Bringing Home to the Hospital: Development of the Reflection Room and Provider Perspectives

Introduction

It is well understood that the majority of pediatric deaths in the United States occur within the walls of a hospital and mostly in an intensive care unit (ICU).^1–4 Often public sentiment is that patients and families would prefer for end-of-life (EOL) care to take place at home,^1,4,5 but achieving this goal remains elusive. Several factors can influence a family's decision to stay in the hospital or not, including the proximity of their home, socioeconomic concerns, ethnic and cultural influences, the complexity of the medical regimen for the child, and how families often regard the hospital as a second home.^2,4,6 In addition, the provision of pediatric hospice services is not universal,^2,4,6 which affects the reality of a safe discharge home for compassionate EOL care.

Families and professional caregivers are, therefore, faced with the ongoing reality of children continuing to die in the hospital. To address this, our hospital, UCSF Benioff Children's Hospital Oakland, implemented a palliative care consult program in 2011 that embedded an alternative homelike location, called the reflection room (RR), for EOL care, with donor funding. Our hospital serves more than 73,000 children per year, is Northern California's only Level 1 pediatric trauma center, and experiences on average 66 deaths per year. A series of small offices adjacent to the primary palliative care offices were converted to a family suite with a designated bedroom, living room, kitchenette, and children's play area (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Reflection room bedroom/living room.

The goal of the RR was to provide an alternative location for EOL care for the last 8–12 hours of life and for up to 24 hours of postmortem (PM) care. Families are given the choice by the medical team whether they would like to use the RR, but can choose to stay in their assigned hospital room with their child if preferred. The time frame of 8–12 hours for EOL care was chosen so as not burden the medical team, who transfers to the RR with the child, away from the original unit/ward of care. This use of the RR heralded a significant change in practice for providers at our hospital.

The RR aimed to bring a homelike environment to the hospital with minimal medical technology, allowing more space for families to feel comfortable to participate in and maintain cultural and religious traditions while keeping the child's inpatient status. For example, cultural traditions for premortem and PM bathing and dressing require the presence of several family members in attendance. This is aided by the child not having multiple monitoring devices and catheters in place. The palliative care team is available for additional support and anticipatory guidance for providers, the child, and their family. After hours, support is garnered from the nursing supervisor and social worker on call. Patient care staffing remains the same, incurring no additional costs.

Objectives

The purpose of this explorative study was two pronged: (a) to analyze the data collected about the patients who used the RR and (b) to obtain feedback from providers who gave care to patients in the RR because of a change in practice. This was the initial part of a more comprehensive review of the impact of the RR on family experiences whose child used the room. The overall goal was to demonstrate whether this new model of care can be a serious consideration for other hospitals that care for children at EOL. For this initial portion of study, we hypothesized the following:

(1) Providers would report an overall positive experience in providing imminent EOL or PM care to children and their families in the RR.

Materials and Methods

Approval was obtained from the IRB at UCSF Benioff Children's Hospital Oakland. Demographics for use of the RR were extracted from the existing palliative care database and compiled into a summary comparative database (EXCEL) by our research assistant (Y.W.). Palliative care patient characteristics were categorized into four broad groups, as defined by ACT and RCPCH (Association for Children's Palliative Care and Royal College of Paediatrics and Child Health). ⁵ Group 1 encompasses life-threatening conditions in which curative treatment is possible but may fail, for example, cancer and cardiac disorders; Group 2 encompasses conditions in which life can be prolonged with periods of intensive treatment but early death is likely, for example, cystic fibrosis and muscular dystrophy; Group 3 includes progressive conditions in which there is no curative treatment and care is primarily palliative, for example, trisomies 13 and 18 and mucopolysaccharidosis; Group 4 encompasses conditions with extensive neurological disability and associated sequelae, whereby patients may decline unpredictably, for example, complicated cerebral palsy and brain injuries.

A quantitative and qualitative survey was developed based on the four components of the model of reflection for children's palliative care: the preliminary, deductive, inductive, and postreflection phases.^5,7 This model specifically deals with debriefing opportunities for healthcare providers working with EOL circumstances and is integral to evaluating provider experiences with the RR. Our survey assessed the impact of a change in location of care of patients at EOL or PM from the usual location within the hospital (e.g., pediatric ICU (PICU), neonatal ICU (NICU), emergency department (ED), and acute care wards) to that of the RR on a different hospital floor.

The preliminary phase of the survey (questions 1–6) aimed to gather demographics of inpatient providers such as professional capacity, length, and clinical area of experience. The format of multiple choice and directed answers was used, the latter applying a 5-point Likert scale (with responses ranging from strongly agree to strongly disagree). The deductive phase (Part 1, questions 7–11) assessed provider knowledge of the RR, regardless of use. Part 2 (questions 12–16) was specifically for providers who had used the RR and evaluated the logistical effectiveness of the practical components involved in transferring a patient and family to the RR. The inductive phase (questions 17–21) used open-ended questions to give providers a chance to describe advantages and barriers to using the RR as well as suggestions for increased use of the room. The postreflection phase (questions 22–23) aimed to assess whether providers who had used the RR would continue to do so in the future and encourage their peers to consider the same practice. Participants were given the opportunity at the end to share additional comments.

The survey was piloted with selected providers with extensive experience in providing EOL care to children at the hospital. After feedback was incorporated, the survey (utilizing Survey Monkey www.surveymonkey.com) was sent to all providers (n = 912) working with children and families in the inpatient setting, including physician assistants, nurse practitioners, clinical nurse specialists, respiratory therapists, social workers, registered nurses, and physicians including fellows and residents. Participation was voluntary and the survey was emailed twice, four weeks apart. The results were summarized descriptively. Spearman correlations were applied to ordinal variables to test whether provider experience at the hospital (length of employment) and prior use of the RR were associated with likelihood of recommending the RR (either as a valuable resource and/or for future use).

Results

RR demographics

Since its inception in June 2011 through 2014, a total of 116 patients and their families have used the RR, 62% (n = 72) for PM care and 33% (n = 38) for EOL care. Figure 2 describes the locations that EOL care occurred (PICU/NICU, non-ICU, RR). Since inception, up to 23% of all EOL care occurred in the RR with a progressive decrease in EOL occurring in the ICUs. Other unique uses of the RR (n = 6, 5%) have included religious ceremonies or other celebrations of life such as birthday parties. The average length of time that families chose to stay in the room was 9 hours (64%), although 19% chose to stay for more than 19 hours. Patients (82%) were largely transferred from the ICUs (Fig. 3), reflecting literature citing these as the predominant location of death for children in the United States.^1–4 The largest age group of patients to use the room was infants (<1 year) at 64%. Specifically, 34% of the total cohort were neonates (newborn–1 month), consistent with national death statistics (Fig. 4).^4,8

OPEN IN VIEWER

FIG. 2.

Location of end-of-life care at UCSF Benioff Children's Hospital Oakland per hospital area: 2010, 2012–2014.

OPEN IN VIEWER

FIG. 3.

Cumulative number of patients transferred from hospital unit to reflection room, 2011–2014.

OPEN IN VIEWER

FIG. 4.

Number of patients utilizing reflection room per age group by year.

Table 1 describes the diagnostic categories of patients using the RR. The largest ACT/RCPH category (63%) included those with life-threatening illnesses. The remainder of patients fell evenly into categories 2, 3, and 4, reflecting life-limiting, incurable illnesses.

OPEN IN VIEWER

Table 1.

Diagnostic Categories (ACT and RCPCH) of Patients Using the Reflection Room

Year	Category 1	Category 2	Category 3	Category 4	Total
2011	12	2	2	3	19
2012	19	2	2	4	27
2013	14	3	3	4	24
2014	28	8	6	4	46
Total	73 = 63%	15 = 13%	13 = 11%	15 = 13%	116

ACT and RCPCH, Association for Children's Palliative Care and Royal College of Paediatrics and Child Health.

Preliminary phase

Respondents represented a broad selection of healthcare providers (Fig. 5) with the overwhelming majority having more than five years experience (14% 5–10 years, 12% 10–15 years, 47%> 15 years). Interestingly, most respondents worked primarily in the ICUs (27% PICU and 37% NICU) with the remainder across oncology (11%), acute care (13%), subspecialty teams (9%), and the ED (3%). In addition, the majority of respondents had cared for a child at EOL multiple times (56%> 10 times, 20% 5–10 times, and 9% 3–5 times). Although most respondents (65%) agreed that the optimal location for EOL and/or PM care for children should be their own home (with 26% choosing a community inpatient palliative care center), the vast majority (90%) supported a preference for use of the RR rather than the hospital unit (71% strongly agree and 19% agree).

OPEN IN VIEWER

FIG. 5.

Percentage of total survey responses by healthcare profession.

Most respondents (86%) agreed that the main purpose of the RR was for imminent EOL and PM care. Despite most respondents knowing about the existence of the RR with hospital-wide announcements and open houses, ∼35% felt that they had an inadequate introduction to the use of the room, and 23% never had the opportunity to use it first hand in an EOL experience.

Discussion

Suggestions for a private home-like space for EOL care are documented^8–11 and use of a “cold bedroom” (or special room kept at <50F to slow physical deterioration after death) in British inpatient palliative care centers are reported as having a positive impact for families. ¹² Importantly, research has recognized the value of allowing families extended time with their child after death, positively impacting grief and bereavement. ¹

Our research has shown the RR to be a viable, alternative model of care. Our primary results demonstrate how use of the RR allowed for a better patient care experience away from an intensive-type environment. Providers identified improved family privacy, increased visitor allowance, and a peaceful environment as the top benefits of the RR. The major disadvantages remain primarily logistic: feeling distant from the unit of transfer, challenging to manage patient assignments in two locations, and cumbersome transportation of medicines necessary for symptom management. Despite these concerns, our results still support the hypothesis that providers would report a positive experience in providing EOL and PM care to children and their families in the RR and, in turn, would recommend this venue to peers. The study's strengths include the number of respondents (>200), the variety of providers who responded (Fig. 5), and their depth of clinical expertise.

The results reflect an increased comfort of providers introducing this location to families coupled with an increased familiarity and practice with managing EOL care and grief in a different location. Involvement of the palliative care team for family and/or provider support may have contributed to this increased comfort. However, ongoing education regarding room purpose and use remains a priority. Familiarity with RR use was a key driver impacting providers' likelihood of suggesting the RR to peers and patients alike. Indeed, our statistical analysis supports a “seeing is believing” aspect to the RR, with the need to have providers experience this space first hand.

Multiple benefits exist. The child and family continue to receive care and support from their original care team, enhancing familiarity and consistency of care, even under extenuating circumstances. For example, Federal and State laws require certain cases to be referred to the medical examiner (ME), for example, suspected nonaccidental trauma or medicolegal concerns. If ME permission is obtained, the RR can give families much needed time and support previously not possible. When a critically ill newborn is emergently transferred before the mother has been discharged from the hospital of birth, transfer of the mother can occur, as ongoing maternal recovery is possible because of the “rooming in” nature of the bedroom. Or, if a child experiences sudden death in surgery, more prolonged psychosocial and grief support can be offered to the family. Families can also receive assistance with funeral planning, allowing the body to be retrieved directly from the RR.

Other proposed benefits include the gift of time to families, including time to:

(a) spend one last night with their child, even PM;

One final consideration is the location of the RR at our hospital, which is on a floor different from the ICU/ED/pharmacy. Future RR type design should consider room size, physical access to the patient, and nearness to the ICU/ED/pharmacy. The largest study limitation was the lack of a validated survey tool. Although our survey yielded 201 respondents, there may have been selection bias, whereby only providers interested in EOL or PM care responded as indicated by responses largely from ICU staff (64% of responses) and an 80% use of the RR by ICU patients. Furthermore, our survey was limited to providers and not families themselves.

Conclusion

This study demonstrates how hospitals could consider establishing an alternative location for EOL and PM care such as the RR. We conclude that a private inpatient EOL and PM care location, such as the RR, can improve provider experiences around these exceedingly stressful times. Future research includes plans to evaluate family responses to the RR and to validate our findings in other hospitals. It is hoped that having a space like the RR will create a culture of increased compassion and care from providers, as well as more positive memories for families who experience the death of their child.