Identifying the Types of Support Needed by Interprofessional Teams Providing Pediatric End-of-Life Care: A Thematic Analysis

Abstract

Background:

Despite the number of interprofessional team members caring for children at the end of life, little evidence exists on how institutions can support their staff in providing care in these situations.

Objective:

We sought to evaluate which aspects of the hospital work environment were most helpful for multidisciplinary team members who care for patients at the end of life and identify areas for improvement to better address staff needs.

Design:

Qualitative thematic analysis was completed of free-text comments from a survey distributed to interprofessional staff members involved in the care of a recently deceased pediatric patient. A total of 2701 surveys were sent; 890 completed. Free-text responses were provided by 306 interprofessional team members.

Setting/Subjects:

Interprofessional team members involved in the care of a child who died at a 348 bed academic children's hospital in the Midwestern United States.

Measurements:

Realist thematic analysis of free-text responses was completed in Dedoose using a deductive and inductive approach with line-by-line coding. Descriptive statistics of demographic information was completed using Excel.

Results:

Thematic analysis of the 306 free-text responses identified three main support-related themes. Interprofessional team members desire to have (1) support through educational efforts such as workshops, (2) support from colleagues, and (3) support through institutional practices.

Conclusions:

Providers who participate in end-of-life work benefit from ongoing support through education, interpersonal relationships, and institutional practices. Addressing these areas from an interprofessional perspective enables staff to provide the optimal care for patients, patients' families, and themselves.

Introduction

Over 60% of pediatric deaths occur in a hospital setting¹ with staff members of differing backgrounds providing care for these children in their last hours. Despite interprofessional provider involvement in end-of-life care, limited evidence exists exploring their experiences.² Nursing perspectives have been analyzed most, with research indicating that both adult and pediatric nurses experience complex, often negative, emotional responses following the death of a patient. These may include startle responses, nightmares, sleeplessness, and loss of objectivity.^3,4 Nurses in these studies noted that institutional support was often unavailable to address these emotional issues and cite the need for further education, team building, and debriefing.⁵ Further characterization of this “gap in support” may enhance response efforts addressing these concerns described in the nursing population, but a knowledge gap exists regarding other members of the care team. Little is known about the needs and experiences of the broader interprofessional group, including physicians, respiratory therapists, and social workers, who are also likely to be impacted by a child's death.²

One response to alleviate the stress associated with end-of-life work is designing and implementing support strategies to bolster individual resilience in the setting of challenging experiences. Devising such an approach is not new, although most such interventions have focused on individual professional backgrounds.^6–9 Our institutional response was the development of an interprofessional workshop focusing on practical aspects of end-of-life care alongside sensitive language and communication approaches. To understand the impact of this training, a survey was designed exploring participants' perceptions of the workshop's effect on care provision and experience with a recent and specific death. Both workshop attendees and nonattendees were surveyed regarding their responses to the training (if applicable), as well as the death experience. Quantitative survey results are reported elsewhere¹⁰; free-text comments are discussed in this study. We aimed to determine both workshop attendees' and nonattendees' perspectives on which support measures facilitated their ability and comfort in providing pediatric end-of-life care, as well as areas for improvement.

Materials and Methods

This qualitative realist thematic analysis was deemed IRB exempt (HUM00116059) as part of ongoing institutional quality improvement work associated with the end-of-life care experience.

Survey tool

The full survey was developed by several members of our institutional Pediatric Palliative Care Team. The survey was reviewed for face and content validity by members of the Pediatric Palliative Care Team not involved in the study, as well as a focus group of physicians and nurses, before being piloted for feasibility and reliability.¹¹ Following a successful pilot, no changes were made. Demographic and situational data were collected, including details on case and participant, type of death, and past experience providing end-of-life care. In addition, respondents were asked about their agreement/disagreement with statements regarding impact of the training and the death experience. This analysis will focus on the last survey question, an optional, general¹² free-text comment section stating “Please feel free to share thoughts, comments, or questions about your experience surrounding the patient's death or about this survey.”

Sample

This study was conducted at a Midwestern academic children's hospital between October 2014 and May 2016. All providers (n = 2701) identified through the electronic medical record as part of the care team or authoring documentation for a patient within the 24-hour period leading up to the patient's death were contacted for participation.

Data collection

The survey was e-mailed using Qualtrics¹³ within one week of the death. Surveys were distributed to all staff members who provided care preceding the deaths of 170 pediatric patients at our hospital. Providers participating in more than one death during the time period were able to respond to the survey for each patient for whom they provided care. Free-text responses ranged from one to nine sentences in length, with a median length of four to five sentences.

Data analysis

Deidentified free-text responses were imported into Microsoft Excel (2016) (P.K.), and descriptive statistics was completed (S.K.). Realist thematic analysis of the data followed Braun and Clarke's approach.¹⁴ In thematic analysis, unlike grounded theory which uses theoretical or thematic saturation as a criteria, themes are considered robust when they work well in relation to the coded extracts and the entire dataset and are coherent and distinct.^14,15 Dedoose¹⁶ was used to organize data and improve transparency and credibility of the analysis through a systematic and comprehensive audit trail.^17,18 C.R. became familiar with the free-text content. Comments related to support were identified deductively (C.R.). An inductive approach was then taken to line-by-line coding of these comments and development of initial themes (C.R.).¹⁴ Codes and themes were identified and refined by C.R. in discussion with members of the research team, S.K., P.K., J.F. Throughout analysis, reflexivity allowed for critical reflection, challenged authors' assumptions and interpretations, and decreased bias.¹⁴

Results

Participant characteristics

Of the 2701 surveys, responses were initiated by 974 individuals (33% response rate); 880 were analyzed after screening for survey completion and accurate subject identification. Three hundred eighty-five respondents (43.8%) had attended at least one end-of-life workshop. Free-text responses were made by 306 (34.8%) of the survey respondents. Of those providing free-text responses, 124 (40.5%) had participated in the end-of-life workshops.

A variety of respondents with different training backgrounds provided comments (Table 1). Each response was treated as an individual end-of-life experience, whether or not one respondent commented on multiple deaths. Deaths in critical care settings were commented upon most often (Table 2). Comments largely related to a perceived planned withdrawal of life support or transition to comfort care (n = 166, 54.2%) and attempted resuscitations (n = 102, 33.3%).

Table 1.

Respondents' Roles

Respondent role	All surveys (N = 880), n (%)	All free text (N = 306), n (%)
Nurse (RN)	210 (24)	101 (33)
Physician (MD)	309 (35)	88 (29)
Respiratory Therapist (RT)	120 (14)	31 (10)
Physician Assistant/Nurse Practitioner (PA/NP)	32 (4)	14 (5)
Social Worker (SW)	79 (9)	25 (8)
Spiritual Care Provider (SC)	56 (6)	12 (4)
Other (i.e., Child Life [CL], ECMO technician [ECMO])	74 (8)	35 (11)

ECMO, extracorporeal membrane oxygenation.

Table 2.

Respondent Locations

Units of responses	All surveys (N = 880), n (%)	All free text (N = 306), n (%)
Pediatric Intensive Care Unit (PICU)	333 (38)	120 (39)
Children's Emergency Services (CES)	15 (2)	4 (1)
Pediatric Cardiothoracic Unit (PCTU)	196 (22)	66 (22)
Hematology and Oncology (H/O)	14 (2)	4 (1)
Operating Room (OR)	17 (2)	4 (1)
Neonatal Evaluation, Stabilization, and Treatment (NEST)	16 (2)	6 (2)
Neonatal Intensive Care Unit (NICU)	265 (30)	93 (30)
Labor and Delivery (L&D)	24 (3)	9 (3)

Themes

Realist thematic analysis resulted in three themes related to types of support providers need when providing end-of-life care: (1) educational efforts, (2) interpersonal relationships, and (3) institutional practices. Each theme is discussed in the sections below illustrated by quotes. Respondents are identified by provider type (see Table 1 for abbreviations) followed by a number.

Educational efforts

Given the intended purpose of the survey, many respondents reflected on the utility of these educational experiences. Respondents attending the workshop were grateful to have had an example on which to model their language and behavior. “I never had to pronounce a patient as deceased before, so I literally used the language that was in the video.” MD75. Participating in the workshop helped respondents feel more comfortable providing end-of-life care. “One of the [ICU] nurses asked me if it's okay to keep giving more PRNs…I said, yes, it is okay, confirmed by my many palliative care led end-of-life sessions.” MD9. Having confidence in how to provide optimal care allowed providers to feel they were acting in the best interest of their patients and performing their job well. Lacking confidence created a suboptimal experience for respondents and may have negatively impacted family interactions. “I felt it took my attention somewhat away from the family because I was so worried about doing it correctly.” RN68.

Respondents who had not attended a workshop often communicated a desire to do so. “I connected well with the family… but felt very unprepared. I would like further education and support.” RN9. Even those participants who had attended the workshops commented on the importance of continued education. “The end-of-life workshop that I attended was so long ago that I can't say it was helpful just because so much time has gone by.” RN38. Respondents spoke of the need for regular programming to reinforce and maintain what they had learned during the workshop. “I feel the end-of-life workshop helped me except that it was a long time since taking the class before having the experiences. So that made it difficult. Maybe offer a yearly refresher or mandatory online.” RN68. The underlying goal voiced by respondents was a desire to provide more than just adequate care; they wished to be sensitive, empathetic, and kind during a heartbreaking time. “I felt unprepared to provide graceful end-of-life care.” RN59. Many also commented on the continued emotional burden of end-of-life work and the need for ongoing support if they were to appropriately respond to the needs of patients and families, as well as their own needs as providers. “Assisting families in death experiences never gets easier. Continued seminars for training of staff in these situations are always valuable and appreciated.” CL5.

Interpersonal relationships

In addition to education, many of the comments across disciplines alluded to the support respondents receive from their coworkers and how this support assists them in providing better care. Less experienced providers often turn to those who have more experience caring for children at the end of life for practical guidance and emotional support. “Although I do not have a lot of experience in withdrawing support of a patient, I did have support of my coworkers and used their experience in my direction of care.” RN53. More experienced coworkers also serve as a source of knowledge about actions or strategies that could be adopted to address difficult circumstances. “I have already reached out to my fellow coworkers with more experience…to gain more knowledge and resources.” CL2. Respondents found that being in a safe environment to ask questions encouraged team mentality and distributed the burden of providing challenging care. The validation provided by coworkers also made respondents more confident about the care they provided to patients at end of life.

Many respondents also commented on the importance of informal and formal opportunities for debriefing with their colleagues about their experiences. “Informal debriefing led by [ICU] attendings after the event was helpful for coping with this [death].” MD84. This need for debriefing reflects providers' desire not to ruminate with their emotions, but rather to connect with others able to relate to their experiences. Respondents also felt that individual discussions with colleagues would be beneficial. “One-on-one counseling/debriefing by an experienced nurse or clinician may be helpful for members of the care team who might be really struggling with this scenario.” MD58.

A lack of support also had a strong influence on how respondents reflected on their experience. “I felt unsupported by the physicians that were working alongside of me. This made dealing with a bad situation and a challenging family even worse.” RN16. If respondents are unable to work together as a cohesive team and share the burden of caring for patients at the end of life, staff's ability to perform their jobs is compromised.

Institutional practices

The third theme is centered around the frustration multidisciplinary team members experience secondary to hospital policies and procedures hindering their provision of end-of-life care. Respondents shared examples of inappropriate patient assignments, delays in obtaining comfort medications, and difficulty following confusing orders. “I felt like I was functioning outside of the order parameters in order to keep the patient comfortable and to help the family in their perception of the process.” RN53. Failing to provide appropriate instruction related to end-of-life procedures left inexperienced providers with questions on the proper steps to take during sensitive tasks. “Instructions on how to properly prepare the child's body after family leaves have been removed from body bag kits…if I had not participated in postmortem care prior to this event, I would not have known what to do.” RN69. Crowding in certain hospital areas also arose as an issue. “I think that from a safety perspective, the [ICU] needs its own overflow area where we could easily move [more stable patients].” RN64. Respondents expressed the desire to provide appropriate care to patients and families, but impeding institutional barriers created significant frustration. Still others commented upon the importance of communication about a patient's death. “If the notification [of a patient's death] could be done differently for staff involved in the patient's care, I feel it would help all of us process and grieve in a more effective way.” RN77.

Discussion

While previous studies have examined how staff members are impacted by pediatric deaths, our survey is the first, to our knowledge, to query providers' attitudes about specific patient experiences to identify effective support strategies in place and opportunities for improvement. Our findings reflect the difficulties of providing care for dying pediatric patients and how the medical community is currently addressing those needs, as well as ways in which further support is warranted. The many unsolicited comments pertaining to the need for multifaceted support highlight the impact of pediatric deaths on those who care for them.

Consistent with previous studies,^7–9,19 respondents reported that the end-of-life workshops were beneficial while also identifying areas of improvement. Similar to other research,²⁰ in the area of education we found that there are additional elements which could enhance existing approaches. In addition, increasing frequency of educational activities to encourage providers across a greater variety of disciplines to participate was also mentioned in extant literature.²¹ Increased frequency may also support team cohesion due to a collective foundation of knowledge. Offering a “refresher” course on a more frequent basis was also suggested to encourage knowledge maintenance and higher level skill development. Improved knowledge foundation helps providers feel more prepared²²; such preparedness breeds confidence and improves patient management.^23,24

Also consistent with available literature,²⁵ numerous respondents commented on the importance of relationships with their colleagues in coping with pediatric death. Respondents commented on seeking support from a more experienced colleague within their specialty, suggesting that they may be more comfortable having emotional conversations with colleagues with similar training and circumstances. Notably, none of the support-related comments were from social workers, who may have additional support resources or education. To our knowledge, interprofessional support programs have not been explored, although this may be useful to examine when creating staff wellness initiatives.²⁶ Having team members from a variety of roles sharing the impact of an experience on their professional and personal lives may improve the team mentality and equalize specialties, which can reduce work stress.²⁷ More effective teams have been shown to optimize healthcare workers' performance, increase job satisfaction, decrease costs, and lead to better quality patient care.²⁸

A practical theme which arose during analysis was the number of elements providers identified as barriers to good patient care. Respondents frequently had suggestions to improve the institution from patient assignment strategies to medication ordering and delivery. The number of comments related to needed improvement dwarfed those discussing adequate measures in place. This imbalance may demonstrate the human tendency to focus on negative elements of our experiences²⁹ potentially biasing respondents' recollections of these events. Nonetheless, there is room to improve current resources and processes. Seeking and following through on feedback from staff members may facilitate a more positive work environment and avoid burnout²³ which has been shown to negatively impact self-reported patient care attitudes and behaviors.³⁰

Strengths and limitations

Close-ended survey questions often serve the investigators' aims, even when developed with input from other stakeholders.¹² Free-text responses, although less lengthy than full qualitative interviews, are valuable for identifying important issues not directly queried by the survey, as well as informing future research.^12,31 The length and quality of the free-text responses in our survey varied widely and gave respondents freedom in what they chose to discuss and raised points that may not have been addressed otherwise. It is not unusual to have a minority of respondents write comments.³¹ Those who comment are likely to be more articulate, have more time when completing the survey, or want to make a negative or critical comment.³¹ Conversely, respondents who have difficulty expressing themselves in writing, have limited time to complete the survey, or do not have a more neutral or positive experience are less likely to comment.³¹ These characteristics may have impacted the free-text response number, length, and level of detail in this study.³¹

In qualitative research, validity rests on the transferability of perceptions and experiences to other settings rather than generalizability to a population, thus comments from even a subset of a survey are valuable.¹² Although our results are limited to one center, they are likely applicable to other hospitals, given that deaths occur across most pediatric hospital environments and largely in an ICU setting.

Finally, 20 of the 306 respondents who gave free-text comments also indicated elsewhere in the survey that they did not directly care for the patient 24 hours before death. These comments were included in the overall dataset because of their substantive content and relevance to the research question; respondents were presumably present for the death based on the description they provided. Discordance between the quantitative and free-text responses may be due to unclear survey wording, differing perceptions of what it means to be involved in a patient's care, or the care they provided was before 24 hours before death. In addition, deidentification of the available data prevented us from knowing whether the same provider responded to the survey for more than one patient death, potentially introducing bias.

Implications for research, practice, and policy

Further research exploring staff experiences following the initiation of one (or more) of our findings may help guide the medical community toward optimal strategies to prepare and support interprofessional staff members caring for patients at the end of life. Surveys may not capture the depth and variety of staff perceptions or needs when caring for pediatric patients at the end of life. Future research could also include a focused, timely in-person interview, which may provide more detailed information although the lack of anonymity may reduce truthfulness. Other researchers have uncovered invaluable data regarding the work environment through qualitative approaches which can highlight subtle changes institutions can implement to better serve staff and improve educational measures and mutual respect between members of the interprofessional team.²⁶ An ethnographic approach, for example, could also provide greater insight into providers' day-to-day work lives and allow for the observation of subtle relationships between staff members that can be difficult to capture in a survey format.²⁰ While we can surmise that implementing changes to workshops suggested by commenters would be likely to provide benefit, further data gathering would be necessary to confirm and reassess over time to achieve optimal impact. Finally, as team cohesion is important for care providers and correlates with improved patient outcomes,²⁸ educational content should be expanded to incorporate role specific content for each interprofessional member of the care team.

Conclusion

Care providers involved at a patient's end of life carry considerable burden related to their role. We and others have demonstrated that interprofessional staff are invested in improving the experience for themselves, their colleagues, and, most importantly, their patients. While many institutions recognize the need for further programming for their staff members, organized interprofessional forums are rare.²⁶ Providers who participate in this difficult work benefit from ongoing support in education, interpersonal relationships, and institutional practices. Improvement in these areas will enable the healthcare team to care well for their patients, patients' families, and themselves.

Footnotes

Acknowledgments

The authors thank James Azim and Melanie Halsey for creation and distribution of the survey and Ken Pituch for his assistance in this project.

Author Disclosure Statement

No competing financial interests exist.

References

Feudtner

, Kang

, Hexem

, et al.: Pediatric palliative care patients: A prospective multicenter cohort study. Pediatrics, 2011; 127:1094–1101.

Pulsford

, Jackson

, O'Brien

, et al.: Classroom-based and distance learning education and training courses in end-of-life care for health and social care staff: A systematic review. Palliat Med, 2013; 27:221–235.

Kellogg

, Barker

, McCune

: The lived experience of pediatric burn nurses following patient death. Pediatr Nurs, 2014; 40:297–301.

Spitzer

, Burke

: A critical-incident stress debriefing program for hospital-based health care personnel. Health Soc Work, 1993; 18:149–156.

Dixon

, Vodde

, Freeman

, et al.: Mechanisms of support: Coping with loss in a major children's hospital. Soc Work Health Care, 2005; 41:73–89.

Bateman

, Dixon

, Trozzi

: The wrap-up: A unique forum to support pediatric residents when faced with the death of a child. J Palliat Med, 2012; 15:1329–1334.

Carter

, Swan

: Pediatric palliative care instruction for residents: An introduction to IPPC. Am J Hosp Palliat Care, 2012; 29:375–378.

Brown

, Ford

, Back

, et al.: Improvement in trainee self-assessment after simulation-based end of life communication skills workshop. Am J Respir Crit Care Med, 2014; 189(A5243).

Erickson

, Blackhall

, Brashers

, Varhegyi

: An interprofessional workshop for students to improve communication and collaboration skills in end-of-life care. Am J Hosp Palliat Care, 2015; 32:876–880.

10.

Pituch

, Halsey

, Keefer

, Azim

: E03-D multi-disciplinary pediatric end-of-life training improves staff preparedness and lessens staff distress. J Pain Symptom Manag, 2016; 52:e40–e41.

11.

Del Greco

, Walop

, McCarthy

: Questionnaire development: 2. Validity and reliability. CMAJ, 1987; 136:699–700.

12.

O'Cathain

, Thomas

: Any other comments? Open questions on questionnaires—A bane or a bonus to research?. BMC Med Res Methodol, 2004; 4:25.

13.

Qualtrics, Provo, Utah, USA. 2016. www.qualtrics.com

14.

Braun

, Clarke

: Using thematic analysis in psychology. Qual Res Psychol, 2006; 3:77–101.

15.

O'Reilly

, Parker

: ‘Unsatisfactory Saturation’: A critical exploration of the notion of saturated sample sizes in qualitative research. Qual Res, 2013; 13:190–197.

16.

Dedoose: Web Application for Managing, Analyzing, and Presenting Qualitative and Mixed Method Research Data. Vol 7.0.23. Los Angeles, CA: SocioCultural Research Consultants, LLC, 2016.

17.

Sinkovics

, Alfoldi

: Progressive focusing and trustworthiness in qualitative research. Manage Int Rev, 2012; 52:817–845.

18.

Fielding

, Lee

: Chapter 4 in Fielding & Lee, Computer Analysis and Qualitative Research. London: Sage, 1998, pp. 86–118.

19.

Corcoran

: Evaluation of an educational workshop to increase comfort levels of professional caregivers with end-of-life care. Medsurg Nurs, 2016; 25:103–109.

20.

Alvarez

, Agra

: Systematic review of educational interventions in palliative care for primary care physicians. Palliat Med, 2006; 20:673–683.

21.

Forsetlund

, Bjorndal

, Rashidian

, et al.: Continuing education meetings and workshops: Effects on professional practice and health care outcomes. Cochrane Database Syst Rev, 2009:CD003030.

22.

Markin

, Cabrera-Fernandez

, Bajoka

, et al.: Impact of a simulation-based communication workshop on resident preparedness for end-of-life communication in the intensive care unit. Crit Care Res Pract, 2015; 2015:6.

23.

Sturgiss

, Haesler

, Elmitt

, et al.: Increasing general practitioners' confidence and self-efficacy in managing obesity: A mixed methods study. BMJ Open, 2017; 7:e014314.

24.

Selman

, Robinson

, Klass

, et al.: Improving confidence and competence of healthcare professionals in end-of-life care: An evaluation of the ‘Transforming End of Life Care’ course at an acute hospital trust. BMJ Support Palliat Care, 2016; 6:231–236.

25.

Lee

, Forbes

, Lukasiewicz

, et al.: Promoting staff resilience in the pediatric intensive care unit. Am J Crit Care, 2015; 24:422–430.

26.

, Tsai

: Making good death more accessible: End-of-life care in the intensive care unit. Intensive Care Med, 2016; 42:1258–1260.

27.

Mahon

: A critical ethnographic look at paediatric intensive care nurses and the determinants of nurses' job satisfaction. Intensive Crit Care Nurs, 2014; 30:45–53.

28.

Epstein

: Multidisciplinary in-hospital teams improve patient outcomes: A review. Surg Neurol Int, 2014; 5(Suppl 7):S295–S303.

29.

Baumeister

, Bratslavsky

, Finkenauer

, Vohs

: Bad is stronger than good. Rev Gen Psychol, 2001; 5:323.

30.

Baer

, Feraco

, Tuysuzoglu Sagalowsky

, et al.: Pediatric resident burnout and attitudes toward patients. Pediatrics, 2017; 139:pii:e20162163.

31.

Garcia

, Evans

, Reshaw

: “Is there anything else you would like to tell us”—Methodological issues in the use of free-text comments from postal surveys. Qual Quant, 2004; 38:113–125.