The moral distress model: An empirically informed guide for moral distress interventions

AIMS AND OBJECTIVES
To explore moral distress empirically and conceptually, to understand the factors that mitigate and exacerbate moral distress and construct a model that represents how moral distress relates to its constituent parts and related concepts.


BACKGROUND
There is ongoing debate about how to understand and respond to moral distress in nursing practice.


DESIGN
The overarching design was feminist empirical bioethics in which feminist interpretive phenomenology provided the tools for data collection and analysis, reported following the COREQ guidelines. Using reflexive balancing, the empirical data were combined with feminist theory to produce normative recommendations about how to respond to moral distress. The Moral Distress Model presented in this paper is a culmination of the empirical data and theory.


METHODS
Using feminist interpretive phenomenology, critical care nurses in the United Kingdom (n = 21) were interviewed and data analysed. Reflexive Balancing was used to integrate the data with feminist theory to provide normative recommendations about how to understand moral distress.


RESULTS
There are five compounding factors that exacerbate/ mitigate nurses' experiences of moral distress: epistemic injustice; the roster lottery; conflict between one's professional and personal responsibilities; ability to advocate and team dynamics. In addition to the causal connection and responses to moral distress, these factors make up the moral distress model which can guide approaches to mitigate moral distress.


CONCLUSIONS
The Moral Distress Model is the culmination of these data and theorising formulated into a construct to explain how each element interacts. We propose that this model can be used to inform the design of interventions to address moral distress.


| INTRODUC TI ON
There has been continued debate about how moral distress should be understood. Some scholars argue for fidelity to Jameton's (1984) original conception, stating that moral distress only occurs when 'one knows the right thing to do but institutional constraints make it nearly impossible to pursue the right course of action' (p.6). Others have argued that Jameton's narrow definition is insufficient. Fourie (2015,2017) and Campbell et al., (2016) offer conceptual arguments for expanding our understanding of moral distress, and Prentice et al., (2018) have argued for the clinical utility of a broader understanding. More recently however, Paley (2021) recently critiqued expanding conceptions of moral distress, arguing that they 'serve no useful purpose' (p.218). We do not engage directly with that debate here, but elsewhere we have provided empirical evidence and empirically informed reasons to support a broader understanding (Morley et al., 2020. This empirical evidence and empirically informed reasons support arguments that existing models of moral distress (such as, Corley, 2002), useful and influential though they are in conceptualising and responding to moral distress to date, underestimate the richness and variety of moral distress, and are insufficiently holistic to serve us effectively in an ever-changing contemporary environment. Here, then, we build on our broader conceptualisation of moral distress (Morley et al., 2020;Morley, Bradbury-Jones, et al., 2021) by presenting our 'Moral Distress Model' (see Figure 1). This model is based upon the triangulation of empirical data, conceptual literature and normative theorising. We suggest this model provides an overarching structure to guide the formulation and development of interventions and resources to mitigate moral distress. Finally, we provide evidence-based recommendations for addressing moral distress (see Table 1). We argue that leaders within healthcare organisations should make steps to integrate these recommendations into areas in order to enhance their ethical climate.

| BACKG ROU N D
The definition underpinning our model is broad: 'moral distress is the psychological distress that is causally related to a moral event' (Morley et al., 2020). This definition comprises three key components: a moral event, psychological distress and a causal relationship between the two. In this paper, we describe our methodology for the overall project, and the specific empirical methods used. We then present the data that support our Moral Distress Model and make recommendations for addressing moral distress arising from that model.

| Design
Our overarching methodology was feminist empirical bioethics (Scully, 2017), combining empirical research with feminist theory to produce normative recommendations regarding how to conceptualise and respond to moral distress. Specifically, data were collected and analysed following feminist interpretive phenomenology, combining feminist approaches from Fisher (2010) and Simms and Stawarska (2013) with Heidegger's (1962) interpretive phenomenology (1962). Reflexive balancing (Ives, 2014) was used to balance the data and theory to provide the normative justification for our broader definition of moral distress (Morley, Bradbury-Jones, et al., 2021). We followed the 'Consolidated Criteria for Reporting Qualitative Research (COREQ)' guidelines (Tong et al., 2007) and the checklist can be found as Supplementry file 1.
This methodology was chosen because we aimed to gain a greater understanding of critical care nurses' lived experiences of moral distress, how it impacts them and possible support mechanisms, so that we could formulate normative recommendations about how to define and respond to moral distress. Feminist interpretive phenomenology enabled the first author to bring her experiences as a critical care nurse to the data collection and analysis (rather than 'bracket' them out), and to integrate considerations of power, hierarchy and sociopolitical contexts when making sense of the data.

| Data collection
Feminist interpretive phenomenology guided the data collection and analysis method. The first author (a PhD student and critical care nurse at the time of the study) conducted interviews with registered nurses (n = 21) working in critical care in two UK NHS Hospital Trusts to explore their experiences of moral distress. Participants were aged between 24 and 54 years old; the majority worked full time (n = 18) and (n = 3) worked part-time; (n = 12) were junior (below band 5 nurses) and (n = 9) were senior (band 6 or above). The majority (n = 12) had worked as a registered nurse for 5-10 years. For more information regarding demographics, these are referenced in Morley et al., 2020. Research ethics approval was obtained from the University of Birmingham (reference: ERN_15-1168S) and the Health Research Authority (IRAS reference: 197577). A substantial amendment was submitted and approved in May 2017 because the first author moved university.
Purposive sampling was used to recruit nurses who held current registration and worked in critical care. Recruitment occurred at

What does this paper contribute to the wider global clinical community?
• Normative recommendations about how to respond to moral distress in nursing based upon empirical data and ethical theory.
• A theoretically and empirically robust model that can be used for the development of nursing interventions to mitigate the negative effects of moral distress.
F I G U R E 1 The Moral distress model two NHS hospitals in UK cities, with busy critical care units. There were slight deviations in recruitment processes due to local preferences. At site one, unit managers sent emails with study information and participant information letters to their nursing teams.
Potential participants then contacted the first author if they were interested in participating. At site two, participants were required to have six months of experience in critical care due to the preferences of a gatekeeper. The Assistant Nursing Director and two Practice Development Nurses sent the participant information letters to nursing teams and potential participants placed their email address in a secure box on the unit to indicate interest. None of the authors had any prior relationships with potential participants. Prior to the interview, participants were given the participant information sheet explaining the risks and benefits of participation, and they were given the opportunity to ask questions. All participants consented to interviews being audio recorded and transcribed verbatim so that direct, anonymised quotations could be used for publications. Faceto-face semi-structured interviews using an interview guide were conducted with participants until data saturation was reached. At the beginning of every interview, the first author described her professional background and motivation to conduct the research. Notes were made during the interview to track key points made or areas for further exploration. Following each interview, the first author also made field notes about initial thoughts and reflections to return to during data analysis. Member checking with participants following each interview was not conducted since our approach was interpretive rather than descriptive. We did however test the formulated model with multiple specialist audiences, which is explained in more detail below.

| Data analysis
Interviews lasted between 120 and 150 minutes and were recorded, transcribed and analysed according to van Manen's (1990) six activities for interpretive phenomenology: (1) turning to the nature of lived experience; (2) investigating experience as we live it rather than as we conceptualise it; (3) reflecting on the essential themes which characterise the phenomenon; (4) describing the phenomenon through the art of writing and rewriting; (5) maintaining a strong and orientated relation to the phenomenon and (6) balancing the research context by considering parts and the whole. This data analysis method enables the researcher to immerse themselves in the 'hermeneutic circle' which requires moving from singular parts of the text, or unique experiences to how they interact with the whole. As van Manen and Adams (2010) state, a phenomenological

Moral Distress Model Normative Recommendation
Psychological Distress 1. Raise awareness of availability and purpose of Employee Assistance Programs; highlight their appropriateness as a resource for psychological distress; work to reduce stigma of accessing these services. 2. Create time and encourage attendance of bedside nurses at multidisciplinary meetings when plan of care discussions are taking place. This creates more opportunity for nurses to contribute to decision-making and understand clinical reasoning. 3. Create time and encourage attendance of bedside nurses at Morbidity and Mortality meetings and case reviews. This provides nurses with the opportunity to discuss and reflect on cases with the multidisciplinary team. This can also enable understanding of clinical reasoning post hoc. 4. Create time and encourage attendance at offerings such as Schwartz Rounds. These sessions create time and space for multidisciplinary dialogue and reflection and attendance can improve staff psychological well-being (Maben et al., 2018). Optimally, nurses should not be expected to go in their own (unpaid) time but should be offered the opportunity during protected (paid) hours. 5. Safe reflective spaces such as clinical supervision may enable nurses to discuss and process the psychological distress that they experience due to moral distress. 6. Where systems or organisational issues are recognised (such as unhealthy work environments), steps should be taken to address these with team leaders and managers.
Moral Failure 1. Acknowledge that healthcare is complex and ethical demands may be such that at times there are no 'good solutions'. In these situations, rather than focussing upon finding the 'right' or 'best' solution, we should teach nurses and healthcare professionals that solutions are likely to be, and feel, messy and unsatisfactory (Morley & Ives, 2017). 2. Greater access to ethics education that is case based could help healthcare professionals better understand how ethics functions in clinical scenarios. The aim of working through case-based scenarios should not be to only identify the morally preferable action, but also highlight the need for compromise and the inevitability of moral distress, moral residue and moral failure (Morley & Ives, 2017).
Intention to Leave Position or Profession 1. The Canadian Nurses Association, the American Nurses Association and the American Association of Critical Care Nurses have published position statements regarding moral distress and the need to take action to mitigate its effects and support critical care nurse. Similar organisations in the UK and elsewhere should consider publishing position statements and take steps to address moral distress in clinical practice. 2. Healthcare institutions should not only recognise the negative effects of moral distress but also take steps to implement and sustain support mechanisms such as psychological support systems, clinical ethics supports and ethics education to support retention.

TA B L E 1 (Continued) text thrives on an irrevocable tension between what is unique and
what is shared, so the aim is to capture both common and unique experiences. Heidegger's hermeneutic approach also requires that the researcher brings their own interpretation or 'historicality' (unique life experiences, sociopolitical understanding of the world and preconceived notions) to the reading of a text.
The first author operationalised the six data analysis steps above by: (1,2) conducting interviews, reflecting and writing field notes; (3) formulating ideas about key themes, capturing possible interpretations and biases in a reflexive research diary (4) each individual experience was rewritten into a single narrative with unique themes (5) reading, rereading and coding the data line-by-line in NViVO 11; (6) the individual narratives were compared and contrasted to identify the shared themes that captured moral distress. This was an iterative process in which the first author kept returning to the data to clarify developing themes. Each step was discussed with the research team who probed and challenged interpretations and biases to support critical, interpretive analysis.

| Trustworthiness
Trustworthiness in data collection and analysis was enhanced by following Van Manen's data analysis steps and maintaining a reflexive research diary. The first author was immersed in the data; conducting interviews, making field notes, probing and reflecting upon each interview.
Thoughts and reflections were recorded in a reflexive research diary which increases trustworthiness by providing transparency in analysis (Rolfe, 2004). Individual narratives, codes and themes were discussed within the research team, and assumptions and conclusions challenged.

| Model construction
The Moral Distress Model was initially constructed from the empirical data as we sought to understand how the themes related to the overall phenomena, and one another. Heidegger argued that to reach true understanding of phenomena, we need to understand it within the context of 'being in the world' (Heidegger, 1962), so once the themes had been developed, the latter part of analysis was concerned with placing the phenomenon back into the world. Following van Manen's (1990) activities five and six, key themes were constructed into a mind map illustrating the concept of moral distress.
The mind map was re-worked several times after discussion within the research team, and eventually became an explanatory model.
As Paley (2017) argues, models are useful for ensuring the phenomenological researcher is moving from a description or interpretation of the phenomenon to an explanation of it, so although the Moral Distress Model is a reduction of a complex phenomenon into its constituent parts, it positions moral distress within the world which 'restores the contextual and existing meaningfulness of the world' (Heinonen, 2015, p.40).
We continued to develop the model as we began the process of 'Reflexive Balancing' and integrated feminist theory and other relevant theories into our understanding of moral distress. We describe our approach to Reflexive Balancing in greater depth in another paper (Morley, Bradbury-Jones, et al., 2021). Developing this model was an iterative process, with multiple versions drafted, challenged and refined (in keeping with reflexive balancing). Many of the revisions were a result of challenges to the coherence of the model following conversations with the supervisory team, experts in the field of moral distress and bioethics, and healthcare professionals.
The model attempts to describe the phenomenon of moral distress from cause to effect, including ways in which the problem may be compounded and mitigated. The model needs to be understood holistically, but it is of course a limitation of the writing process that we must present it here linearly.

| RE SULTS
In this section, we provide an explanation of each theme identified in the data and explain how it fits within the model. We provide

BOX 1 Definition of moral distress with causal criteria
Moral distress is the combination of: (i) the experience of a moral event The moral event could be any of the following: moral tension, moral conflict, moral dilemma, moral uncertainty or moral constraint.
(ii) the experience of 'psychological distress' The term 'psychological distress' is an umbrella term that captures a variety of different negative emotions that may be expressed differently by each individual, but will often include anger, frustration, guilt, regret, sadness/upset, powerlessness, symptoms associated with stress and feeling torn. And (iii) a direct causal relation between (i) and (ii) This causal relationship may be explained using the following 'Causal Criteria': 1. There is a feeling of either: other-regarding or self-directed empathy for the individual(s) involved in the moral event; and/or recognition and acceptance of a feeling of personal/professional responsibility to those involved in the moral event, including towards oneself. 2. The nurse has a proximate relationship to the moral event in time and space. 3. The nurse experiences a combination of emotions that may be regarded as falling within the umbrella emotion 'distress' following involvement in the moral event.
illustrative verbatim quotations from participants (using pseudonyms to maintain anonymity and confidentiality) to support each theme. We have provided additional verbatim quotations and interpretations to support each theme in Table S1.

Compounding factors
Participants described various factors that impacted the moral events they encountered and either exacerbated or mitigated the moral distress they experienced. We named these 'compounding factors' and this is represented in the top right of the model with an arrow pointing to the different moral events showing the contributory relationship. Some compounding factors were regarded as 'avoidable' (e.g. poor communication) and others 'unavoidable' (e.g. scarce resources). We have discussed other avoidable and unavoidable factors in the context of austerity policies elsewhere (Morley et al., 2019). In this section, we present the themes that we interpreted in the data as compounding factors. In other circumstances, some of the compounding factors might also be moral events in their own right, but in our dataset, they were described as impacting moral events in various ways, which either heightened or mitigated nurses' experience of moral distress.

| Compounding factor I: Epistemic injustice
Nurses described a perception that when co-workers did not seem to value their nursing expertise that they were more frequently excluded from decision-making when a moral event occurred. Although participants did not use this specific terminology, their descriptions suggested that they were recipients of 'epistemic injustice'. This is a specific type of injustice that is directed towards an individual in their capacity as a knower. Fricker (2007) suggests there are two forms of epistemic injustice-testimonial and hermeneutic-both of which are evidenced in the narratives.

| Testimonial Injustice
Testimonial injustice can take the form of either a credibility excess or deficit, where the speaker either receives more credibility than they should, or less. Participants described experiences in which they were frequently recipients of a credibility deficit. Senior nurse Grace describes her experience trying to discuss the resuscitation status of a rapidly deteriorating patient, and she describes how the physician dismissed her opinion: '…when we were speaking and I was trying to explain to this

Consultant [senior physician] like why I really don't think we should
[provide cardiopulmonary resuscitation] -why it's not appropriate and how it's not particularly fair and he said to me like, 'Oh, I normally respect your opinion but today I don't, so stop talking' and he was so rude and he said that in front of like the whole ward round and it was just like, "Right, okay. Great".' (Grace).
Many participants reported similar experiences, feeling their contribution was disrespected and ignored, and describing the belief that they worked in environments in which this was common. It would be an unfair representation of the data to state that all participants felt this way but this was an experience shared by the majority. In Grace's narrative, experiencing testimonial injustice seemed to exacerbate the moral conflict she was already engaged in with the consultant, heightening her negative emotions and moral distress. Fricker (2007) argues that a credibility deficit is more harmful than an excess because it is less likely to be a systematic injustice that tracks an individual through various spheres of life, thereby increasing their vulnerability to other forms of injustice (see Fricker, 2007, p.27). However, many participants also described the perception that physicians were recipients of a credibility excess, and this affected nurses' position as epistemic agents.

| Credibility excess
For participants such as Grace (illustrative quote in the previous section), experiencing a credibility deficit resulted in feeling disregarded and ignored which created additional anger and frustration.
However, other participants believed this was justifiable because they did not have the same level of scientific and medical knowledge as physicians. In the following quotation, junior nurse Natasha suggests her lack of knowledge rightfully excluded her from decisionmaking, suggesting she should not question consultants' decisions.
This willingness to adhere to the decision-making hierarchy might be one reason Natasha did not express the negative emotions associated with moral distress.
'I don't understand how every cell works in the body, I don't have that physiology background, I have some experience but it's nowhere near as much as the consultant so if they tell me something and that they are doing something for a certain reason like I haven't really got a reason to question that….' (Natasha).
We suggest that it is through systematic credibility deficits and excesses that many nurses come to believe that they do not have valuable knowledge to contribute to patient care discussions.
Although Fricker (2007) argues that a deficit is more immediately harmful than an excess because credibility is an infinite good (there is enough to go around), Medina (2013) highlights the interactive nature of credibility such that some groups of individuals come to be regarded as more, and others less, credible. This always affects clusters of people, such as persons of colour and women, who are then considered less credible epistemologically as a group (Medina, 2013).
Although Participants who recognised the value of their 'specialist holistic knowledge' (as we coined it) often said that whilst consultants had medical and prognostic expertise, this did not mean they would make the morally 'right' decisions. They recognised the value of their knowledge, the importance of integrating patient preferences into decision-making, and were more willing to question consultants' decisions and reasoning.

| Hermeneutical injustice
Hermeneutic injustice is a structural identity prejudice defined as, 'the injustice of having some significant areas of one's social experience obscured from collective understanding owing to a structural identity prejudice in the collective hermeneutical resource' (p.155).
Fricker (2007) suggests that hermeneutic injustice follows a collective lack of understanding which prevents the knower from understanding their own social experience.
For example, Olivia expressed frustration about her perceptions of medical colleagues' understandings about the complexity of the nursing role and nurses' proximity to pain and suffering.
'Maybe we don't see things as they see them and we shouldn't because we're not doctors, we're nurses, but I don't think that they get it, they might say that they do but I don't think they understand what our role actually is, and they don't get that 12 hours… could you imagine spending 12 hours in a side room with that one patient? And I don't think they do, no matter how much they say they do, I don't think they do' (Olivia). Most often the epistemic injustice towards families seemed to be in the form of hermeneutic injustice-participants described working within teams that would not share a prognosis until it was verified. This placed the nurse participants in difficult situations as they often knew of a likely diagnosis or poor prognosis but when family members inquired they did not feel in a position to provide that information. This exacerbated moral events by creating an additional constraint and amplified experiences of moral distress because participants felt they were engaged in deceit.

| Compounding factor II: Roster lottery
Interwoven with participants' reports that consultants made the most important decisions, were reports that these decisions were highly variable. Participants explained how certain consultants would delay decisions or were known to have a reputation of never making the decision to withdraw life-sustaining treatment (LST (Joyce).
The term 'roster lottery' is from Wilkinson and Truog (2013) who discussed some of the ethical implications of physician variability in end-of-life decision-making, recognising that this could be a source of moral distress. Participants in our study seemed to be less concerned with the ethical implications of variability itself, and more concerned by the effects of the variability, namely, a perceived increase in suffering for the patient because of delays.

| Compounding factor III: Professional vs. personal responsibilities
Participants reported feeling responsible for patients in a professional capacity as their designated nurse and also on a personal level.
These feelings constituted the causal connection between the moral event and the psychological distress required for moral distress to occur (this will be discussed further below). However, when there was a conflict between participants' sense of personal and professional values, this served as a compounding factor which exacerbated the moral event and heightened participant's moral distress, for example, when participants described experiencing moral uncertainty (a moral event) because of an inner conflict between their personal and professional responsibilities.
When there was a conflict between these responsibilities, participants reported prioritising professional responsibilities over their personal values. Jenna described fulfilling her professional duties despite personally feeling that it was not the right thing. Elsewhere, this may have been interpreted as participants acting contra to their moral integrity (for example Hamric (2014)), but participants did not frame their experiences in this way. Jenna compared her personal beliefs to her emotions and described putting these aside to be professional. Conflict between personal and professional values is an unavoidable compounding factor because nurses are obligated to follow patient's preferences and engage in the provision of care even when this conflicts with their own values (except in very specific situations in which there might be a conscientious objection provision that applies).

| Compounding factor IV: Nurse as advocate
The duty to be a patient advocate was a deeply held professional responsibility that motivated participants to act. For example, Kayleigh describes how this advocate role motivated her to 'protect' the patient from other healthcare professionals.
'I always feel a little bit protective of the family because you're the patient's advocate and they're in bed and can't make any decisions and you want to protect them and be there for them and it's one element that would slightly annoy me if healthcare professionals think they know something better and it's just like you were there for ten minutes, try being there for 12 hours, it's a totally different attitude you get of people and you can't judge people in a ten- '… some of the newer consultants… they will always turn to the bedside nurse and say are you happy with that plan? What do you think? What do you think they would want to do? And X said to me before when I was a bedside nurse, 'I don't think there is anything we can do, this patient is going to die, we have tried everything, do you agree? I think at this point we should just cap the treatment and let nature take its course.' And I've been 'I'm not sure' … Those kinds of people, they will listen and say, 'if that's what you think will make a difference then let's give it a try' and for that patient it really didn't but he was willing to give that time because the patient wasn't in distress or anything but they were also not going to live and so what's two hours in terms of making a decision. But I feel it just goes to show good consultants against bad consultants and whether or not they include the bedside nurse and care about their opinion too'. (Kayleigh).
These compounding factors describe the kinds of conditions where moral distress is heightened and can therefore provide some indication of how environments might be improved to mitigate moral distress.
They also illuminate the causal pathway that leads to moral distress.

Criteria' as a causal story
Themes of space, proximity and personal/professional responsibility (as described above) emerged from participants' narratives and in combination with an element of tort law provided criteria to explain the causal relationship required between a moral event and 'distress' for an experience to be regarded as 'moral distress' rather than simply distress or distress simpliciter.
We draw upon criteria from tort law in Alcock v Chief

| Criterion 1a: Other-regarding or selfdirected empathy
In the 'Alcock criteria', it was determined that for a claimant to be a secondary victim, they must have a close tie of love and affection with the individual killed/injured or imperiled. This criterion, along with proximity, was put in place to differentiate between close family members of victims and bystanders who had witnessed scenes of the disaster on television (Hewitt, 2015). The criteria for moral distress does not need to be as stringent because the aim is not to determine financial compensation but rather to meaningfully determine whether an individual can be said to be experiencing 'moral distress'. Our participants' narratives suggest the connection between the experience of distress and the moral event is typically either an emotional or a professional connection. We first discuss the emotional aspect.
Participants described feeling emotionally invested in patient care and outcomes, to the point that they perceived this as having a negative effect on their own mental well-being. Participants discussed grieving for patients, continuing to think about them after work and coming in early to check on their progress. These feelings seemed to go beyond the bounds of professional caring and became a 'feeling with'. In the next quotation, Isabelle describes the grief and guilt she felt following a moral event (moral dilemma). Logically, she seemed to feel like she had done the right thing, but she continued to feel guilty, and her feelings of empathy appeared to connect her to the moral event.
'I'm sort of covering it but you know it's painful. Now I know it was the right thing to do but it's never going to leave me and I have thought about her and she was more than a patient. In many ways, I wish I could have seen… I really wanted to go to the funeral to talk to him and say, well just apologise if it wasn't the right way or I don't know, they didn't invite us…' (Isabelle).
Other participants explicitly mentioned empathy. In the following quotation, Grace discusses the importance of empathy, suggesting that it is vital for nurses to be empathetic.
'…so much about being a nurse, I think, is about having empathy.
Like I think you have to be really empathetic towards your patients but, at the same time, you have to sometimes recognise that some-times… you've not had that experience or… how they're feeling and I couldn't put myself in that man's shoes'. (Grace). it just really appealed to me. I just love the idea of this one person and their outcome is your focus, and their "your person", and you're "their person"'(Amelia).
Campelia (2017) argues that empathy is an epistemic practice capable of generating reliable knowledge that has utility within a social epistemology. If it is accepted that empathy has a causal role to play in moral distress, and that it has epistemic value, this may also emphasise the value of moral distress as a phenomenon that can alert us to moral problems.

| Criterion 1b: Recognition and acceptance of a feeling of personal or professional responsibility
Empathy is a sufficient, but not necessary criterion, to fulfil the causal criteria. Other participants instead framed their connection as a feeling of professional and/or personal responsibility.  Gorin (2016) and Dudzinski (2016) suggest that moral responsibility helps us to distinguish between moral distress and distress simpliciter. Dudzinski (2016)suggests that moral distress is accompanied by a heightened feeling of moral responsibility and often the feeling that responsibilities are conflicting.

The Alcock Criteria Moral Distress Criteria Explanation of the Moral Distress Criteria Data to Support
1. Close tie of love and affection with a person killed, injured or imperiled and; 1. There is a feeling of either:a) Other-regarding or self-directed empathy for the individual(s) involved in the moral event and /or b) Recognition and acceptance of a feeling of personal/professional responsibility to those involved in the moral event, including towards oneself and; 1. The data suggest the connection between the experience of distress and the moral event is typically either an emotional or a professional connectiona) Participants discussed feeling emotionally invested and attached to patients and their outcomes which seemed to imply that feelings of empathy causally connected them to moral events. b) Other participants framed the connection in terms of professional and/or personal responsibility rather than through an empathic emotional connection. 2. The moral agent has a proximate relationship to the moral event in time and space 2. Participants discussed the intimate connections they formed with patients and families because they shared emotional experiences with them, and because they spent so much of their time at the bed space.
2.Isabelle described the feeling of being the nurse at the bedside when faced with a moral event: 'because you spend so much time at the bedside you end up getting to know the patient more than the doctors often, or the rest of the team. And then you have to stand up for people, for patients, I find and sometimes that is, you know you can be torn thinking you know, is this right? Have I gone too far? Am I just going crazy? Am I just tired? … I genuinely think it leaves a mark on you in some way so it does affect you in some ways that you couldn't quite explain…' 3. Claimant directly perceived the incident, rather than via a third party and; 3. The moral agent experiences a combination of emotions that may be regarded as falling within the umbrella emotion 'distress' following involvement in the moral event.
3. The third criteria establish that the individual experiences negative emotions that can be regarded as falling within the umbrella of 'distress'.
3. Jenna discussed the distress she experienced which seemed to be intensified being in a side room in which she felt 'trapped': 'I just wanted to cry with the daughter and be like no I think you're right but also I felt really trapped because physically I was in that side room and I couldn't have anyone to be like 'look come and look at him he's dying; let's stop this now…'' 3.10 | Criterion 2: Proximate relationship to the moral event in time and space As mentioned in criterion 1a, participants discussed how proximity caused them to feel 'wrapped up' in patients' and families' experiences, and they struggled to find emotional distance-and this appeared to be a necessary pre-cursor to experiencing distress. Indeed, many participants discussed how sustained proximity heightened their emotional distress. Beth discussed her belief that bedside nurses can be more susceptible to psychological distress because of their proximity to patients. Beth suggests that proximity makes it difficult to think in a 'practical' and 'logical' way.
The suggestion that proximity has a role to play in the conception of moral distress has also been mentioned by Peter and Liaschenko (2004) who theorised that proximity to the patient creates a heightened sense of moral responsibility. Other healthcare professionals are able to walk away, whereas the bedside nurse remains in place, enacting the plan of care that may be causing suffering and bearing the burden of that proximate moral responsibility. This seems to be supported in our data, as participants frequently discussed the difficulty of sustained proximity to patients experiencing pain and suffering. In addition to causing distress, moral dilemmas were also associated with residual feelings of guilt, regret and anger, which suggests that moral residue may be a type of moral distress that extends beyond this initial causal pathway.

| Criterion 3: Moral agent experiences a combination of emotions regarded as 'Distress'
By establishing these casual criteria, we can rule out instances of distress simpliciter as moral distress. This is not to dismiss other kinds of distress but, rather, allows us to distinguish between them and enables us to establish mechanisms to respond to the different experiences. If there is a moral catalyst, knowing the type of moral event could help tailor responses, which might involve ethics support as well as any appropriate psychological support.

| Responses to moral distress
Participants indicated four possible responses to moral distress: withdraw, fight, satisfactory resolution and acquiesce. From the data, we also suggest three possible ways in which moral distress might be avoided altogether: lack of awareness, being fully withdrawn, satisfactory resolution reached.

| Withdraw
Some participants suggested that after repeated exposure to moral distress and failed attempts to change the outcome or impact decision-making, they began to withdraw. Withdrawal seemed to be active, whereas acquiescence-discussed later in this sectionseemed to be more passive. Holly described actively withdrawing herself through avoidance and distancing behaviour, she describes avoiding one patient whose LST she believed was futile and refers to the patient as 'that'. Other participants described patients as 'corpses' that were 'dead already'.
In the following quotation, Isabelle describes consciously distancing herself by focussing on the technical equipment and practical tasks so she could forget about the patient. By emotionally distancing themselves, participants seemed to be trying to break or weaken the causal connection between themselves and the moral event by reducing their proximity to it. Through distancing, they hoped to reduce feelings of empathy, or by focussing on the technical aspects of their professional responsibilities, they could dampen their emotional response. Although they tried to suppress feelings of empathy and personal responsibility, the causal pathway often remained intact because of their sense of professional responsibility and the need to fulfil their nursing tasks. For many these avoidance and distancing behaviours were then a further source guilt.
'I didn't ask too many questions, I didn't want to get to know them too much. I just felt it was already very difficult. Like you know you sort of withdraw because you know there is only so much you can handle, I could feel that I couldn't take very much already for some reason and I didn't want to build too much of a relationship with them…. …I know this is awful but you've got to focus on the equipment and I guess I was too busy to think much about the actual patient but that's, to be honest that's what saved me that day; I was busy enough not to think too much because I just felt like there was too much emotion in that bed-space. It's awful, really awful when you think about it, like it's not many people you do say these things to because it sounds like you're driven by the equipment and nothing else like you don't care for the person, the human being but I remember thinking this is the only way to handle this for me at the particular moment' (Isabelle). Peter and Liaschenko (2004) have also discussed the perilous nature of proximity, suggesting that whilst proximity to the patient can call the nurse to act, it can also cause them to turn away when they are unable to fulfil their perceived moral responsibilities. This can create a dilemmatic situation for the nurse as both options feel equally bad and can culminate in feelings of moral failure (see Table 3 for a description of moral failure).
On the top left-hand side of the model, we suggest that some individuals may be involved in a moral event but could be 'withdrawn fully' and therefore avoid moral distress completely. These nurses have broken the causal pathway, possibly because they are so removed from their feelings of empathy and sense of personal or professional responsibility that they conduct their duties without becoming emotionally involved-they are physically, but not emotionally, proximate. No participants in this study seemed to describe being fully withdrawn, and this is unsurprising considering they were a self-selected group of participants discussing moral distress.
However, the possibility that some nurses could be fully withdrawn can be extrapolated from our data.

Concept Suggested Relationship Data to Support
Compassion Fatigue (CF) 1. The causal relationship between moral distress and CF is unclear due to conceptual ambiguity. 2. We tentatively suggest CF can be understood as 'an acute onset of physical and emotional responses that culminate in a decrease in compassionate feelings towards others because of an individual's occupation' (Sinclair et al., 2017, p.10). 3. With this conceptualisation, we suggest that moral distress did seem to result in CF. 4. Other authors suggest nurses experience moral distress due to CF because they struggle to fulfil their professional obligations (Ledoux, 2015), the participants in this study did not seem distressed by their reduced ability to empathise with patients but rather saw it as necessary to fulfil their professional obligations. Participants seemed to suggest that they purposefully tried to put up mental barriers to protect themselves from intimate relationships and emotional attachments as a way to cope and decrease distress. CF seemed to be another form of distancing.
'I think I have a way of sort of blocking it out…our job isn't an easy job and you have to be able to build walls in places where you suppose you probably wouldn't really want them. Otherwise you'd spend all day in a bed space crying. There's not -like this is where I say I think my level of sympathy and empathy might have -not gotten less but hardened up, because I can't stand next to a patient's bed and cry to a consultant and say, 'This patient shouldn't be alive. Why are you doing this?' cause that's not the right way to go about it'. (Chloe) --'…your level of experiencing rubbish, horrible, horrible things we experience as ITU nurses, that's really high level isn't it. We see it all day, day in day out and so probably empathy is a lot less than it would be if we didn't see that all the time but I think it's a natural progression you can't help. If you got so emotional about every single small thing you would never be able to do the job that you do, you would just fall apart, you just wouldn't be able to do it, you would have to leave because you'd be an emotional wreck, you wouldn't be able to do your job'. (Rachel) Burnout 1. Burnout can be differentiated from moral distress because a moral event or catalyst is not necessary (but may be present) for burnout to occur. 2. Some participants explicitly self-identified as burnt out. 3. As with CF and resilience, this may be due to the popularity of burnout in the nursing literature and mainstream media. 4. Burnout seems most often to be understood according to Maslach et al., (1997) as a 'psychological syndrome of emotional exhaustion, depersonalization, and reduced personal accomplishment that can occur among individuals who work with other people in some capacity ' (p.192).
'I talked to one friend about it in depth and that's it really and I, for some reason with lots of other factors, I just had a burn out last June and when I went through therapy I think that was one of the things that just, and I still feel emotional talking about it to be honest….I looked for other jobs actually. It was the fact that it was a nightshift as well. I didn't sleep very well, I was losing my sleep and all of that so it was not just that but I felt, the following month I felt so bad. I wasn't coping with things in general so that I did look for other jobs. So, I guess my practice was affected in that respect. I didn't want to be there, I didn't like my work. I was feeling grim about things in general like work and then life I guess as well, so I did look for other jobs. I did get another job and actually I think it was at the point where I started to work through things and then I remember I do like this job I don't want to leave it'. (Isabelle) Resilience 1. Participants seemed to align resilience with distancing and dampening their emotional responses. 2. Whereas in the context of nursing, resilience is usually conceived as something that allows the nurse to engage and restore their caring relationships with patients. 3. For example, Jackson et al., (2007) refer to resilience as the 'ability of an individual to adjust to adversity, maintain equilibrium, retain some sense of control over their environment, and continue to move on in a positive manner' (p.3). 4. Participants seemed to view resilience as a tool to enable them to regain some control and continue with their professional obligations but it did not allow them to deepen or strengthen their relationships. Rather 'being resilient' involved them distancing themselves from deep emotional involvement, which they seemed to perceive as ultimately for the good because it allowed them to continue in their role.
'I think it's weird but over your time as a nurse you build up a resilience and you adjust yourself so that you don't have that emotional connection with people sometimes whereas definitely when I was a student nurse I found things much more emotional and harder to deal with'. (Natasha) --'…you do become very immune to families' pain and you cannot give your whole heart to everybody because I don't think you would be able to live. You couldn't go home; you would stay in the hospital all day because you would retain nothing for yourself. And I think there is something about resilience it's you give what you can but you know to keep something back for yourself, to keep yourself strong and to keep yourself able to carry on and to come back in the next day. Because you have to, number one because it's your job but because these people need you and so you have to be resilient in a way that they don't have to be… because their family member is in a critical state. So yeah, it's keeping something back for yourself…because you can't nurture someone if you haven't nurtured yourself. You have to take care of yourself first to stay healthy and strong. I think recognising that, I think that's what resilience is. I just think you see too many things to not build that resilience. I think everyone has it in their own way and I don't know how other people do it. I only know how I take care of myself'. (Amelia) (Continues)

| Acquiesce
The second response to moral distress is to simply acquiesce. The moral agent is aware of the moral event, and they feel distressed but they simply accept the outcome without objecting. In the next quotation, Rebecca seems to be rather passively withdrawing herself by fading into the background, almost trying to dismiss the moral event-whereas withdrawal seemed to be more of an active process in which participants consciously tried to distance themselves.
'There are times when you feel something needs doing and you just carry on because you're carrying on. Again, that really depends on the consultants. We have got some who on the ward rounds will say 'carry on with this because of this; look out for this and if this happens tell me' and you know some of them will explain things a lot more than others but, for the most part they just say carry on because that's what we do…' (Rebecca).

| Fight
A third response to moral distress was the determination to fight and engage with the moral issues. Some participants saw value in their moral distress experiences as a learning opportunity and were determined to improve their practice. In the quotation below, Max described his concerns that a patient was entering the dying phase.
He felt moral distress because he worried that the continuation of aggressive care was wrong. After his shift, he went home and considered his options, and the next day he spoke to a senior nurse who helped him raise his concerns with the medical team who reviewed

| Satisfactory resolution
Although, as in Max's case above, satisfactory resolution is possible, some participants also felt the negative emotions associated with moral distress even when they believed the right thing had happened

Concept Suggested Relationship Data to Support
Crescendo Effect 1. Epstein and Hamric (2009) suggest that moral distress leaves a 'moral residue' such that after repeated experiences of moral distress, one's feelings of distress, rather than return to baseline, accumulate-creating a crescendo effect over time. 2. We conceptualise of moral residue as originally suggested by Bernard Williams (1965) and Ruth Marcus (1980) as signalling the presence or experience of a true moral dilemma. 3. Disagreement over labelling aside, there was evidence in the data that morally distressing experiences having a cumulative effect. 4. Many participants spoke about anticipating a point at which they would not be able to face the moral and emotional challenges of their work and would need to leave their role, or the profession.
'I think that there's only so long you can feel like this and then either you, you put up barriers and you think that, 'This is normal, this is a normal way to treat people', or you leave. And that's why ICU nursing has such a high turnover because people can't cope with the, the sadness. And the fact that I think a lot of the time people feel like they, the nursing staff aren't listened to and that we are sort of not like pushing for withdrawal or you know, end-of-life care but we, we clearly see things differently to the medical staff'. (Danielle) Moral Failure 1. Tessman (2015) argues that because we face nonnegotiable moral requirements and moral dilemmas, the demands of morality are so great that moral failure is inevitable and inescapable. 1 2. This concept was not explicitly mentioned by participants, but many participants seemed to suggest-due to the complexity of ICU-that not only was moral distress inevitable, but so was a feeling of moral failure.
'… if it was the right thing why does it feel so hard and so painful, because often if you do the right thing you go home satisfied thinking I have done what I am supposed to do which is on many days working in ICU that is what you are going to get often, you leave and even if it's a little thing you think okay I have done this and this and this and that's great. That day was probably the worst day where I didn't feel like I had done the right thing and looking back I am convinced it was the right thing, it just did not feel like that and for a long time it still didn't feel like that'. (Isabelle)

TA B L E 3 (Continued)
1 To accept this notion of moral failure, one must also accept the existence of genuine moral dilemmas. and a satisfactory resolution had been reached. We suggest that in these circumstances, the lingering feelings of guilt and regret signal that the moral event was experienced as a moral dilemma, and the individuals are left with a moral residue. Amelia describes feeling like they did the right thing by withdrawing LST from a patient, but despite feeling it was morally right she describes feelings of moral uncertainty (moral event) and of guilt, regret and sadness (moral distress). These feelings lingered, as Amelia notes 'I still think about it, it's not left me', and we suggest these enduring emotions signal moral residue-moral distress that is experienced as lingering feelings of upset, regret and guilt, caused by a moral dilemma (Williams, 1965).
A satisfactory resolution will not always be accompanied by moral residue and this is indicated both in the bottom half of the model and in the top left-hand side which indicates that moral distress can be avoided altogether.
Another way that moral distress can be altogether avoided is through lack of awareness-applying to those individuals who may not have sufficient moral sensitivity or ethical awareness to have noted the moral event. Christen and Katsarov (2016) suggest that moral sensitivity is likely to be a pre-cursor to moral distress because a level of ethical awareness is needed to recognise that you are responsible within a certain context and to recognise the moral salience of that context. If a feeling of moral responsibility is required for ethical awareness, then this further supports the suggestion that personal/professional responsibility is part of the causal chain required for distress to be regarded as moral distress. Christen and Katsarov (2016) propose that individuals may intentionally become less morally sensitive as a way to reduce their moral distressperhaps by hardening their emotions and/or engaging in strategies to reduce proximity (described above).

| DISCUSS ION
We have outlined above how participants' experiences of moral distress informed the construction of our Moral Distress Model, which we suggest can be used to further inform strategies, responses and interventions to address moral distress.

| Clinical decision-making and team communication
Participants described experiences in which they were excluded from decision-making (sometimes due to logistics, such as rounding occurring in a boardroom) or were recipients of a credibility deficit and therefore felt de-valued. Whilst some participants seemed to accept the norm of privileging scientific knowledge above 'values knowledge', the majority of participants did not. They were angry that their patient knowledge and nursing expertise were not incorporated into decision-making. This seemed to exacerbate the moral event and heighten feelings of moral distress. We cannot determine from our data whether the medical team undervalued holistic knowledge, failed to recognise how this knowledge could offer rich information for clinical-ethical decision-making, or whether they simply undervalued the nursing. Peter et al., (2014) found similarly that nurses experienced moral distress because they felt that the biomedical knowledge held by medics was privileged, and therefore, medics dominated decisions about initiating and continuing lifesustaining treatment.
All this suggests that one way to address moral distress is to involve nurses in clinical decision-making discussions, recognising the value of their specialist holistic knowledge and the insight they can offer into patient preferences. This requires a team approach to care planning, ensuring that all stakeholders understand the importance of taking time to address nursing, psychosocial and ethical concerns.
In North America, ethics rounding is a more common practice than in the UK, and in some institutions, professional clinical ethicists and nurse ethicists join the healthcare team to provide ethics expertise and promote dialogue. Wocial et al., (2017)

| Advocacy
This also ties into the notion of advocacy, which was found to be a compounding factor that exacerbated moral distress in our data.
Historically, the notion of advocacy was adopted to empower nurses and free them from subservience (Kuhse, 1997). This reflects the way in which participants employed the concept-advocacy provided a reason and justification for engaging in decision-making and empowered them to enter into moral conflict with other healthcare professionals. Despite participants feeling empowered by the imperative to advocate, their attempts were often thwarted because their agency and decision-making power remained constrained. This resulted in tension between their sense of duty and their expectations of themselves-exacerbating moral distress.
In an environment, however, in which one does not feel valued or listened to, advocating for your patient against people with more epistemic power is an act of courage-and this is reflected in common parlance. We speak, for example, of having the 'courage of one's conviction' or the bravery of those who have the 'courage to speak out'. The notion of requiring 'courage' to advocate has recently been critiqued. Hamric et al., (2015) argue that calls for 'moral courage'-the virtue to speak out against clinicians when you believe a wrong is being committed-have become overly burdensome. They quote Tessman (2005), who argues that calls for courage can be oppressive and deflect the responsibilities of those whose job it is to create environments that do not require courage (Hamric et al., 2015). Indeed, the same can be said for advocacy. If environments were such that nurses could be part of decision-making, courage would not be required for advocacy. Building these cultures requires undergraduate, postgraduate and ongoing professional education that highlights the value of different roles, knowledge contribution and promotes inter-professional collaboration and communication. Maintaining these cultures requires strong leadership in which dismissive behaviours are addressed, and collaboration-and epistemic justice-is embraced and promoted.

| Ethics education
As our Moral Distress Model shows, a moral event is the primary cause of moral distress. As suggested by other authors, there is an extent to which moral distress should be regarded as a natural response to a moral problem and in fact has some positive value as the warning signal of a moral problem (Tigard, 2018;Gallagher, reported in Morley, 2016). To embrace this positive aspect of moral distress, nurses need to be equipped with ethical knowledge and skills which enable them to effectively engage with and address ethical problems. Of our participants, those who felt able to 'fight' and address the ethical problem (for example Max), described feeling moral distress less intensely.
Focussed educational interventions have shown some promise with regard to mitigating moral distress. In a recent review of the literature, the first author identified seven studies that designed educational interventions to address moral distress amongst nurses , four of which reported statistically significant reductions in moral distress (Molazem et al., 2013;Abbasi et al., 2019;Monteverde, 2016;Robinson et al., 20162014). In two of those studies, mean moral distress scores at two months were at their lowest which suggest the effect of education increased over time, perhaps due to the ability of participants to practice skills learned in practice (Molazem et al., 2013;Abbasi et al., 2019). Many of these authors developed educational content focussed on defining and recognising moral distress, developing strategies to address moral distress, and formulating action plans. This kind of intervention is no doubt important and useful, but functions as a 'treatment' as opposed to a prophylaxis. We suggest that educational interventions to enhance nurses' ethical awareness and competency, both pre-and post-registration, are likely to guard against harmful moral distress, as opposed to mitigating it when it occurs.

| Provision of safe moral spaces
Our participants described experiencing conflicts between their personal and professional values. We suggest that this is an unavoidable compounding factor because of the nature of the nursing role: nurses are obligated to fulfil patient's preferences, or physician implemented care plans which have the potential to conflict with their own values. Though unavoidable to a large extent, the potential harm of this should still be mitigated, and we suggest two important responses. First, the provision of safe moral spaces in which nurses and other professionals are able to be vulnerable, share their perspectives and reflect upon challenging cases. In the UK and the USA, Schwartz Rounds may provide this safe space. A recent longitudinal study found Schwartz Rounds resulted in a statistically significant improvement in staff psychological well-being (Maben et al. 2018).
However, it should be noted that although ethical issues may be discussed in these sessions, this is not a formal component. Within clinical ethics, Moral Distress Reflective Debriefing has been offered as one way to fully integrate the ethical and the emotional to enable nurses to better understand their ethical experiences (thus addressing hermeneutical injustice), engage in perspective-taking and receive psychological distress support .
A second response is to teach nurses emotional regulation skills.
Thinking about the causal criteria outlined above, participants talked about putting up walls and barriers to weaken their sense of empathy or responsibility for patients and families. This coping mechanism of withdrawal and distancing should not be condemned because it enabled them to continue caring for patients, but it is not clear what the long-term effects of this may be both personally and professionally. Rushton (2018) suggests that moral resilience is one way in which individuals can embrace and combine skills in self-regulation, self-care and ethical reflection to overcome the negative effects of moral distress. One component of moral resilience in which we see particular promise is that of relational integrity: the ability of nurses to act compassionately, and in the best interests of the patient despite their own feelings or beliefs (Rushton, 2018).

| Limitations
Our interviews provided rich data but the sample size was small and composed of mostly white, European female participants.
Nurses from different cultures, contexts and backgrounds may report different moral distress experiences and different emotional responses. However, because we suggest that 'distress' should be understood as an umbrella emotion, this allows for a variety of different emotional responses. To increase the representativeness of the sample, the project would have benefited from more diverse participants. Three participants self-identified as male, and the rest female; therefore, the gender ratio was more than the average number of men on the NMC register (which also includes midwives and nursing associates), at 14%, compared to 10.7% nationally (Nursing & Midwifery Council, 2020). Due to the location of the recruitment sites, all the participants lived in large multi-cultural cities and none of the participants expressed their own religious beliefs, although some participants did express an understanding that this might affect others moral beliefs. The findings may also be critiqued based upon our interpretation of the participants' experiences. We have tried to maintain trustworthiness in this process by carefully explaining interpretations and offering verbatim quotations to support interpretations.

| CON CLUS ION
We have presented the empirical findings that support the construction of the Moral Distress Model and provided recommendations for effective responses and interventions to address moral distress that can be developed and integrated into clinical nursing practice.

| Relevance to clinical practice
Moral distress is a complex concept and this study provides empirically informed normative recommendations about how moral distress can be responded to in clinical practice, which can inform future nursing interventions aimed at mitigating moral distress.

CO N FLI C T O F I NTER E S T S TATEM ENT
The authors have no conflicts of interest.