Aim of the Review
The aim of this systematic review was to explore, synthesise, and analyse the views of GPs on end-of-life care learning preferences
Review Question
What are the views of GPs on end-of-life care learning preferences?
The end-of-life care learning preferences of GPs were reviewed using the Population, Phenomenon of Interest, and Context (PICo) framework [14]. The population studied was GP, the phenomenon of interest was views on learning preferences and the context was end-of-life care.
The philosophical paradigm and theoretical framework of research
Constructivist grounded theory was the philosophical paradigm underpinning this review that explored the GP’s experiential views, their actions and interactions, and the implicit meanings attributed to their learning preferences [15]. Social constructivist learning theory was applied to understand how individuals construct knowledge in a social context [16-20]. It is built on three premises: cognitive processing of knowledge, self-directed learning, and social construction of knowledge.
Review Design:
A scoping search was conducted to determine the breadth of the evidence available on the phenomenon explored. The search showed a heterogeneous mixture of quantitative and qualitative studies. Popay’s narrative synthesis method enables synthesis of the data from a mixed typology of studies into themes [21]. Moreover, it allows using a theoretical framework for the interpretation of review findings and provides flexibility in choosing the methods within each step of the synthesis relevant to the review.
Search strategy:
The review question was divided into search concepts, which were further used to conduct an initial scoping review. Scoping review helped derive key search terms relevant to each concept of the review questions. The search identified three papers to test the sensitivity of the search [22–24]. The search terms identified from these studies were further expanded into thesaurus terms and free text terms [25]. A search was conducted using electronic databases (MEDLINE, EMBASE, CINAHL, and PsychINFO) to identify articles published in English between 01/01/1990 and 31/05/2021 (Appendix 1), as the first article on palliative care was published in MEDLINE in 1993 [26]. The search was conducted using thesaurus and free-text terms specific to the database, and the terms were combined using Boolean operators [25]. Additionally, searches were conducted using SCOPUS, the Web of Science, and the Cochrane database using free texts. A list of twelve journals was hand searched for additional citations (Appendix 2). The bibliographies of the full-text articles were screened using the Google Scholar database for any new articles that could be added.
Study eligibility:
Selection criteria of studies included is provided Table 1.
Assessing methodological rigor of the studies included in the review:
As the review included a mixture of qualitative and survey studies, Hawker’s tool was used to assess its methodological rigour [27]. There are a growing number of palliative care systematic reviews that have used this tool [28–30]. Hawker’s tool allows systematic appraisal of the study by analysing the title and abstract, introduction and objectives, method and data, sampling, data analysis, ethical aspects, results, transferability/generalizability, and implications of the study [27]. These criteria were scored between 1 and 4 (1=very poor and 4=good). A score of 9 was considered a minimum score and a score of 36 as a maximum (Appendix 3) [27]. Although Hawker does not mention a cut-off score, based on the previous studies [29, 31], a cut-off was set at 19. Three studies were excluded from the review as they had a score of less than 19 [32–34]. The minimum score of the studies included in the review was 19 and the maximum score was 32. Furthermore, the studies were classified into the following grades: "high quality" (A), 30–36 points; "medium quality" (B), 24–29 points; and "low quality" (C), 19–23 points [35]. Six studies were of high quality [22–24, 36–38], ten studies were of medium quality [39–48] and seven studies were of poor quality [49–55].
Data Extraction:
Screening, quality appraisal, and data extraction were conducted independently by two reviewers. The third reviewer helped resolve the conflicts. The initial section of the data extraction sheet had information regarding the country and year of publication. The second section focused on the type of study, that is, survey, qualitative, or mixed-method. In this section, study objectives, population, and study setting were also described. The study sample, participants, inclusion and exclusion criteria, research design, and methods were elucidated in the third section. The fourth section provided information on the study findings and conclusions.
Data Synthesis:
The review findings were synthesised using Popay’s narrative synthesis [21]. The first step of the narrative synthesis is to identify a theoretical framework, and social constructivist learning theory was used to interpret the findings of the review [17]. It was followed by developing a preliminary synthesis that involved a brief description of the studies in the review. The data gathered was classified into countries, the year of publication, type of population, and the factors involved in constructing knowledge. The words and texts extracted helped the reviewers familiarise with the study findings before analysis. Patterns were identified from the preliminary synthesis, and the reviewers explored the relationships within and between studies in order to generate meaningful categories and themes. The reviewers were mindful of gleaning the similarities and differences in the data generated. The fourth step was to assess the robustness of the synthesis. To ensure this, the reviewers critically reflected on the synthesis process and identified possible sources of bias [21].
Review findings:
Overview of the studies:
Out of 10037 articles identified from the database searches, 23 articles were included for synthesis (The PRISMA flow diagram-Figure1). Eleven studies were qualitative, eleven were surveys, and one was a mixed-method study. Fifteen studies were from Europe (eight from the United Kingdom, two from Belgium, and one each from Denmark, the Netherlands, Austria, Germany, and Ireland), six from Australia, and two from Canada. The majority of the surveys were from a single centre or spanned different centres in the country of study (Table2). The qualitative studies were single centric (Table2). Two studies included combination of GPs and paediatricians involved in general practice [41,47].
Review themes
Five themes were generated in the review. These themes were: motivation for end-of-life care learning, end-of-life care learning needs, preference for a learning style, perceived facilitators of learning, and perceived barriers to learning. Refer to Appendix4 and Appendix 5 for the table narrating the themes and thematic diagram.
Theme 1: Motivation for end-of-life care learning
Motivation is defined as an intrinsic trigger that allows for sustained goal-directed activity [56,57]. Self-directed learning enhances the learner’s competence, and belongingness facilitates motivation [58,59]. Perceptions of value addition, peer recognition, previous learning experiences, and individual goals all trigger learning [60].The need to provide palliative care, self-actualization needs, relevance to practice, sense of responsibility, and therapeutic bond motivated them to learn about end-of-life care.
A perceived need to improve the quality of end-of-life care and advance care planning [37,38,42,55], as well as having a larger clientele of geriatric and sicker patients, piqued GPs' interest in end-of-life care learning [42,44,47,49,55]. Unmet self-actualization needs included a desire to integrate palliative care into routine care [36,46,47,54], a perceived need for self-transformation [46], coping with own bereavement [37], and a perceived inability to manage symptoms [38], all of which instilled a sense of powerlessness, helplessness, and emotional burden of caring [22, 36-38,46] for which they needed a recourse [46].GPs accessed training only if the training [22,39,52] and the trainer’s skills were in alignment with their needs [52] and helped address complex end-of-life care needs [22,24]. A therapeutic bond developed with their patients over a prolonged period of caring [45,49,51], at various stages of their illness [36,37,44,45,47], instilled a sense of responsibility towards their patients [23,37,38,47]. They believed in helping a patient as a whole person and not as a disease entity [47] and felt the need to address the complex sufferings of patients while honouring their dignity, expectations, and wishes [39,42].
Theme 2: End-of-life care learning needs
Humans are constantly on a quest to evolve their self-identity [61]. In the process, they encounter new experiences that may not be in consonance with their past experiences [62]. The unrest arising out of this is tackled in three ways: by preserving the existing schema and ignoring the new contradictory knowledge; by preserving both the contradictory knowledge and appropriating it as the situation requires; or by constructing a new concept that aids in resolving the contradictory knowledge [62]. Following were the learning needs: accessing palliative care, learning about pain and symptom management, communication skills, compassionate care, addressing the caregiver’s needs, the ethical and medico-legal aspects of end-of-life care, and teamwork.
GPs expressed the need for a framework for accessing local specialist palliative care resources [22,36,37,45,47], guidelines on out-of-hours care [22], timing and criteria for referral to specialist services [38,45] and knowledge of a multidisciplinary team approach [22,37]. Pain management [37,40,41,46,48,50,53], psychosocial issues [24,37,44,50,54], fear of death and dying [37,42,46,55], anger and demanding behaviour [53], nutrition [41,46], spiritual needs [23,41,55], and religious concerns [23] were all perceived as important. GPs expressed the need for training in communication skills and breaking bad news that enabled them to make decisions on chronic illnesses [47] and facilitate transitions to end-of-life [46,48]. They appreciated training that might enable them to have the right attitude, diplomacy, and skills [37,46] to resolve conflict [37,42,46], denial, and bargaining [36, 37, 46] and handle vulnerability around death and dying [47]. They expressed the need for training in handling the emotional distress of family caregivers [46,50], and anticipatory grief [37], and bereavement support [47,50,54,55]. GPs felt a need for guidelines to enable them to provide care in the best interests of their patients [22,42]and techniques to discuss ethical and medico-legal aspects of care with their patients, families and colleagues [45,46]. They expressed the need for skills in developing a team and team work, garnering mutual support from colleagues in a multidisciplinary team [41,47] and for debriefing sessions to cope with emotionally burdensome situations [36,41,47].
Theme 3: Preference for a learning style
A learning style is defined as a set of "characteristic cognitive, affective, and psychosocial behaviours that serve as relatively stable indicators of how learners perceive, interact with, and respond to the learning environment" [63]. When individuals become aware of their strengths and weaknesses, they become motivated to learn and become lifelong learners [64]. It may be essential to match learning styles with learning preferences and tailor the style to the content and context of learning in order to enhance learning [65,66,67]. GPs preferred learning styles that were experiential, pragmatist, self-directed, and didactic.
According to GPs, expertise grows with experience [22,23,37,39,54] and through interaction with patients/families [22,23,24,39]. They preferred work-based learning under a mentor [22,24,36,38] or co-management with a palliative care team [22,45]. They wished to reflect on their practice under a mentor [23,24,38,43,47,48] either through virtual interaction [22,52], by the patient’s bedside [45,53,55], or through audited data [55]. They also expressed the need for problem-based learning by patient follow-up and learning from mistakes [43,46]. GPs preferred self-directed learning through web-based or printed resources [24,36-38,41,43,44,47,48,53]. Some of them were computer-based learning [55], correspondence learning [4], guidelines-based learning [52,53], E-learning [22] and learning by researching [55]. Other forms of learning included lectures [47,54,55] and workshops on team building and addressing religious and psychological needs [23,38].
Theme 4: Perceived facilitators of Learning
Facilitation of learning is an essential process that empowers learners to evolve in the learning process through self-evaluation and interchange [68]. According to Brockett, attending, responding, and understanding are important for effective facilitation, which builds a foundation for a constructive and meaningful learning environment [68]. An amicable learning environment, mentorship, appropriate timing, and feedback facilitated end-of-life care learning.
GPs felt the need for an amicable, safe, non-intimidating, non-judgmental, trustworthy and respectful learning environment where their deficiencies were not exposed to their patients [23,24]. Mentorship is a bidirectional process [69] and provides a safe space for the learners to interact [70]. GPs expressed the need for mentors from a multidisciplinary team [23,46,48,54] and a small group [23,36] interactive session with multidisciplinary peers [40,48]. Due to time constraints, solo practice, remote practice, and multiple concurrent responsibilities, GPs are unable to attend educational programs during office hours [71,72].GPs felt that learning was a lifelong process [23,36] and preferred to attend half-day workshops [52], quarterly topical in-service training [53] and quarterly case rounds [53], courses out-of-business hours [36,40] over the weekend [40,41]. Goal-directed feedback keeps the learner motivated and engaged in order to strike a balance between one’s goals and the expectations that one can achieve these goals [73]. GPs preferred feedback from specialists and peers in the team [23] and patients and families if it was completed in a time bound manner [22, 38]. They preferred to avoid feedback from dying patients or their families, as this might cause discomfort to the respondents [22].
Theme 5: Perceived Barriers to Learning
The learner’s decision to participate in learning is influenced by constraining factors that could be dispositional, situational, institutional, or academic in nature [74, 75]. End-of-life care learning was hampered by a lack of resources, individual and professional challenges, disempowerment, conflicts, and training that was not aligned with clinical practice.
GPs had apprehensions about initiating discussions around end-of-life care as this was perceived as taboo by their patients and families [39,47]. They feared treading into their patients’ or family’s private spiritual space [39] and felt guilty about their perceived lack of competence in addressing end-of-life care concerns [39]. Other limitations included fear of medico-legal recrimination [43,46] and the emotional burden of care [37]. GPs felt a sense of insecurity when broaching their deficiencies in knowledge [43]. Younger GPs with fewer years of experience had less autonomy to focus on palliative care as a choice of specialty [39,45,49]. Some felt they had enough knowledge to address their patients’ palliative care concerns [52] and that there were no new developments in the field that warranted up-gradation of knowledge [37,38]. Additionally, family and personal commitments gave them less time for palliative care provision [44]. GPs who had solo practice or practiced in remote areas had fewer support systems that precluded access to training [47,49,51]. Excess work pressure [39,52,55] with the resultant lack of time [22,23,36,37,44,46,52,55], poor remuneration [36,44], and the temporary nature of the job [37] were hindrances to training. They were also limited in their access to training programs because they had to self-fund their courses or provide compensation to organise locum GPs in their absence [55].When specialists took over care [22,23,44], GPs felt a loss of control over their patients [37].They were disillusioned when specialists or patients lacked trust in them [22], or when they were not acknowledged [36,46]. Lack of clarity in roles and responsibilities led to conflict in care coordination [23,38,52]. GPs were apprehensive about confronting specialists for fear of being reprimanded [52]. Furthermore, previous negative experiences and delays in receiving responses from specialists limited future interactions with specialists [2].Most end-of-life care training programs focused on oncology, providing fewer opportunities for them to learn non-malignant end-of-life care skills, which accounted for a significant portion of their practice [51].Training occurred mostly during office hours in the hospital setting [43,55] and skills gained could not be replicated in the community due to differences in the infrastructure in both settings [23].