Key elements and checklist of shared decisionmaking conversation on life-sustaining treatment in emergency: a multispecialty study from China

Shu Li ,Jing Xie ,Ziyi Chen ,Jie Yan ,Yuliang Zhao ,Yali Cong ,Bin Zhao,Hua Zhang,Hongxia Ge,Qingbian Ma,Ning Shen

1 Department of Emergency Medicine, Peking University Third Hospital, Beijing 100191, China

2 Department of Infectious Diseases, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing 100730, China

3 Department of Neurology, the First Affiliated Hospital, Sun Yat-sen University, Guangzhou 510080, China

4 Department of Obstetrics and Gynecology, Peking University First Hospital, Beijing 100034, China

5 Department of Nephrology, West China Hospital, West China School of Medicine, Chengdu 610041, China

6 Institute of Medical Humanities, School of Foundational Education, Peking University Health Science Center,Beijing 100191, China

7 Department of Emergency Medicine, Beijing Jishuitan Hospital, Fourth Medical College of Peking University, Beijing 100035, China

8 Clinical Epidemiology Research Center, Peking University Third Hospital, Beijing 100191, China

9 Department of Pulmonary and Critical Care Medicine, Peking University Third Hospital, Beijing 100191, China

BACKGROUND: Shared decision-making (SDM) has broad application in emergencies.Most published studies have focused on SDM for a certain disease or expert opinions on future research gaps without revealing the full picture or detailed guidance for clinical practice.This study is to investigate the optimal application of SDM to guide life-sustaining treatment (LST) in emergencies.METHODS: This study was a prospective two-round Delphi consensus-seeking survey among multiple stakeholders at the China Consortium of Elite Teaching Hospitals for Residency Education.Participants were identified based on their expertise in medicine,law,administration,medical education,or patient advocacy.All individual items and questions in the questionnaire were scored using a 5-point Likert scale,with responses ranging from “very unimportant” (a score of 1)to “extremely important” (a score of 5).The percentages of the responses that had scores of 4-5 on the 5-point Likert scale were calculated.A Kendall’s W coefficient was calculated to evaluate the consensus of experts.RESULTS: A two-level framework consisting of 4 domains and 22 items as well as a ready-touse checklist for the informed consent process for LST was established.An acceptable Kendall’s W coefficient was achieved.CONCLUSION: A consensus-based framework supporting SDM during LST in an emergency department can inform the implementation of guidelines for clinical interventions,research studies,medical education,and policy initiatives.

KEYWORDS: Shared decision-making;Life-sustaining treatment;Emergency;Checklist

INTRODUCTION

Shared decision-making (SDM) has broad application in emergencies,[1-3]such as medical decisions on acute coronary syndrome,thrombolysis in stroke,and CT scan strategies.[4,5]Patients in an emergency department (ED) are very vulnerable,with distinct characteristics and tremendous decisional stress. SDM in an ED,the “serious illness conversation”,[1]is in fact challenging and creates quite a few barriers from both the provider side and patient side.[6]Given that a patient’s situation might be indeterminate and life-threatening on arrival,the clinical uncertainty of prognosis and absence of prior provider-patient relationships increase the decisional conflict of patients/surrogates,which makes medical decisions difficult.[7-9]The conversation must be initiated immediately and should be a process that occurs on a continuum.[7]On the one hand,some critically ill patients may be difficult to participate in medical decisions because of suffering or cognitive impairment.[5]On the other hand,emergency physicians usually have insufficient SDM training and are busy dealing with a large number of daily visits;thus,they have limited time for conducting SDM processes in individual cases.[5]

At the same time,initiatives to improve SDM for life-sustaining treatment (LST) are increasing in the clinical,research,and public sectors.In 2017,the Society for Academic Emergency Medicine Consensus Conference released a consensus statement on SDM in EDs,[5]emphasizing that excellent emergency care depends on enhanced SDM.Common elements included process over time through conversations on patient’s values,medical information disclosure,prognosis evaluation,and strategies for medical treatment,which were described as advanced directives (AD),advanced care planning (ACP),[10]goals of care (GOCs)conversation,[1,11]or code status conversations.[12]Prior descriptions and studies have mostly emphasized def initions,goals,and advantages.[8]

Most published studies have focused on SDM for a certain disease,or future research gaps published as expert reviews.[1,6-9,12,13]Most of these studies are limited by a small sample size or confined vision[12]without revealing detailed guidance for clinical practice.Moreover,there may be certain variation in bioethics principles in Asian countries based on the Eastern socioeconomic background.Different conceptualizations and the absence of shared quality standards make it impossible to measure and improve the quality of such conversations between clinicians and patients.Many pending questions surrounding SDM for LST and its practice remain,with new questions constantly being formulated.

We aimed to establish a unified consensus to guide clinical initiatives,research,and medical education on SDM for LST in emergencies.Thus,we leveraged the expertise of our multidisciplinary panel with the Delphi technique to explore the essential elements during the process in China.

METHODS

Study design and setting

This study was a prospective Delphi consensus-seeking survey to define essential elements of physician practice during the SDM process for LST in emergencies.The Delphi process was conducted from March to April 2022.This study was approved by the Institutional Review Board of Peking University Third Hospital (IRB-2021-591).

Selection and qualification of participants

We aimed to have a diverse representation of the expert panel,including clinicians with expertise in a variety of disciplines,researchers,policy-makers,and patient volunteers.Based on a literature review of studies by Chinese authors,we identified three research experts for the Delphi process.We then used snowball sampling(i.e.,referred by initial invitees) to recruit the remaining nine experts from frontline faculties through the China Consortium of Elite Teaching Hospitals for Residency Education.The expert panel has been deeply involved in clinical,research,teaching or administrative work related to SDM,including emergency physicians,cardiologists,respiratory intensivists,epidemiologists,medical science researchers,bioethicists,clinical medical ethicists,experts in medical-legal affairs,representatives of educational administration,and hospital policy-makers.We also included two representatives of patients and their surrogates.Experts’familiarity with the survey was divided into “very familiar,familiar,general,unfamiliar,and very unfamiliar”,with 1,0.8,0.6,0.4 and 0.2 points,respectively.The judgment basis assignment of indicators is shown in supplementary Table 1.The authority of experts was the arithmetic mean of the familiarity and judgment basis coefficients.

Delphi methods

Literature research was performed,and potential key elements during the SDM process for LST were identified.[12]A questionnaire with several domains and a subsequent related series of questions were developed and revised by a working group consisting of emergency clinicians,academic researchers,and administrative authorities experienced in LST and SDM.A statement was on the title page of the survey for consent to participate.Then,items in the questionnaire were further modif ied and finalized through a Delphi process.All individual items and questions were scored using a 5-point Likert scale,with responses ranging from “very unimportant” (a score of 1) to“extremely important” (a score of 5).At the same time,the experts were encouraged to make their own comments to refer to unmentioned items that they thought were of great value,or direct edits could be made to all items.After review of the data from the first round,the working group created a second-round questionnaire,removing items that received low scores,clarifying terms and uncertainties in wording,and adding new domains or items suggested by participants.The consensus and changes were presented as iterative feedback at the beginning of the secondround questionnaire with revised and clarified items.Panel members received feedback regarding the overall opinion of others before they made further suggestions.The demographic characteristics of all participants were collected through the first-round questionnaire.Data were collected using an online questionnaire tool,WJX (https://www.wjx.cn/vj/YCJ1pwU.aspx).Source data were exported as one SPSS file.

Primary and secondary outcome measures

As the area of SDM remains poorly investigated and unclear,we chose a modified Delphi process that allowed conversation and discussion among experts after two rounds of survey to achieve final consensus.The primary objective was to def ine essential elements of physician practice during the SDM process for LST in emergencies.Consensus was achieved on an item if 70% or more of the responses had scores of 4-5 on the 5-point Likert scale.Disagreement was considered to occur if 35% or more of the responses fell within either extreme range of possible options on the Likert scale (1 and 5).A Kendall’s W coefficient of 0.7 represented a valid cutoffcriterion for the Delphi process or when no further insights or improvements in answer quality could be derived from additional Delphi rounds.

Statistical analysis

Individual survey responses served as the unit of analysis.Median with interquartile ranges (IQRs) were used for continuous variables,and counts and frequencies were used for categorical variables.Kendall’s W coefficient was calculated to evaluate the consensus of experts (＜ 0.2 as poor;0.2-0.4 as fair;0.4-0.6 as moderate;0.6-0.8 as good;0.8-1.0 as excellent).AllP-values were two-tailed,and the results were statistically significant at aP-value ＜0.05.Data were analyzed using IBM SPSS version 22.0 for Windows.

RESULTS

Characteristics of participants

The first round of the Delphi exercise lasted from March 28 to April 8,2022,and the second round was performed two weeks later on April 22,2022,and lasted 7 d.All 14 participants finished the two questionnaires,with a response rate of 100%.The median age of the experts was 41 years(IQR 37,45),and there were 11 (78%) female participants.The demographic features are listed in supplementary Table 2,and the authority of experts is listed in supplementary Table 3.

Main results

After a review and investigation of the literature,the questionnaire was built.The questionnaire explored key elements during the SDM process between doctors and patients or their proxies in emergencies.The questionnaire in the first round consisted of 4 domains with 44 items (comprising 7-20 questions each),including communication skills,disclosure of medical information,patient autonomy and decisional assistance,and summary and process improvement.

In the first round of the questionnaire,the frequency of questions with 4 or 5 points was 57%-100%.No expert scored 1 point for any of the items.The Kendall’s W coefficient of the first round of the survey was 0.351-0.607,with a Chi-square testP＜0.05,indicating that the results were acceptable.After the first round of responses,there were no new domains or items suggested by the participants.The four domains were agreed upon by experts,and the working group removed four items that received low scores,such as“use open-ended questions to start a conversation”,“express gratitude at the end of the conversation”,“provide advice for financial support if necessary”,and “ask the patient/surrogate for their opinion and suggestion on the communication process,and make use of a questionnaire when feasible”.Based on comments from the expert panel,the items related to common communication skills were subsequently removed,such as “dress neatly,be vigorous and energetic”,“give a warm welcome and use proper name etiquette”,“smile,maintain eye contact,and listen attentively”,“pay attention to logicality”,and “express gratitude at the end of the conversation”.Some of the items that conveyed similar information were integrated.In the second round of the questionnaire,the frequency of questions that had 4 or 5 points was 86%-100%.The Kendall’s W coefficient of the second-round survey was moderate,between 0.397 and 0.591,with a Chi-square testP＜0.05.All the items reached consensus in the second round (Table 1 and supplementary Table 4).

For ease of reference during doctor-patient/family conversations on LST,the items were further designed into a checklist (supplementary Table 5).

DISCUSSION

By means of a Delphi process involving two rounds,we identif ied domains and items of extreme importance during SDM and informed consent on LST in emergencies.A checklist was created as well.

Most of the key elements that have been proven and well developed in previous studies[1,5,7,8,11-16]were conf irmed in our study.Communication skills are necessary to ensure adequate decision support,including translating information into plain language,maintaining an open attitude,and showing empathy in dialog.[8,11,15]Disclosure of medical information and decisional assistance are the most important domains in the main SDM process.With regard to medical information disclosure,“rationale of diagnosis”,“disease severity and prognosis”,“risk and benefits of various LST,alternatives,and treatment-related costs”,and “medical uncertainty” all entered the final version of the checklist as standards of medical informed consent.[1,5,7,11,16]T he novel framework and checklist might differ in several ways from those in previous studies.

First,although SDM has broad application in emergencies,[13]China and several other Asian countries have specific cultural and socioeconomic backgrounds,including religious beliefs,health care insurance systems,social interaction patterns and so on.The Western principle of autonomy demands self-determination,assumes a subjective conception of the good,and promotes the value of individual independence,while the East Asian principle of autonomy requires family determination.[17]Regarding patient autonomy in our study,the “patient’s value” or“patient’s attitude toward LST” was not emphasized and was represented by family members’s attitude during LST conversations.The health problem is taken as a major crisis for the whole family in China,while the medical decision istaken as “the decision of the family” rather than a personal thing,even though our patients and their families might hold a totally different view of life and death. Patients usually receive financial and emotional support from family members even if they are insured or live alone.At the same time,the patient may be in pain,panic,and anxiety in the context of a life-threatening illness and might not be prepared to make any decisions at all,not to mention decisions on LST.Many patients prefer to leave medical decisions to their providers[18]or are accustomed to referring to their families and friends’ opinions.Directly asking the patient’s wishes was understood as an action that might cause emotional stress to the patients and thus lead to refusal of necessary treatment.This is also written in the Civil Code of the People’s Republic of China[19]:“Where it is impossible or inappropriate to do so,the medical staffshall explain it to the patient’s close relatives and get their express consent.” A considerable number of surrogates will seek aggressive intensive care as their default choice to avoid “giving up hope” on their loved ones.They felt stressed and guilty about the decisions they made and doubt whether they had made the right decisions.[20]Our study was set in such a different background,which made it impossible to indiscriminately imitate the theory and practice of SDM from the Western countries. We amended and modif ied the existing framework to improve its feasibility and adaptability.Attention was given to addressing issues when the whole family,as a medical decisional unit,plays a more important role.Therefore,we designed and conducted this study to identify factors that have an essential impact on SDM in China.

Second,we used a well-designed modified Delphi method and reported data statistics,producing qualitative results,whereas currently most of the research in the SDM field is based on interviews,questionnaires,and observer evaluations.The Delphi technique is widely applied for policy-making,coding of best practice and theoretical framework construction in healthcare systems.[7,8,21]The strengths of this study include a robust protocol-based Delphi method and representation of participants from diverse groups.All 14 participants finished the two questionnaires with a response rate of 100%.The panelists are regarded to be with high enthusiasm.A total of 50% of the respondents provided remarks for the modification of certain items,which indicated that the quality of the questionnaires was considerably high.The median age of the twelve experts was 41 years,and 11 experts served in the health care system for more than 10 years,and nine experts owned the senior title of chief physician or professor.The results revealed that the experts were substantially adept in their respective fields.The Kendall’s W value for indicators tended to lean toward 0.6 with a significant difference.Thus,we concluded that the experts’ results in scoring the index system ultimately achieved consistency.

Third,we not only developed a framework for the SDM process but also provided a ready-to-use checklist for clinicians who had not received formal clinical ethics training or for clinical medical educational purposes.Emergency physicians report using SDM in approximately half of the encounters when they believe it appropriate but cite multiple barriers to its widespread use.[2]The ED code status conversation guide allows emergency physicians to succinctly make patient-centered recommendations for LST.[22]Our checklist might facilitate resident training by enhancing professionalism,patient care,and communication,and help emergency physicians build doctor-patient relationships rapidly and successfully Additionally,the key elements and checklist may offer a few commonplace remarks by way of introduction so that clinical researchers and educators could get the opportunity to develop their own socioeconomically adapted system.

Limitations

Our study has several limitations.First,all Delphi panelists were selected by using snowball sampling,which may have resulted in selection and information bias,but all the experts came from four famous medical colleges located in China.Second,the Kendall’s W coefficient index during the Delphi process was not optimal.Several large-scale Delphi surveys implemented in the health domain concluded that the Kendall’s W consistency coefficient in the last round is generally prone to f luctuate by approximately 0.5,which indicated that the results of the present study were in accordance with those of domestic and foreign research.Third,the feasibility and applicability of the checklist remain to be tested in further studies.

CONCLUSION

Emergency physicians are expected to provide goalconcordant care during acute health decompensation for seriously ill patients.There may not be a universal precept of autonomy shared by both Asian and Western bioethics. The innovatively constructed system created by a multidisciplinary panel of experts was considered to be of good reasonability and scientificity and can be used to optimally support SDM during LSTs in emergencies.The key items and checklist will probably change as new evidence emerges and our understanding of SDM continues to evolve.

ACKNOWLEDGMENTS

We thank all participants,particularly the patients and their families,who contributed their time and expertise to this Delphi exercise while experiencing severe medical conditions.We also thank Professor Mark Siegler (University of Chicago,USA) for his historical work,conscientious faceto-face guidance and for opening the gate of clinical medical ethics for all of us.

Funding:This study was supported by the China Medical Board-Open Competition Program (20-378) and Peking University Third Hospital Fund for Returned Scholars (BYSYLXHG2020004).JX was supported by the Peking Union Medical College Fund for Informatization of Postgraduate Courses (2021YXX001).YLZ was supported by the Sichuan University Graduate Education Reform Project (GSSCU2021046).

Ethical approval:This study was approved by the Institutional Review Board of Peking University Third Hospital (IRB-2021-591).

Conf licts of interest:There are no conf licts of interest.

Contributors:SL conceived and designed the study,and obtained research funding.HXG and QBM supervised the conduct of the study and data collection.SL undertook recruitment of participating panelists and managed the data,including quality control.HZ provided statistical advice on study design and analyzed the data.NS and QBM chaired the data oversight committee.SL drafted the manuscript and all authors contributed substantially to its revision.NS and QBM takes responsibility for the paper as a whole,they are co-corresponding authors.

All the supplementary files in this paper are available at http://wjem.com.cn.