eTable 1. Differences Between Enrolled Patients and Nonenrolled Patients
eTable 2. Keyword Search Terms Used in Identifying Physician Initiation of Religious Conversation
Customize your JAMA Network experience by selecting one or more topics from the list below.
Ernecoff NC, Curlin FA, Buddadhumaruk P, White DB. Health Care Professionals’ Responses to Religious or Spiritual Statements by Surrogate Decision Makers During Goals-of-Care Discussions. JAMA Intern Med. 2015;175(10):1662–1669. doi:10.1001/jamainternmed.2015.4124
Although many patients and their families view religion or spirituality as an important consideration near the end of life, little is known about the extent to which religious or spiritual considerations arise during goals-of-care conversations in the intensive care unit.
To determine how frequently surrogate decision makers and health care professionals discuss religious or spiritual considerations during family meetings in the intensive care unit and to characterize how health care professionals respond to such statements by surrogates.
Design, Setting, and Participants
A multicenter prospective cohort study was conducted between October 8, 2009, and October 24, 2012, regarding 249 goals-of-care conversations between 651 surrogate decision makers and 441 health care professionals in 13 intensive care units across the United States. Audio-recorded conversations between surrogate decision makers and health care professionals were analyzed, transcribed, and qualitatively coded. Data analysis took place from March 10, 2012, through May 24, 2014.
Main Outcomes and Measures
Constant comparative methods to develop a framework for coding religious and spiritual statements were applied to the transcripts. Participants completed demographic questionnaires that included religious affiliation and religiosity.
Of 457 surrogate decision makers, 355 (77.6%) endorsed religion or spirituality as fairly or very important in their life. Discussion of religious or spiritual considerations occurred in 40 of 249 conferences (16.1%). Surrogates were the first to raise religious or spiritual considerations in most cases (26 of 40). Surrogates’ statements (n = 59) fell into the following 5 main categories: references to their religious or spiritual beliefs, including miracles (n = 34); religious practices (n = 19); religious community (n = 8); the notion that the physician is God’s instrument to promote healing (n = 4); and the interpretation that the end of life is a new beginning for their loved one (n = 4). Some statements fell into more than 1 category. In response to surrogates’ religious or spiritual statements, health care professionals redirected the conversation to medical considerations (n = 15), offered to involve hospital spiritual care providers or the patient’s own religious or spiritual community (n = 14), expressed empathy (n = 13), acknowledged surrogates’ statements (n = 11), or explained their own religious or spiritual beliefs (n = 3). In only 8 conferences did health care professionals attempt to further understand surrogates’ beliefs, for example, by asking questions about the patient’s religion.
Conclusions and Relevance
Among a cohort of surrogate decision makers with a relatively high degree of religiosity, discussion of religious or spiritual considerations occurred in fewer than 20% of goals-of-care conferences in intensive care units, and health care professionals rarely explored the patient’s or family’s religious or spiritual ideas.
Religion and spirituality shape surrogates’ and health care professionals’ ideas about sickness, health, and the end of life.1,2 Many patients and family members wish to speak with their health care professionals about religious and spiritual concerns, particularly when patients get sicker and as they approach the end of life.3 Patients who discuss such concerns with their health care professionals report more satisfaction with their clinical care.4-7 For some, such discussions foster appropriate levels of hope and help guide medical decisions.8 Evidence suggests that these patterns are found in the intensive care unit (ICU),9 where attention to spiritual concerns may affect clinical decisions and outcomes for both patients and surrogates.10,11
Experts recommend that health care professionals attend to the spiritual concerns of patients and their family members9,12,13 by using open-ended questions, asking patients to say more, providing empathy,14 and involving spiritual care appropriately.15 However, little is known about the extent to which the health care professionals of critically ill patients follow such recommendations. Understanding how frequently discussions of spiritual concerns take place—and what characterizes them—is a first step toward clarity regarding best practices of responding to spiritual concerns in advanced illness.
We therefore analyzed audio-recorded goals-of-care conversations from a multicenter observational study of family meetings in 13 diverse ICUs across the United States to determine the frequency and characteristics of religious or spiritual statements by health care professionals and surrogate decision makers.
We conducted a prospective cohort study in 6 US medical centers from October 8, 2009, through October 24, 2012, in which we audio-recorded, transcribed, and qualitatively coded physician-family conversations. The primary purpose of the project was to examine how health care professionals discuss prognostic information with surrogates during family conferences in the ICU. This study examines how religious or spiritual considerations manifested during such conversations.
We conducted study activities in medical and medical-surgical ICUs at the University of Pittsburgh Medical Center; University of California, San Francisco, Medical Center; University of California, Fresno, Medical Center; Baystate Medical Center; Harborview Medical Center; and University of North Carolina–Chapel Hill Medical Center. The institutional review board at each site approved the study procedures. The research staff obtained written informed consent from all enrolled surrogates and health care professionals; surrogates provided proxy consent for patients. For study participation, surrogates were paid $20 and health care professionals were paid $10.
We recruited the surrogate decision makers of patients who met the following inclusion criteria: (1) age older than 18 years, (2) lack of decision-making capacity, (3) respiratory failure requiring mechanical ventilation, (4) acute lung injury diagnosed by conventional criteria,16 and (5) either a score of 25 or higher on the Acute Physiology and Chronic Health Evaluation II or more than a 50% chance of long-term severe functional impairment (as judged by the patient’s attending physician). Exclusion criteria were (1) lack of an English-speaking surrogate decision maker who was older than 18 years and able to complete a written questionnaire and (2) being on a waiting list for organ transplantation. eTable 1 in the Supplement highlights the differences between individuals who were enrolled and those who where not.
Surrogates provided basic demographic information, including their religion or spirituality and degree of religiosity, for themselves and the patients via paper questionnaires. Demographic information was also obtained from the health care professional who conducted the family meeting. Clinical details were prospectively collected from the patient’s medical records.
After obtaining permission from all participants, research staff audio-recorded the meetings and the recordings were transcribed verbatim.
To characterize the content of surrogates’ and health care professionals’ religious statements, we used constant comparative methods of qualitative description,17 an approach that is well suited for in-depth detailed analysis of data when conceptual frameworks in the field are not fully formed.18
Two coders (N.C.E. and 1 other person) reviewed all 249 transcripts of goals-of-care conferences and identified all religious or spiritual concepts and terms that were voiced by surrogates. To ensure comprehensiveness, we then searched for statements using a list of religious and spiritual terms (eTable 2 in the Supplement). Statements were excluded from analysis if they contained only colloquial use of a word with alternative religious or spiritual connotations (eg, “Thank God”).
To develop the coding scheme, we used qualitative description to characterize the content of surrogates’ and health care professionals’ religious and spiritual statements. We then supplemented our inductive framework with a review of publications to identify the types of religious and spiritual topics that were previously reported by patients as well as those that were proposed as part of best practices for health care professionals in discussing religious and spiritual considerations.1,3,9,12,13,19-23
Through a series of investigator meetings, we developed consensus on the coding framework (N.C.E., F.A.C., and D.B.W.), which we then modified iteratively when interviews yielded new themes or ideas. Using the final coding framework, one investigator coded all transcripts of family meetings in the ICU that contained at least 1 religious or spiritual statement, as identified in the initial round of coding (N.C.E.). To assess interrater reliability, another individual trained on the coding framework independently coded statements from 10% of the conferences (N.C.E. and 1 other person). Coders were unaware of the demographic characteristics of the conference participants and of the other coders’ results. Overall, the coders had excellent interrater reliability as evidenced by an overall κ of 0.83. We used ATLAS.ti software (Scientific Software Development GmbH) for qualitative data management.
We identified 405 eligible patients during the study period, of whom 275 had surrogates who consented to participate (enrollment rate, 67.9%); 109 patients were not enrolled because their surrogate declined to participate; 21 patients were not enrolled because their attending physician declined. Enrolled and unenrolled patients did not differ in age, sex, or race. A total of 546 surrogates and 150 health care professionals participated in the study. The characteristics of enrolled patients, surrogates, and health care professionals are summarized in Table 1. Surrogates were significantly more likely to be Christian compared with health care professionals (74.0% vs 44.1%; P < .001). Roughly one-third (35.2%) of health care professionals were atheist, agnostic, or had no affiliation.
Of the 275 patients who were enrolled in the study, 259 had physician-family meetings and 249 (96.1%) of these meetings were audio-recorded (mean [SD] duration of recordings, 24.0 [13.8] minutes). The number of health care professionals present in the meetings ranged from 0 to 9 (mean, 1.8). Thirty family meetings had a social worker present; only 2 had a spiritual-care provider present. The mean number of family members (combination of enrolled and not enrolled) per conference was 2.6.
Religious or spiritual considerations were raised in 40 of 249 (16.1%) family meetings. Surrogates were the first to voice a religious or spiritual concept in 26 meetings (10.4%); health care professionals were the first in 14 meetings (5.6%). Surrogates’ religious or spiritual statements arose in diverse contexts, including discussions of prognosis, discussions of the medical situation, when the physician assessed their impressions of the medical situation, when health care professionals inquired about patient values, in response to health care professionals’ expressions of empathy, and in response to health care professionals voicing religious or spiritual ideas.
In the 40 conferences in which religious or spiritual concepts were voiced, surrogates made 59 statements with religious or spiritual content. The content centered on the following 5 broad themes: their religious or spiritual beliefs (n = 34), practices (n = 19), or community (n = 8); their belief that the physician is God’s instrument to promote healing (n = 4); and their reinterpretation of the end of life as a new beginning for their loved one (n = 4). Some statements contained more than 1 theme. Table 2 shows examples of surrogates’ use of each of the themes.
The most common theme in surrogates’ religious or spiritual statements was the belief that God is ultimately responsible for physical and spiritual health (15 of 40 meetings), including statements about miracles. One surrogate said, “Let’s just be clear. It’s a miracle that we’re looking for, for him to come out of this. It’s a miracle.”
Some surrogates expressed the need to participate in religious practices such as engaging in ritual (eg, administering last rites), reading or watching religious or spiritual material (eg, reading the Bible), and praying (10 of 40 meetings). One surrogate stated about prayer, “Just pray that’s all. That’s the best you can do.”
With respect to religious or spiritual community members, surrogates spoke of wanting to involve them in caring for the patient. This included having members of the community visit the patient (5 of 40 meetings). One surrogate noted, “Mother and Daddy’s minister has been here several times already. And they have a very supportive church family. So, we are very appreciative of that.”
Some surrogates indicated their belief that God acts through the physician to cure illness or that the physician is God’s instrument to promote healing (4 of 40 meetings). One stated, “But I know my God’s a big God. And I know He can even guide your hands to do the right thing.”
Surrogates also used religious language to reinterpret the end of life as a new beginning for their loved one. One surrogate said of her mother, “She’s becoming an angel.”
Four themes arose most frequently in health care professionals’ responses to religious or spiritual statements by surrogates. Health care professionals most commonly responded to surrogates’ religious or spiritual statements by speaking about the medical plan for treatment or goals of care, including terminal-event planning and implications for care (15 of 40 meetings). For example, in response to a surrogate who expressed that she was praying for the patient not to require a tracheostomy, one physician stated, “The long-term question is how to prevent the pancreatitis from happening again. It’s not a question for now, but it’s gonna be a question pretty soon, I think.”
In some instances, health care professionals responded with empathic statements (13 of 40 meetings). In response to one surrogate’s statement, “Prayer’s not gonna work,” the physician responded, “Hang in there. I know it’s hard. I know.” As in this example, physician responses rarely directly addressed surrogates’ spiritual or religious language.
The third most common response by health care professionals was acknowledging surrogates’ religious or spiritual statements with closed-ended responses, such as, “Mhmm,” or, “Okay” (11 of 40 meetings).
Few health care professionals responded to religious statements by emphasizing their commitment to high-quality medical treatment, reassuring surrogates regarding their own dedication to the patient with a statement containing more emotional support (4 of 40 meetings). After one surrogate said, “I know my God’s a big God. And I know He can even guide your guys’ hands to do the right thing,” a physician responded, “We’ll do the best with what we’ve got.”
Other themes arose less frequently. In response to surrogates’ religious or spiritual statements, health care professionals in 2 conferences made efforts to explore the patient’s or surrogate’s personal spirituality. In response to one surrogate’s request to continue medical treatment until God asks for the patient, a physician noted, “Okay, so he has some religious beliefs.” In response to another surrogate’s statement that God is the most powerful entity, a physician voiced his own religious beliefs by saying, “I agree. He is more powerful.”
In response to surrogates’ statements about hoping for a miracle (4 conferences), health care professionals either changed the subject (n = 3) or expressed empathy (n = 2; in 1 conference the physician did both). Likewise, in response to surrogates’ statements that the physician is God’s instrument to promote healing or that God acts through health care professionals to cure illness (4 conferences), health care professionals offered a closed-ended response (n = 3) or reassured the family (n = 1).
Table 3 shows the themes and examples of health care professionals’ responses to surrogates’ religious or spiritual statements.
Table 4 shows the themes and examples of health care professionals’ statements initiating conversation about religious or spiritual concerns.
The most common ways in which health care professionals initiated conversations about religious or spiritual concerns was by acknowledging sources of religious or spiritual support, such as offering the hospital chaplain or social worker (13 of 40 meetings), for example, “Would you like me to call our hospital chaplain?”
Some health care professionals initiated conversations about religious or spiritual concerns by asking about the patient’s or surrogate’s religion or spirituality (6 of 40 meetings). This included questions about the importance of religion to the patient or surrogate. One physician asked, “Are there any things, in particular, that are really important now? Does he have religious beliefs?”
Physician’s occasionally spontaneously voiced their own religious or spiritual beliefs to surrogates (3 of 40 meetings). One physician stated, “I do believe in miracles.”
Among a cohort of surrogates with a relatively high degree of religiosity, discussion of religious or spiritual considerations occurred in fewer than 20% of goals-of-care conferences and health care professionals rarely attempted to explore the patient’s or family’s religious or spiritual ideas.
Although we are aware of no other studies that have addressed this topic in the context of surrogate decision making, several studies of physician-patient encounters have found similarly low rates of discussion of religious and spiritual considerations.5,24,25 For example, Phelps et al24 found that three-fourths of oncology patients at the end of life thought spiritual care would have a positive effect on their care and only one-fourth had actually received spiritual care.
There are several potential explanations for the low rates of discussion of religious or spiritual considerations. Health care professionals may avoid these topics because they feel unprepared to engage patients and surrogates.26 Ellis et al27 and Ford et al28 found that a central barrier to discussing such concerns is a lack of training for discussing religion and spirituality. Another possibility is that health care professionals may place less value on such discussion either because of their own attitudes toward religion and spirituality or because they perceive the decisions as depending solely on medical facts instead of including diverse values.26
We also found that when surrogates did raise a religious or spiritual consideration, health care professionals rarely made attempts to explore the topic. Curlin et al29 found that many health care professionals reported changing the subject when patients raise spiritual concerns. By not engaging such concerns, health care professionals may miss opportunities to better understand patients and surrogates, including how they make medical decisions. Studies have found that patients want to discuss their religious and spiritual concerns, and when practitioners do speak with patients about spiritual concerns, patients are more satisfied with the care they have received.8 Furthermore, the discussions are associated with increased referrals to hospice, suggesting that the process of health care professionals discussing religious and spiritual values and concerns may have a tangible effect on patient outcomes, potentially in a more patient-centered direction.5,7,16
Our findings suggest that religious considerations—viewed as important to a large proportion of Americans—are often absent from end-of-life conversations. This may signal a need for changes in health care delivery in ICUs. One possible approach is to redesign health care processes to increase the involvement of spiritual-care providers in end-of-life discussions that involve patients with high degrees of religiosity. Chaplains, who have expertise that is complementary to health care professionals’, may be able to help navigate issues at the interface of religion and end-of-life decision making. In light of existing shortages of chaplains and the high prevalence of patients who cite religious considerations at the end of life, health care professionals may need enhanced training on how to navigate such issues.30 We suggest the development of scalable approaches to teach health care professionals how to explore patients’ religious and spiritual concerns, how to respond to frequent topics, and when to involve spiritual-care providers.
Our study draws from the largest existing cohort of audio-recorded physician-family encounters in ICUs, drawn from 13 ICUs at 6 sites across the United States.
Although we had high interrater reliability, it is possible that other coders would have interpreted the statements in somewhat different ways. It is unclear what contributed to the infrequency with which religious and spiritual topics were discussed in this sample, and we cannot exclude the possibility that surrogates and health care professionals spoke about religious and spiritual considerations outside the audio-recorded goals-of-care conversations. Although there were no differences between enrolled surrogates and those who declined to participate in terms of basic demographic information, we cannot exclude the possibility that there were unmeasured differences that resulted in a sample that may not generalize to all family members in ICUs. In addition, 3 of 4 surrogates were Christian, potentially limiting the generalizability of the study to other religious groups.
Although many patients wish to have their religious values incorporated in end-of-life decisions, our research indicates that religious and spiritual considerations are infrequently discussed during physician-family meetings. Developing strategies to ensure adequate exploration and integration of religious and spiritual consideration may be important for improving patient-centered care in ICUs.
Accepted for Publication: June 11, 2015.
Corresponding Author: Douglas B. White, MD, MAS, Clinical Research, Investigation, and Systems Modeling of Acute Illness (CRISMA) Center, Department of Critical Care Medicine, University of Pittsburgh School of Medicine, 3550 Terrace St, Scaife Hall, Room 608, Pittsburgh, PA 15261 (firstname.lastname@example.org).
Published Online: August 31, 2015. doi:10.1001/jamainternmed.2015.4124.
Author Contributions: Dr White had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Ernecoff, White.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Ernecoff, Buddadhumaruk, White.
Critical revision of the manuscript for important intellectual content: Ernecoff, Curlin, White.
Statistical analysis: Buddadhumaruk.
Obtained funding: White.
Administrative, technical, or material support: White.
Study supervision: Ernecoff, White.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grant R01 HL094553 from the National Institutes of Health (Dr White).
Role of the Funder/Sponsor: The National Institutes of Health had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Kaitlin J. Shotsberger, RN, MSN, and Natalie Gessert Hall, MA, MDiv, Clinical Research, Investigation, and Systems Modeling of Acute Illness (CRISMA) Center, Department of Critical Care Medicine, University of Pittsburgh School of Medicine, provided support during the qualitative analysis. Ms Shotsberger received salary support from National Institutes of Health grant R01 HL094553. Rev Gessert worked on the project as a student intern and was not financially compensated.