Growing a Community-Academic Partnership: Lessons Learned in Forming a Qualitative Interview Team for the Community Partners in Care Study

Abstract

By engaging, partnering, and building trust with community members, research on vulnerable populations may offer opportunities to improve population health in communities that suffer from health disparities. While the literature on participatory and partnered approaches offers techniques and strategies for forming community-academic partnerships, less information is available about how partnerships can grow and evolve over time. In this article, we describe the expansion of a long-standing partnership that uses principles of community partnered participatory research (CPPR), a variant of community-based participatory research (CBPR). We outline the preparation and executive phases of conducting qualitative interviewing with highly vulnerable study participants who have already been participants in a longitudinal survey. We describe the challenges and concerns at each phase of the research and summarize some lessons learned. To grow and evolve, the partnership must constantly be reaffirmed in the experiences of new members.

Introduction

Participatory approaches to research, including community-based participatory research (CBPR)1-6 and community partnered participatory research (CPPR), may offer opportunities to improve population health in communities with disparities in health outcomes. CPPR describes a collaborative research approach that engages academic researchers and community members into partnerships to collaboratively implement and evaluate evidence-based approaches to health.7 Goals of this orientation to research include: to effect change in community health, systems, programs, or policies; and to build capacity for health improvement, particularly within minority communities. Like other participatory approaches, CPPR promotes respect, two-way knowledge exchange, and equal authority among partners through transparency and trust-building.8-11

Participatory approaches like CPPR are commonly used to study sensitive topics among vulnerable populations. The CPPR literature offers techniques and strategies used to form community-academic partnerships. However, less information is available about how partnerships can best grow and evolve over time to build research capacity.12 In particular, few articles describe preparing a community-academic research team to use qualitative methods to address difficult and complex study aims.

In this article, we describe the training of a small community-academic team to conduct complex qualitative interviewing in the context of an expansion of a long-standing partnership. The parent partnership, Community Partners in Care (CPIC), had successfully used a CPPR approach over almost a decade to study depression in under-resourced communities, then sought to quickly assemble and train a new partnered team to conduct approximately 100 interviews with vulnerable clients. This article features reflections from community and academic members of the qualitative interviewing team. We describe several key challenges in pursuing partnered, team-based qualitative interviewing and in expanding an existing partnership. We provide suggestions for the preparation and implementation phases of a partnership aiming to expand its capacity to conduct qualitative interviews.

Methods

Context

The Community Partners in Care (CPIC) partnership dates back to 2003 when Healthy African American Families (HAAF) partnered with Charles Drew University of Medicine and Science, the RAND Corporation, and the University of California, Los Angeles (UCLA) to create Witness for Wellness (W4W), a program aimed at developing community-based approaches to improve depression outcomes in minority communities. The W4W efforts led to CPIC, a cluster-randomized comparative effectiveness trial led by the W4W investigators along with 25 additional agencies. CPIC screened and enrolled more than 1000 depressed clients drawn from 133 programs in 60 agencies.12 The clients were primarily African American and Latino and the majority were living at the federal poverty level. Approximately half of the clients were uninsured and at high risk for homelessness.13,14

In 2013, CPIC received funding to conduct 100 qualitative interviews with clients at 3-year follow up in order to understand their preferences for depression care. CPIC investigators planned to interview a subsample of 100 minority clients (ie, African American, Latino) of both sexes who scored in the highest or lowest quartile on the 9-item Patient Health Questionnaire, a standard, brief, self-report measure of depression.15 The interview aims were: 1) to learn about the experiences of clients whose depressive symptoms did or did not improve over time; and 2) to document the range of clients’ needs and identify which needs were a priority. The interview included a structured portion, with interviewers asking a set of closed-ended questions about needs (eg, finances, employment, housing, physical and mental health, completing daily tasks) and an exploratory, open-ended portion eliciting help-seeking experiences related to three urgent needs. Interviewers also discussed with clients how providers and agencies could better address these needs.

Conceptual Framework

The parent CPIC study used the CPPR model of partnership, and CPIC investigators articulated the Vision, Valley, Victory model for understanding the phases of partnered team development. In this model, a team develops strategies and goals for the project (Vision), carries out the activities necessary to implement the project (Valley), and then celebrates success and completes and disseminates products (Victory).10 Yet, because the CPIC study had not undertaken formal qualitative interviewing with clients, no qualitative researchers had participated in all phases of CPIC partnership development. Moreover, CPIC included no established process for training a community-academic team to conduct reliable and ethical interviews. In addition, CPIC investigators considered these interviews to be complex because of the vulnerability of the client sample, the sensitivity of the interview topics, and the goal of completing a large number of interviews in a short timeframe. Thus, a new team had to quickly prepare to conduct interviews with vulnerable clients. Interviewers would need to understand CPIC aims thoroughly in order to improvise follow-up questions during interviews. Furthermore, the team members had to quickly grasp the CPPR ethos and work together within the framework of community participation.

Data Sources

Below we describe the challenges and concerns at each phase of the research (Building the Team, Refining Instruments and Training, Addressing Interviewing Challenges, and Learning from the Process). All the quoted material presented here emerged from phone conversations and email correspondences during the final phase and while writing this article (see Phase 4). Team members used phone conversations to reflect on the team-building process and then shared their reflections in written form with the team by email, which became the source of quotes for this article. This study was approved by the RAND Human Subjects Protection Committee.

Results

Phase 1: Building the Team

Before forming the new team, CPIC investigators, with assistance from the CPIC Council, developed a preliminary interview guide and a template for generating a “report card” about each client to be interviewed. The report card summarized select items from the client’s 3-year survey responses, such as recent stressful life events and depressive symptoms. Council members had raised concerns about the focus on deficiencies involved in inquiring about multiple needs and were concerned about building trust with clients unless interviewers were culturally concordant. CPIC investigators raised concerns about the reliability and quality of the data collected by a relatively large number of interviewers with no prior experience conducting qualitative interviews. From both perspectives, training was considered important.

The qualitative interview team ultimately consisted of three community members, four academic staff members, and four individuals who were both community members and academic students. The students were in programs for clinical psychology and public health; academic team members had training in social sciences, psychiatry, and health policy; and community members had expertise in community advocacy for health. Three of the team members were new to the CPIC project, and while other team members had some involvement with some CPIC events, none had collected qualitative data for CPIC. All team members served in various roles at different points in the project; however, two of the three academic members were involved in the project during the drafting of the interview protocol, as well as the training binder. The team met in four two-hour training meetings over four weeks to develop and practice the procedures for interviewing clients by phone. Each person was given a binder with their name on it containing a training agenda, a draft of the client interview guide, a sample report card, a copy of the 3-year client survey, and other information. The group meeting began with icebreakers to get to know each other and build rapport as a team.

Initially, many community partners who agreed to work on the qualitative team were uncertain of the direction the study would take and whether they would feel comfortable with it. Some had reservations because of a lack of trust in, or negative experiences with, academic research. One academic student stated, “I came to the work a little nervous about possible exploitation of suffering disproportionately experienced by people of color by UCLA researchers for career fame, grants, etc.” A community member voiced, “I was very uncomfortable – intimidated; I had to learn to trust academia.” One of the academic staff stated, “As soon as I identified as an academic, I realized that it put me in a privileged category, for which I then felt uncomfortable... I did not feel that I could safely share my knowledge without creating resentments or my life experiences without undermining my precarious status as a professional.” The parent CPIC study had well-established procedures for openly managing conflict and reaching consensus, which the qualitative team reviewed together. In retrospect, power differentials and cultural differences may have initially undermined the evolving partnership but, as described below, the team found creative solutions to address differences and share decision-making power. Yet some areas of disagreement about the team process were not shared until after the data collection was complete because team members felt focused on ensuring the quality and consistency of the data collection.

Phase 2: Refining Instruments and Training

As they prepared to conduct ethical research, team members reviewed and refined the interview guide together. Two members role-played as interviewer and interviewee in front of the group and then everyone broke into dyads for further role-playing. The process revealed some possible roadblocks and additional areas that needed to be addressed, such as a need for alternate language use or clarification of the questions being asked of the interviewee.

HAAF staff delivered a two-hour training on ethical interviewing and cultural competence, and an academic investigator trained the team on study safety protocols and the management of adverse events, like suicidal ideation or reported abuse of an elderly person or child. This training included participation from both the community and academic partners. Team members learned useful information, such as the suggestion to use Miss/Mrs./Mr. when addressing interviewees, as is common among African Americans.

In week three, the team reviewed logistical matters and developed step-by-step instructions detailing the use of the recording equipment. The team learned how to securely store recorders, electronic audio files, and forms that contained client’s identifiers.

The team agreed upon various modifications intended to make the interviews more culturally sensitive and effective. One participant found discussions of the interview guide “a little overwhelming” because of the “implicit cultural dynamics that were hard to articulate,” and sensed a cultural bias to the interview guide because its frequent use of the word depression and other academic jargon. The participant was appreciative of not being the only community stakeholder on the team as another team member was “there to help with the translation” of cultural differences.

After interviewers completed one or two interviews, the group listened to the recordings and provided feedback to refine skills and improve the quality of the interviewing. Feedback often touched on interview opening and rapport building and techniques for following up on key topics. The interviewer was given an opportunity to discuss the client’s circumstance, review what happened during the interview, and consider how she did or did not ask follow-up questions. Each interviewer also participated in one-on-one feedback sessions with an academic team member who listened to interview recordings and offered recommendations for improving interview technique. A student team member commented on the difficulty of the feedback process: “Going into the interviews, I thought I had a good grasp of what information we were trying to get and why; therefore I was a little taken aback when I got feedback that I wasn’t exactly getting the type of information that the data team wanted.” This process was seen as key to data quality, but in retrospect, it became clear that the team also improved in its ability to collaborate through this process. After these initial feedback sessions, the team came together to revise the interview guide again by expanding the list of needs to elicit and improving the wording of some questions.

Phase 3: Addressing Interviewing Challenges

While conducting interviews, the team continued to meet as a group and one-on-one to debrief and receive feedback about the interviews. While some interviewers learned to adapt to the story being told and to ask follow-up questions, others had more difficulty assessing clients’ responses and pivoting the dialogue to follow-up on what could be an important thread. Researchers found it difficult to train the interviewers on how to use the guide loosely—to diverge from it to follow the client’s story. One academic providing feedback felt trapped between competing expectations to ensure data quality and not act authoritatively. She endeavored to be sensitive to novice researchers, and so hesitated when critiquing interviewers or suggesting that a particular interviewer was not meeting standards of excellence. Data collection proceeded according to schedule and review of the data indicated its quality was acceptable.

Challenges emerged in engaging with clients and responding to their needs. Interviewers built trust with clients to help them feel comfortable sharing personal information about their most urgent needs and frustrations seeking help. Interviewers used information provided on the report card at the start of the interview to build rapport (eg, “I understand from your survey responses that you’ve been feeling less blue in the past few months”). Interviewers realized immediately that these were emotionally difficult interviews to conduct. One community team member needed to completely stop conducting interviews and reflected: “When it was time to do the actual interviews, I decided to not be a part of it because I felt the pain of the clients.” In response to this team member’s emotional reaction to conducting interviews, the team developed a debriefing process to encourage reflection immediately after the call. Interviewers filled out a worksheet that asked how the call had proceeded, what was learned, and if any problems arose. The debriefing process seemed to help interviewers process their feelings and maintain a safe distance from the client, as well as assess the usefulness of the data collected, thereby increasing their capacity to think like an investigator.

Team members also developed a strategy to provide some ancillary care to clients when they realized they could assist with clients’ unmet needs. Several team members knew of community resources that could be helpful and wanted to share this information with clients. As one interviewer said, providing this help seemed an appropriate exchange for their participation: “people really opened up about problems impacting them now and it would have felt harsh to listen to those problems, know of resources in my head to share-but not have a way or it not being part of the protocol to offer them as a part of the reciprocity in participating in the interview.” The team decided to give all clients information about a free phone line that provides information about community resources. The team further developed a catalog of resources to share with clients as needed. For clients who needed it, the referral information along with a personal note from the interviewer was included in the mailer that sent their gift card payment. This personalized act felt like it acknowledged for both interviewer and client the importance of their conversation.

In the end, the team met its goals in data collection. The team completed 104 client interviews in 4 months with a small amount of missing data and good-quality, in-depth interview data. Community interviewers (n=3) completed 48 of 104, while academic interviewers (n=4) completed 36, and the community members/academic students (n=4) completed 20. The team disseminated preliminary findings to community groups through a large community conference.

Phase 4: Learning from the Process

In many ways, it was only after the interviews were complete that the team was able to acknowledge the profound degree of bi-directional learning that took place through the process. In bi-weekly and monthly calls after data collection, the team reflected upon and acknowledged these lessons. These conversations continued over email and through the writing of this article. Community members learned new research methods and cited the benefits of capacity building while the academic partners cited improved cultural competence and sensitivity.

Among other issues, the team reflected on the fact that power differentials may always be present in relationships, but these power dynamics can be minimized to create reciprocal partnerships that sustain trust and engagement. The research team itself developed a mature partnership that openly shared conflicts and valued team members’ diverse perspectives. As one community member stated, “I can tell you that I have learned about humility being a part of this research project. I learned that I was working with a group of individuals who understand what it means to collaborate, be supportive, and move together toward a unified goal.”

While an emphasis was placed overtly on inclusion, mutual respect, and bi-directional learning and sharing, some community team members (but not all) did not perceive that the process was as collaborative as it could have been. For many, it felt as if the project were an academic endeavor in which they, the community members, were merely providing assistance. It did not feel that the interviewer or the coder always understood what the researchers were looking for. As another non-academic interviewer put it, “At times, I was afraid of letting them [the researchers] down if I could not accomplish the goals of the interviews.” In retrospect, opportunities for working more thoroughly toward equity and shared leadership in all study procedures were recognized.

Discussion

In this article, we described the expansion of a long-standing community-academic partnership in which a new team was organized to collect qualitative data from a vulnerable group of clients. Our experience highlights that bringing new partners into an established community partnership can be challenging. In many ways, members of this team expressed that the foundational work of relationship-building should have been repeated, with each new team member exploring assumptions, welcoming new perspectives, and being open with concerns. Yet, the strength of the existing partnership may have led some team members to assume that the work of building relationships was already complete. Several team members suggested more sharing of community and academic experience at the beginning of the project.

Moreover, it may have been a disadvantage for the qualitative team that the larger partnership conducted some preliminary preparation for interviews before the qualitative team came together. A preliminary interview guide, a report card, and some expectations about team members’ roles were in place at the first team meeting. It may have been more supportive of relationship-building that all members of the qualitative team be involved in all phases of the project.

We learned that qualitative data collection may entail unique challenges for a new community-academic team. Data collection by community interviewers with lived experience of the condition under study may help build trust with interviewees. Yet, these new interviewers also face challenges of managing the emotional experience of interviewing and learning the techniques of follow-up and value-neutral interviewing through feedback on their performance. A climate of trust on a research team is thus especially important. On this team, opportunities for debriefing and providing ancillary care were essential to help team members tolerate the emotional challenges of interviewing highly vulnerable research participants.

In an ongoing partnership, the partners may cycle through stages of Vision, Valley, and Victory multiple times,10 yet when new members join the team, the stages may not be worked through successively by all members at the same time. New team members who have not experienced all these stages will require time to learn about the partnership, buy into this approach for themselves, and learn to partner together with specific team members. On our team, some initial conflicts did emerge but were worked through before data collection began, and other conflicts remained latent until after data collection was completed when they could surface for reflection because team members had achieved the goal of data collection. New team members can be brought into an established partnership by having experienced partners give them specific information about the ground rules and norms of behavior within the partnership. Yet, each new team will also need to work through the unique, first-hand experience of learning to partner together. Group members should be composed of representative community partners who are comfortable with each other and who are all comfortable having their voices heard.

Conclusion

Our experience suggests that, in order to grow and evolve, the partnership must be reaffirmed in the experiences of new members. Techniques for growing an academic partnership should include sufficient time and the creation of a safe space for learning about the history of the project and individuals on the team.16,17 Much of our discovery aligns with the findings and recommendations for qualitative research teams in general regarding team building and group process. However, the uniqueness of our journey may lie in that, in forming new team configurations and new projects, even within a well-established partnership, earlier team developmental stages may need to be revisited and reflected upon because of the dynamic nature of the academic-community partnership.18 Introductions, storytelling, and reflections about each person’s experiences on the team at the beginning of every meeting may help both veteran and novice team members learn from one another. In this way, new team members can successfully overcome biases, learn about cultural differences, and garner new research skills. In these ways, as this experience demonstrates, community and academic partnerships become opportunities for capacity-building, cross-cultural learning, and for the improvement of health in under-resourced communities.