Fighting Fires and Battling the Clock: Advance Care Planning in Family Medicine Residency

Family doctors are ideally positioned to initiate advance care planning (ACP) to help patients with serious illness ensure that the medical care they receive is concordant with their values and wishes. ¹ Patients with serious illness engaging in ACP with their doctors benefit from better knowledge of disease, ² mood, ^{3, 4} positive family outcomes, ^{4, 5} documented goals, ^{4, 5} and shared decision-making, ⁶ and may benefit from care that is more concordant with their goals ^{5, 7} and location preferences ⁸; additionally, they experience less decisional conflict, ⁹ nonbeneficial medical care near death, and lower overall health care expenditure. ^{3-5, 8, 10, 11} Unfortunately, while family physicians play the most central role in assessing and treating patients at the end of life, ¹² fewer than one in 10 patients have conversations about their values and care preferences with their family physician, ¹³ despite an interest from patients in initiation of such conversations. ^{14, 15}

Physicians attempting to engage their patients in ACP encounter a multitude of logistical, cultural, and educational resistive forces. Logistical barriers, such as insufficient time and poor electronic medical information transfer, appear to impede physicians’ ability to have frequent, iterative, and well-informed ACP conversations. ^{8, 11, 16-18} Social barriers, including a perceived aversion to death and dying from both patients and providers, also appear to play an important role. ^{11, 19, 20, 21} Physicians have reported low confidence in their ability to overcome these barriers. ²² Recent efforts to address this confidence gap have focused primarily on the use of structured communication guides ^{23, 24} to facilitate ACP. ²⁵ But this combination of communication guides and workplace education have found limited success. ^{3, 26} This lack of success might be explained by cultural resistance hampering effective implementation of such interventions outside of the education research setting. ^{6, 18, 22, 27} More recently, though, a new line of ACP scholarship has suggested that this incomplete traction for current communication guides may stem from lack of scholarly attention to innovations that support learners to address illness understanding and elicit patient values in ACP rather than focus purely on having better goals of care (GOC) decisions about potential interventions. ²⁸

Rethinking advance care planning would benefit from a deeper understanding of workplace practice. Despite low clinician confidence and engagement with ACP, few studies have deliberately examined how postgraduates approach ACP during workplace-based learning. ^{11, 29, 30} Despite formal expectations that residents graduate with individual competence in facilitating advance care planning, ³¹ family medicine researchers remain unclear on how residents approach ACP conversations in clinical encounters, experience role-modeling of ACP, or encounter barriers to their learning and skill development. ^{30, 32} Future efforts at systemic improvement of ACP face risk of continual failure if proceeding with ongoing attempts at intervention without deep understanding of the cultural and educational issues at play. ^{28, 33}

To meet this call, our research team deeply explored how family medicine residents learn to approach advance care planning during their residencies. We focused on eliciting the social processes family medicine residents encounter in their efforts in the hope that a rich understanding of the sociocultural aspects of this problem can inform future research, teaching, and curriculum development.

We designed this study of social processes based on a constructivist-grounded theory approach and aspects of autoethnography to frame the analysis. ^34-36 We recruited from a large urban site (80-90 residents) comprised of two academic family medicine clinics (38 residents each) and a few smaller distributed clinics from a single university in Ontario, Canada. The study was approved by the Hamilton Integrated Research Ethics Board (HiREB #8320).

The data included in the study consisted of nine semistructured interviews conducted from August 2020 to March 2021 with a convenience sample of family medicine residents (three year-1 and six year-2 postgraduates); reflective memos that focused on the interviewer’s experience of receiving each interview; 37 autoethnographic field notes; and member-checking feedback from five of the nine participants. The interviews were semistructured in nature, telephone-based, 35.6 minutes in mean duration, audiorecorded, and transcribed verbatim. We used NVivo (Lumivero) qualitative analysis software to facilitate coding. We began with collecting three interviews and then analyzing the interviews, memos, and field notes using sentence-by-sentence open coding to create a preliminary framework that we then refined as an analytic team. We then collected and analyzed three more interviews using focused coding to explore gaps in the initial analysis. The primary author (T.A.) then synthesized two composite narratives ³⁷ of participant experiences using the detailed open-coding categories and emergent focused-coding categories. Finally, we ended data collection after three additional interviews exploring lingering questions in the analysis through theoretical coding. Our approach to the completion of data collection and analysis relied on saturation of the theoretical framework to the satisfaction of the research team and member-checking feedback ³⁸ (five of the nine participants responded) to ensure participants’ lived experience had been richly and sufficiently represented. ^{35, 39, 40}

Our team consists of a departmental research associate (O.V.); a family doctor and palliative care specialist working at one of the academic clinics in the study (E.G.); the senior author, who is a palliative care scientist (M.H.); and the lead author, who is a medical education researcher and was a resident physician in one of the academic clinics included in the study (T.A.). Our analysis is based on the constructivist premise that the findings of the study are inextricable from the perspective of the researchers. ⁴¹ So, rather than focusing on generalizability and reproducibility of analysis and coding, we chose an autoethnography-informed approach to grounded theory that uses the narrative influence of the authorship team to crystalize the lived experience of the participants through their stories. ^42-44 The lead author’s lived experience as a resident in the program under study inextricably shaped the content and tone of the analysis. Through stepwise, repeated, and thorough collaborative analysis, we brought the analytic perspectives of the research team and the lived experience of the participants themselves to the data to ensure that the narratives resonated with the participants’ experiences of advance care planning. This constructivist approach to rigorous qualitative inquiry focuses on ensuring credibility through richly detailed storytelling and situational resonance of the findings rather than on positivist criteria of rigor such as generalizability and reproducibility. ^{45, 46}

Our findings detail the sociocultural and logistical complexity family medicine residents face when they attempt to engage patients in advance care planning. At the core of our findings lie two composite narratives depicting a process our participants characterized primarily using the language of struggle and conflict. While the narratives stand as the primary analytic contribution, ⁴⁷ in support of the narratives, we also have included an analytic description of the coding structure as well as additional representative quotes as supplemental material.

The composite narratives tell the story of learners struggling to initiate advance care planning conversations and encountering cultural and logistical obstacles. In the first narrative, about Phillip and Anna (Figure 1), residency training-specific logistical barriers act as rate-limiting factors, including lack of control over schedules, truncated relationships with patients, lack of deliberate practice, and fractured service blocks. In the second narrative, about Suzanna and Robert (Figure 2), unspoken cultural norms regarding the care of patients with life-limiting illness, such as supervisory aversion to prognosis, lack of autonomy in practice, and lack of supervisory encouragement, make prioritizing advance care planning over other aspects of patient care difficult for residents.

As depicted in the narratives, our participants described six unique social processes in approaching ACP: (a) feeling underprepared and cautious; (b) confronting overwhelming odds; (c) asking for more guided practice; (d) developing patience, comfort, and confidence; (e) acknowledging complexity; and (f) jumping in. These processes are not linear steps; instead, they are acts residents participated in as they navigated the sociocultural complexity of treating patients with serious illness. Additional representative quotations are available as supplemental material. Subscripted numbers following quotations denote the deidentified participant number.

The primary underlying vocabulary that emerged in our analysis was one of struggle and conflict. Residents described feeling underprepared and cautious as a function of a “lack of training”_(R2) in how to have ACP conversations and a lack of clinical knowledge around “navigating some of the questions that might come with an ACP conversation . . . [about] prognosis if they do have a terminal illness.”_(R4) They described confronting overwhelming odds stacked against their efforts to initiate ACP conversations. Residents described fighting to start ACP conversations due to multiple logistical barriers, including “battling the clock,”_(R5) “lack of information in the chart, lack of ability to actually find any dictations,”_(R2) and competing clinical problems that render their clinic days “more firefighting than anything.”_(R5) Residents also described complex psychosocial and cultural barriers such as “aversion to consider[ing] one’s own mortality,”_(R1) “the feeling that these aren’t my patients . . . and [that] you’re just parachuting into someone’s care”_(R6) while “the [attending physician] might approach it differently”_(R4) than they themselves hoped to.

Our participants found themselves asking for more guided practice where a preceptor looks “at their patient roster and then identifies people that would benefit from having some time to have one of these conversations, takes the initiative to book them an appointment with a resident . . . [and for] clinics to encourage their clinicians to do that or to build that into the structure of the way that appointments are scheduled.”_(R1) Despite their struggles, some residents found roundabout methods of developing patience, comfort, and confidence by using off-service acute care rotations where patients present to “hospital for symptom management and were having this conversation . . . [because] in the hospital there’s more flexibility”_(R5) than in the family medicine clinic where “it just feels like a harder conversation.”_(R5)

Residents found themselves acknowledging complexity in ACP, which can feel “hypothetical”_(R4) and dynamic because “people’s circumstances change, they change their opinion, or their disease trajectory changes . . . [and] the outcome of different interventions changes”_(R2) ensuring that “it’s not always possible”_(R5) to clearly delineate care plans ahead of actual decisions. And, despite all that, our participants found ways to jump in and try to “be proactive about it, be the primary educator for patients: plant those seeds so that if and when . . . they’re faced with life and death, those [previously made] decisions are followed through”_(R5) despite obstacles encountered.

The findings of our exploratory qualitative grounded theory study support previous research identifying structural barriers to advance care planning in health care settings. Quantitative survey-based, observational, and experimental research has long shown that doctors struggle to engage in ACP because they do not have enough time in their days, enough flexibility in their schedules, or enough information transfer between doctors. ^{8, 16, 17, 18} By looking closely at the way residents are taught to approach care of patients with life-limiting illness, we also identified avoidant cultural norms that lead to hesitance toward ACP from the earliest days of family medicine training.

All of our participants talked broadly and at length about their attempts to overcome their own resistance and personal aversions to advance care planning in inflexible clinical and pedagogical environments. Such barriers are not unique to resident doctors. For example, previous research on practicing clinicians identified cognitive barriers to initiating ACP such as difficulty finding the right moment, trouble with emotions related to ACP, and frustration with care plans that are too simple for complex medical issues. ¹⁸

Our findings suggest that resident physicians encounter an additional set of barriers unique to their roles as postgraduate medical learners. Residents battled the clinical priorities of their preceptors, their lack of authority to shape clinic flow, their limited exposure to deliberate practice, and a fear that the ACP conversation was not theirs to have. Ultimately, these unique logistical barriers conspired together to create an avoidant approach to ACP. Pedagogical focus on fighting the crisis of the day in the form of adjustments to symptom management or lab results led to an absence of support for examining the overall big picture, mapping the trajectory of the disease, or planning for the future.

Residents collided with cultural avoidance of advance care planning. Many physicians in training are introduced to the reality that metastatic cancer (with a small number of exceptions), late-stage organ failure, frailty, and dementia are progressive life-limiting diagnoses. ^{48, 49} In the moments where our participants expressed their hope to talk with patients about their illnesses, they found themselves guided by preceptors toward avoidance, other priorities, or deferral to other physicians.

This culture of avoidance appeared to create a gap in understanding about the purpose of advance care planning. Because many of our participants were discouraged from engaging their patients in ACP in the primary care setting, they shifted their focus to off-service acute care settings to hone these skills. With practice and supervision, they learned how to talk with sick patients in a moment of crisis about their medical options, such as invasive resuscitation or consent for life-prolonging procedures. For example, rather than learning how to slowly, gently, and iteratively discuss the downward trajectory of heart failure, they reported learning how to dissuade patients from invasive and nonbeneficial treatments at the time of admission to hospital.

Astutely, some of our participants identified the difference between such acute goals of care conversations and ACP conversations best suited to primary care when a patient is chronically ill. But they did so only insofar as to say that they missed opportunities to learn how to approach ACP in the primary care setting. This educational failure may play a role in creating the poor understanding in practicing clinicians ²⁸ of the difference between GOC conversations at the time of a clinical decision and ACP conversations. ^{28, 50} Given low clinician confidence in ACP, clarifying the differences between ACP and GOC conversations—such as those about code status, for example—may benefit both residents and practicing clinicians. Addressing this confusion may help improve clinical teaching around ACP and rectify the low rates of ACP in clinical practice. ^{13, 50, 51, 52}

Family medicine residents are uniquely positioned to learn how to iteratively discuss patients’ personal values, wishes, and understanding of their illnesses. ¹¹ Resolving the unfortunate avoidance of advance care planning at the core of postgraduate family medicine training will require sustained attention, research, and education. Previous primary care postgraduate education research has examined didactic teaching, chart audit, reflection, simulated patient interactions, and role play. ^{32, 53, 54} Future family medicine researchers may wish to more robustly test novel educational interventions like flexible clinic schedules, dedicated ACP time, ⁵⁵ and deliberate observed practice as in keeping with the principles of workplace-based learning inherent to competency-based medical education. ⁵⁶

Most importantly, family medicine educators may wish to consider directly teaching residents and preceptors about the crucial differences between goals of care conversations focused on interventions and ACP focused on values and illness understanding. When offered, patients with serious illness tend to accept and appreciate conversations about the progressive course of their disease. ^57-60 Teaching residents that such ACP is most effective when iteratively and longitudinally focused on values and illness understanding may both address this patient desire and also help resolve the cultural avoidance that limits residents’ attempts to engage in ACP. ^{28, 61}

Constructivist qualitative research seeks to describe complex social experiences that resonate with educators. This methodological approach inherently seeks out tensions and sociocultural complexity rather than claiming with certainty what can be generalized to all learners across different programs, places, and preceptors. ^{45, 46} As such, the limitations of this work are not related to sample size (9 interviews, 9 memos, 37 fieldnotes, and 5 member-checking letters) or absence of triangulation in coding but rather are related to the inherent tension between the space and voice limitations in scholarly publication and transparency of autoethnographic reflection. Given that these findings are highly situated in both their local context and in the subjective experience of the authorship team, we recommend that family medicine educators use the narratives to stimulate conversations in their local setting and design future research about advance care planning rather than use this research as a singular basis for decision-making about curriculum design or learner assessment.

In our study, we found that family medicine residents struggled in their attempts to engage their patients in advance care planning. Logistical barriers specific to postgraduate medical education and pedagogical avoidance of the natural history of life-limiting disease led residents to conflate ACP focused on values and illness understanding with goals of care conversations focused on specific interventions. Testing educational interventions specific to ACP and targeted education around the culture of avoidance of death and prognosis may be required before significant improvements in ACP in family medicine will be realized.