We recruited pediatric oncologists RNs and APPs to form a dyad of one physician and one RN or APP from 8 institutions: MD Anderson Cancer Center (TX), Children’s Hospital Colorado (CO), Children’s Hospital of Atlanta (GA), SSM Health Cardinal Glennon Children’s Hospital (MO), Riley Hospital for Children (IN), Seattle Children’s Hospital (WA), Children’s Wisconsin (WI), and Nemours Children’s Hospital (DE). The dyads were randomized to the intervention or an attention-control group by clinician practice (non-central nervous system solid tumor or neuro-oncology); size of practice (large or medium); and whether the clinicians had previous formal PC/EOL CST. Dyads received 3days of training (in-person or virtually). In-person training occurred at Bradford Woods, a 2500-acre retreat center of Indiana University in November 2019. Dyads recruited subsequently and unable to participate in-person due to the COVID19 pandemic were trained virtually. There were 5 virtual trainings between July 2020 and January 2023. Virtual training was adapted from in-person training and used the same agenda, didactic presentations, videos and case-based role plays with simulated actors.
Our PC/EOL communication intervention consisted of a conversation guide based on VitalTalk™ REMAP [10] modified for use at different timepoints along the continuum of care with parents of children with cancer. Table 1. In addition, two novel features incorporated into the intervention were the use of dyads to co-deliver the intervention instead of single clinicians, and the development of two visual aids (Diagrams I and II) to use alongside the guide. Diagram 1 depicted hope, dual GOC (QOL and best-possible treatment) and team support with two overlapping circles. See Fig. 1. Diagram II illustrated 3 GOC: “cure the cancer”, “slow the cancer”, and “comfort” with 3 overlapping circles. See Fig. 2.
Table 1. REMAPS / COMPLETE Intervention
Our training method was based on the VitalTalk™ course with dyads trained as a team and interprofessional team principles added to the curriculum. Evidence-based VitalTalk™ skills such as “Ask-Tell-Ask”, silence after giving bad news, “I wish/I worry” statements, and expressive use of empathy described by the acronym NURSE (naming, understanding, respecting, supporting, exploring) were taught via didactic PowerPoint presentations to learners [13]. The curriculum included dedicated time to practice using diagrams I and II with simulated patients. VitalTalk™ videos depicting difficult conversations between patients, family members, RNs, APPs, and doctors were shown and the unique roles of interprofessional team members were emphasized in the training. Sample videos were created with the VitalTalk™ facilitators (LV, JH, KM) portraying the dyad with a simulated patient/parent/family member for the specific pediatric oncology scenarios of breaking bad news, discussing GOC at time of disease progression, and EOL planning at the terminal stage of disease. These videos served as resources and “Booster Training” for participants to review as needed throughout the study.
Four pediatric-oncology cases were created from the experience of one of the MPIs (KM) and literature review. Each case depicted a child with a poor cancer prognosis and included three different time points: new diagnosis, progression of disease, and approaching end of life. Actors were hired to portray an adolescent patient (in one case), parent(s), and family members. The actors were given extensive backgrounds about their illness, their mindset, family dynamics and their cultural context. During actor-training, actors were given time to practice their roles with the facilitators (KM, LV, EN). A key component of the actors’ job was to respond realistically when the clinicians presented the PC/EOL intervention. For example, if the dyad partners did not exhibit empathy when giving poo prognostic information the actors might respond by disengaging from the conversation or becoming frustrated [13].
Each dyad was assigned to a simulated patient and family member(s) to practice the communication intervention in the context of the 3 clinical time points. Dyads role-played with the same family over the 3-day course to simulate a longitudinal relationship. A brief synopsis of the case (including key medical and social details of the patient and their parent[s] and family member[s]) was given to the dyad in temporal order, immediately prior to each role play. During the role plays, the other dyads watched as one dyad performed the communication intervention and then provided feedback about what the dyad did well. This process enabled participants to learn from each other and observe multiple examples of the intervention delivery.
If learners felt stuck in their interactions or the facilitators sensed the learners needed assistance, a “Time out” was called by the learner or facilitator, respectively. During the “Time out” the learning dyad could “phone a friend” or ask observers for suggestions for what to say or do to move the conversation forward. To ensure standardized and consistent delivery of our PC/EOL communication intervention during the study period, dyads were trained to perform quality assurance (QA) on themselves. This QA process required the dyads to fill out a check-list documenting which tasks they did and did not complete during the intervention and included items such as “used NURSE statements” and “reviewed Diagram I with family”. If the dyad did not complete the checklist within 48 hours of the intervention, they would have to complete the check list after listening to the intervention's audio recording (all intervention sessions were audio recorded). Study intervention fidelity was operationalized as the percentage of tasks completed during the intervention session, with 0 percent = none and 100 percent = all.
A post-training survey included closed-ended demographic questions and questions regarding the value of several aspects of the course using 5-point Likert-like scale, with 5 representing the most favorable response. Four open-ended questions queried opinions on the learning environment (virtual and in-person); novel concepts learned; general course feedback; and how participants wanted to improve their skills when they returned to work. Participants were asked how committed they were to make this change. The survey was sent electronically and anonymously to the participants. An informed consent statement was included in the invitation email. Participants were invited to read the informed consent statements and proceed to the survey study if they agreed with them. The MD Anderson Cancer Center IRB approved the post-training survey study in accordance to the U.S. Department of Health and Human Services regulations for the protection of human subjects in research. The Indiana University single IRB approved the IMPACT study in accordance to the U.S. Department of Health and Human Services regulations for the protection of human subjects in research. QA data is presented to illustrate intervention fidelity during the study to date.
Data Analysis
Descriptive statistics were used to describe the frequencies, median (M), and interquartile range (IQR), of the learner’s responses to the closed-ended survey questions. Manual qualitative analysis on the open-ended text responses was performed by K.M. Text responses were de-identified and reviewed for common themes and patterns. Responses were classified into categories based on similarities and differences within and across questions. Frequencies of categories were calculated. Notable quotes for categories were added for context. Categories were incorporated into themes, where appropriate [14]. A second author (AH) manually reviewed the qualitative data and resulting categories and themes to reduce bias. Differences in qualitative data analysis results were discussed among the two reviewers until reaching consensus.