Improved understanding of determinants of nurse behavior

This review identified a range of personal, organizational, and patient-related factors influencing the nurse provision of spiritual/existential care of end-of-life patients. Palliative care managers can use this information as a checklist to gauge a unit's conduciveness to nurse provision of spiritual/existential care and to identify areas requiring attention. Understanding determinants of nurses’ spiritual/existential care practices is a first step to improving the quality of patient care (Grol and Grimshaw, Reference Grol and Grimshaw2003). A useful aspect of the TDF is that relevant interventions for behavior change in each domain have been identified (Michie et al., Reference Michie, Johnston and Francis2008).

The factors identified in this review as influencing nurses’ spiritual/existential care practices incorporate many factors identified in previous reviews. Edwards et al. (Reference Edwards, Pang and Shiu2010) identified several facilitators, including reflection on an individual's own spirituality, ample time, support of team members, and life experience; and several barriers, including high patient turnover, high workload, low staffing, lack of privacy and nurse continuity, task focus, lack of confidence, and feelings of ill-preparedness. They also identified the importance of training staff to recognize spiritual issues of religious groups and noted complexities in assessing and documenting spiritual distress. Gijsberts et al. (Reference Gijsberts, Liefbroer and Otten2019) identified factors including feelings of incompetence, training, self-reflection, differences in the needs and convictions of patients and family members, weak integration of spiritual care in palliative care, and emphasis on patient physical well-being. Because our review collated evidence from nurses only — rather than from patients and health care providers, as in the Edwards et al. and Gijsbert et al. studies — the results provide greater precision regarding these factors. For example, similar to Edwards et al. (Reference Edwards, Pang and Shiu2010), we identified that while time was a factor influencing practice, different facets of time also influenced care. These include temporal demand, duration of a patient's time in the unit, duration of nurses’ time spent with patients, and time available for self-reflection. Also similarly to Edwards et al., we found that professional and personal life experience with loss was beneficial, but our review additionally found a qualification to this factor: that personal experience of loss can affect spiritual care when personal self-disclosure supersedes awareness of client needs (Pittroff, Reference Pittroff2013). As another example, Gijsberts et al. (Reference Gijsberts, Liefbroer and Otten2019) found that feelings of incompetence influenced spiritual/existential care; we also identified skills and beliefs about capabilities as factors, as well as aspects of felt competence, such as courage and learning from others and reflection on one's own existential issues.

Our study did not support the findings of previous reviews entirely, however. For example, we did not identify the potential barriers found in the Edwards et al. review of loss of human touch, and formal spiritual care training and education. These discrepancies may have arisen because the Edwards et al. review included mixed-sample studies (e.g., participants who “… performed a variety of roles: chief executive, manager, nurse, medical director, therapist, artist, volunteer and chaplain” (Wright, Reference Wright2002)). The discrepancies illustrate the value of a profession-specific review, because the findings of systematic reviews are often used in health policymaking and training design.

Our review adds to the findings of existing reviews, such as beliefs about consequences, intentions and goals, reinforcement, and memory. From a psychological perspective, the emergence of these factors is not surprising as they have been well studied in the organizational behavior field. These findings likely arose because the present study did not restrict itself to nurses’ explicit statements of perceived barriers/facilitators (in response to an interviewer's explicit question), but widened the search to include studies presenting statistical findings or nurses’ statements referring to factors influencing their behaviors.

Contribution to nurse competence frameworks

If competence is broadly defined as the ability to do something well (Cambridge Dictionary, 2021), then factors that help or hinder the “doing” of those behaviors may also affect competence in that behavior. To the extent that this premise is true, our study contributes to frameworks of nurse competence in spiritual/existential care, thus answering Selman et al.'s (Reference Selman, Young and Vermandere2014) call for more research into this area.

The extant nursing literature generally views competence in spiritual/existential care as a set of knowledge, skills, and attributes possessed by a nurse. Broad lists of competence items have been developed (van Leeuwen et al., Reference van Leeuwen, Tiesinga and Middel2009; Attard et al., Reference Attard, Ross and Weeks2019; McSherry et al., Reference McSherry, Ross and Attard2020). Many of these items appear to be congruent with factors identified in this review that are intrinsic to nursing (e.g., knowledge, skills, and capability beliefs). For example, an item in Attard et al. (Reference Attard, Ross and Weeks2019) is “[a]cknowledge personal limitations in providing spiritual care and consult other members of the multi-disciplinary team … as deemed necessary”; this factor seems concordant with an item that emerged in the beliefs about capabilities domain “accept/know limits of their expertise and [be] ready to work with other team members.” Another example is an item in Van Leeuwen et al.'s (Reference van Leeuwen, Tiesinga and Middel2009) study: “I have an accepting attitude in my dealings with a patient (concerned, sympathetic, inspiring trust and confidence, empathetic, genuine, sensitive, sincere and personal)”; this item seems concordant with items appearing under the behavioral regulation domain: “be open, honest, caring, respectful, compassionate, and show genuine desire to care and love the patient.” An example from McSherry et al.'s (Reference McSherry, Ross and Attard2020, p. 63) framework is “ … [awareness] of the different world/religious views …”; this factor seems concordant with an item that emerged in the knowledge domain: “knowledge of different religious practices/beliefs and spirituality.”

These concordances, among others, support the idea that factors uncovered in our review can be viewed as aspects of competence, and could therefore be used to elaborate on competence frameworks already developed. For example, our study found courage as a sub-theme in beliefs about capabilities. Attard et al.'s (Reference Attard, Ross and Weeks2019) competence list also refers to courage but only in a vague sense, directed toward all people with whom the nurse interacts, including clients, their families, and colleagues. This vague description of courage could be made more specific by adding details uncovered in our study of the many ways that nurses display courage toward patients: to encounter vulnerability, suffering, and death in patients; to be emotionally intimate; and to ask difficult questions and hear difficult answers. Future research could explore more fully the concordance between extant competence lists and our list of factors.

Another way that our study contributes to the understanding of nurse competence in spiritual/existential care is by proposing the notion of “environment competence.” The notion of “work environment competence” emerged as a category of death work competence among helping professionals (Chan and Tin, Reference Chan and Tin2012) and was defined as a supportive working environment that included appropriate supervision, teamwork, and organizational support. This notion supports our starting premise that factors contributing to the enactment of appropriate nurse behaviors also contribute to nurse competence. It also broadens the view of nurse competence in spiritual/existential care from being a purely individual characteristic, to being an interaction of individual and organizational characteristics. Support for an interactional view of competence is found in the organizational behavior field that considers employee performance (and hence competence) as being shaped by interacting individual and organizational characteristics (Kozlowski and Klein, Reference Kozlowski, Klein, Klein and Kozlowski2000).

The interaction between individual and organizational characteristics is present in extant nursing spiritual care competence lists, but it is overlooked due to item wording that ignores barriers within the nurses’ environment. For example, two items, one in Van Leeuwen et al. (Reference van Leeuwen, Tiesinga and Middel2009, p. 2868) stating “ … I can in a timely and effective manner refer [patients] to another care worker (e.g., a chaplain …)” and another in Attard et al. (Reference Attard, Ross and Weeks2019, p. 100) stating “[f]acilitate … privacy … to maintain clients’ dignity,” assume that spiritual care providers and privacy are readily available at the nurses’ behest. Our study explicitly identifies “availability of spiritual care providers” and “privacy” as factors in the environment domain that influence care practice and thus contribute to spiritual/existential care environment competence.

Suitability of the TDF to study spiritual/existential care practices

Our analysis showed that there were no factors derived from the review findings that could not be accounted for by one of the TDF domains. This indicates that the TDF framework is broadly relevant to nurse behavior in spiritual/existential care, albeit with some qualifications.

One qualification to the use of TDF to study spiritual/existential care is that the frequency of reporting of domains and factors does not necessarily reflect the relative importance or impact of identified factors on care practice. Discordance between reported frequency and effect on spiritual care behavior has been observed empirically by Neathery et al. (Reference Neathery, Taylor and He2020) and Balboni et al. (Reference Balboni, Sullivan and Enzinger2014). In our study, several domains of the TDF had few or no factors (e.g., only seven studies reported factors related to memory, attention, and decision-making, and no study reported factors related to optimism). But we know that they must affect behavior. Nurses are not automatons; they must be attentive to patient cues and use knowledge stored in memory to decide the most appropriate care actions in particular situations. The infrequent reports of some domains in our review could be because (i) the primary studies did not question participants directly about these domains and/or (ii) individuals find it difficult to recall affective attitudes (i.e., emotions) (Thomas and Diener, Reference Thomas and Diener1990) and may not even be conscious of factors affecting their behavior (e.g., Shantz and Latham, Reference Shantz and Latham2009). Future research could therefore explore which factors have the greatest effect on spiritual/existential care practices.

Another qualification to the use of TDF to study spiritual/existential care is that the domains are not distinct, and relationships between domains are not explicit. The TDF identifies constituent domains but not the causal processes linking domains in a coherent explanation of behavior (Michie et al., Reference Michie, Johnston and Abraham2005). This limitation in our review means that the frequency analysis should not be considered in isolation. During coding, we found that some factors could be represented by more than one domain. For example, patient–nurse boundaries could be categorized under social/professional role as well as under patient–social influence; this overlap is not surprising because professional boundaries are defined in terms of limits in social relationships, and palliative care clinicians are susceptible to such boundary challenges with patients (Lawton et al., Reference Lawton, Lawton and Stevens2019). Another example is the difficulty in distinguishing between intentions and goal domains using the data available. While intentions are intertwined psychologically, they are generally more proximal determinants of behavior than goals (Castelfranchi, Reference Castelfranchi2014). Also we noted that optimism was not ostensibly identified as a factor in any study; however, factors that were the outcome of optimism/pessimism could be manifested in the data as beliefs about positive/negative consequences of care practices. This is supported by research that assesses clinician optimism by measuring their expectations (or beliefs) about treatment outcomes (e.g., Byrne et al., Reference Byrne, Sullivan and Elsom2006). These examples illustrate how domains could form a causal network of distal and proximal factors influencing spiritual/existential care behavior. Future research could develop this network for spiritual/existential care.

Strengths/limitations and future research

One strength of the present review is that it is a mixed-methods review. Most of the studies investigating nurse spiritual/existential care practice were single, qualitative, interview-based studies. Individually these studies were not intended to be generalizable and used small samples; but together they provide a more complete depiction of factors influencing spiritual/existential care. Conversely, the few quantitative studies included in the review did not capture all domains, but did allow measurable investigation of factors not normally perceived by nurses. For example, the Doorenboos et al. (Reference Doorenbos, Wilson and Coenen2006) study showed statistically that ethnic culture influences whether nurses focused on religious rituals, but not whether nurses encouraged patients to talk about dying.

Some limitations of our review should be noted. Firstly, as the search was limited to peer-reviewed journal studies on end-of-life care published in English, the included studies were not representative of all cultural or work settings. Even though the generalizability of findings was not an aim of this review, this drawback might reduce the applicability of the findings to some work/country or healthcare contexts. Most included studies involved lengthy, face-to-face interviews with a nurse researcher, which may have introduced bias by self-selection of nurses who valued spiritual/existential care. Future research should set out to overcome these limitations.

The current study provides ample opportunity for future empirical work. Some possibilities have been mentioned, but we will comment on three additional areas. One involves the investigation of discrepancies and gaps in findings. Some factors were reported as both barriers and facilitators (e.g., participation/identification with faith tradition was identified both as a barrier (Kisvetrová et al., Reference Kisvetrová, Klugar and Kabelka2013), facilitator (Pittroff, Reference Pittroff2013), and unrelated (Johnston-Taylor, Reference Johnston-Taylor2013) to spiritual/existential care practices); discrepant findings suggest that a contextual variable may be operating. One gap identified in the frequency analysis is that few studies were conducted in Middle East, Africa and South America, probably because only English language studies were included. Future research could include studies in other languages, which might better capture culture-specific aspects of spirituality (Schultz et al., Reference Schultz, Lulav-Grinwald and Bar-Sela2014). Another gap is the relative absence of studies identifying the emotion domain in the North American region; this might be due to North American nursing research generally lagging in emotion work (e.g., in a 2017 review of emotional labor in nursing work, of 16 relevant international empirical studies, only two were North American (Delgado et al., Reference Delgado, Upton and Ranse2017)). Moreover, even though the domain-by-time-period analysis showed how studies identified barriers over time, all studies were snapshot studies and therefore did not capture how barriers and facilitators — real or perceived — changed over time with individual nurses or their organizational milieus.