Expanding the 3 Wishes Project for compassionate end-of-life care: a qualitative evaluation of local adaptations

The 3 Wishes Project (3WP) is an end-of-life program designed to attenuate human suffering by cultivating connections and deepening relationships among patients, family members, and clinicians in the intensive care unit (ICU) [1‐3]. These relationships are fostered through a focus on the preferences and legacy of the dying patient [3‐5]. The 3WP begins with clinicians asking how they might honor the patient in the last moments of life and bring comfort to the family, then evolves to working with other clinicians, family members, volunteers, and the community to implement these wishes. The program began in 2013 at St. Joseph’s Healthcare in Hamilton, Ontario. Today, the 3 Wishes Project has not only become embedded into the culture of end-of-life care in the original unit, but thrives in over 20 ICUs across North America in a variety of formats ranging from research project to clinical program. Through a variety of mixed-methods studies, we have demonstrated the impact of the 3WP as experienced by families, clinicians, and trainees [6‐11]. Recently, we published a multicenter formative program evaluation of the 3WP in 4 hospitals, demonstrating it is a transferrable, affordable, and sustainable end-of-life care program that provides value to families, clinicians, and the institution [9].

As 3WP continues to expand, the original 3WP site receives numerous requests for advice about initiation in new centers. Many requests relate to how new sites might implement 3WP with limited resources, and how the program might operate without the research elements and the associated infrastructure of an academic tertiary care hospital. With the goal of assisting new centers considering local implementation of 3WP, the objective of this study is to describe various designs of the 3WP across 9 ICUs in North America, to identify essential elements for success, and to describe local adaptations.

We conducted a qualitative descriptive [12] study of the process of implementing this semi-structured end-of-life clinical program in 9 ICUs across 7 hospital systems within Canada and the United States. This research was aligned with a pragmatist epistemological paradigm [13]. A qualitative descriptive approach is appropriate for this type of applied health research question because it allows analysis to stay closely connected to participant words and ideas, without significant interpretive inference.

The expansion of the 3WP was guided by implementation support from the original center in the form of a Start-Up Guide [14], publications [7, 9], periodic teleconferences, email, and/or site visits. As the program spread, the original center encouraged new sites to adapt and refine the program to address their units’ identified needs with available resources, so long as the program continued to be guided by conversations with dying patients (if able) and their families about what would be meaningful at the end of life. These “wishes” were then implemented by the care team, other hospital staff, family members, and friends.

Participating centers were sampled from amongst centers where 3WP was successfully implemented and data was shared back to the original center. We allowed each center to self-define “successful implementation”, because we wanted a sample of centers with different levels of experience with the program, from different sizes of hospitals so imposing a definition of success related to the number of patients enrolled or longevity at the center would not be helpful. We sampled participating centers to represent diversity in the model of 3WP implemented, duration of implementation, and organizational and geographical context. All centers collected data describing patient characteristics and wishes implemented. At some centers, quantitative data were supplemented with qualitative data about the experiences of family members and clinicians. Supplemental Figure A describes different implementation models, including an approach that did not involve data collection (a model not represented in this study, as this report is data-based).

Ethics approval for this study was obtained from the Hamilton Integrated Research Ethics Board and from each institutional review board.

From 2016 to 2019, we conducted interviews and focus groups with individuals active in the initiation or day-to-day leadership of 3WP at each of the 7 hospitals. Focus groups were composed of individuals from multiple centers to facilitate conversations about local variations. Interviews included one or two key personnel from each center to generate an in-depth description of the program at their site. Interview and focus group guides were designed and piloted with members from the original center and further refined as the project progressed. All interviews and focus groups were conducted by trained qualitative researchers who do not have clinical roles (MV and MES) or bedside implementation experience, but who have long standing familiarity with the program. Interviews and focus groups were audio-recorded and transcribed verbatim; all participants provided written informed consent.

Interview and focus group data were supplemented with process documents produced by the original center. After preliminary analysis, we returned to center leads at each hospital, asking for additional input on several key concepts identified during analysis where data were incomplete. Review of findings and provision of supplemental information where necessary served as a form of member checking to ensure resonance and credibility [15]; and also an opportunity for leads to provide updated information about the evolution of the program at their hospital. Data sufficiency was determined by the ability to create a rich description of the program at each hospital, and local affirmation that no key elements were missing.

Data were analyzed using an iterative, step-wise approach adapted from Grounded Theory [16]. Preliminary coding was completed by two independent coders (MS and MES), who met with a larger multidisciplinary team (MV, DJC, FJC, NH) to review key ideas and decide on a coding schema for subsequent rounds of analysis. Focused coding followed this trajectory, with two coders working independently and a larger group reviewing the work and making decisions about ideas to investigate in future rounds. Findings were triangulated among different sources of data (e.g. interviews, focus groups, process documents) [17] and in the member-checking exercise [15]. N-Vivo 12 was used to manage data.

Figure 1 displays the timeline over which each center implemented the program and collected data. Table 1 describes the features of the 9 ICUs located in 7 hospital systems. Table 2 outlines the roles of the 85 interview and focus group participants, who represented nurses, physicians, spiritual care clinicians, recreational therapists, and others.

All centers identified a common goal for project initiation: a desire to improve end-of-life care. In addition to identifying essential elements related to this goal, we also describe variations across sites, mainly related to organizational context and resources. Finally, we describe a consistent pattern of program evolution, wherein local adaptations and refinements were facilitated through cross-pollination of ideas fostered by a network of like-minded individuals with shared purpose.

In this paper we describe the implementation of 3WP in 9 ICUs located in 7 hospital systems, detailing variations in the program across sites, identifying essential elements, and reporting potential adaptations that new centers may find helpful when considering whether or how to implement 3WP. Examining the factors facilitating multicenter implementation reveals the importance of program flexibility and the empowerment of frontline staff. These related elements are both fundamental to successful expansion of 3WP.

Each environment in which the 3WP is implemented is unique, with different organizational features, core team composition, administrative support, academic affiliation, and available resources. The semi-structured nature of the program allows for adaptations that consider the culture and context of each setting [18, 19] and thereby encourages uptake and dissemination [20]. The involvement of frontline clinicians ensures that the program is adapted in a way that would be unlikely with a strictly protocolized intervention implemented in a top-down manner by administrators. The empowerment of frontline clinicians to “own” the program encourages staff to take action to positively influence the culture of their workplace, which may account for our prior finding that 3WP enhances professional satisfaction [7‐9]. A strictly protocolized program would likely inhibit individuals from identifying creative ways of relating to each other, potentially reducing the value of human connections realized through the program. Current and previous research about 3WP has found the meaning of these human connections to be one of the most important strengths of the program [6, 7, 9‐11, 21].

Another key consequence of frontline clinician ownership of the program is the team approach to end-of-life care. This may supplement the typical patient-physician dyad by explicitly empowering other healthcare staff to recognize and respond to the needs of both the dying patients and their family members. Although an interprofessional approach is generally acknowledged as an important component of high quality end-of-life care in ICU [22, 23], other studies have suggested that this approach is impeded by staff discomfort due to lack of training or uncertainty in their perceived role when caring for dying patients [24]. The open, visible nature of 3WP and its frequent focus in workplace conversations helps to maintain and sustain this team approach, allowing the contributions of all clinicians to be recognized and valued. This democratization of leadership opportunities united clinicians in a common focus on patient and family experiences in end-of-life care.

In this report, we describe the adaptations and refinements to the 3 Wishes Project that enabled successful implementation and fostered viability in variable ICU settings despite different contexts, needs, and resources. Commitment to high quality end-of-life care was an essential motivator, inspiring frontline clinicians to initiate the program, creatively resource it, and maintain enthusiasm. As the program became more established, it transitioned from an “add on” to existing end-of-life programs to become “embedded into” the unit’s approach to end-of-life care. This evolution was initially facilitated and later fueled by the alliance of like-minded individuals within each site, among several communities, and across the 3WP network.