Physician-patient boundaries in palliative care.

Background: Nurturing effective physician-patient relationships is essential to the provision of patient-centred care. Palliative care physicians may apply boundary-crossings or breaches in professional standards to nurture effective physician-patient relationships. Being highly individualized and shaped by the physician's narratives, clinical experience, and contextual considerations, boundary-crossings are susceptible to ethical and professional violations. To better appreciate this concept, we employ the Ring Theory of Personhood (RToP) to map the effects of boundary-crossings on the physician's belief systems. Methods: As part of the Tool Design SEBA methodology, a Systematic Evidence-Based Approach (SEBA) guided systematic scoping review was employed to guide the design of a semi-structured interview questionnaire with palliative care physicians. The transcripts were simultaneously content and thematically analysed. The themes and categories identified were combined using the Jigsaw Perspective and the resulting domains formed the basis for the discussion. Results: The domains identified from the 12 semi-structured interviews were catalysts and boundary-crossings. Boundary-crossings attempt to address threats to a physician's belief systems (catalysts) and are highly individualized. Employ of boundary-crossings depend on the physician's sensitivity to these 'catalysts', their judgement and willingness to act, and their ability to balance various considerations and reflect on their actions and their ramifications. These experiences reshape belief systems, understandings of boundary-crossings and may influence decision-making and practice, underscoring the potential for greater professional breaches when unchecked. Conclusion: Underlining its longitudinal effects, the Krishna Model underscores the importance of longitudinal support, assessment and oversight of palliative care physicians, and lays the foundation for a RToP-based tool to be employed within portfolios.

Keywords: Boundary-crossings; Palliative care; Physician-patient relationship; Doctor-patient relationship; Boundaries; Professional identity formation; Personhood; Professionalism

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1186/s12904-023-01161-0.

Caring for patients with terminal illnesses relies on the nurturing of respectful, open, trusting, and enduring connections between physician and patient [[1]]. In such complex and often difficult circumstances, maintaining the physician-patient relationship places unique demands upon the physician. Amidst evolving conditions, changing care goals and demands and difficult discussions about prognosis, care, and treatment plans, some physicians opt for "deliberate departures from customary practice to benefit their patients" [[2]].

These boundary-crossings often breach parameters that describe the mutually understood, unspoken, physical, emotional, role-related, time-related and physical limits of a fiduciary relationship between the trusting patient and the caring physician or provider [[3]–[5]]. The complex and personalized nature of physician, patient, sociocultural, clinical, contextual, psycho-emotional, existential and relational considerations see attempts at boundary-crossings occupy a range of practices including the sharing of "personal disclosure, explore common ground and shared interests, and share humorous exchanges" [[6]] to infringements on personal, moral, ethical and practice standards [[1]]. With unique and diverse considerations associated with each boundary-crossing, few physicians are adequately equipped to draw effective physician-patient boundaries. This may impact clinical objectivity, and result in moral distress, burnout and greater ethical and professional violations [[2]–[5], [7]].

By acknowledging that palliative care offers an ideal setting to examine boundary-crossings, a study to examine the lived experiences of these physicians and determine the nature, features and impact of boundary-crossings is proposed [[8]–[11]].

Informed by Gutheil and Gabbard [[12]], Vig and Foglia [[2]] differentiate boundary-crossings from egregious boundary-violations such as sexual misconduct and financial exploitation by characterising boundary-crossing as "harmless or benign, sometimes beneficial departure from customary or traditional clinical practice". Whilst this characterization is increasingly contentious, this definition is selected given its wide use in current practice.

To study boundary-crossings as a sociocultural construct, a holistic perspective is adopted. This includes consideration of the physician's regnant practice, social, cultural, familial, relational, existential, and clinical considerations (henceforth narratives); personal values, beliefs, and principles (henceforth belief system); clinical experiences, knowledge, skills and competencies (henceforth clinical insights), regnant practice, social, cultural, familial, relational, existential, and clinical considerations (henceforth contextual considerations); and moral, ethical compass, judgement and attitudes towards the employ of boundary-crossings [[8], [13]].

Interactions between the belief system and the physician's narratives, clinical insights and contextual considerations are also scrutinized on the premise that the physician's practice, conduct, thinking, and patient care are informed by their belief system. Therefore, appreciating how these various experiences influence the physician's belief system would improve the assessment, guidance of practice and direct timely, appropriate, personalized and longitudinal support to physicians. With recent reviews suggesting that interactions between the belief system and physician's narratives, clinical insights and contextual considerations may be mapped using the Radha Krishna and Alsuwaigh [[15]]'s Ring Theory of Personhood (RToP) [[13], [16]–[21]], we adopt the RToP to study current data on boundary-crossings.

Use of the RToP premises that a physician's belief system is informed by their concepts of identity and personhood. Thus, mapping changes in self-concepts of personhood would highlight shifts in the belief system. The RToP suggests that personhood is comprised of the Innate, Individual, Relational, and Societal domains [[22]]. The Innate Ring encompasses the physician's Divine connections and/or their genetic propensity of being human [[15]]. The Individual Ring encompasses their conscious function which are manifested in the physician's thinking, decision making, actions and conduct. The Relational Ring comprises important personal relationships while the Societal Ring consists of the physician's societal and professional obligations.

When new experiences, beliefs, values, principles, insights, and reflections are consistent with regnant belief system within the rings, there is 'resonance' [[17], [19], [21], [23]–[25]]. When the contents of the belief system are reprioritized to better fit with practical considerations - 'synchrony' is created. Conflict between current values, beliefs, and principles and those being introduced within one of the rings, creates 'disharmony' whilst 'dyssynchrony' occurs when conflicts occur between rings [[19], [23]]. Adaptations to resonance, synchrony, disharmony and dyssynchrony shape self-concepts of identity and personhood.

To determine, "how boundary-crossings impact palliative care physicians' practice and identities?", semi-structured interviews were proposed to capture the often complex ethical and moral deliberations that shape physician-patient interactions, their practice, and decision-making processes [[26]].

Adapting Krishna's Systematic Evidence Based Approach (SEBA), the two-staged Tool Design SEBA methodology (Fig. 1) was used [[28]–[32]]. In the first stage, a systematic scoping review was employed to guide the design of the tool. In the second stage,the data accrued from its application was analysed.

Graph: Fig. 1 The Tool Design SEBA Process

An expert team consisting of a medical librarian from the Yong Loo Lin School of Medicine (YLLSoM) and local educational experts and clinicians at the National Cancer Centre Singapore (NCCS), the Palliative Care Institute Liverpool, YLLSoM and Duke-NUS Medical School ensured a balanced, accountable and trustworthy evaluation of each stage.

Having posited that boundary-crossings are evolving and context-specific, the research and expert teams focused on the need to create, apply, and analyse the data accrued from evidence-based semi-structured interview questionnaires. A scoping review was thus conducted to better understand existing literature on boundary crossings in palliative care (Appendix A).

Data from the scoping review was then used to design the semi-structured interview questionnaire (Appendix B).

Ethics approval (2021/2176) was obtained from the SingHealth Centralised Institutional Review Board.

Eligible participants were palliative care physicians at the Division of Supportive and Palliative Care (DSPC) at the National Cancer Centre Singapore (NCCS).

The semi-structured interviews were conducted via video conferencing, with resident physicians, associate consultants, consultants and senior consultants at the Division of Supportive and Palliative Care (DSPC) at the National Cancer Centre Singapore (NCCS), between July and August 2021. Purposive sampling was conducted and email invitations containing participant information sheets, study information, and details on the nature, duration and aims of the audio-recorded interviews. The invitations stressed participants' anonymity in the audio-recorded semi-structured interviews and highlighted their right to withdraw from the study at any point and without prejudice. Two trained members of the research team (ASIL and MC) sought verbal and written consent before conducting the interviews which lasted about 45 minutes. These interviews took place in quiet offices to ensure privacy and facilitate an in-depth exploration of personal beliefs, experiences, and practices. Audio recordings were transcribed verbatim using the NVivo 12 Software, anonymised, and 'member checked'.

Braun and Clarke [[33]]'s approach to thematic analysis was employed to independently construct 'codes' from the 'surface' meaning of the transcripts. Overseen by the expert team, the 'codes' were discussed online and categorized into groups. A coding framework and code book was agreed upon using 'negotiated consensual validation'. The remaining transcripts were independently coded and categorized into sub-themes and themes.

Hsieh and Shannon's approach to directed content analysis was also employed. Chan et al. [[23]] and Kuek et al. [[19]]'s studies, "Extending the Ring Theory of Personhood to the Care of Dying Patients in Intensive Care Units" and "The Impact of Caring for Dying Patients in Intensive Care Units on a Physician's Personhood: A Systematic Scoping Review", provided the categories.

The Jigsaw Perspective employed Phases 4 to 6 of France et al. [[34]]'s adaptation of Noblit and Hare [[35]]'s seven phases of meta-ethnography [[36]] to combine overlapping areas within the themes and categories.

A total of 13 palliative physicians, labelled p1 to p13, comprising five senior consultants, three consultants and two associate consultants and three resident physicians, were interviewed.

Thematic analysis of the data revealed four themes: motivations, boundary-crossings, reflections and adaptations. Directed content analysis revealed three categories: belief system, catalysts for boundary-crossings and the impact of boundary-crossings along the RToP.

The Jigsaw Perspective gave rise to two final domains: (1) catalysts and (2) boundary-crossings.

Beyond traditional notions of boundary-crossings being confined to specific physician-patient relationships, catalysts can present as general threats to overall care provisions. Physicians p1, p3, p5, p7 and p9 breached personal and relational beliefs and responsibilities to sustain their palliative care involvement and their contribution to a field struggling with shortages (Table 1).

Table 1 Regnant Belief System and Catalysts through the Lens of the RToP

Recognizing a catalyst, determining its significance, balancing its potential risks and benefits of attempts to ameliorate this threat gives way to a decision to apply boundary-crossings. Boundary-crossings maybe adaptive or proactive (Table 2).

Table 2 Adaptive and Proactive Boundary-Crossings

Adaptive boundary-crossings or adaptations are changes to the belief system in response to the effects of catalyst. Proactive boundary-crossings occur when the physician acts pre-emptively to negate the effects of a catalyst.

The decision on whether adaptive or proactive boundary-crossings are to be employed is determined by the physician's self-awareness, clinical insights and reflections. The impact of boundary-crossings includes 'ring-fencing' an experience as a guide or warning against future boundary-crossings or similar situations. This is exemplified by Physician p5's 'ring-fencing' the loss of a patient with whom they shared close personal ties to as a guide for future conduct and boundaries for relationships with patients. Physicians p1, p6, p9 and p11, on the other hand, 'compartmentalized' their 'difficult' experience in order to continue their 'usual' practices, unencumbered by the need for changes to their practice.

Our research addresses several misconceptions. To begin, boundary-crossing are not limited to preserving physician-patient relationships but may be focused on the desire to sustain a career in palliative care. In the latter, boundary-crossings are deliberate departures from understood, unspoken, physical, emotional, role-related, time-related, and physical limits of a role as a palliative care physician. Familial, relational and personal roles, physical, emotional, existential considerations and 'working time' directives are compromised as a result. Unsurprisingly boundary-crossings can undermine clinical objectivity and result in moral distress, burnout and greater ethical and professional violations [[2]–[5], [7]].

Two, it is also unsurprising that data suggests that boundary-crossings are more common than previously believed and often unrecognized in its various forms even amongst experienced palliative care physicians.

Three, the desire to provide holistic care and nurture longitudinal relationships with patients, their families, and caregivers may propagate the perception of boundary-crossing as a 'benign' occurrence [[4], [37]].

Four, the notion that belief systems evolve with increasing 'boundary-crossing', precipitating a relaxing of guiding principles that confine practices and creating 'slippery slopes' to greater breaches in professional, ethical, personal, moral and legal standards is a concerning finding. Unobserved and unchecked, the possibility of these breaches increases sharply.

Five, the physician's individualized decision to employ boundary-crossings reflects their sense of identity and its influence on their practice. It is preservation of this identity that guides their 'sensitivity' to the presence of catalysts, informs their 'judgement' to determine the need to respond to a catalyst, informs their 'willingness' and ability to employ these boundary-crossings effectively to advance their objectives and their knowledge on how to 'balance' practical and personal considerations with their skills, abilities and experience to address the catalyst and sustain their existing belief systems [[28]].

These findings allow the proffering of the Krishna Model for boundary-crossings (Fig. 2). This model recognizes the physician's narrative, contextual considerations and clinical insights in shaping the belief system and the processes behind boundary-crossings. As a result, the belief system depicted by the RToP is encircled by three rings.

Graph: Fig. 2 The Krishna Model for Boundary Crossings

In this model, boundary-crossings are cast as a means to attenuate the effects of a catalyst. When a catalyst is detected, an internal system of decision-making involving the physician's 'sensitivity', 'willingness', 'judgement' and 'balance' is set in motion and informed by a continuous process of review or 'reflections-in-action'. This results in either a boundary-crossing or no response. If a catalyst is not detected until after its effects are felt or if the response is suboptimal, reflection-on-action ensues. Similarly, adoption of a boundary-crossings sees reflections-in-action initiated. The insights gained from reflections in both circumstances reshape the belief system (Fig. 3).

Graph: Fig. 3 Potential Effects of Boundary Crossings

It is here that its relevance to clinical practice becomes clear. The frequent presence of boundary-crossings and their impact upon a physician's belief system, professional identity and patient care underlines the need for access to longitudinal and accessible support. This will help them as they confront boundary-crossings and the closely related issues of moral distress and emotional, ethical and practical issues surrounding care of dying patients [[17], [28], [30]–[32], [39]]. This model also echoes concepts behind reflective practice and the teaching of empathy.

These insights underscore the need for longitudinal assessment-driven support of physicians. Consequently, it also underlines the potential for a RToP-based tool, designed to evaluate the physician's needs, coping and progress. Both considerations coalesce to highlight the need for a portfolio-based program as an additional source of reflection and to guide support and oversight of physicians [[40]–[46]].

Drawing upon our findings, we propose four measures to support physicians in navigating professional boundaries.

One, incoming junior physicians should be oriented to the concept, prevalence and ill-effects of boundary-crossings and be equipped to identify the risk factors and early signs of boundary violations.

Two, the Krishna Model may be used to facilitate faculty development programs on identifying 'at risk' physicians such as those identifying strongly with a saviour complex and differing goals of care which make them especially prone to boundary-crossings [[4], [47]]. In each case, faculty members should be able to evaluate the physician's level of self-awareness, willingness to seek help, ability to reprioritize and balance work-life commitments, delegate work to a wider team, and their ability to self-care. These evaluations must be accompanied by individualized assessments of the work environment, the availability and effectiveness of peer, counselling, spiritual, psychological, and psychiatric support and the potential to take 'time off' either for a leave of absence or to focus on non-patient facing work.

Such a comprehensive assessment and support mechanism for physicians may be structured around a portfolio-based program [[40], [46]] and supported by a RToP-based assessment tool. This portfolio-based program echoes Fronek et al. [[48]]'s calls for "a model of training that can meet multi-level learning needs of practitioners" that "is delivered along the professional life span[ing] and incorporate[ing] external reinforcement, inclusive of peer support and supervision opportunities". This portfolio-based approach must be supplemented by safe communication channels and a structured mentorship program within a psychologically 'safe' practice environment [[49]].

Three, faculty should be equipped to support physicians with distress tolerance skills such as compassion-based and mindfulness-centred interventions [[13], [51]], self-soothing relaxation techniques, guided imagery, biofeedback, exercise, and mindfulness [[47], [52]]. Faculty, too, should have access to assess, coordinate, and supplement their assessment and support with peer, counselling, spiritual, psychological, and psychiatric support programs.

Four, underpinning the efficacy and sustainability of the three recommendations made must be a host organization capable of designing and supporting effective recruitment and training of physicians and faculty, creating, and overseeing the portfolio-based approach and nurturing a supportive and secure work environment.

This study is not without its limitations.

Perhaps prime amongst these is the attempt at establishing causality or linking both events to adaptations to belief systems and sustained changes in professional identities. Neither can be accurately made even amongst senior palliative care physicians within a single tertiary healthcare cluster. Multisite studies across the Southeast Asian context and beyond could provide greater insights and enhance the applicability of these findings.

Other limitations include the use of purposive sampling which draws attention to physicians who already are aware of boundary-crossings and their effects. In addition, although analysis of the interviews was carried out by two investigators through triangulation of data, the omission of member checking is a concern. Additionally, whilst inspired by studies on dignity, professional identity formation, moral distress, empathy, and caring for the dying, use of the RToP to guide the analysis has not been evidenced and should be evaluated further.

Reframing boundary-crossings invites scrutiny of the often complex psychosocial, financial, practical, cultural and relational settings that occurs within palliative care. This is especially so in Confucian-inspired societies such as China, Singapore, Hong Kong, Taiwan, Korea, and Japan [[8]–[10], [54]–[57]] where the role of the physician is seen as part of the family unit inviting boundary-crossings and highlighting its sociocultural roots. Here focus for future studies should be upon the design and employ of a RToP-based tool [[13], [16], [23], [50], [58]] and on the experiences of all members of the palliative care multidisciplinary team and healthcare professionals in fields such as intensive care and specialities involved in the treatment of chronic illnesses.

The authors would like to dedicate this paper to the late Dr. S Radha Krishna whose advice and ideas were integral to the success of this study, the late A/Prof Cynthia Goh whose mentoring and insights guided the analysis of this data and Thondi and Maia Akhlil whose short lives continue to inspire us to better support and care for others.

The authors would like to thank Alexia Sze Inn Lee (ASIL) for her help with the interviews and Drs Stephen Mason and Ruaraidh Hill for their feedback and guidance on the analysis and in advancing the SEBA methodology. The authors would also like to thank the anonymous reviewers whose advice and feedback greatly improved this manuscript.

CYH, NL, NDAR, MC, JXZ, GLGP, EKO, CL, ARC and LKRK were involved in research design and planning, investigation, analysis, reflection, manuscript writing and review, and administrative work for journal submission.

CYH, NL, NDAR, MC, JXZ, GLGP, EKO, CL, ARC and LKRK were involved in reviewing and editing the manuscript.

All authors have read and approved the manuscript for submission.

This work was supported by the 2021 SingHealth Medical Student Talent Development Awards – Project, Singapore. The funding body played no role in the design of the study and collection, analysis, interpretation of data, and in writing the manuscript.

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

All experimental protocols were approved by the SingHealth Combined Institutional Review Board [Ethics approval (2021/2176)]. All methods were carried out in accordance with relevant guidelines and regulations. Informed consent was obtained from all participants.

NA.

All authors declare that they have no competing interests.

• DSPC

• Division of Supportive and Palliative Care (of NCCS)

• NCCS

• National Cancer Centre Singapore

• RToP

• Ring Theory of Personhood

• SEBA

• Systematic Evidence Based Approach

Below is the link to the electronic supplementary material.

Graph: Appendix B. Interview Questions

Graph: Appendix A. SEBA guided in Scoping Reviews

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.