Electronic supplementary material
Teaching medical students to talk about death and dying in the ICU: feasibility of a peer-tutored workshop
Lena Dorner,1,2 Daniel Schwarzkopf,2 Helga Skupin,1 Swetlana Philipp,3 Katrin Gugel,1 Winfried Meissner, 4 Stefan Schuler,5 Christiane S. Hartog1,2
1Department for Anesthesiology and Intensive Care Medicine, Jena University Hospital,
Erlanger Allee 101, 07743 Jena, Germany
2 Integrated Research and Treatment Center, Center for Sepsis Control and Care (CSCC), Jena University Hospital, Jena, Germany
3 Institute of Psychosocial Medicine and Psychotherapy, Jena University Hospital, Friedrich-Schiller University, Jena, Germany
4 Clinic for Palliative Care, Jena University Hospital, Erlanger Allee 101, 07743 Jena, Germany
5SkillsLab, Jena University Hospital, Bachstraße 18, 07743 Jena, Germany
Supplementary Method Section
Peer Teaching
The Medical School of the Jena University Hospital teaches about 250 medical students per year. A “Skills Laboratory” (SkillsLab) was established in 2010 to provide practical training. The SkillsLab has a structured program to recruit and train tutors. Peer-tutored skills training takes place in small workshops of 4 - 6 participants. Content and design of these workshops is jointly developed by experienced clinicians and pedagogically trained SkillsLab staff who then train medical students to act as peer tutors. Peer tutors learn to teach medical skills and how to provide structured feedback to participants. They must pass a practical exam in front of clinicians and SkillsLab staff to demonstrate that they can successfully conduct a workshop. Tutors for the communication skills workshop were recruited through the SkillsLab core facility. Each workshop was accompanied by one tutor who was specifically trained for this workshop at the beginning of the semester (academic half-year). One additional tutor (LD) accompanied all workshops to conduct the surveys and videotape role plays.
The workshop
The workshop “talking about death and dying” was based on nine communication skills, five of which were developed previously and shown to effectively reduce the relatives’ emotional burden [1]. These skills are known by the mnemonic “VALUE” which stands for “value family statements”, “acknowledge family emotions”, “listen to the family”, “understand the patient as a person”, and “elicit family questions”. The remaining 4 skills were developed by discussion with experienced intensivists in our ICU who regularly hold end-of-life family conferences. These skills related to sharing of information (create a common information base, share medical information effectively, explain death and dying in the ICU, make clear arrangements). All skills are depicted in Table 1.
The workshop was divided into a theoretical and a practical part. The theoretical part was developed to a) help students experience the emotional needs of people in states of anxiety and stress and b) understand the theoretical framework of communication in this context. In order to understand special requirements in states of anxiety or shock, students were shown pictures of a natural disaster which had occurred some years ago nearby (a sinkhole had destroyed parts of a village). Students were asked to imagine the emotional situation of the village residents who woke up and found that their garden had disappeared overnight into a large hole in the ground, that their house was threatened by collapse and that they had to pack and leave their house within the next hour. Students could well imagine the state of shock these people were in. They then listened to the verbatim expertise of a geological expert who explained some possible mechanisms. Students understood that in such situations technical explanations were not helpful at all and “passed completely over their heads”. In this way, students could experience the reduced cognitive capacity of people in states of crisis and their increased need for emotional support. The tutor explained that talking to people in shock (i.e. to relatives of dying ICU patients) required special skills and introduced the nine communication skills.
In the practical part, students were presented with a fictive story of a young patient who had received life-threatening injuries in a bicycle accident and was being treated in the ICU for sepsis and multi-organ failure. His condition had been steadily worsening and he was now found to be non-responsive to ICU therapy and expected to die. The students’ task was to break the news to the family. A written handout contained the physician perspective (“case history”), the family perspective (i.e. the “information” they had received until then), and a structured outline of the talk built on the nine communication skills with model phrases which the students were encouraged to use (see Supplementary Table 1). In addition to the 9 communication skills, the handout also included more general items such as introducing one-self, creating a private atmosphere, and being authentic, i.e. honest and open. However, in order to reduce the complexity of the workshop and considering that these more general items are also addressed in communication skills training in other clinical contexts, our workshop only mentioned these and then focused in more detail on specific skills. The students were asked to enter into pair-wise role-play and alternate between the roles of physician and relative to experience both sides of the conversation. They were encouraged to use the handouts. Afterwards, students received structured feedback and discussed their experience.
Outcomes
Rating of communication skills
Students rated their ability to perform the nine communication skills before and after the workshop on a 6-point Likert scale ranging from 1 (“not true”) to 6 (“very true”). Videotaped role plays were analyzed by an external expert. This was an experienced intensivist (SS) who was not involved in the development or conduct of the workshop. He was instructed in how to perform the ratings by one of the authors (LD), then viewed the available videotapes and rated students’ skills during filmed role plays by the same Likert scale which the students had used. One item was considered not to be applicable for rating by a person who was not involved in the workshop (“I can understand the patient as a person”).
Qualitative analysis of students’ language
One of the communication skills was to give medical information effectively. Students had the task to break the news of imminent death. It was explained to them that relatives are in a state of shock or anxiety and therefore need to be given the news in a clear, succinct language. Students were encouraged to speak out the words “death”, “die” or “dying”. However, in the past tutors observed that students often avoided these words and used other expressions instead. Therefore, we decided to also perform a qualitative analysis of verbatim transcripts to identify alternative expressions.
Data analysis
Quantitative data
Overall scores of students’ self-ratings of their communication skills before and after training were calculated as the mean of the respective items if ratings of all items were available. The ratings on all items as well as the overall score before training and after training were compared. Only cases with the respective rating available before and after training were used, ratings were described by mean and standard deviations and difference was tested by Wilcoxon signed-rank test. Students’ self-ratings and expert ratings after training were compared in the same way. All tests were done bidirectional at alpha level 0.05 using the statistical software R [2].
Qualitative data
Written transcripts were analyzed inductively using content analysis to generate categories [3].
The software RQDA [4] was used to code written transcripts. One reviewer (LD) noted if words like “death”, dying” or “die” were used or whether participants used other expressions. She developed categories and cross-checked them with a second, uninvolved reviewer. In case of disagreement, consensus was reached by discussion. The senior author (CSH) cross-checked and approved all categories.
Supplementary results
External expert rating
18 role plays with 29 participants were videotaped and analyzed. Role-plays were conducted simultaneously, but only one camera was available which limited the total number of videotaped encounters. The mean expert ratings mostly did not differ from the students’ self-rating of their workshop performance (i.e. after training), but the correlation was weak throughout (Supplementary Table 1).
Analysis of statements used by medical students
Verbatim transcripts of filmed role plays contained 95 coded statements in which “physicians” informed “relatives” about imminent death. These 95 statements were grouped into categories; cross-checking by the second reviewer led to re-coding of 5 of the 95 statements. The third reviewer (author CSH) recommended to omit one category with 5 statements altogether because it contained questions on “how to proceed” which did not directly refer to death or dying. Therefore, 90 statements remained for the final analysis. Eighteen statements (20%) contained the words “death”, “dying” or “die”. The remaining 72 statements (80%) did not; they contained paraphrasing which was grouped into 7 categories. Five denoted lack of medical options or hopelessness (Nothing can be done, No cure, No survival, No hope, Won’t leave the ICU). One category was “Forced admission of death”. The seventh category contained miscellaneous statements (“Stop treatment” - 2 statements, “Will not wake up again” – 1 statement, “Change to palliative care” - 1 statement, or “Say good-bye” - 1 statement) (Supplementary Table 2).
Supplementary References
1. Lautrette A, Darmon M, Megarbane B, Joly LM, Chevret S, Adrie C, Barnoud D, Bleichner G, Bruel C, Choukroun G, Curtis JR, Fieux F, Galliot R, Garrouste-Orgeas M, Georges H, Goldgran-Toledano D, Jourdain M, Loubert G, Reignier J, Saidi F, Souweine B, Vincent F, Barnes NK, Pochard F, Schlemmer B, Azoulay E (2007) A communication strategy and brochure for relatives of patients dying in the ICU. NEJM 356: 469-478
2. R Core Team (2013) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria
3. Pope C, Ziebland S, Mays N (2000) Qualitative research in health care. Analysing qualitative data. BMJ 320: 114-116
4. Huang R (2011) RQDA: R-based qualitative data analysis. R package version 0.2-2. http://rqda.r-forge.r-project.org/.
Supplementary Table 1
Expert versus student rating
Expert versus student ratingCommunication skill / Na / Expert / Student / P valueb / Corre-lation c / P valuec
Skills 1 - 4
Sharing information
“I know how to –
1. - create a common information basis” / 24 / 4.2±0.7 / 4.8±1 / 0.028 / 0.02 / 0.942
2. - share medical information effectively” / 27 / 4±0.9 / 4.1±1.3 / 0.662 / -0.19 / 0.335
3. - explain death and dying in the ICU” / 22 / 4.2±0.6 / 4.1±1.2 / 0.694 / 0.21 / 0.346
4. - make clear arrangements” / 13 / 4±0.4 / 4.2±0.8 / 0.429 / -0.29 / 0.341
Skills 5 – 9
Providing Emotional Support
“I know how to -
5. - be respectful and value family statements even in difficult situations“ / 29 / 4.8±0.5 / 4.8±0.8 / 0.674 / 0.23 / 0.239
6. - acknowledge even covert family emotions” / 13 / 4±0.8 / 3.9±1 / 0.829 / -0.12 / 0.701
7. - listen to the family” / 29 / 4.4±0.6 / 4.2±1.1 / 0.35 / 0.16 / 0.408
8. - speak with relatives and understand who the patient was and what he/she valued” / na / na / na / na / na / na
9. - elicit family questions even under time pressure” / 29 / 4.2±0.8 / 4.5±0.9 / 0.122 / 0.23 / 0.238
The external expert (an experienced intensivist) assessed students’ performance by watching videotaped role plays. Ratings were on a Likert scale from 1 – 6 (not true – very true). Ratings are compared with students’ self-ratings after the training.
na – “not analyzable” because the external expert considered he could not analyze this skill from watching videotaped role plays
a Column gives number of cases with ratings available for pairwise comparison.
b P-value obtained by Wilcoxon signed-rank test.
c Correlation calculated as Spearman correlation. Correlations indicate the degree of agreement between the self-ratings of students and expert ratings. A correlation of 1 indicates that the expert ratings and student ratings show the same order, a correlation of -1 indicates that expert ratings show the reversed order of student ratings, correlations near 0 indicate that expert ratings show no relation to student ratings.
Supplementary Table 2
Statements used by medical students who avoided the word “death”
Categories / Examples / FrequencyNothing can be done / “There is nothing left to do…“
“We have simply exhausted all possible measures…”
“We have done everything possible…” / 21
No cure / “There is going to be no cure at the end of therapy.”
“We do not think that function can be restored…”
“We believe it does not make sense to continue.” / 19
No survival / “She cannot live without machines.”
“…no hope for survival.”
“Organ failure is irreversible”. / 16
No hope / “We see no hope for your mother” / 4
Won’t leave the ICU / “We will not be able to discharge him.”
“He will never leave the ICU.”
“Won’t come out as he went in.” / 4
Forced to be clear / Relative: “Does this mean he will die?” Doctor: „Yes, exactly.” / 3
Miscellaneous / “we therefore will stop further treatment” / 2
“From the doctors’ view, he will not recover and wake up again.” / 1
“…no way around saying good-bye to your friend.” / 1
“We’ll simply start palliative care.” / 1