching principles, and implementation in complex systems. It also included facilitated planning sessions for conceptualizing Capacity Coaching implementation in practice.
Following the workshop, sites independently planned and implemented the program in their settings; one of the authors (K.R.B.) provided limited remote support. Following the closure of the implementation period, we returned to site 1 to understand the program's implementation successes and challenges. While returning to both clinics would have been ideal, due to pragmatic time and funding constraints, we sought to collect data from site 1, as it was where the workshop occurred and housed 4 of the 5 PACTs trained. We used focused ethnographic methods to accomplish our aims. Focused ethnography is similar in its methods to traditional ethnography but instead focuses on answering specific questions in microcultures that exist within larger cultures. 11 Given our study's primary aim to understand the feasibility of implementation, we employed our qualitative methods using an implementation science perspective. 12 Sample and Data Sources Participants were employees of site 1 and were notified 1 month before the site visit via email from clinic leadership that a researcher would be visiting. Four sources of data were collected: observations, interviews, a focus group, and artifacts. Participants were allowed to partake in as many or few data collection activities as schedules allowed. Some participants took part in all activities, whereas others only took part in one activity. All participants provided oral consent for participation in the study. Ethics approval and oversight were provided by the Mayo Clinic Institutional Review Board.

Data Collection
The data collection period took place over 3 days on site (January 2-4, 2018), as well as a number of preparatory phone meetings and email exchanges. Fourteen people were invited to participate in interviews by email based on their high level of involvement in the Capacity Coaching program and role diversity; ultimately, 10 chose to participate including current and past directors of the Women's Health Clinic, a social worker, a program manager, a nurse, a pharmacist, and a women's health fellow. We also conducted a focus group with the current Women's Health Clinic team and 9 hours of observation. Some interviewees also participated in the focus group; the focus group was also enriched by team members who had more minor roles in the implementation of the program. All data were collected by the lead author (K.R.B.).
Observations occurred during clinic on and off hours; observations during clinic hours included patient interactions with health professionals and health professionals with each other. No patient protected health information was collected. Those observed were members of the PACTs that participated in the pilot. Observations were conducted in the Women's Health Clinic, in the mental health clinic, and in another primary care clinic where pharmacy consultation services were provided. Observations were conducted in areas beyond the Women's Health Clinic, the primary pilot location, because all Women's Health Clinic staff members are not located in the same area of the center. Detailed observation notes were collected in the field, and at the end of each day, further written notes summarizing data collection activities were completed.
Focus group data were collected using a semistructured interview guide informed by the NPT from the team perspective. 8 Individual interview data were collected using an unstructured interview format, beginning with a "grand tour" of the reason for the interview: to fully understand the experience of implementing Capacity Coaching in their practice. 13 After letting the participants talk as long as they liked, follow-up questions probed at experience, such as asking about the influence of the program on their practice following the pilot period, detailed information about changes in day-to-day work, and sensitive group dynamic information, which was unlikely to emerge in the focus group discussion. Guides are included in Supplemental Appendix 2 (available online at http://www. mcpiqojournal.org). All focus groups and interviews were audio recorded and were transcribed verbatim. Artifacts provided by the team to the lead author included the final version of the implementation toolkit, workbook materials for coach-patient interactions, and template tools for documenting capacity coaching in the medical record. The implementation toolkit contained patient stories told in the first person, as recreated by staff who delivered the intervention. We did not have institutional review board approval to interview patients for this study, and therefore, patient quotes were derived exclusively from the toolkit artifact.

Reflexivity
The practice of reflexivity "involves being deliberately aware of oneself, one's responses, and one's internal state in relation to a specific situation." 11 The lead author kept a reflexivity journal during time on site and during the analysis period, tracking initial impressions, state of mind, and thoughts about ideas that need to be subjected to follow-up interview questions, observations, and discernment.

Qualitative Analysis
All data sources including transcripts, artifacts, and observation notes, hereafter referred to as "source documents," were imported into Nvivo 11 Pro for Windows (QRS International Pty) for analysis. The analysis process used procedures guided by Roper and Shapira's process for ethnographic analysis. 11 First, the lead author and 2 trained research assistants (P.O., A.T.) listened to the audio-recorded data and read all source documents, then coded source documents inductively using line-by-line coding to develop a code book. We used this process on 3 source documents, meeting 2 to 3 times weekly to discuss newly emerging codes and codes that should be combined or reconciled. We deemed the codebook complete after coding the third source document, as no new codes were emerging. We continued to meet twice weekly to address discrepancies or potentially new codes as we coded remaining source documents using the codebook; 2 new codes emerged from the later data and were added to the codebook. In addition to our inductive codes, we included a priori deductive codes related to NPT: coherence, cognitive participation, collective action, and reflexive monitoring.
After coding, the lead author began to aggregating codes into themes using grouping and matrix functions in Nvivo 11. Coders reviewed results from the analysis to ensure that interpretations of the data remained close to the data. The lead author summarized all data into key themes found in the culture of the clinic, using CuCoM, NPT, and TPC. This step represents the use of well-fitting existing conceptual models and theories to organize the data after initial analyses in a manner that is useful for the translation of interventions into practice and is respected as particularly useful for meeting challenges in implementation of complex interventions. 12,14

Summary of Capacity Coaching Intervention Delivered
A core group of individuals championed the capacity coaching intervention, including a primary care physician, social worker, and PACT nurse. The social worker served as the primary project coordinator, and a research coordinator kept up with data elements used to generate the implementation toolkit. These individuals met regularly, and other clinicians and health professionals met with them periodically, particularly in the planning phases. The planning phase lasted approximately a year, and the pilot implementation lasted 6 months (April 1, 2017 through October 31, 2017).
At the time of the workshop, site 1 had planned to have the social worker serve as the capacity coach, and site 2 had a PSS as the capacity coach. During planning and the early pilot, it became apparent that site 1 could also be best served by a PSS as the capacity coach because patients placed value on the shared life experience of a female veteran living with chronic illness. Therefore, the PSS was brought on board by the team for site 1. The social worker then trained the PSS and supervised her throughout the intervention period.
During the 6-month implementation at site 1, 21 patients went through the program. Patients interacted with the coach between 3 and 6 times, depending on their needs. This process comprised a total of 21 in-person encounters and 42 telephone encounters. At the conclusion of the pilot, the team developed a freely available Capacity Coaching toolkit to help other VAs implement the program (available through VA Pulse or minimallydisruptivemedicine.org) Core components of the intervention delivered appear in Table 1.

Alignment With CuCoM
There was evidence that the pilot capacity coaching program was aligned with the CuCoM. Specifically, there was considerable evidence that those most involved in the pilot focused on reducing patient work. For patients, the primary work added by the program was engaging with the coach. However, there was evidence that even this work was carefully engineered to be minimized, using strategies such as video cameras and iPads, patient-coach visits off site, and warm handoffs to the coach so that the first visit could occur immediately.
Although formal outcomes were not collected from the program's implementation, interviews and artifacts reflected success stories in patients' ability to access and use health care and enact self-care, as well as their outcomes and quality of life. Despite staff members' description of how implementing capacity coaching impacted their practice (eg, how they consider MDM in their everyday interactions with patients), there was no clear evidence from ethnographic observations that the program had impact elsewhere in the center.

Alignment With TPC
Participants' stories and artifacts illustrated that the implementation of the Capacity Coaching program positively acted upon each construct of the TPC (Table 2).

Implementation Successes and Challenges
Beyond the impact of the program and its alignment with the principles of MDM, the remainder of the data were primarily focused on stories that were illustrative of the success and the challenges of implementation. These factors can be broken down into the 4 constructs of NPT: coherence, cognitive participation, collective action, and reflective monitoring ( Table 3). The primary success of the implementation was seen in the domains of coherence and collective action, and challenges occurred in cognitive participation and reflexive monitoring.

Transferability of Capacity Coaching Curriculum
A surprising finding that emerged outside of the conceptual models and theories was the applicability of the capacity coaching curriculum beyond patients living with chronic conditions. One participant highlighted that she took the skills from the workshop and applied it to working with nursing students.

DISCUSSION
Key findings are summarized in Table 4.

Limitations and Strengths
This evaluation was conducted after the pilot concluded, and no continued Capacity Coaching activities occurred during the time the data were collected. This means that the robustness of the evaluation relies on participants' memories of the previous 2 years and may only capture key highlights of what occurred. Second, the evaluation was entirely qualitative, which means the impact of the program on patients' health outcomes and quality of life relies entirely on anecdotal cases documented by staff. This limitation exists because of the narrow scope of the grant funding for the program, as well as the heterogeneous nature of the small patient population that participated in the pilot.

Resources
The Capacity Coach and social worker worked collaboratively, with the social worker supervising the coach and the two meeting weekly to discuss cases. This pairing worked well, and they were able to connect patients to resources in the VA or their community "There was one person who she broke her leg and she needed a wheelchair and I was like 'Okay. Hmm. I can come and help a little bit'. And they were able to get the things that they needed and they were like 'oh, that was so helpful.' I recommended, you know, going to a senior center to that same person, and her partner really was appreciative of the things that I was coming to them with." e Staff member 2, capacity coach

Social
One of the toughest challenges patients encountered in caring for their health that the staff highlighted was balancing self-care with caregiving for others. The Capacity Coach was often able to support them in working through this balance as well as working productively with their social network.
"My sister, God bless her, is always needin' my car for this, that, and everything. I don't mind helping her out, but now I can't get to the grocery store when I need to, so I just pop in at the convenience store on my block for stuff, and all they've got is junk food. My capacity coach is helping me work out some ways to talk nicely to my sister about getting my car back." e Toolkit, Patient Success Story "So just it seems like there's always like some other outside influence. There was another lady who she had lots of medical issues and a spouse who was not doin' what he needed to be doin'. And she was so worried about tryin' to keep things intact that her care fell off and right now. and she's in like a rehabilitation center." eStaff member 2, capacity coach discussing the biggest barriers she had to help patients overcome  "Cause the infrastructure wasn't there. If we did six months now, it would look completely different because you would be comin' out of the gate running. Because things needed to-we used to go back and we'll say 'Okay, let's uh, strengthen this piece here and do somethin' different here." Staff member 1, clinician d Balancing planning logistics and focusing on the big picture of the program's intended impact. "On the one hand, it's nice to just sort of be in the presence of people that are sort of big picture thinkers, but on the other hand, it's like okay, at some point, we have to just, you know, decide and do something." e Staff member 5, project coordinator d Co-location, visibility, and marketing of the capacity coaching program. "I think [peer] needed to be in the clinic or like have space so that-for me, a lot of things are out of sight out of mind." eStaff member, social worker. "Yeah 'cause it woulda been nice if you woulda e you know, if we would've e if you woulda walked in the nurses and say 'hey, have you been doing this?" eStaff member 12, nurse Collective action d Appropriate patients were referred to the program. d The referrals to the program were primarily driven by a few champions of the program. "I'm surprised we didn't get more referrals is the other thing. Um, because I know how many people that I see whether they're male or female could've been helped." Staff member 6, pharmacist d Five hours Monday and Tuesday were dedicated for the capacity coach to be in the primary care clinic.
d Limited flexibility of the capacity coach's time due to the fact she was shared with another program. "Her supervisor was really strict on, like, 'You're there Monday and Tuesday morning from like, 10 to 12' kind of deal, but that doesn't work. Like that is a really small window, so she was able to give herself some more flexibility, which I really appreciated, and she would kind of weave people into her other schedule." eStaff member 4, social worker d Individual's practices with patients did change because of the intervention. "I think it was just having a little bit more focus on, you know, 'cause my practice I kinda of had to figure out those barriers, I needed to focus on those, um, before we had the training. But like how to approach focusing on them, and being more approachable to my patients, the veterans, um, on how to you know, get that out of them. And be a little bit more nurturing so to speak about how we get to that point, and maybe even having a better structure of how we facilitate doing that." eStaff member 6, pharmacist d When the program transitioned the capacity coaching role from social worker to peer, the coach and social worker had a productive working relationship with each other and with patients. "[Peer] is awesome at coaching and has, like, all of those skill sets. So she, I think, did a lot better job and really brought the program to life more than I could." eStaff member 4, social worker Continued on next page  15 This finding suggests that Capacity Coaching may deserve broader testing to understand its impact on patient health outcomes, quality of life, and health care utilization.
Positive changes in a variety of outcomes have been elusive when testing team-based management interventions for multimorbidity. 16 In studies, multiple interventions saw no effect on measures of utilization, health outcomes, or caregiver-or patient-reported outcomes. [17][18][19] One of these interventions was even specifically implemented within VA PACT teams. 19 However, participants exposed to the Capacity Coaching program could clearly articulate the difference between it and other programs at the VA, including one for intensive management of patients with chronic conditions and numerous programs offered through mental health services. This distinguishability suggests it should not be immediately lumped with past programs but rather compared. Importantly, the fidelity of the program's adherence to MDM principles should be monitored closely in future research, as it appears to be a distinguishing factor.
The HWC literature has lacked clarity in defining coaching. 20 However, a recent systematic review of HWC in chronic conditions indicated statistically significant changes in patients' psychological, behavioral, physiologic, and social outcomes across 11 of 13 studies examined. 21 Furthermore, a recent compendium of HWC interventions indicated mostly positive results for patients living with cancer, diabetes, heart disease, hypertension, and obesity. 22 Finally, in a recent systematic review of HWC for cancer survivors, patient capacity was supported across 4 of the 5 constructs of the TPC. 23 These previous studies also point to usefulness in fully testing Capacity Coaching.
It is important that future research of Capacity Coaching and MDM-driven interventions incorporate the implementation learnings of this study. First, implementation was a complex process of integrating a new way of working into the primary care team, which filled much of the 6-month pilot period. There was little time to enroll others and to plan expansion or evaluation. Second, critical challenges in the cognitive participation of the full primary care team could potentially have been better addressed with a longer-term cultural change approach. Curriculum to consider includes Leadership Saves Lives (LSL), which was previously applied to address the problem of posteheart attack mortality. 24 In testing LSL, researchers found significant changes in culture across hospitals and decreased posteheart attack mortality rates in hospitals that had the greatest changes in their culture. 25 Therefore, there is potential to use LSL as a facilitation strategy to overcome implementation challenges faced when piloting Capacity Coaching to test its impact on patient health outcomes and quality of life.

CONCLUSION
This study represents the first implementation and evaluation of Capacity Coaching. It found The program's implementation strengths were in participants making sense of the intervention (coherence) and working collectively to enact the program in the pilot period (collective action) 4 The program's implementation challenges were in planning the work of implementation and enrolling a diverse coalition of clinical staff to expand referrals to the program (cognitive participation) and in evaluating the impact of the program on outcomes that upper leadership was interested in to continue the program beyond the grant funding period (reflexive monitoring) that the program was feasible to implement in team-based primary care and adhered to principles of MDM. We saw substantial success of the implementation in participants making sense of the intervention (coherence) and working collectively to enact the program in the pilot period (collective action). Implementation challenges occurred in planning the work of implementation and enrolling a diverse coalition of clinical staff to expand referrals to the program (cognitive participation) and in evaluating the impact of the program on outcomes that upper leadership was interested in to continue the program beyond the grant funding period (reflexive monitoring). These challenges suggest the potential positive impact of incorporating existing culture-change curriculum, such as LSL, to facilitate implementation.

ACKNOWLEDGMENTS
The content of this work is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

SUPPLEMENTAL ONLINE MATERIAL
Supplemental material can be found online at http://mcpiqojournal.org. Supplemental material attached to journal articles has not been edited, and the authors take responsibility for the accuracy of all data.