Introduction

Interprofessional education (IPE) is a critical component of medical education that prepares trainees to work effectively in healthcare teams. Academic primary care clinics are sites where robust interprofessional coordination is needed to deliver high quality care; they are also settings where much fundamental medical student and resident education takes place [1]. However, clinical teams might not demonstrate successful interprofessional collaborative practice (IPCP) by virtue of not employing the habits of communication and teamwork that are taught to trainees [2]. These conditions are partly due to social and cultural divisions between professions that hinder IPCP and the potentially harmful hierarchies of power that compound these divisions [3,4,5]. The resultant lack of formative interprofessional modelling for interprofessional students and residents has resulted in calls to enhance teamwork training in workplace settings [2, 6].

Narrative approaches to clinical care, with their focus on allowing space for the diverse voices on interprofessional teams, have the potential to improve interprofessional education and collaborative practice (IPECP) [7]. Narrative medicine, a branch of health humanities that provides training to practice medicine with ‘narrative skills of recognizing, absorbing, interpreting, and being moved to action by the stories of illness’ can be an innovative approach for improving IPECP [8]. Narrative medicine is a discipline that deploys knowledge and methods from humanities and the arts to improve the effectiveness of healthcare. By probing the meanings of oral, written, or visual narratives, narrative medicine equips learners with interpretive and interpersonal skills to recognize the tellers of stories. The methods used include working with groups to: 1) engage with creative works to develop skills of attention, 2) write reflectively to discover and explore personal narratives, and 3) share writing with others to strengthen relationships [8]. Narrative medicine programs have been successfully implemented in educational settings [9]; however, there are few reports of such programs in clinical settings. Furthermore, these programs have been either relatively short-term in duration, not interprofessional, or not situated in primary care settings [10, 11].

We implemented a narrative medicine program in academic primary care clinics with the long-term goals of improving interprofessional communication and relationships. In this report, we focus on: 1) description of the program structure, 2) barriers and catalysts to program implementation, and 3) process-oriented program evaluation.

Methods

Program setting and participants

We implemented a 1-year (April 2016 to March 2017) narrative medicine program in three New York Presbyterian Hospital academic primary care clinics that serve as training sites for medical students and resident physicians in family medicine, internal medicine and paediatrics.

We met with clinic leaders before implementation to discuss how narrative medicine might improve teamwork; given shared interest in team development, leaders allotted 30 minutes during monthly required interprofessional team meetings for narrative medicine sessions. In this paper, ‘staff’ refers to attending physicians, resident physicians, nurses, medical assistants, social workers, administrators, and many other clinic personnel. Staff were required to attend the narrative medicine sessions but could opt out of program evaluation. The program evaluation was approved by the Institutional Review Board at the Columbia University Irving Medical Center (protocol # AAAP1351). The work was carried out in accordance with the Declaration of Helsinki including, but not limited to, there being no potential harm to participants, that the anonymity of participants was guaranteed, and that informed consent of participants was obtained.

Session description

Each 30-minute session consisted of four stages: engaging with a creative work, writing to a prompt, sharing writing in pairs, and sharing writing in a large group [12]. Table 1 details the activities and functions of each stage. Co-author DG facilitated the majority of the sessions. Facilitators of narrative medicine sessions require skills in textual interpretation, group facilitation and application of specific methodology; all program facilitators had masters-level training in narrative medicine. We maintained a consistent structure for each session, but adjusted timing to meet clinical needs. We also adapted creative works used based on expressed staff interest (e. g. a Spanish-language song used as a creative work in clinics staffed by native Spanish speakers).

Table 1 The four stages of the narrative medicine session
Full size table

Program evaluation

Evaluation data were derived from: 1) observation notes from narrative medicine sessions by either TC, FJ, or AK; 2) longitudinal semi-structured individual interviews with interprofessional staff from each clinical site at program start, midpoint, and end conducted by either DB, UD or MM (to allow open and critical dialogue, session facilitator DG did not conduct interviews); and 3) 12-question questionnaire exploring barriers and catalysts to participation, and staff’s overall reactions to the program.

Qualitative data analysis

We used a general inductive approach for qualitative analysis, which asks, ‘What are the core meanings evident in the text, relevant to evaluation objectives?’ [13,14,15]. DG and AK led the data analysis. They partnered with the team to create codes (words acting as labels for core concepts). They applied codes to data from interview transcripts and observation notes, and managed data in Dedoose.com. At weekly team meetings, DG and AK reviewed the codes to establish links between the data and evaluation objectives. Emergent findings informed iterative interview guide development and real-time program improvement. Given the evaluation focus and the rich and varied sources of data, the team deemed the data sufficient to address the evaluation objectives. Quantitative data from the questionnaire were descriptively analyzed.

Results

Over a 1-year period, 36 narrative medicine sessions took place at three clinics. We analyzed 32 observation notes and 33 interview transcripts. Fifty out of the 104 unique staff members who attended the sessions during the year completed an end-of-program questionnaire. The questionnaire respondents across sites included 14% attending physicians, 16% resident physicians (representing all 3 years of training), 26% nurses and nurse practitioners, 16% medical assistants, and 28% other staff. Medical students were not involved in program evaluation. Age was reported by 45 respondents; 47% of respondents were less than 40-years-old and 53% were 40-years-old and over. Average duration of employment was 7.3 years (median 6 years and range 4 months to 23 years).

Observation notes and attendance sheets indicated that sessions were attended by an average of 11 staff (82% women and 18% men). Session composition was 21% attending physicians, 9% resident physicians (representing all 3 years of training), 11% nurses and nurse practitioners, 12% medical assistants, and 47% other staff. One or two medical students on their primary care clerkship participated at some sessions at the family medicine site. Fifty-six percent (28/50) of staff surveyed at program conclusion reported they had attended over half the offered sessions.

Engagement with narrative methods

We observed that staff participated actively during each of the four stages of the narrative medicine sessions. While exploring creative works (Stage 1), staff actively commented on works and listened and responded to others’ observations. Staff remained open to differing viewpoints on creative works. When asked to write to a prompt (Stage 2), staff wrote attentively and without hesitation. Many continued writing, even after a call was made to stop and reconstitute the larger group. When sharing writing in pairs (Stage 3), staff members sat closely and leaned in to listen to one another. We commonly observed laughter and sometimes witnessed expressions of sadness during sharing of writing. During Stage 4, staff routinely volunteered to read their writing to the group, and other staff listened closely and responded to the reading. Staff sometimes stayed after the conclusion of the session to continue conversations.

Acceptance of program

Based on questionnaires, 94% of staff (47/50) indicated that they would recommend the narrative medicine program to other clinics and 74% (37/50) of staff reported they would continue participating in the program even if it were optional and offered during their free time (before/after clinic or during lunch).

One program participant recalled that at program start, some were hesitant to ‘share personal information,’ particularly with staff from different professions. Despite this initial hesitation, staff actively engaged in the sessions and explored personal stories that often invoked strong emotional responses from both reader and listener (e. g. laughter, tears). This occurred during both pair-sharing and group-sharing; as the year-long program progressed, sharing of personal stories sometimes occurred even earlier in the session during discussion of the creative work.

We observed that the participation in discussion and sharing during sessions was distributed evenly across professions and hierarchy. Several staff members stated that one of the most striking aspects of the program was the communication that opened up across professions and levels of hierarchy that had not occurred with other prior team-building activities at the clinic. Indeed, 40% (20/50) of respondents rated the program’s ability to facilitate interprofessional dialogue as excellent, 26% (13/50) rated it as very good, and 24% (12/50) rated it as good. One staff member said, ‘The best thing I think about narrative medicine … is that it really involved everyone.’ In an interview, a staff member noted that the conversations that occurred at the sessions allowed staff to ‘take our hats off’ and ‘really get to know one another on a different level’.

Barriers and catalysts to participation

Staff most frequently expressed time constraints and the pull of clinical duties as barriers to program success. A staff member said, ‘I think the activity is great. It’s just, where do we carve out the time that doesn’t affect access for patients?’ Thus, buy-in and support from clinic leaders appeared crucial to program success. Support took the form of protecting time for the meetings, arranging coverage for clinical duties, and clinical leaders’ attendance at sessions. At one clinic, leaders even delayed clinic start time by 30 min with a corresponding reduction in patient load to accommodate the sessions.

Discussion and implications

This process-oriented program evaluation revealed that a year-long narrative medicine program with modest monthly exposure was successfully implemented in academic clinical settings, as evidenced by staff member engagement and acceptability. Staff members expressed positive regard for the program and supported its ongoing implementation.

Although attendance was required, staff were engaged during sessions and quickly took part in activities. Narrative medicine methods are designed to encourage exploration of one’s personal and professional life. Thus, emotions such as joy and sadness were explored and openly expressed during sessions, in keeping with a recognition that the expression of emotions may be important for clinical care and IPCP [16, 17]. Such expressions occurring earlier in sessions as the program progressed suggests increasing comfort with narrative medicine methods, increasing levels of trust with one another, or both. A prior narrative medicine report argues that safety in groups improves trust and creativity [18]. Indeed, psychological safety in medical and educational contexts may facilitate core behaviours like asking for help or reporting mistakes [19].

In interviews, staff felt that personal connections made across professions and levels of hierarchy were particularly striking. One staff member said that the sessions allowed ‘the whole care team’ to ‘just talk to each other like human beings.’ This points to the potential of art and literature to serve as a democratizing space where persons can gather to explore life experiences as relative equals [7]. Indeed, this effect of narrative medicine may have great value in improving collaborative practice.

We report a narrative medicine program that serves as a feasible model for implementing health humanities in clinical settings. The program’s feasibility was reliant on support from clinic leaders and the availability of a facilitator trained in narrative medicine. Sustainability was not an objective of the project, but it is notable that only the paediatrics clinic, led by a physician with narrative medicine experience, continued monthly narrative medicine activities for an additional year. Staff in that clinic expressed interest in further narrative medicine training. Future narrative medicine interventions will attend to staff training to achieve sustainability. Short, intensive facilitator training programs are under development by the program team to maintain quality while allowing wider dissemination.

Narrative medicine has the potential to improve IPCP by strengthening interprofessional communication and relationships while softening clinical hierarchies; this, in turn, could enhance IPE through positive modelling of interprofessional teamwork. With feasibility and acceptability established in this report, the team’s next steps will focus on more rigorous evaluation of narrative medicine’s impact on outcomes relevant to IPCP such as communication, hierarchy, cohesion, and clinically situated behaviours, in addition to the impact on trainees’ behaviours and attitudes. A limitation of this report is that the number and duration of sessions needed to achieve and sustain benefits of narrative medicine programs remains unknown; this will also be investigated in future work. Finally, research in narrative medicine will increasingly strive to measure patient-oriented outcomes such as patient satisfaction, clinical processes, and clinical outcomes.