Nurses’ models of spiritual care: A cross-sectional survey of American nurses

Abstract

Objectives

Despite there being many models for how spiritual care should be provided, the way nurses actually provide spiritual care often differs from these models. Based on the premise that the way a person enacts their work role is related to how they understand that role, this study aims to describe the qualitatively different ways that nurses understand their spiritual care role.

Methods

A convenience sample of 66 American nurses completed an anonymous, online questionnaire about what spiritual care means for them and what they generally do to provide spiritual care. Their responses were analyzed phenomenographically.

Results

Four qualitatively different ways of understanding emerged: active management of the patient’s experience, responsive facilitation of patient’s wishes, accompaniment on the patient’s dying journey, and empowering co-action with the patient. Each understanding was found to demonstrate a specific combination of 5 attributes that described the spiritual care role: nurse directivity, the cues used for spiritual assessment, and the nurse’s perception of intimacy, the patient, and the task.

Significance of results

The findings of this study may explain why nurses vary in their spiritual care role and can be used to assess and develop competence in spiritual care.

Introduction

Several models of spiritual care for patients have been formulated for use by clinicians, including nurses (Barss 2012; Delgado 2007; Emblen and Pesut 2001; Ghorbani et al. 2020a; Govier 2000; Puchalski and Ferrell 2010; Ross and McSherry 2018; Royal College of Nursing 2011; Skalla and McCoy 2006; Smith 2006). These models describe how spiritual care should ideally be provided to patients and generally include key elements, such as the development of trustful relationships; exploration of patients’ spiritual perspective about the meaning of life, pain, suffering, or death; and sharing of self and common existential experiences. Despite the preponderance of these prescriptive spiritual care models, nurses provide spiritual care that often departs from these models: literature reports behaviors that appear close to ideal (Pittroff 2013), less than ideal (e.g., providing a chaplain only at patient request (Egan et al. 2017)), and unacceptable (e.g., imposing personal beliefs upon patients (Narasayasamy and Owens 2001)). The term “ideal” is used here, following Narayasamy and Owen’s (2001) usage, but we recognize that nurse behavior can be constrained by several factors, including those related to the patient or environment (Mascio et al. 2021).

The way a person enacts their work role in a particular situation is related to how they understand that role (Sandberg and Targama 2007). It follows then that the way a nurse provides spiritual care is a manifestation of their understanding of their role in spiritual care. Several studies exploring the meaning that nurses ascribe to spiritual care have adopted one of 2 approaches. One approach, labeled here the “uniform” approach, aims to arrive at a single, comprehensive meaning of spiritual care (Britt and Acton 2022; Ramezani et al. 2014; Tirgari et al. 2013; Wisesrith et al. 2021). These studies presume that there is a meaning common to a majority (at least) of nurses and may be motivated by the need to develop health and educational policies related to spiritual care. A risk with this approach, however, is that the ironing out of variations in meanings among individuals is also likely to iron out variations in practice, which could lead to misinformed policies and ways of implementing them.

In contrast, studies in the other approach, labeled here the “pluriform” approach, are motivated by the recognition that the meaning of spiritual care is “subjective and personal and based on each individual’s own spiritual belief system” (Dell’Orfano 2002, 384). This approach thus produces a “kaleidoscope of understandings” (MacLaren 2004, 457), assorted medleys of meanings of spiritual care (e.g., Dell’Orfano 2002; Kang et al. 2021; Shin et al. 2020) that overlap only partly. For example, one meaning appearing in Shin et al.’s (2020) medley, “Considering the perspective of the patient” (p. 156), is faintly similar to the one appearing in DellOrfano’s (2002) medley (“It is different for every person and depends on their values, beliefs, and their religion” (p. 383)); yet another meaning in Shin et al.’s (2020) medley, “Taking countermeasures against difficult medical situations” (p. 156), does not appear – even faintly – in Dell’Orfano’s (2002) medley. These medleys of meanings led to a recent critical review of nurses’ understanding of spiritual care to conclude that “…confusion still exists among nurses in understanding [spiritual care]” (Cooper et al. 2020, p. 116). A risk with the apparent confusion produced in the pluriform approach can prompt some to assume that spiritual care is not something substantive and stabile and therefore has no legitimate value in health care (Swinton and Pattison 2010).

Phenomenography

Yet perhaps there is more order in the confusion than meets the eye. Ference Marton, a learning psychologist, observed that whatever phenomenon people encounter, there seemed to be a limited number of qualitatively different ways in which that phenomenon was understood, which is in between the common and the idiosyncratic (Marton 1981). Marton pioneered a qualitative approach called phenomenography that elucidates the different ways a group of people make sense of phenomena around them. This approach has been applied extensively in education settings to map the different ways that students understand various concepts. Because learning in this setting involves a qualitative change in the way a phenomenon/concept is understood, teachers can use the resulting maps to assess students’ current level of understanding and to prompt/guide students to see concepts in new and more powerful ways (Johansson et al. 1985).

The phenomenographic approach is now used in broader work settings (research supervision (Franke and Arvidsson 2011), engine optimization (Sandberg 2000), and construction management (Chen and Partington 2006)) to elucidate how workers understand their work by analyzing how they describe their work, e.g., what the work means for them (Blomberg et al. 2015) or what they would do in a typical situation (Strand et al. 2017). The use of phenomenography in health care is particularly informative because the way health-care professionals enact their role can impact patient health. Several phenomenographic studies have mapped the different ways that practitioners understand anesthesia (Larsson et al. 2004), telenursing (Kaminsky et al. 2009), asthma management (Lundborg et al. 1999), and ambulance nursing (Forsell et al. 2020). These maps of understandings can be ordered in terms of complexity or completeness, with more complete or broader understandings being associated with better job performance (Sandberg 2000). The resulting maps can therefore be used to assess practitioners’ current level of competence and to guide them to see their role in a new and more effective way (Dall’alba and Sandberg 1996) (e.g., in diabetes care (Holmstrom et al. 2000)). This study will use phenomenography to map the different ways that nurses understand spiritual care.

A particular context: spiritual care at end of life

The discussion above mentions that an understanding of work is related to the way a person enacts their work in a particular situation (Sandberg and Targama 2007). As a means of specifying the situation in which spiritual care is embedded, our empirical study will focus on care of dying patients. This focus was chosen not just because of researchers’ interests but also because we follow Lundmark’s (2006) view that spiritual care studies of an “extreme” situation of caring for dying patients can provide a sound knowledge of “non-extreme” care situations.

Two factors make care of dying patients “extreme.” One is the need of the patient. Spiritual issues come to the fore when patients are dying since they are losing independence, relationships, role, future hopes, etc., and may experience feelings of meaninglessness, demoralization, anguish, loneliness, anger, or hopelessness (Best et al. 2015). The other factor is nurse-related, in that nurses are more sensitive to spiritual needs when patients are terminally ill (Abu-El-Noor 2016; Strang et al. 2002) and face many difficulties in providing appropriate spiritual care for these patients (Browall et al. 2014).

Lundmark’s view regarding extreme situations is actually an example of a principle that operates in social research generally: extreme cases provide more information than non-extreme cases, making them worthy of intense observation and sampling (Stinchcombe 2005). This idea has spurred studies of extreme organizations (Marti and Fernandez 2013), extreme business systems (Del Vecchio et al. 2021), and extreme migration (Burley 1982) to illuminate how mechanisms in “ordinary” cases operate. The principle also forms the basis of the critical incident technique (Flanagan 1954), widely used in health care, which gathers data about people’s experiences and behaviors in extreme or critical situations, as a “considerabl[y] efficien[t]” (Flanagan 1954, 338) way of exploring experiences and behaviors in less extreme situations.

Aim

Our study aimed to use phenomenography to describe the different ways of understanding spiritual care among nurses. Following other phenomenographic studies of health-care work (e.g. Kaminsky et al. 2009; Larsson et al. 2004), the term “understanding” when used in relation to nurses’ understanding of spiritual care refers to a combination of the meaning and practice aspects of spiritual care, as these 2 aspects are intimately intertwined (Sandberg and Targama 2007). These descriptions will contribute to our knowledge of nurses’ models of spiritual care, in the sense that a model is “something … that is used to show what something else is like or how it works” (Cambridge Dictionary 2022). The results will be a step toward a better understanding of nurse behavior regarding spiritual care and will contribute to competence frameworks, thus answering a call for more research in these areas (Selman et al. 2014).

Method

Study design and participants

A cross-sectional survey was conducted with a convenience sample of nurses in the United States. This group was targeted due to ease of access and budgetary limitations. Nurses who were members of a commercial, online panel (SurveyMonkey Audience®) were invited by email/text to respond to an online questionnaire administered by SurveyMonkey® (rebranded as Momentive® in 2021). SurveyMonkey Audience is a professional online platform with over 20 million volunteer members from which researchers can purchase a number of responses with desired demographic requirements. This panel categorizes members according to job function and has been used in organizational research of other groups of workers (e.g., project managers (Brandon et al. 2022); senior managers (Baffoe and Luo 2021), manufacturing employees (Skelton et al. 2019)). We invited participation from members whose job function was “nurse.” Nurses who stated they did not provide spiritual care were excluded.

Participants were asked 2 open-ended questions: (1) “What does ‘spiritual/existential care of dying patients’ mean for you personally?” and (2) “How do you approach the work of providing spiritual/existential care of dying patients, i.e. what sort of things do you do generally?” The questions referred to “spiritual/existential care” because a review found that caregivers used the terms “spiritual” and “existential” interchangeably when caring for dying patients (Edwards et al. 2010). Participants were not limited in the length of their responses. Background questions (e.g., sex, age) were also asked to develop profiles of nurses typically subscribing to different models; those results will be reported elsewhere. The questions specified the care of dying patients, as a means of specifying the work and the situation in which it is embedded. An incentive (gift card, donation to charity, and sweepstakes) equivalent to $20 AUD was offered for completing the questionnaire.

Analysis

The sample was randomly split into a classification sample of 50 participants, which was used to develop draft descriptions of understandings of spiritual care and a verification sample of 16 participants to test the completeness and replicability of these descriptions. Fifty participants’ responses were considered a sufficient initial classification sample, as previous phenomenographic studies using written responses reached saturation in variation with a comparable sample size (e.g., Bruce 1994 [n = 41]; Gerber and Velde (1996) [n = 52]; Marton and Saljo (1976) [n = 40]). A sample of 16 participants’ responses was considered sufficient for verification, also based on precedent (Sandberg 2000).

In the first stage of analysis of the classification sample, responses were read slowly several times by the first author. Interesting or significant comments were highlighted, and notes were made of any revelations about the participant’s view of spiritual care. The aim of this stage was to become aware of the participants’ views, rather than to determine how understandings were structured. Participants’ statements were treated as equally important, regardless of whether they would be considered appropriate according to prescriptive spiritual care literature. No predetermined categories or theories were used at any stage of the analysis.

After a general grasp of responses was obtained, the second stage of analysis aimed to sort responses into categories according to participants’ understandings of spiritual care. The focus was on the essence of each response rather than merely on its constituent phrases. Individual responses were considered in comparison with other responses, and significant quotes were compared to find sources of agreement or variation. While each participant’s understanding of spiritual care was somewhat idiosyncratic, there were underlying similarities. Responses that seemed similar were then grouped according to commonalities in the representative quotes.

In the next stage, a first attempt was made to describe the essence of similarity within each group of responses. Within each group, responses were read to examine and attempt to understand what each participant conceived of as spiritual care. Again, the focus was on the essence of the response rather than its constituent phrases. Different responses were then compared within the group and then compared between groups. This comparison resulted in some responses being moved from one group to another to enhance similarity within groups and dissimilarity between groups. Four groups emerged representing 4 different understandings of spiritual care. These category descriptions are reduced descriptions of key features that distinguish ways of understanding from each other, rather than rich descriptions of the countless variations of individual nurses’ “lived experience” of spiritual care.

In the next stage, the responses were analyzed in terms of the “how” of spiritual care. The aim was to describe a common set of attributes that identify the operational aspects of spiritual care or how the nurse participant practiced spiritual care. The operational aspects expressed in each response were summarized using representative quotes; quotes from responses in each of the 4 groups were pooled, and commonalities were sought. Quotes were also compared across the 4 groups to identify differences between groups. This stage resulted in 5 distinct attributes being identified.

Although the second and third stages are described separately here, in practice they formed an iterative process, resulting in some regrouping of responses. As well, both of these stages examined a participant’s responses to both questions, as some participants mixed meaning and operational aspects among their responses to the 2 questions. We distinguished meaning and operational aspects by how generalized or specific the response was: generalized actions were taken to be related to meaning and specific actions were taken to be related to operational aspects. For example, one respondent answered the first question with “Doing whatever I can to meet patient’s needs for comfort. If patient wants, talking about what to expect over time” and responded to the second question with “Praying with family, arrange for their pastor to come, help with final wishes.” In this example, the first sentence described a generalized action, so it was taken to be the meaning, and the remaining sentences described specific actions that illustrated the generalized action, so these sentences were taken to be operational aspects.

Some responses contained aspects of 2 categories, which is not uncommon in phenomenographic studies of work (e.g., Larsson et al. 2004; Sandberg 2000). In such studies, responses were categorized according to the more comprehensive understanding because – as will be demonstrated in the Discussion – more comprehensive understandings encompass less comprehensive understandings, but not the reverse. We followed suit in our analysis.

While the first author was responsible for identifying and describing the initial set of preliminary categories of understandings, she then explained to other researchers how and why those categories developed. Other researchers in the team familiarized themselves with the data and then discussed, challenged, and debated categories. Where differences of opinion occurred, discussion and recourse to data occurred, until agreement was reached. When each difference was resolved, the team refined the category descriptions.

Strategies to enhance trustworthiness

While the original formulation of categories in phenomenographic analysis is a form of discovery that does not have to be replicable, once the categories have been found, there should be high degree of intersubjective agreement concerning their presence or absence to be useful to researchers and practitioners (Sjostrom and Dahlgren 2002). Intersubjective agreement was important, as the categories of spiritual care that emerged might simply have been an artifact of the way the data were analyzed: just as nurses can frame a concept like spiritual care based on personal knowledge/beliefs/experience, researchers can also interpret data based on preexisting knowledge/beliefs/experience. Even though every attempt was made to ensure the researchers’ opinions did not bias the results (e.g., by treating all respondent statements as equally important), prior knowledge generally does influence the structure of concepts and the formation of categories (Murphy and Medin 1985). Table 1 describes the strategies undertaken to enhance trustworthiness of the category descriptions.

Table 1. Strategies used to improve the trustworthiness of interpretations

Results

Of the 149 nurses who responded to the invitation to participate, 66 completed the survey. Of these, 89% were female and ranged in age (18–60+ years), nursing experience (up to 20+ years), education levels (from vocational/practical to Masters), and health settings (e.g., general hospital, ICU, and community). The combined length of responses to both questions ranged from 18 to 152 words and averaged 47 words. Table 2 provides examples of responses and the 4 models of spiritual care that emerged from the analysis. These models represent the range of qualitatively different ways of understanding spiritual care among study participants: Model A – active management of the patient experience; Model B – responsive facilitation of patient’s wishes; Model C – accompaniment on the dying journey; and Model D – empowering co-action with the patient.

Table 2. Models of spiritual care

^a Abbreviated attributes are as follows: “Nurse directivity” refers to Attribute 2 – Nurse directivity relative to patient; “Patient data” refers to Attribute 3 – Type of patient data used for spiritual assessment; “Purpose of encounter” refers to Attribute 5 – Task/purpose of the spiritual care encounter.

^b Minor spelling mistakes have been corrected.

The 4 models are marked by variation in understanding of 5 key attributes, shown in Table 3, which both link and separate the different models:

1. (1) Intimacy: the degree and type of the nurse’s personal involvement in the patient relationship.

2. (2) Nurse directivity relative to patient: the relative influence of the nurse in specifying the patient’s spiritual needs and how those needs are to be fulfilled.

3. (3) Type of patient data used for spiritual assessment: the cues that the nurse uses to determine the patient’s spiritual condition and inform the nurse’s own response.

4. (4) View of the patient: the degree to which patients are personalized and ascribed agency regarding spiritual needs.

5. (5) Task/purpose of the spiritual care encounter: what the nurse aims to achieve in specific encounters with patient.

Table 3. Key attributes that run across the models, outlining similarities and differences in operational aspects of spiritual care

Together, Tables 2 and 3 suggest that movements from Model A to Model D involve increasing expansiveness of understandings of spiritual care and increasing breadth of awareness of its attributes.

Discussion

This study identified 4 different models of spiritual care among a group of nurses. Each model demonstrated a specific combination of 5 attributes that described the spiritual care role: nurse directivity, the cues used for spiritual assessment, and the nurse’s perception of intimacy, the patient, and the task. Elucidation of these models has several implications for research and practice.

Improved understanding of nurse behavior

This study adds to our understanding of the factors that influence nurses’ spiritual care behaviors. A recent systematic review identified scores of personal, organizational, and patient-related factors influencing these behaviors (Mascio et al. 2021), yet none of those factors captured the nurses’ understanding of their role. This additional factor is important because a person’s understanding of the work role influences the enactment or behaviors within the role (Sandberg and Targama 2007). That understandings form a range of models of spiritual care incorporating different sets of spiritual care practices may also help explain departures from ideal spiritual care behaviors. Some models (Models A and B) depart significantly from prescriptive ideal models (e.g., Barss 2012; Emblen and Pesut 2001; Smith 2006) that include facets such as establishing a partnership with the patient, exploring the patient’s perspective about the meaning of life, and sharing of self.

The set of models uncovered in the present study may provide a way of organizing the apparent “confusion” (Cooper et al. 2020) among nurses about the meaning of spiritual care because a phenomenographic approach produces a set of understandings that lie between the common and the idiosyncratic or between the uniform and pluriform. Some support for the notion that the set can be used to harmonize extant medleys of meanings lies in the observation that semblances of all models can be found in other studies of nurses’ spiritual care meanings and experiences. For example, Model A is concordant with “…we advise our patients … to worship and pray to Allah…” (Abu-El-Noor 2016, 4) and “Taking countermeasures against difficult medical situations” (Shin et al. 2020, 156). Model B is concordant with “Providing specific needs catered to the religious beliefs” (Kang et al. 2021, 968) and “arranging for rituals to be performed” (Ross 1997, 141). Model C concords with a nurse who states “…I’ve always allowed them [patients] to initiate anything and then if they say something then … I just go with wherever they lead. I let them direct wherever they want to go…” (Deal 2014, 858). Model D is concordant with nurses being personally involved in relationships based on mutuality and equal partnership, and in which they could use a counseling approach to support patients (Narayanasamy and Owens 2001). A substantive and stable set of understandings of spiritual care may reduce the risk, identified by Swinton and Pattison (2010), of delegitimizing the spiritual care that nurses deliver within health care.

As well, our study adds to findings of existing studies of nurses’ experiences of spiritual care by showing that different understandings of spiritual care are yoked with different bundles of care activities. One implication can be understood by considering a statement typical in the spiritual care literature: “…for spiritual care, meanings included showing empathy, spending time listening and talking to the patient, providing comfort, support of religious and cultural practices, smiling and singing to patients, and referral to religious workers and chaplains” (Cooper et al. 2021, 3). Our study suggests that some of these activities may figure more prominently in some nurses’ minds than others. For example, nurses working with a Model D understanding would likely describe their activities as spending time listening to patients and showing empathy, and they would be unlikely to talk only about activities such as referral to religious workers, which is typically a Model B activity. However, this does not mean that these nurses would never make such a referral. This tendency for some activities to be more salient than others has implications for quantitative studies that present nurses with lists of spiritual care activities (Kisvetrová et al. 2013): our results suggest that these activities do not form “flat” lists of independent activities but rather have a geography in the sense that they form bundles of activities. Those studies should therefore consider the appropriateness – even after data has been collected – of determining whether clusters of nurses of different understandings/practice reside within the sample.

One reviewer asked whether it was possible for nurses to be flexible and “switch” between models according to the situation. While our study did not delve into situational influences on understandings, Marton (1981) has observed that understandings are context-sensitive. As well, other phenomenographic studies of work have observed that individuals who have a more comprehensive understanding of work are also aware of a less comprehensive understanding, but the reverse does not apply (e.g., Akerlind 2005; Sandberg 2000). Translated to our study, these observations suggest that while individual nurses may tend toward a particular understanding of spiritual care in “normal” circumstances, they can understand spiritual care differently in different circumstances. Thus, as an example, a nurse who can – or prefers to – work with a Model D understanding can switch to a lower model when the situation requires (e.g. when patients are very unwell/unresponsive), but a nurse who normally – or prefers to – work with a Model A understanding cannot work with a Model D understanding when no situational constraints are present, even when the situation might require it. It is this unidirectionality of movement between understandings that makes understandings useful as indicators of competence (Sandberg 2000). Incidentally, that some nurses “switch” between models is probably why some responses could be grouped into more than one category in the analysis. The issue of how spiritual care models are influenced by situational factors is worth further study.

Contribution to competency frameworks

Using phenomenography to explore how nurses understand spiritual care allows us to contribute to frameworks of nurse competence in spiritual care. Several pieces of evidence suggest that the models can be ordered in terms of competence. First, literature acknowledges that intimate relationships are needed for provision of good spiritual care (e.g., Carroll 2001; Ghorbani et al. 2020b; Kociszewski 2003; Veloza-Gómez et al. 2017) and that greater levels of nurse–patient intimacy allow deeper levels of spiritual care to be provided (e.g., Carr 2008; Cockell and McSherry 2012; Keall et al. 2014; Ross 1997). Thus, ordering the models in terms of increasing intimacy (i.e., A, B, C, D) also orders them in terms of increasing capability to provide deeper levels of spiritual care. Second, the literature acknowledges that the ideal model of spiritual care is one characterized by intimate personal involvement by the nurse, equal partnership and mutuality between nurse and patient, the use of a coaching approach and use of self by the nurse, reciprocal interaction based on shared humanity, and common existential experiences (Narayanasamy and Owen 2001; Pesut and Thorne 2007). This ideal model seems concordant with Model D.

Third, the work pedagogy literature posits that workers with a broader way of understanding their work perform better (Sandberg 2000). Our results suggest that movement from Model A to Model D involves increasing expansiveness of understandings of spiritual care and increasing breadth of awareness of its attributes. Fourth, the expertise development literature observes that experts represent situations differently than do novices or intermediates. Expert representations incorporate values and emotions within a situation rather than simply surface-level perceptual features seen by novices (Lord and Hall 2005); include more variables, such as aspects of themselves (e.g., strengths and emotions); and meet a larger and more subtle range of task requirements (Bereiter and Scardamalia 1993). In our study, movement from Model A to Model D involves deeper awareness of patients’ values and emotions, greater use of nurses’ own strengths and emotions as resources, and a more sophisticated task purpose. Putting these 4 pieces of evidence together suggests that arranging the models as A-B-C-D also orders the models in terms of increasing competence.

To the extent that our premise regarding the ordering of model competence is true, then the models provide an alternative way of assessing and developing competence. The customary way of defining competence in spiritual care is as a set of knowledge, skills, and attitudes possessed by a nurse, as illustrated in various lists of competence items (e.g., “To collect … information about the patient’s customary spirituality (van Leeuwen et al. 2009) “…[awareness] of the different world/religious views…” (McSherry et al. 2020); “[v]alue importance of a psycho-social approach to care…” (Attard et al. 2019)). This customary view implies that competence comprises a specific set of components that workers use to accomplish their work and that more competent workers possess a superior set of components (Sandberg 2000). In turn, competence development involves “filling up” the worker with relevant components (Dall’alba and Sandberg 1996). This notion of incremental “filling up” of competence components is exemplified in the use of Likert scales (e.g., ranging from disagree to agree) on discrete spiritual care competency items (e.g., SCCS, van Leeuwen et al. 2009).

An alternative to this customary perspective of competence is a perspective of competence development as a change in meaning structure of work and its associated attributes (Dall’alba and Sandberg 1996). Transitioning from a low-competence model to a higher-competence model would mean adopting a broader way of understanding the work as a whole (Sandberg 2000), involving more like a quantum jump between models rather than incremental accretions of understanding. For example, moving from Model B to Model C involves a discrete shift in understanding from one in which the nurse’s contribution is the performance of concrete tasks to fulfill patient’s request (somewhat like a customer service officer) to an understanding in which the nurse’s contribution is personal involvement in a trustful relationship with the patient. Examples of extant professional development based on shifting understandings of work include training in diabetes care (Holmström et al. 2000), teaching (Prosser and Trigwell 1997), and hyperlipidemia management (Wahlström et al. 1997). In a similar vein, spiritual care educators can use our results to help trainees become aware of the different ways that spiritual care can be understood and ask trainees to assess their own way of understanding. Educators might also induce a shift in understanding by adapting methods of conceptual change that articulate understandings, introduce anomalies in low-order understandings, and present understanding that resolves anomalies (West 1988). Specific aspects of understandings that could be probed in the articulation phase, for instance, could be nurses’ understanding of nurse–patient intimacy, nurse directivity, views of the patient and task, and cues used for spiritual assessment.

Limitations and future research

Some limitations of our study should be noted. One limitation is that the representativeness of respondents is unknown (e.g., convenience sample and racial/ethnic data not obtained) so the transferability to other nurses is unknown. Future research could assess the models across a larger representative sample of nurses. As well, this study sampled nurses residing in a single country, the United States, and working in a range of health-care settings. Future research could investigate the applicability and usefulness of the models in other cultures and for other spiritual care providers. Because we gleaned data from brief online text responses, the model descriptions are only reduced descriptions of key features of spiritual care models; the use of interviews would add depth and complexity to the models.

Conclusion

Using a phenomenographic approach, this study found 4 models of spiritual care that collectively represent the range of qualitatively different ways that nurses understand spiritual care. The models help to explain why nurses vary in how they accomplish the role of spiritual care and can be used to assess and develop competence in spiritual care.