SEARCH METHODS AND INCLUSION CRITERIA

For inclusion within this literature review, studies had to be empirically based and published in a peer-reviewed journal or dissertation database. Initially, selected literature had to address both family communication and decision making as observed constructs and had to be within the context of terminal illness or end-of-life care. Due to varying structures of healthcare options at the EoL across countries, reviewed literature was limited to within the United States. After discovering a scarcity of empirical literature that addressed both family communication and decision making at the end-of-life, the criteria were expanded to include secondary articles relating to each of these constructs separately, yet still within the context of EoL care. These secondary areas were included to provide greater depth within this review but are not exhaustive of the current literature in these separate areas. Instead, a synthesis of the relevant literature related to the interrelationship of family communication and decision making is presented. The articles in the secondary areas were similarly identified with the initial articles using the search methods described directly below.

Articles were identified using systematic keyword searches within the following relevant databases: Academic Search Complete, CINAHL Plus, Communications and Mass Media Complete, ERIC, PsychINFO, MEDLINE, SocINDEX, and ProQuest (searches conducted during June of 2013). No timeframe limitations were implemented within the searches; however, resulting articles were published within the past 30 years, with the majority in the past decade. Key search terms included “family communication,” “communication,” “family,” “end of life,” “end-of-life care,” “decision making,” “patient family communication,” “health communication,” and “terminal illness.” Initially, articles were screened by abstract only and saved into files by topic. After search completion, full articles were screened using the above-identified inclusion criteria. Reference lists of key articles were also explored for additional relevant literature. The three bodies of relevant literature that emerged during this review include: (1) the importance of family communication at the EoL, (2) family decision making at the EoL; and (3) the interrelationship of communication (both within the family and with healthcare professionals) and decision making at the EoL. Primary research methodologies that are useful in understanding these relationships are also described.

IMPORTANCE OF FAMILY COMMUNICATION AT THE END OF LIFE

Within palliative care, the patient and family are recognized as the unit of care (Connor, Reference Connor2009), versus focusing on the patient alone. One approach, family-focused grief therapy, recognizes the importance of intervening with families preventatively during palliative care based upon types of family functioning, resulting in reduced distress for bereaved family members (Kissane et al., Reference Kissane, McKenzie and Bloch2006). Outside of palliative care, some argue that family-centered care or collaborative approaches that involve family members are more appropriate than the traditional client-centered approach (Hardwig, Reference Hardwig1990; Hidecker et al., Reference Hidecker, Jones and Imig2009). While patient autonomy is a core value in Western medicine, other cultural groups may not share this value, suggesting an even greater need for involvement of family members (Ballard-Reisch & Letner, Reference Ballard-Reisch and Letner2003; Volker, Reference Volker2005). Research has shown that whites are more likely to prefer autonomous decision making than blacks (Bullock, Reference Bullock2011) and are likely to be more exclusive regarding whom to include in discussions, whereas blacks are more inclusive, often extending participants beyond family to close friends and spiritual leaders (Shrank et al., Reference Shrank, Kutner and Richardson2005). A number of studies have also shown family-centered models of decision making as preferred by Mexican and Korean Americans (Ballard-Reisch & Letner, Reference Ballard-Reisch and Letner2003). Merging person-centered planning with family-focused care is offered as one suggestion for enhancing EoL care and transitions for patients with intellectual disabilities (Kirkendall et al., Reference Kirkendall, Waldrop and Moone2012). Family members are most likely to be the designated person within caregiving; in addition to this involvement, family influence plays a dominant role in a person's development of values, health attitudes, and behaviors (Pecchioni et al., Reference Pecchioni, Thompson, Anderson, Turner and West2006), all of which are key variables in decision making (Shrank et al., Reference Shrank, Kutner and Richardson2005).

With the prominent role of family during illness and at the end of life, family communication is an important consideration. Segrin and Flora (Reference Segrin and Flora2005) define communication as “a transactional process in which individuals create, share, and regulate meaning” (p. 15). Though multiple definitions of family communication exist, the one accepted in the context of this review is “the act of making information, ideas, thoughts, and feelings known among members of a family unit” (Olson & Barnes, Reference Olson and Barnes2003, p. 1). Much of the literature on family communication provides insight into its importance at the end of life. This portion of the review examines the content of EoL conversations, important outcomes of family communication, and the challenges within family communication.

CONVERSATIONS AT THE END OF LIFE

Both patients and caregivers desire more communication at the end of life (Fried et al., Reference Fried, Bradley and O'Leary2005). Even though multiple studies suggest the absence or difficulty of family communication at the EoL (Boehmer & Clark, Reference Boehmer and Clark2001; Fried et al., Reference Fried, Bradley and O'Leary2005; Gotcher, Reference Gotcher1995; Zhang & Siminoff, Reference Zhang and Siminoff2003), others provide insight into the content of final conversations between family members. Final conversations are defined as any communication between the time of diagnosis and death (Keeley, Reference Keeley2007). Badr and Taylor (Reference Badr and Taylor2006) characterized four different areas of relationship talk that emerged from separate interviews with patients and their spouses. The quality of the relationship, relationship memories, planning for the future, and problem solving all became topics of conversation during the cancer experience (Badr & Taylor, Reference Badr and Taylor2006). Messages about love, personal and relational identities, faith, routine interactions, and the difficulty of past relationships are conveyed in final conversations between dying patients and their family members (Keeley, Reference Keeley2007). Conveying love, gratitude, forgiveness, and farewell in conversations at the EoL are also considered important (Byock, Reference Byock1996).

The impact of conversations directly related to EoL topics was a significantly stronger indicator than agreement between patient and family members about the need for communication (Abbey, Reference Abbey2009). In other words, whether patients and families agree on the desired amount of talking does not matter as much as whether the conversation occurs at all. Results from Wittenberg-Lyles and colleagues (Reference Wittenberg-Lyles, Goldsmith and Demiris2012) support this finding in that “conformity in family communication [does] not equate with family agreement or open communication” (p. 25).

OUTCOMES OF FAMILY COMMUNICATION

A number of studies have considered the impact of family communication on patient and/or family functioning through a variety of approaches. Increases in the frequency of, level of honesty in, and satisfaction with family communication along with emotional support had positive impacts on effective adjustments of terminally ill patients (Gotcher, Reference Gotcher1993). Other studies considered the positive effect of family communication on psychosocial adjustment resulting in increased communication about the relationship (Badr & Taylor, Reference Badr and Taylor2006). In the opposite direction, a lack of communication between patients and family members was found to present an increased risk of poor adjustment for prostate cancer patients (Boehmer & Clark, Reference Boehmer and Clark2001). The ability to express emotion, like within the definition of family communication provided above, was associated with a decrease in pain for cancer patients (Dalton & Feuerstein, Reference Dalton and Feuerstein1989). Increased discussion of EoL topics was also associated with less distress and enhanced quality of life (Abbey, Reference Abbey2009).

Communication constraints, which developed due to a desire to protect patients or family members, has been documented as a predictor of family conflict (Kramer et al., Reference Kramer, Kavanaugh and Trentham-Dietz2010). Many studies have considered the impact of family communication on the caregivers instead of on the patients themselves. EoL conversations have multiple positive functions for bereaved family members, such as affirmation of their relationship with the patient, sense of closure of the relationship, validation of beliefs, and reconciliation (Keeley, Reference Keeley2007). Self-efficacy and increased length of caregiving were associated with greater perceived levels of open communication by caregivers, while emotional exhaustion and depression of caregivers were associated with lower perceived levels of open communication (Bachner & Carmel, Reference Bachner and Carmel2009).

CHALLENGES IN FAMILY COMMUNICATION

In addition to positive outcomes, challenges in family communication at the end of life cannot be ignored. Across multiple studies, findings show a large prevalence of patients and families that experience communication difficulties. In a study that surveyed 193 terminally ill patients and their caregivers, 39.9% of caregivers desired more communication along with 20.2% of patients (Fried et al., Reference Fried, Bradley and O'Leary2005). Within the same study, 37.3% of caregivers and 22.3% of patients reported that communication was difficult. Two thirds of a sample of 26 lung cancer patient–caregiver dyads reported communication problems (Zhang & Siminoff, Reference Zhang and Siminoff2003), while in a sample of 78 breast cancer patients and their significant others, one fourth reported strained communication (Lichtman et al., Reference Lichtman, Taylor and Wood1987).

The nature of terminal illness itself causes the potential for communication difficulty due to the physical progression of illness, which diminishes the ability of patients to communicate normally (Planalp & Trost, Reference Planalp and Trost2008; Stone et al., Reference Stone, Mikucki-Enyart and Middleton2012). Patient impairments, alongside the geographical distance between family members, are situational factors that influence difficult communication (Stone et al., Reference Stone, Mikucki-Enyart and Middleton2012). Other natural responses to terminal illness, such as a wide range of emotions (Planalp & Trost, Reference Planalp and Trost2008; Zhang & Siminoff, Reference Zhang and Siminoff2003), denial (Planalp & Trost, Reference Planalp and Trost2008), and family conflict (Kramer et al., 2012; Planalp & Trost, Reference Planalp and Trost2008) also make family communication more challenging. Relational influences—for example, role changes and the influence of multiple family members—are cited as increasing communication difficulty (Stone et al., Reference Stone, Mikucki-Enyart and Middleton2012). Avoidance, belief in the power of positive thinking, and psychological distress (Zhang & Siminoff, Reference Zhang and Siminoff2003) along with the belief that talking about concerns might hasten death or reoccurrence (Lichtman et al., Reference Lichtman, Taylor and Wood1987) are other variables that contribute to complexity in communication.

FAMILY DECISION MAKING AT THE END OF LIFE

Research on decision making at the end of life has invariably discovered family members to be at the center of things (Cohen et al., Reference Cohen, McCannon and Edgman-Levitan2010; Hiltunen et al., Reference Hiltunen, Chase, Medich, Rantz and LeMone1999; Karasz et al., Reference Karasz, Sacajiu and Kogan2010; Tschann et al., Reference Tschann, Kaufman and Micco2003). Though Hiltunen and colleagues (Reference Hiltunen, Chase, Medich, Rantz and LeMone1999) anticipated studying accounts of patients' decisional conflict in their review of narrative accounts from the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT), they found that the majority of decisions were actually made by family members. The presence of family members at the time of death when compared to patients who died alone is an indicator for greater presence of do-not-resuscitate (DNR) orders, documentation of withdrawn treatments, and use of pain medication prior to death (Tschann et al., Reference Tschann, Kaufman and Micco2003). These findings suggest that family involvement increases the use of comfort care within the hospital for dying patients, also implying that family members play a role in making these decisions.

While it is often assumed within the American healthcare system that patients will make care decisions when they are alert and able to, multiple cultures prefer family decision making, or collectivism and interdependence, versus individualism and independence (Candib, Reference Candib2002; Johnstone & Kanitsaki, Reference Johnstone and Kanitsaki2009). In a milestone study regarding culture and the end of life, Blackhall and colleagues (Reference Blackhall, Murphy and Frank1995) found great variation across cultures within the United States (Korean, Americans, Americans, and African Americans) about whether or not a patient should be told of their diagnosis (47, 65, 87, and 88%, respectively) or about a terminal prognosis (35, 48, 63, and 69%). Ultimately, suggestions for providing culturally competent care at the EoL involve the inclusion of family and assessment on the role of cultural preferences related to decision making and other values and beliefs (Bullock, Reference Bullock2011; Stein et al., Reference Stein, Sherman and Bullock2009).

Eight informal roles for family members across four differing intensive care units emerged during an ethnographic study spanning from 2001 to 2004: primary caregiver, primary decision maker, family spokesperson, out-of-towner, patients' wishes expert, protector, vulnerable member, and healthcare expert. Each of these roles was tied to the decision-making process for families at the end of life, often creating a complicated dynamic (Quinn et al., Reference Quinn, Schmitt and Baggs2012). Within this context of decisional conflict (Hiltunen et al., Reference Hiltunen, Chase, Medich, Rantz and LeMone1999), a decision-making process was identified involving several stages: recognition of a dilemma, a period of vacillation, moving to a turning point, and letting go. Additional challenges in decision making for families included the need for multiple decisions throughout the process, the multiple perspectives of family members, and the presence of family members at different stages within the decision-making process at any given time (Hiltunen et al., Reference Hiltunen, Chase, Medich, Rantz and LeMone1999). A second identified process for family decision making at the EoL includes cognitive, affective, and interpersonal steps guiding family members to understand a patient's condition, identify a patient's preferences for treatment, and continue familial roles, respectively. These processes ultimately help family members to understand a patient's unlikelihood of recovery, enabling them to make a decision (Swigart et al., Reference Swigart, Lidz and Butterworth1996).

PROXY DECISION MAKING

As noted above, family members are often called upon to make decisions with or on behalf of patients with advanced illness (Meeker & Jezewski, Reference Meeker and Jezewski2005). Though historically medical decision making was in the hands of physicians who were sworn to the Hippocratic Oath (Meeker & Jezewski, Reference Meeker and Jezewski2005), current models are based upon a model of rational choice, which emphasizes a patient's stated wishes first, followed by substituted judgment about what a patient would decide, and, last, the patient's best interest (Drought & Koenig, Reference Drought and Koenig2002; Karasz et al., Reference Karasz, Sacajiu and Kogan2010; Meeker & Jezewski, Reference Meeker and Jezewski2005). One of the challenges with this model is the overwhelming lack of implementation of “advance directive” and “living will” paperwork, designating someone's choices at the EoL (Bomba et al., Reference Bomba, Morrissey and Leven2011; Fagerlin & Schneider, Reference Fagerlin and Schneider2004; Glass & Nahapetyan, Reference Glass and Nahapetyan2008; Lang & Quill, Reference Lang and Quill2004). Minorities are often cited as completing advanced directives at even lower rates than Caucasians (Bullock, Reference Bullock2011; Cohen et al., Reference Cohen, McCannon and Edgman-Levitan2010; Volker, Reference Volker2005). In observation of the decision-making process between staff and family members of 26 patients, Karasz et al. (Reference Karasz, Sacajiu and Kogan2010) suggested that consideration of best interest took precedence over a patient's wishes. Additional arguments against the use of living wills in decision making include the unpredictability of complex conditions and potential choices, ambiguous understanding of one's wishes prior to facing the need for decisions, and the ability of proxies to interpret their loved ones' documented choices (Fagerlin & Schneider, Reference Fagerlin and Schneider2004). An evaluation of published reports from the SUPPORT study showed that the likelihood of a surrogate understanding their loved ones' preferences was barely better than chance (Covinsky et al., Reference Covinsky, Fuller and Yaffe2000).

Prior to the need for proxy decision making, multiple studies have considered whether or not patients and their designated proxy matched with regard to EoL preferences (Ditto et al., Reference Ditto, Danks and Smucker2001; Parks et al., Reference Parks, Winter and Santana2011; Schmid et al., Reference Schmid, Allen and Haley2010). In a systematic review examining family decision-making experiences, Meeker and Jezewski (Reference Meeker and Jezewski2005) found that proxies demonstrated low to moderate accuracy in predicting choices. Proxy accuracy was significantly higher when a spouse or partner served as proxy versus an adult child or other, and increased accuracy across proxy types was associated with lower family conflict (Parks et al., Reference Parks, Winter and Santana2011). Other variables that increased accuracy included conditions of forced choice and directed use of substituted judgment (Meeker & Jezewski, Reference Meeker and Jezewski2005).

THE INTERRELATIONSHIP BETWEEN COMMUNICATION AND DECISION MAKING

The literature on health communication is heavily focused on the relationship between patients or their family members and healthcare practitioners (Zhang & Siminoff, Reference Zhang and Siminoff2003). Though the purpose of this review is to explore how interaction within a family relates to decision making, it is important to examine the more general context of the relationship between health communication and decision making as well. This section of the review considers the relationship between communication with healthcare providers and end-of-life decision making, the relationship between advance directives (as an instrument for decision making) and family communication, and, finally, literature that links family communication and decision making at the EoL.

COMMUNICATION WITH HEALTHCARE PROVIDERS

One study found that inadequate communication with physicians was the most significant barrier influencing quality at the end of life (Eues, Reference Eues2007). In a study examining 100 patients already enrolled in a hospice program, 30% were not aware of their diagnosis and 62% had no awareness of their prognosis (Andruccioli et al., Reference Andruccioli, Montesi and Raffaeli2007). Other studies have discussed physicians' inability to communicate clearly about diagnosis and prognosis with patients for a variety of reasons. Hospice experts identified physicians' personal discomfort with death and lack of training and experience in EoL care as barriers to communication and work with dying patients and their families (Friedman et al., Reference Friedman, Harwood and Shields2002). Another suggested that the reason is that physicians are reluctant to inform patients about impending death due to anxiety about their own mortality (Feifel, 1969, as cited in Neimeyer et al., Reference Neimeyer, Wittkowski and Moser2004). Higher death anxiety in physicians has been related to greater difficulty disclosing prognoses (Kvale et al., Reference Kvale, Berg and Groff1999) along with consideration of more factors before breaking bad news to patients (Eggerman & Dustin, Reference Eggerman and Dustin1985). An additional concern is that minority patients may receive this information at an even lower rate. Thornton (Reference Thornton2008) found that, after reviewing data from the SUPPORT study, physicians reported having prognostic conversations with only 41% of African Americans compared to 58% of white patients.

Other barriers that contribute to problematic communication between practitioners and patients include our death-denying culture, patient denial, and the focus within the American healthcare system on cure (Hickman, Reference Hickman2002). Ultimately, patients and families cannot talk or make decisions about something they are not even aware of. Hickman (Reference Hickman2002) suggested that inclusion and involvement of family members is one way to overcome barriers and enhance decision making at the end of life.

In contrast to the literature that documents physicians' lack of communication, many empirical studies have attempted to evaluate the effectiveness of the communication that does in fact occur. In interviews with family members who experienced loss of their loved one in an acute hospital, Radway and colleagues (2009) found that decision making in family meetings was one of the three salient experiences that related to emotional burden for caregivers. Additional identified barriers to effective provider–family communication include multiple or alternating physicians in acute settings (Baggs et al., Reference Baggs, Schmitt and Prendergast2012), inadequate time conversing with family, inconsistent use of multidisciplinary teams, and lack of frequent communication (Ahrens et al., Reference Ahrens, Yancey and Kollef2003). Promoting excellent communication between physicians and family members is one way to lessen caregiving burdens (Rabow et al., Reference Rabow, Hauser and Adams2004). Normal communication typically happens in an informal fashion and not until care providers have already reached a consensus that restoration or survival is unlikely (Lilly & De Meo, Reference Lilly and De Meo2000) or until a decision has to be made (Hines et al., Reference Hines, Babrow and Badzek1997). The use of communication interventions involving interdisciplinary team meetings shortly following admission and daily communication from consistent team members has been shown to decrease the length of stay in critical care, to lower hospital costs, and allow for earlier access to palliative care (Ahrens et al., Reference Ahrens, Yancey and Kollef2003; Lilly & De Meo, Reference Lilly and De Meo2000).

ADVANCE PLANNING AND END OF LIFE PREFERENCES

While “advance directives” are meant to be useful in decision making, the complementary variable of communication provides additional benefits. In fact, planning for care at the EoL and making decisions are “fundamentally relational” (Jennings & Morrissey, Reference Jennings and Morrissey2011). Meeker and Jezewski (Reference Meeker and Jezewski2005) found that group decision making among a family was preferred over use of an individual surrogate, indicating the need for communication between family members. Open family communication was found to be an indicator for improved congruence in care preferences between terminally ill patients and their caregivers by Gardner and Kramer (Reference Gardner and Kramer2009). Family members who discussed patients' advance directives with them also reported greater confidence in their ability to serve as a proxy, even though it did not improve accuracy in predicting patients' choices (Ditto et al., Reference Ditto, Danks and Smucker2001). Overall, family communication about EoL preferences serves a greater purpose for decision making than completion of advanced directives alone. Barriers to these discussions, also discussed above under challenges in family communication, include fear of death, trust in others to make decisions, and challenging family dynamics. Prior experiences of death, acknowledgement of the reality of death, and spirituality are facilitators of family communication about EoL preferences (Elliott et al., Reference Elliott, Gessert and Peden-McAlpine2009).

FAMILY COMMUNICATION AND DECISION MAKING

Unfortunately, there are no consistent patterns in how family communication and decision making are studied in relation to one another, though both family and communication are regularly considered in relation to decision making. Babcock and Robinson (Reference Babcock and Robinson2011) present a model of “seven core components of communication and decision making” that includes specific interventions for counselors to assist the family unit in navigating the process of decision making. The seven components include getting ready to work with the family and other systems, assessing the situation, managing conflict, providing information, identifying roles, processing familial responses, and follow through. While this model provides a framework for helping families to communicate during the decision-making process, the processes specific to how the patient and family members communicate among themselves were not considered.

Waldrop and Meeker (Reference Waldrop and Meeker2012) investigated the role of family in their study on understanding decision making surrounding hospice enrollment. In order to be enrolled in the study, patients had to score 40 or above on the Palliative Performance Scale (PPS), indicating an ability to communicate. Family members were also involved in the interviews, due to the authors' understanding of family involvement within the decision-making process. Using the conflict theory model (CTM) of decision making as a framework to guide the study, explained further below, the authors discovered that the decision-making process prior to hospice encounter varied among cancer diagnoses and other chronic illnesses. Their study's results provide a useful framework for understanding differences in the process of family decision making for hospice by diagnosis, but the role of communication within the family surrounding this process was not explored.

Scott (Reference Scott2010) employed a theoretical perspective of multiple goals to consider how the quality of family communications about EoL decision making is more important than its quantity. Outside ratings of communication from observing family communication were positively associated with reports from patients and their loved ones on satisfaction with the conversation and decision-making efficacy. More research related to informal communication between older adults and their family members (Hopp, Reference Hopp2000) is certainly needed.

METHODOLOGY AND RESEARCH DESIGN

Research at the end of life presents multiple challenges. One of the first is in regard to the vulnerability of terminally ill patients. Worries about terminally ill patients' participation in research include their debilitating physical and psychosocial symptoms, dependency on others for care, and potential cognitive impairment or inability to provide consent (Fine, Reference Fine2003; Reyna et al., Reference Reyna, Bennett, Bruera and Addington-Hall2007). However, the concern over patients' vulnerable state and inability to participate in research has stunted the growth of evidenced-based knowledge about EoL care (Casarett, Reference Casarett2005; Gysels et al., Reference Gysels, Evans and Higginson2012). In a systematic review of 20 studies considering the impact of EoL research on participants, positive outcomes were seen across each of the different studies, and very few experienced any distress (Gysels et al., Reference Gysels, Evans and Higginson2012). While certainly the focus on potential risk to terminally ill patients should not be minimized, it is also important to recognize the benefit of such research to dying patients, their families, and the field of EoL care at large when appropriate measures are taken to limit risks (Fine, Reference Fine2003).

In addition to discussion about vulnerable patients, inclusion in research is a secondary consideration about the challenges faced with participation that can impact the research design. For example, recruitment and retention of participants, attrition, and poor health status impacting data collection are all challenges (Addington-Hall, Reference Addington-Hall and Addington-Hall2007). Due to the relative newness of palliative care research and the subjective constructs often studied, optimal measurement techniques are not agreed upon, making validity a particular challenge (Casarett, Reference Casarett2005). Longitudinal designs are difficult due to high rates of attrition or dropouts, and quasi-experimental and experimental designs are difficult as they may be “too regimented for clinicians accustomed to flexible, patient-centered care” (Head & Ritchie, Reference Head and Ritchie2004, p. 40). Some clinicians even worry whether randomization in relation to EoL care is even ethical (Connor, Reference Connor2009). Though pure experimental models are always desirable, other approaches are also appropriate and valuable (Head & Ritchie, Reference Head and Ritchie2004). In fact, with the nature of palliative care utilizing a team approach, it is natural that EoL research also be approached across interdisciplinary fields using a combination of qualitative and quantitative designs (Addington-Hall, Reference Addington-Hall and Addington-Hall2007; Lloyd et al., Reference Lloyd, White and Sutton2011).

The SUPPORT study, described earlier, demonstrated that collection of large amounts of quantitative data utilizing a randomization design is possible in EoL research. The feasibility of such a study, though, is an important consideration. While funding was obtained through the Robert Wood Johnson Foundation, a huge funding source within EoL research, the study was their most expensive venture, costing more than $29 million (one of their normally funded projects generally involves less than $500,000) (Schroeder, Reference Schroeder1999). Another feasibility concern in research at the end of life is with hospice patients, where the average length of stay is short, causing concern over involving them in a process of informed consent and data collection (Head & Richie, Reference Head and Ritchie2004). In prospective studies, even patients who are able to consent upon enrollment may lose that ability at a later point within the study (Casarett, Reference Casarett2005). These challenges are all aspects that need consideration within the methodology of a research study.

Though there has been recent growth in EoL research efforts, they are still overall considered to be relatively small in comparison to other topic areas, leaving many “unanswered questions” (Addington-Hall, Reference Addington-Hall and Addington-Hall2007, p. 2). Much of the research on decision making and family communication at the end of life is exploratory and descriptive, demonstrating there is still much we do not know about these constructs. Qualitative research methods are common, utilizing focus groups (Bullock, Reference Bullock2011; Cohen et al., Reference Cohen, McCannon and Edgman-Levitan2010; Elliott et al., Reference Elliott, Gessert and Peden-McAlpine2009) and guided or semistructured interviews (Badr & Taylor, Reference Badr and Taylor2006; Gardner & Kramer, Reference Gardner and Kramer2009; Keeley, Reference Keeley2007; Stone et al., Reference Stone, Mikucki-Enyart and Middleton2012), with grounded theory (Badr & Taylor, Reference Badr and Taylor2006; Bullock, Reference Bullock2011; Radwany et al., Reference Radwany, Albanese and Clough2009; Wittenberg-Lyles et al., Reference Wittenberg-Lyles, Goldsmith and Demiris2012) and/or content analysis (Bullock, Reference Bullock2011; Waldrop & Meeker, Reference Waldrop and Meeker2012) in the data analysis. Cross-sectional designs are common (Kramer et al., Reference Kramer, Kavanaugh and Trentham-Dietz2010; Waldrop & Meeker, Reference Waldrop and Meeker2012), as longitudinal data are difficult to obtain due to the short prognoses of terminally ill patients. Quinn and colleagues (Reference Quinn, Schmitt and Baggs2012) utilized a prospective ethnographic approach to study trends in decision making within four ICUs over an extended period of time from 2001 to 2004. Some studies are retrospective and utilize bereaved family members to portray the EoL experience (Stone et al., Reference Stone, Mikucki-Enyart and Middleton2012). Critiques of this approach include that it does not capture the experiences from the perspective of the terminally ill patients themselves. The amount of time passed from the experience may also have influenced how bereaved family members remembered it.

Due to the exploratory and descriptive nature of studies on decision making and/or family communication at the end of life, measurement tools and scales are not extensively used. The Life Support Preferences Questionnaire (LSPQ) has been employed in quantitative studies about decision making as a measure for treatment preferences when considering proxy accuracy in predicting patients' choices (Parks et al., Reference Parks, Winter and Santana2011; Schmid et al., Reference Schmid, Allen and Haley2010). In both quantitative and qualitative designs, questions about and the presence of advanced directives were often used as an indicator for decision making (Bullock, Reference Bullock2011; Schmid et al., Reference Schmid, Allen and Haley2010; Tschann et al., Reference Tschann, Kaufman and Micco2003; Young & Rodriguez, Reference Young and Rodriguez2006). Qualitative studies have also considered the process of decision making from the multiple perspectives of patients, family members, or clinicians (Elliott et al., Reference Elliott, Gessert and Peden-McAlpine2009; Gauthier & Swigart, Reference Gauthier and Swigart2003; Quinn et al., Reference Quinn, Schmitt and Baggs2012). In studies considering family communication at the end of life, self-report measures are most commonly used (Harris et al., Reference Harris, Bowen and Badr2009). While these tools provide flexibility to family researchers, they collect general impressions of communication versus actual occurrence of behaviors (Metts & Lamb, Reference Metts, Lamb, Turner and West2006). The use of previously validated scales to study family communication at the end of life is uncommon.

GAPS AND IMPLICATIONS

While family communication is recognized as a barrier to transition to hospice care, the research surrounding this connection is limited. Research related to the role of familial communication within the decision-making process is warranted, along with further research about how families communicate about their illness in general (Harris et al., Reference Harris, Bowen and Badr2009). Little is known about the decision-making process for hospice (Chen et al., Reference Chen, Haley and Robinson2003). While a number of studies describe family members' feelings that communication and information are insufficient, what is meant by communication is not adequately discussed. How this fits together with the way families communicate among themselves seems to be an important consideration that remains overlooked. Also, much of the literature considered within this review is specifically related to cancer and does not apply across the total spectrum of terminal illnesses at the end of life. While cancer patients continue to make up close to half the patients served in palliative programs, these studies do not provide an inclusive picture of other types of patients served by EoL care. A methodological gap includes the use of a recognized measurement scale on either decision making or family communication that is validated for use with terminally ill patients.

Understanding the role of family relationships and types of communication in decision making at the end of life will help clinicians identify and provide support to families who may be struggling to make decisions. Failing to recognize the role of family communication within this process could cause further delay in access to end-of-life care and lessen the overall quality of and satisfaction with the experience.