Background

There is growing evidence that religion and spirituality play an important role in medical decision making. Several studies have found that patients who rate themselves higher on various dimensions of religious experience tend to favor more aggressive care at the end of life (Balboni et al. 2007; Phelps et al. 2009; Sullivan et al. 2004; Van Ness and Larson 2002). For example, studies of cancer patients have found that positive religious coping is associated with a higher use of life-sustaining care in the last week of life (Phelps et al. 2009), and religious support from one’s faith community or the medical team may impact the decision-making process in complex ways (Balboni et al. 2013).

During hospitalization, older adults may face difficult decisions such as when to continue or withdraw life-sustaining therapy at a time when they have lost decision-making capacity due to acute illness, delirium (Inouye 2006), or dementia (Hebert et al. 2001). It is estimated that 40 % of all hospitalized adults (Raymont et al. 2004) and 47 % of those 65 and older (Torke et al. 2014) are unable to make decisions. Close to death, 30 % of older adults who require medical decisions lack capacity to make them (Silveira et al. 2010). In such cases, family members or other surrogate decision makers, are often asked to participate in making major medical decisions.

Given the importance of religion to patient medical decisions, it is possible that religious and spiritual beliefs will play an important role in surrogate decision making as well. Previous studies have noted that beliefs about the value of life, patterns of religious coping, and/or the support and guidance of a religious community are important to surrogates (Braun et al. 2008; Elliott et al. 2007) and have identified faith as one of several important factors in decision making (Boyd et al. 2010; Zier et al. 2008). One study found that over 50 % of the public believes that divine intervention could save a family member from a major trauma even when physicians have determined care is futile (Jacobs et al. 2008). A study of cancer patients found that both patients and their caregivers rank faith in God as the second most important factor in decision making after the physician’s recommendation (Silvestri et al. 2003). In spite of its importance, we know little about how surrogates incorporate religion and spirituality into the specific decisions about medical treatment. We used open-ended interviews to better understand the role of these factors for family member surrogates making major medical decisions on behalf of hospitalized older adults.

Methods

Study Design

We conducted individual, semi-structured interviews with surrogate decision makers of hospitalized older adults. The study was approved by the Indiana University-Purdue University Institutional Review Board.

Setting and Participants

Participants were recruited through a larger, prospective observational study of surrogate decision making (Fritsch et al. 2013; Torke et al. 2012, 2014). We recruited adults age 65 and over from the hospitalist and medical intensive care unit services of an urban, public hospital, and a university-affiliated tertiary referral center in the Indianapolis area. Research team members were notified of new admissions via the electronic medical record for all hospitalized adults age 65 and over. A brief interview was then conducted with the patient’s primary inpatient physician to determine study eligibility, defined as a hospitalization of at least 24 h, medical team consideration of a major medical decision during the first 48 h of current hospital admission, and presence of a surrogate decision maker due to patient impaired cognition from any cause. For this study, major medical decisions were defined as those concerning (1) procedures and surgeries; (2) life-sustaining treatments including code status and intubation; and (3) hospital discharge to a skilled nursing facility or similar institution (Torke et al. 2008). Informed consent was obtained from the surrogate decision maker or the patient, based on the treating physicians’ judgment of capacity to consent.

Data Collection

We approached the primary surrogate decision maker for patients who were judged by their physician as being partially or totally unable to make medical decisions. Surrogate interviews were conducted within 1 month of the patient’s hospital admission. If the patient died before the interview, we waited 2 months before approaching the surrogate, as has been done in other studies (Mitchell et al. 2006). A semi-structured interview guide (Table 1) was used and included questions about the experience of decision making and communication within the hospital, as well as specific questions relating to religion and spirituality. Interviews were conducted by one of the investigators (AMT, KM) or by a trained research assistant (SB), in the hospital or the surrogate’s home.

Table 1 Interview guide
Full size table

Our approach to data collection followed an iterative process that interspersed data collection with analysis, and modification of the sampling and data collection strategy based on initial findings. This iterative approach is consistent with standard qualitative research methods (Giacomini and Cook 2000; Lindlof and Taylor 2002). The first 35 interviews asked generally about factors important in decision making but did not specifically mention spirituality and religion. We found that religion and spirituality were often briefly mentioned but not addressed in detail by participants. Because of prior research evidence and our clinical experience suggesting that faith is often an important factor in medical decision making, we revised our interview guide to include specific questions about these topics.

After seven additional interviews, we further modified our sampling process to include only surrogates whom physicians identified as being the sole decision maker in order to have representation from this group. Based on our preliminary finding that religion played a much larger role on decisions related to life-sustaining treatments, we then limited the types of medical interventions to those pertaining to life-sustaining treatments. We conducted 4 additional interviews until reaching theme saturation, for a total of 46 interviews.

Data Analysis

Interviews were digitally audio-recorded and transcribed verbatim. Four investigators (AMT, KM, KG, SB) read and independently coded each transcript using approaches to coding guided by grounded theory (Strauss and Corbin 1998). All investigators met to review the emerging categories, organize codes into major themes, and discuss any discrepancies until a consensus was reached. At each meeting, theme saturation was assessed and eventually obtained when new interviews no longer yielded new themes.

We took the following measures to ensure credibility within our study: All data were independently coded by multiple investigators; an interdisciplinary team was used for analysis that included a practicing physician with bioethics training (AMT), a medical student with a public health background (KG), and a theologian with leadership experience within a large university healthcare organization (SI); and finally, we used an iterative process of data collection and analysis that continued until no new themes emerged.

Results

We approached 119 eligible surrogates for our interviews and conducted a total of 46 interviews (Table 2). The most frequent reason for refusal was not having enough time. Our participants were 76 % female, 50 % white, and 50 % African-American, and most commonly daughters. Only 20 % of surrogates had a living will documenting prior preferences for care at the end of life. Religion was discussed in 21 of the first 35 interviews and in all subsequent interviews. Forty-one were conducted within 30 days of admission, and five were delayed due to patient death (range 98–117 days from hospital discharge).

Table 2 Characteristics of surrogate decision makers (n = 46)
Full size table

We identified three major themes related to the role of religion and spirituality in decision making: (1) religion as a guide to decision making, (2) control, and (3) faith, death, and dying.

Religion as a Guide to Decision Making

Many individuals described the importance of their own religious beliefs and faith during the decision-making process. However, even people who self-reported as “religious” only regarded certain decisions as requiring spiritual guidance, particularly those that the surrogate perceived to be related to life and death. One man described his approach to making a decision about his mother’s code status (Table 3, Quote 1):

Table 3 Themes and representative quotes
Full size table

[I] went to church, early service, and, uh, I asked to speak with one of the pastors. And, uh, so my wife and I went and talked to them about, you know, making those life or death decisions.

He then went on to describe a previous decision to put his mother in hospice as “separate” from religion (Quote 2):

You know, it’s just to provide care.

During this interview, the surrogate mentioned that he felt pressured by the hospital staff to make a decision about code status before he was able to have a religious consultation.

One woman struggled with how to reconcile her Catholic faith with decisions about her mother’s pacemaker (Quote 3). This surrogate perceived the initiation of the pacemaker as morally different than turning it off, and she saw the latter as inconsistent with her Catholic faith. While she perceived stopping the pacemaker as generally unacceptable, she also could imagine circumstances, such as great suffering, in which stopping the pacemaker may be consistent with her faith.

Those who felt strongly about using religion during their decision making often sought guidance from clergy to make sure their choices were consistent with their faith. It seemed that surrogates most often consulted their personal religious leaders. However, some involved the hospital chaplains in their discussions.

Surrogates who defined themselves as spiritual rather than religious did not invoke faith beliefs directly when making decisions. They discussed their spirituality more generally (Quote 4):

I think spirituality has a lot to do with healing.

Control

Surrogates’ perspectives varied on whether they perceived God or human beings as being in control of the patient’s outcome. We found several examples of how this sense of control impacted the decision-making process.

God in Control

Many surrogates felt that a person’s time of death was under God’s control, and some expressed a belief that this was predetermined by God (Quote 5, 6). Some surrogates felt that God answers prayer and appeals through prayer could impact the health outcomes of their loved ones. Others talked about God’s purpose in their role as caregiver (Quote 7).

Shared Control

Some surrogates ascribed control to a combination of God and human efforts. One woman commented (regarding her mother; Quote 8):

She’s strong so God will pull her through.

This seemed to describe belief in both the patient’s strength and the will of God. Similarly, another surrogate described a conversation with a physician as she contemplated ventilation for her mother (Quote 9). In this case, letting God take over seemed to involve stopping certain medical therapies. The surrogate also described the physician’s discomfort with making the decision about when to stop aggressive care.

The same surrogate who struggled with her Catholic faith and decisions regarding her mother’s pacemaker perceived that two priests who affirmed her decision were providing confirmation from God (Quote 10). The surrogate took ownership of the decision, while at the same time perceiving that the priests’ visits provided confirmation that her decision was consistent with God’s wishes.

Surrogates often prayed for help as they tried to make decisions for their loved one, expressing a sense of their own role in decision making as well as a hope for divine guidance (Quotes 11, 12).

Human Control

Other surrogates felt that the patient was in control of the situation based upon the previous expression of desired healthcare wishes. Even surrogates who placed an emphasis on the role of God seemed to ultimately defer to the patient’s wishes when they were known. For example, after several consultations with his own clergy about disconnecting his mother’s ventilator, a surrogate described the following (Quote 13):

And I says, well, Mom, are you telling me that you’re ready to die? Is that your decision? And then she shrugged her shoulders. So, then I felt pretty confident that she really understood what I was saying.

In the end, this surrogate justified the decision to disconnect the ventilator based on his mother’s own preferences.

Faith, Death, and Dying

We found two important but contrasting themes about death and the dying process. Some people spoke about acceptance of the eventual death of their loved one based on their faith. A surrogate spoke about accepting the death of his father, a church deacon, in the following manner (Quote 14):

Well, I just know that…or I feel that he’s lived a life where…if he passes, then he will be with the Lord.

Others described how this mind-set helped them as they made difficult medical decisions. This surrogate decided to pursue a DNR order for her mother (Quote 15).

…I feel that that will be the best decision for her and if her heart was to stop beating, I feel like that God was calling her home… To me, that’s God’s doing so I wouldn’t want to mess with God’s plan.

A surrogate in a similar situation was focused on death as an end to her mother’s suffering (Quote 16). In contrast, other surrogates described how their faith led them to focus on the survival of their family member. Some surrogates felt that being faithful meant hoping for the patient’s recovery (Quote 17). For many, recovery was attributed to God’s will (Quote 18). One surrogate felt that withholding care was in opposition to her religious beliefs (Quote 19). She relayed a discussion where her family was considering choosing a DNR status for her mother. The surrogate challenged her family’s desire to change a code status to DNR by asking them, “Don’t you believe in miracles?” She expressed that changing the code status was a sign of inadequate faith.

Discussion

This interview study of surrogates making medical decisions for hospitalized older adults found that religious considerations were important elements of decision making for many surrogates, especially when making life and death decisions regarding code status or withdrawal of care. Our findings add to the growing literature suggesting that religion and faith may be important for surrogate decision making (Boyd et al. 2010; Zier et al. 2008). We also found that allowing time and guidance from religious leaders is important to the decision-making process of some surrogates. This builds on prior work suggesting spiritual care is important to ICU family members (Abbott et al. 2001; Wall et al. 2007) and should be a part of quality care for critically ill patients (Clarke et al. 2003; Davidson et al. 2007).

An important theme that emerged from our interviews was the perceived role that God played in the outcomes of medical care. Specifically, surrogates varied in whether they ascribed control of the situation to God or to people, including the patient, clinicians and themselves. Surrogates expressed belief in varying levels of control, ranging from full belief in God’s ability to intervene to a belief in human control. This variation in beliefs about control can be understood using Pargament’s typology of religious coping styles. Some surrogates perceived both that God played a role and that their actions mattered, consistent with Pargament’s collaborative religious coping; others expressed that the time of a patient’s death was entirely in God’s hands, suggesting passive religious deferral (Pargament et al. 2000).

We found several instances where the surrogate expressed preferences for the patient’s care that were based on the surrogate’s own religious beliefs, but deferred to the patient’s own preferences when those were known. This suggests that surrogate beliefs are often balanced against or superseded by patient preferences. This is consistent with standard ethical guidelines for surrogate decision making that advocate relying on substituted judgment, in which the surrogate attempts to make the decision the patient would have made (Buchanan and Brock 1990). Based upon our study, it would seem that patient autonomy and substituted judgment is influential even for those who are concerned with aligning their decisions with their faith. In many cases, the surrogates expressed that patient and surrogate had similar religious beliefs, such that the surrogate’s statement of beliefs may have been endorsed by the patient. However, our semi-structured interviews found evidence that surrogates’ own beliefs about religion and spirituality have an impact on some medical decisions. Surrogate religious belief is a source of value that is not included in standard frameworks for surrogate decision making (Buchanan and Brock 1990). It is important for ethicists and clinicians to consider the appropriate role for these beliefs in major decisions, given our finding that they are a part of the decision-making process for some surrogates, and may lead to different decisions than if the surrogate were relying only on the patient’s wishes or best interests.

We found that religious beliefs impacted decisions about end of life care in complex ways. Two of our surrogates who attributed a high degree of control to God had very different interpretations of this control; one hoped for a miracle and wanted to pursue more aggressive treatment, while the other perceived stopping aggressive treatments as shifting control from physicians to God by “letting God take over.” Although prior studies have demonstrated that those who are religious are more likely to desire more aggressive medical care overall, there is also evidence for very diverse perspectives in prior studies. In a study of cancer patients, the vast majority of all patients, including those with high religious coping, received hospice and avoided life-sustaining interventions at the end of life (Phelps et al. 2009). Another study found patients with a high degree of religiosity or spirituality were equally likely to have an advance directive or do not resuscitate order as their less religious peers (Karches et al. 2012). A qualitative study found that some patients use religious language to describe how they would reject further life-sustaining treatments or desire hospice or palliative care for themselves because of their belief in heaven or their belief that avoiding medical technology is consistent with a divine plan (Branch et al. 2006; Torke et al. 2004).

Most participants in this study were Christian, a religion for which death and the afterlife are central elements of theology. Christianity offers the contrasting themes of the inherent value of human life and the death of the body as the beginning of eternal life (Rees 2001). Our largely Christian sample described both of these views. This variation has also been noted among religious African-Americans who may either struggle against death or perceive death as going “home,” two contrasting views of death that are both consistent with their faith (Crawley et al. 2000). Although there is evidence that religious individuals prefer more aggressive care on average, the evidence for diverse perspectives should lead clinicians to inquire individually about a surrogate’s religious views and how they might influence decision making (Sulmasy 2009).

There were limitations to our study. It was conducted in two hospitals in a single metropolitan area, and their identified religion was almost exclusively Christianity. Our population only consisted of those with an African-American or white race. Fewer than half of approached subjects agreed to be interviewed; it is possible that there were differences in the religious and spiritual beliefs of eligible surrogates who declined. This qualitative research focuses on identification of important themes described by participants, but we were not able to establish the most common or prevalent beliefs.

In conclusion, our qualitative study of surrogate decision makers found that surrogates making medical decisions for a close family member or friend often invoke their religious beliefs, especially when facing end of life decisions. Further study is needed to determine if allowing time and support for religious considerations would improve the quality of medical decisions. Second, preferences for aggressive or palliative care cannot be anticipated based on the presence of strong religious beliefs. This may be challenging for the clinician as it may require more time and more detailed communication; however, it is worth noting that this mainly occurs when life and death decisions are to be made. We believe further interventions are needed to study the impact of religious counseling and support to families of seriously ill patients.