If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence to: Dr. David Shih Wu, MD, FAAHPM, Palliative Care Program, Division of General Internal Medicine Johns Hopkins Bayview Medical Center/Johns Hopkins University School of Medicine, 5200 Eastern Avenue, Mason F. Lord Building, Suite 2300 Baltimore, MD 21224.
Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USAHealth Policy and Management, Bloomberg School of Public Health, Baltimore, Maryland, USA
Innovative patient-centered approaches to goals of care (GOC) communication training are needed. Teaching a narrative approach, centered on the patient's unique story, is conceptually sound but has not been evaluated with respect to objective skills attainment. We developed a curriculum based on a novel, easily-remembered narrative approach to GOC, the 3-Act Model, and piloted it with a cohort of internal medicine (IM) interns.
Objectives
To describe the development of the 3-Act Model curriculum and to assess its impact on the GOC communication skills of IM interns.
Methods
The curriculum was developed with input from multidisciplinary experts, IM residents, and patient/family representative. Notable elements included instrument development with validity evidence established, determination of proficiency standards, and creation of role-play scenarios. In two three-hour workshops, interns participated in role-plays as both providers and patients, before and after teaching (which included narrative reflection, didactics, and video demonstration).
Results
22 interns played the role of provider in five unique scenarios; 106 proficiency ratings were analyzable. Interns objectively rated as proficient increased from 30% (pretest) to 100% (final role-play). By the end of the training, 96% of interns strongly agreed or agreed that they felt ready to independently lead basic GOC discussions and the percentage who strongly agreed increased with successive role-plays. All interns indicated they would recommend the training.
Conclusion
This pilot demonstrates that the 3-Act Model is teachable and appreciated by learners. This GOC curriculum is the first based on a narrative approach to demonstrate objective skills improvement.
This article describes the development of a curriculum to teach a novel narrative-based approach to goals of care discussions, the 3-Act Model, and its impact on the communication skills of a cohort of interns. The results indicate that the curriculum led to objective skills improvement and was appreciated by learners.
Introduction
A goals of care (GOC) discussion explores a seriously ill patient's values, beliefs, and priorities in the context of the clinical situation, with the aim of providing goal concordant care.
Negative consequences of these deficiencies include worse quality of life, burdensome health care utilization, and distress for patients and their families.
GOC discussions may be challenging and uncomfortable for health care providers (including resident physicians), who often lack the experience, knowledge, skills, and confidence to conduct them effectively.
“I just felt like I was stuck in the middle”: physician assistants' experiences communicating with terminally ill patients and their families in the acute care setting.
The effect of communication skills training for generalist palliative care providers on patient-reported outcomes and clinican behaviors: a systematic review and meta-analysis.
The effect of communication skills training for generalist palliative care providers on patient-reported outcomes and clinican behaviors: a systematic review and meta-analysis.
Effect of a patient and clinician communication-priming intervention on patient-reported goals-of-care discussions between patients with serious illness and clinicians: a randomized clinical trial.
Evaluating an intervention to improve communication between oncology clinicians and patients with life-limiting cancer: a cluster randomized clinical trial of the Serious Illness Care Program.
However, research on using a narrative approach to GOC is limited to the COMFORT (Communication, Orientation and options, Mindful communication, Family, Openings, Relating, Team) curriculum,
and thus may be complex to learn. The COMFORT curriculum has not been objectively evaluated with respect to its effects on communication skills or behaviors,
underscoring the need for innovative methods. We developed a novel method for teaching GOC discussions based on an easily-remembered, patient-centered narrative approach: the 3-Act Model. Based on our needs assessment, we learned that most internal medicine (IM) interns had not received structured GOC communication skills training. The primary aims of this pilot study are 1) to describe the development of the 3-Act Model curriculum and 2) to assess its impact on the GOC communication skills of IM interns, through both objective and self-assessment.
Methods
Curricular Development and Content
The 3-Act Model consists of 1) understanding patient's story, including perspective of the medical situation within the larger personal narrative with sensitivity to language and values; 2) discussing medical opinion in simple language and big picture terms; and 3) making shared decisions within context of patient's story, language, values, and receptivity. Other key elements are the prologue (preparing beforehand, including clarifying issues and options, and establishing rapport) and the epilogue (concluding the meeting and coordinating afterward). Of note, we teach this as a framework that can, and should, flex to the situation in real time. One of the authors (D. S. W.) conceptualized this approach in 2013, inspired mainly by the teaching of a masterful mentor, Dr. Stuart J. Farber. Subsequently, our team refined curricular content and developed educational methods under the auspices of the Johns Hopkins Longitudinal Program in Curriculum Development (CD).
—and serial meetings with multidisciplinary experts (including palliative care physicians, social worker [SW], and chaplain, as well as internists specializing in medical education).We also incorporated input from members of our institution's Patient and Family Advisory Council (PFAC) and IM residents.
Role-plays of GOC discussions were a core part of the curriculum. Ten unique clinical scenarios were written for the role-plays (see Appendix for one example), and each was rated by a clinician panel (palliative care physicians and SW and one IM PGY-2 resident) to be either of low, medium, or high difficulty. Scenarios were revised until all had an average rating of low or medium difficulty, to reflect the consensus opinion on what rating would constitute a “basic-level” GOC discussion.
Instrument Development
Our CD team developed a novel rubric for assessing GOC discussions, incorporating feedback from multidisciplinary colleagues (palliative care attending physicians and SW, plus multiple experts in medical education and CD) and a member of the PFAC. The product of this iterative process was the Goals of Care Communication Assessment Tool (GCAT), which included 16 items grouped into five domains congruent with the 3-Act Model: Prologue, Acts I-III, and Epilogue (Table 1). Each item pertained to a task/behavior and could be scored from 0 to 2 points (“not done,” “needs improvement,” or “effective”). With support from the Johns Hopkins Office of Assessment and Evaluation, we established internal structure validity evidence for the rubric, as below.
Table 1Factor Analysis of GCAT Items and Percent Variance Contributing to Overall Score
Rotated component matrix: Rotation was converged in six iterations. Extraction method was principal component analysis. Rotation method was Varimax with Kaiser normalization.
Items
Component (Factor)
1
2
3
4
Act 3: Facilitated care aligned with patient/family values
0.807
Act 3: Adapted discussion to the needs of patient/family
0.672
0.423
Act 3: Effectively involved patient/family in decision-making process
0.666
Act 2: Tailored content to the needs of patient/family
0.613
Act 3: Centered discussion of options on patient's story/values/goals
0.595
0.455
Act 1: Elicited patient's story (values, goals, patient as a person)
0.706
Act 1: Acknowledged emotions/values shared by patient/family
0.689
Act 1: Listened attentively to patient/family (silence, reflective questions)
0.601
Act 2: Focused on big picture (broad overview of medical condition)
0.476
0.447
Act 2: Used terms patient/family could understand (avoided medical jargon)
0.451
0.426
Prologue: Established rapport (introductions, bedside manner)
0.802
Prologue: Discussed purpose of meeting
0.655
Act 1: Elicited patient/family perspective of condition early in meeting
0.409
0.594
Act 2: Encouraged patient/family to ask questions
0.342
0.742
Epilogue: Ensured patient/family understanding throughout
0.729
Epilogue: Summarized discussion and next steps
0.582
Eigenvalue
4.66
1.68
1.42
1.11
% of variance
29.12
10.49
8.87
6.96
a Rotated component matrix: Rotation was converged in six iterations. Extraction method was principal component analysis. Rotation method was Varimax with Kaiser normalization.
Before curricular implementation, a panel (palliative care attending physicians and SW and one PFAC member) formulated the proficiency standard for trainer GCAT ratings, through an approach inspired by the dominant profile judgment method.
Each panelist individually identified and wrote a policy regarding what score would reflect the barely proficient learner. These policies were shared within the group. In discussion, each panelist explained the rationale for his/her policy, and then the group came to a consensus on the standards. The panel decided on a noncompensatory model, meaning that for any given role-play, the learner in the provider role must meet the targets for all five domains, as rated by the trainer, to pass as proficient. The proficiency standards are shown below:
1.
Prologue: 2 out of possible 4 points
2.
Act I: 6 out of possible 8 points
3.
Act II: 5 out of possible 8 points
4.
Act III: 5 out of possible 8 points
5.
Epilogue: 2 out of possible 4 points
Before any of the sessions with learners, trainers (palliative care physicians, SW, and chaplain, plus one IM PGY-2 resident and one PFAC member) participated in a three-hour workshop to delve into the 3-Act Model and develop shared frames of reference for assessment and feedback. Trainers watched and rated six role-play videos, discussing their individual scores and coming to consensus about grading after each one.
Curricular Implementation
With residency program leadership support, we integrated our teaching sessions into a mandatory clinical skills course for Johns Hopkins Bayview Medical Center IM interns (n = 22) in August-September 2018. When considering which learners would be best for this pilot, we decided that focusing on interns would be preferable to senior residents who might be more set in their ways. Each intern participated in two small group sessions (three hours each), which incorporated narrative reflection and best practices for skills training, including didactic description of the approach, video demonstrations, learner practice, and detailed feedback.
Role-plays involved one intern acting as provider and the other, as patient, with a trainer assigned to each pair. In total, each intern acted as provider in five role-plays and as patient in another five. Within the pairings, each intern took turns as provider and as patient, using two unique scenarios. After every two scenarios, we rotated the interns and trainers to create different combinations. Each intern's first role-play as provider was preintervention, that is, took place before teaching of the 3-Act Model; the remainder were postintervention. After each postintervention role-play, trainers facilitated structured 360-degree reflection on practice
and feedback for each resident in the provider role. We arranged scenarios in the sessions so that early role-plays occurred with easier scenarios and later role-plays with more difficult scenarios, to increase the likelihood that any trend of improving rubric scores over time might be due to increased skill level, rather than change in role-play difficulty.
Evaluation Design
The overall evaluation plan for this pilot involved a pre-post study. The primary metrics of interest were trainer GCAT ratings of interns acting as provider for all role-plays, as compared to the predetermined proficiency standard. Ratings were done immediately after each role-play. Learners in both provider (self-assessment) and patient roles also completed GCAT assessments of the provider for each postintervention role-play. The GCAT included a global item regarding the readiness of the learner “to independently lead basic GOC discussion in clinical setting”; answer choices were “disagree,” “agree,” or “strongly agree” and were not factored into the proficiency standard. At the end of the second session, learners were asked to complete a brief course evaluation survey, which included the question: “Would you recommend the first two sessions on GOC discussions to colleagues?” Answers were on a five-point scale, from 0 = “no” to 5 = “yes, as outstanding.” The surveys also gave respondents the opportunity to free text observations about course strengths, areas for improvement, and other comments. Data were collected via Qualtrics from August 16, 2018, to September 26, 2018, and deidentified and password-protected to protect learner anonymity. The study was deemed exempt by the Johns Hopkins Institutional Review Board in July 2018.
Data Analysis
Factor analysis of all GCAT items except for the global rating was performed. Extraction method was principal component analysis; rotation method was Varimax with Kaiser normalization. Descriptive statistics and analyses were performed for the proficiency measurements. Intraclass correlation coefficients were calculated across all trainer raters. Nonparametric testing was performed to check the statistical significance of proficiency ratings between role-play 1 and 2, 2 and 3, and so on. Proficiency scores did not follow normal distribution, so t-tests were not appropriate. Statistical Package for the Social Sciences was used for all analyses.
Results
The 22 interns studied had a mean age of 29 years (standard deviation = 6.0), and 55% were female. Interns identified their race as Asian (32%), White (27%), Black/African-American (18%), or Hispanic/Latino/of Spanish origin (18%); one intern did not identify race. 50% of interns were in the categorical IM track, whereas 27% were in the primary care track and 23% were preliminary interns. While all interns were evaluated in the provider role by a trainer through five separate role-plays (for a total of 110 trainer ratings), 106 unique trainer ratings were appropriate for analyses after removing unidentifiable and incomplete assessments.
Factor Analysis
Factor analysis identified four factors based on eigenvalues >1. Each of the 16 items in the GCAT contributed variably to the overall percent of the variance (range: 1.4–29%). According to the variables mapping to each factor (Table 1), they were named as follows—factor 1: shared decision-making; factor 2: patient-centered communication; factor 3: setting the stage; and factor 4: ensuring understanding of discussion. Items clustered within factors according to the individual components of the 3-Act Model, with the exception of Act II items that were distributed among multiple factors. Act II items, which pertain to the provider's skill in delivering the medical overview, were maintained as a separate GCAT domain with the rationale that this skill is conceptually distinct from the four identified factors.
Objective Proficiency Ratings
For proficiency ratings across all trainers, average intraclass correlation coefficient was 0.78 (95% confidence interval 0.71–0.84), suggesting good interrater reliability. Cronbach's alpha of 0.78 for the five domain scores and the total score supports strong internal consistency for the GCAT.
In the pretest role-play, 30% of interns were rated proficient (Figure 1). Over successive role-plays, the scores trended upward, culminating in 100% interns rated proficient in their final role-play.
Fig. 1The number (percentage) of interns in the provider role who were rated by trainers as proficient. aBy nonparametric testing, the difference between Role-Plays 4 and 5 was statistically significant (phi = 0.36, Cramer's V = 0.36).
By the end of the training, 96% of interns agreed or strongly agreed that they felt ready to independently lead a basic GOC discussion in the clinical setting. 41% of interns strongly agreed that they felt ready after their final role-play, compared with 18% after their first postintervention role-play. Of the 21 interns who completed the course evaluation, 100% indicated they would recommend the two teaching sessions to colleagues as “outstanding” (67%) or “very good” (33%). 18/21 (86%) respondents thought the number of role-plays in the sessions was “about right”; the remainder would have preferred fewer role-plays. In written and verbal comments, interns indicated they appreciated the 3-Act framework, variety of the sessions, opportunity to practice, and coaching. Several commented on the “realistic” and “nuanced” role-play scenarios, and the perception that playing the role of patient was beneficial, even “critical,” for genuinely understanding the patient's perspective. One intern wrote: “Amazing opportunity to learn a framework for palliative discussions, with carefully crafted cases that allowed for application and practice.” Another commented: “I'm going to do more GOC discussions because of this.”
Discussion
This pilot study demonstrates that the novel 3-Act Model is learnable, producing objective improvement in communication skills among a cohort of medicine interns. In the final role-play, 100% interns were deemed proficient according to predetermined standards. In addition, intern feedback and self-assessments were very positive.
Narrative medicine has been used as a paradigm to improve communication skills and empathy in learners.
To our knowledge, the 3-Act Model is the first based on a narrative approach to objectively demonstrate improved GOC communication skills and adds to existing evidence that narrative-based GOC training is valued by learners.
One example is a study of the COMFORT curriculum in which educators trained interprofessional palliative care team members through several communication skills modules over two days and achieved positive results per learner self-report.
Our intervention measured objective skill proficiency using an internally validated tool congruent with our methods; to our knowledge, no gold standard comparator exists.
The 3-Act Model equips learners with a simple framework and narrative skills that are highly adaptive to a patient's individualized story and conducive to natural conversation. No two GOC discussions will ever be the same because no two people nor their medical circumstances and life stories—their “personomics”
—are the same. In this age of rapidly progressive health care technology, the timeless need for clinicians to prioritize the patient as a person is as acute as ever. Education grounded in the humanities, like the 3-Act Model (whose name is inspired by the three-act dramatic structure often credited to Aristotle), holds great promise in helping clinicians do just that.
Several limitations of this study should be considered. First, the study was conducted at a single institution. Second, the GCAT scoring was nonblinded. However, the GCAT items measure whether discrete behaviors/tasks were done using a simple 3-point rating; such a scale is less subject to bias than more subjective or complicated measures might be. Third, the intervention required significant faculty/staff time, but the finding of 100% proficiency by the final role-play suggests that the resources translated into positive return on investment.
The positive outcomes of this pilot study support our hypothesis that the 3-Act Model is teachable, appreciated by learners, and leads to objective skills improvement. As such, curricular implementation should help our IM residents attain mandated goals, including the Accreditation Council for Graduate Medical Education's Milestones pertaining to “communication” and “decision-making.”
We have permanently embedded this training into the residency curriculum. Future areas of study may include persistence of use in clinical practice, blinded assessment of communication skills, adaptation and implementation for interprofessional teams, and patient outcomes with attention to diverse patient populations. By enabling clinicians to attentively listen and adaptively respond to the individualized stories of patients, the 3-Act Model can lead to artful patient-centered communication, with the ultimate aim of care aligned with patient values.
Disclosures and Acknowledgments
This study is dedicated to the late Dr. Stuart J. Farber.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The authors have no conflicts of interest to report. Dr. Wu is supported by the Hearst American Academy of Hospice and Palliative Medicine Leadership Scholars Fund. Dr. Wright is the Anne Gaines and G. Thomas Miller Professor of Medicine and supported through the Johns Hopkins Center for Innovative Medicine.
The authors are indebted to Will Bevins, MDiv, MA; Anna Diller, MD, MPH, MBA; Linda Gorman, MLS; Rev. Peter Heikkinen, MDiv; Erica Leonard, MD; Lingsheng Li, MD, MHS; Jane Schindler, LCSW-C, APHSW-C; and Seonho Shin, PhD, for their help and guidance.
Appendix: Example Role-Play Case
Information for Provider (Primary Team)
Diagnosis: End-Stage Renal Disease (ESRD) on hemodialysis (HD)
Setting: Inpatient.
Medical history: Mr./Ms. Holloway is a 54-y/o patient with ESRD, directly admitted from HD center due to a near-syncopal episode while getting dialysis. Patient is anuric and has been dialysis dependent for 12 years, with increasing episodes of hypotension limiting dialysis sessions in last few months.
Patient was admitted to intensive care unit three days ago with the diagnosis of sepsis in the setting of pneumonia and has remained borderline hypotensive, despite intravenous antibiotics and recent addition of two pressors, now on continuous dialysis. You and your team are very concerned about acute prognosis, but chance remains that sepsis might resolve. You decide to explore goals of care, including if/when patient might consider comfort care.
Information for Patient
Diagnosis: End-Stage Renal Disease on hemodialysis (HD)
Setting: Inpatient, admitted for almost passing out in dialysis center
Medical Story: I am a 54-y/o patient, Mr./Ms. Holloway, who has been on dialysis for 12 years. Recently, they have had to stop the sessions earlier than usual because I get light-headed and nauseous in setting of low blood pressures. I was sent to the intensive care unit because I almost passed out during my last session. Now I've got intravenous antibiotics and monitors everywhere, and they are doing a different kind of dialysis. I know I have pneumonia and a bad infection, but other than still feeling light-headed, I feel pretty normal. I expect I'll be able to go home soon and restart outpatient dialysis, like I did before.
Social History: Divorced. I live alone in a first-floor apartment. I finished the eighth grade and have two children who are both tattoo artists.
Occupation: I work in food services at a local hospital, delivering meal trays to patients.
Hobbies: I like to gamble and play the lottery. I have been saving some money from my paychecks to go to the casino.
Spiritual: Baptist.
Mood/affect: Cheerful, positive.
Responses to provider: I want to know all the details about my health and illnesses, so that I know what to expect and will ask questions about blood pressure and lab numbers, etc. I want to understand well enough to tell my children myself what is going on. I hate being in the hospital and feel frustrated that I'm not getting better. I am open to discussing big picture prognosis, if provider raises. I feel my quality of life at home is good, and I see my dialysis sessions as part of my regular routine and community. I am hoping to get back to baseline, go to rehab, then back home. However, I will be open to discussing comfort focus in case I don't get better. However, I am not ready to seriously consider stopping dialysis, as I would see that as “giving up.”
References
Kaldjian L.C.
Curtis A.E.
Shinkunas L.A.
Cannon K.T.
Review article: goals of care toward the end of life: a structured literature review.
“I just felt like I was stuck in the middle”: physician assistants' experiences communicating with terminally ill patients and their families in the acute care setting.
The effect of communication skills training for generalist palliative care providers on patient-reported outcomes and clinican behaviors: a systematic review and meta-analysis.
Effect of a patient and clinician communication-priming intervention on patient-reported goals-of-care discussions between patients with serious illness and clinicians: a randomized clinical trial.
Evaluating an intervention to improve communication between oncology clinicians and patients with life-limiting cancer: a cluster randomized clinical trial of the Serious Illness Care Program.
30509-3&id=gr1.jpg){kind=link}