Article Text

Download PDFPDF

Psychological perspective on compassion in modern healthcare settings
  1. Michelle Rydon-Grange
  1. North Wales Forensic Psychiatric Service, Llanfairfechan, Conwy, UK
  1. Correspondence to Dr Michelle Rydon-Grange, North Wales Forensic Psychiatric Service, Llanfairfechan, Conwy LL33 0HH, UK; michellerydongrange{at}


Compassionate care is a foundation of the National Health Service (NHS). However, several high-profile inquiries into healthcare failures in the NHS suggest compassion is often absent in our hospitals. Ensuing policies mandate healthcare professionals to ‘show more compassion’ but, as the psychological evidence-base indicates, this instruction neglects the complexity of this social emotion. This paper applies the psychological research on compassion to modern healthcare settings with the aim of creating a better understanding of the pathways leading to uncompassionate care. A review of the empirical evidence suggests a range of psychological factors modulate compassion. In particular, the psychological literature indicates the human compassion system is adaptive, highly attuned to its environment. As such, a healthcare professional’s ability to experience and display compassionate behaviour will be, in part, determined by the environment in which they practise; that is, aspects of the organisational environment will either facilitate or inhibit compassion. This paper argues that the typical organisational set-up of a modern healthcare setting seriously undermines compassionate care. Organisational features frequently associated with uncompassionate care include the understaffing of hospital wards, excessive working shift patterns and the dogged focus on achieving service efficiencies—each has been identified as contributing to the alarming breakdown of compassionate care at the centre of several healthcare failings. Policies focusing on culture change in the NHS neglect the growing psychological evidence base on compassion, but by applying a psychological understanding of compassion to healthcare settings, we can begin to adequately understand and address the real causes of uncompassionate care.

  • psychology
  • quality of health care
  • behavioural research

Statistics from

Compassion ought to be at the centre of modern healthcare. Indeed, several government policies now mandate compassion by healthcare professionals: ‘Clinical practice… must be… compassionate’ according to the Department of Health’s ‘Culture Change in the NHS’ document1; nurses must adhere to the six Cs (of which compassion is one) in ‘everything’ they do2; and the National Health Service (NHS) ‘should have compassionate leadership at its heart…’.1 Put simply, compassion in our hospitals should be unfaltering from ward to board. However, the results of several recent high-profile inquiries into serious failings in patient care within the NHS suggest it is not nearly that simple: a failure to meet basic standards of care at Mid-Staffordshire Hospital due to a ‘harrowing… lack of compassion…’ leading to ‘unavoidable’ deaths,3 examples of undignified and uncompassionate care within Tawel Fan psychiatric ward at Betsi Cadwaladr University Health Board4 5 and a disturbing lack of compassion towards families accessing maternity services at Pennine Acute Hospitals NHS Trust.6 That the majority of healthcare professionals are dedicated to providing compassionate care for their patients is assumed, so how can we understand how such egregious lapses in care occur? Compassion, so it seems, is a tricky business; ‘impossible to mandate’7 try as we might. The outcomes of these healthcare inquiries raise important questions about compassion: are there circumstances under which individuals are strongly motivated to avoid compassion? What are the psychological processes that modulate the experience of compassion, and based on our understanding of such processes, should we expect fluctuations in compassion?

The aim of this paper is to apply findings from the psychological research on compassion to modern healthcare settings with the aim of creating a better understanding of the pathways leading to uncompassionate care. Based on a review of the psychological literature, this paper argues that three key features relating to the organisational ‘set-up’ of the modern healthcare setting significantly undermine compassionate care. Organisational features identified as playing key roles in inhibiting compassionate care include: (1) the chronic understaffing of hospital wards, (2) long working shift patterns and (3) an unyielding focus on achieving service efficiencies: each feature prominently in the inquiries into what went wrong in several of the hospitals at the centre of high-profile healthcare ‘scandals’. While compassion, like all other emotions, is an intrapersonal experience, recently the interdependency between a clinician’s ability to experience compassion and the environment within which they operate has been recognised: Chadwick and Lown8 note, ‘Compassionate…care is dependent not only on the individual clinician, but also on…the organizational context within which he or she practices’. It is this interaction that leads to a more, or less, compassionate response. While there has been increasing recognition of the value of psychology in helping understand how serious failures in healthcare occur,9 there has yet to be a review of the psychological underpinnings of compassion and application to healthcare settings. Following a brief definition of compassion, this paper is presented in three sections, each focusing on one of the three organisational factors noted above, and relevant findings from the psychological literature on compassion are applied with the aim of creating an improved understanding of why we observe uncompassionate care in our hospitals.

Compassion: a definition

Examples of uncompassionate care in our hospitals often make front-page news: we certainly notice its absence. But what exactly is compassion? Goetz and colleagues10 define compassion as the ‘… feeling that arises in witnessing another’s suffering and that motivates a subsequent desire to help’. In line with several other definitions, this conceptualisation suggests compassion involves multiple components: a subjective ‘feeling’, which involves taking another’s perspective, along with a motivation to alleviate suffering. Moreover, unlike other related emotional states typically activated in response to witnessing another’s suffering (eg, sadness and distress), compassion appears to motivate approach related (eg, caretaking and helping), as opposed to escape behaviours.10 Other, as opposed to self-oriented behaviours, also appears to be uniquely associated with a compassionate response to another’s distress suggesting the emotion has ‘… distinct motivational underpinnings…’.10 Furthermore, a growing body of empirical data indicate a range of autonomic changes uniquely associated with the experience of compassion, including heart rate deceleration, reduced skin conductance and greater heart rate variability, all of which suggest activation of the parasympathetic nervous system. Recent neuroimaging studies also show that the ventromedial prefrontal cortex—a region involved in perspective taking—is activated during compassion but not during other prosocial states like love.10 Overall, the literature suggests compassion is a prosocial emotion with specific motivational, behavioural and neurobiological underpinnings.

Compassion killers

Inadequate staffing levels

Inadequate staff to patient ratios feature prominently across several healthcare failures. The Tawel Fan psychiatric ward in North Wales4 5 ‘had struggled to maintain appropriate staffing levels and subsequent patient safety’. Similarly, in Mid-Staffordshire, there was ‘… a shortage of skilled nursing staff. As a result…a completely inadequate standard of nursing care was offered…’.3 In a report into improving the safety of patients in England, Berwick noted, ‘organisations should…protect patients and staff against the dangers of inadequate staffing’.11 On these hospital wards, there was a clear relationship between chronic understaffing and uncompassionate care. Other than the obvious difficulties associated with inadequate staffing—clinicians not having adequate opportunity to build therapeutic relationships with patients, increased workload leading to stress and burnout—precisely how do the working conditions in a typical hospital lead staff down the pathway to ‘callous indifference’3 towards their patients? How does compassion ‘operate’ under such environmental conditions? Research by Cameron and Payne12 indicates that specific characteristics of modern hospitals—in particular poor staffing levels—may dull the capacity for a compassionate response to those in need. In a series of experiments, they showed that as the number of people in need of help increased, the degree of compassion reported by participants decreased but only under conditions where there was an explicit expectation to help ‘victims’. Conversely, when there was no explicit expectation to help, participants showed more compassion towards eight victims than one, as would be expected by normative models. Furthermore, this effect was due to changes in compassion towards eight victims rather than by changes in compassion towards one victim, suggesting that compassion towards eight victims might be seen as particularly costly. Termed the ‘collapse of compassion’,12 Cameron and Payne argue this effect could be due to an active attempt to suppress emotions that are seen as costly. If this is the case, they propose that under high-cost circumstances—such as when people anticipate compassion will be overwhelming due to a larger number of ‘victims’—the motivational conditions are created for a downregulation in compassion in order to protect against potential emotional exhaustion.

These results suggest that the human compassion system is adaptive; that is, an individual may be more or less compassionate depending on environmental conditions. If the human compassion system is adaptive, then modern hospitals create the perfect conditions for a downregulation in compassion—they are characterised both by high patient numbers and high levels of suffering. According to Cameron and Payne’s findings, it is possible that a busy clinician faced with a large number of distressed patients requiring help will, in order to prevent burnout, engage in the downregulation of compassion. This may explain why ‘conditions of appalling care were allowed to flourish’3 in the hospitals at the centre of several high-profile inquiries. According to Cameron and Payne’s findings, inadequate staffing levels may have been key contributors to the uncompassionate care observed: in the context of chronic understaffing, healthcare professionals working on the wards at these hospitals may simply have been overwhelmed at the expectation to alleviate the suffering they observed. Under these circumstances, compassionate care collapsed: on Tawel Fan ward, patients were nursed on the floor and sworn at by staff, while requests from families for medication reviews or swallowing assessments went ignored.4 5 At Pennine Acute NHS Trust, a patient’s death due to a haemorrhage that went unidentified for days because the change in her behaviour was misattributed to mental health problems and not the catastrophic hypoxia that led to her eventual death.6

Crucially, Cameron and Payne’s findings highlight the importance of taking into account the motivational conditions under which people ‘allow’ themselves to feel compassion. In particular, important motivational factors preventing the unfolding of a compassionate response include expectations to intervene to alleviate the suffering of multiple ‘victims’. Furthermore, empirical findings suggest that under such conditions, individuals appear to ‘block’ a compassionate response earlier (rather than later) in the trajectory of the emotion. The motivated emotion regulation account suggests that an overwhelmed healthcare professional expected to care for a large number of patients may, under such circumstances, dull their capacity for a compassionate response in an effort to prevent the ‘cost’ associated with intervening to alleviate suffering.

Shift patterns

While poor staffing levels may have contributed to the ‘collapse of compassion’ observed across several high-profile healthcare failures, additional organisational factors that might inhibit compassionate care include the extended working shift patterns common in modern healthcare settings. In particular, long shift patterns (ie, 12+ hours) are associated with higher levels of emotional exhaustion and burnout.13 While the nature of the relationship between compassion and burnout is unclear, when healthcare professionals report increased levels of fatigue, compromised emotion regulation capacity and reduced coping abilities—all frequently associated with extended shift work13—the capacity to experience compassion may be reduced. Based on Gilbert’s evolutionary model,14 which proposes human beings have a tripartite emotional regulation system resting on ‘threat’, ‘drive’, and ‘soothing’ states, regularly working extended shifts may push staff into threat-based states. According to Gilbert’s model, the ‘threat’ system is associated with distinct neurochemical, emotional, attentional and behavioural patterns. Indicators of threat system activation include high levels of anxiety, anger and disgust, a narrowed attentional focus on threat detection and behaviours characterised by self-protection (eg, fight, flight or submission).

What is the impact of operating within a threat-based state on healthcare professionals’ ability to cope effectively with the high levels of distress observed in modern healthcare settings and offer care that is compassionate? A model of compassion drawing on ideas from evolutionary psychology focuses on the critical role of coping appraisals in mediating compassionate responses.10 In their model, Goetz and colleagues suggest that high coping appraisals will be positively related to the experience of compassion and negatively related to the experience of distress, while the converse is true for low coping appraisals. This differential pattern of compassion is, according to their model, a function of the coping appraisals individuals make: those who report high coping appraisals will show more compassion towards a ‘victim’ due to the belief they feel capable of coping with the ‘victim’s’ suffering, while appraisals of low coping ability will likely activate distress in response to another’s suffering, which counteracts compassion-related tendencies when resources are low.15 To date, there have been no studies exploring the relationship between coping appraisals and compassion. However, Goetz and colleagues10 draw on the literature relating to emotional regulation and empathic self-efficacy, which reliably demonstrates that individuals with trait-like tendencies to regulate emotions16 and those with higher levels of empathic self-efficacy17 and secure attachments18 report more compassion-related tendencies rather than distress in response to others suffering. As such, it is possible that when emotional regulation capacities are compromised, such as under situations of high cognitive load, emotional fatigue and burnout—all associated with extended shift work—the capacity to experience compassion is reduced. Under such circumstances, an individual is more likely to feel anxious, distressed and unable to cope effectively with the demands of the situation, which according to Goetz and colleagues’ model,10 increases the likelihood of a non-compassionate response to suffering.

Long shifts are increasingly recognised as contributors to occupational burnout among healthcare professionals,19 and burnout has been associated with increased judgmental thinking and depersonalisation (ie, negative, cynical attitudes about one’s patients19;). Not much is known about the relationship between the moral judgments clinicians make about their patients and the type of care a patient receives, including whether that care is compassionate or not. However, given that clinician’s typically work with the most unwell (and often most stigmatised) people in society, within organisational environments that provide fertile ground for burnout, it seems reasonable to intuit that under these conditions, judgmental thinking may become ‘the norm’, making a compassionate response towards patients less likely. Indeed, there is evidence to support the idea that compassion appraisals elicit moral judgments relating to deservingness and justice.20 For the burnt out clinician, this process may be intensified and they may notice ever-increasing negative judgments about their patients, particularly those belonging to stigmatised groups. Evidence to support the idea that compassion appraisals operate within a moral framework comes from the stereotype content model (SCM20), a psychological theory that proposes that social groups are judged along two dimensions: warmth and competence. These judgments then lead to certain affective responses (eg, disgust, envy, pride and pity) and predict distinct behaviours, which are either facilitative or harmful. For example, groups stereotyped as both low in warmth and competence (eg, people who use illicit substances and people who have offended) typically elicits disgust and contempt due to perceived moral violations, while individuals belonging to the high warmth low competence stereotype (eg, older people) typically engender pity and sympathy. According to the SCM, stereotypes belonging to the ‘low warmth low competence’ group ‘uniquely captures dehumanization’20; individuals falling into this group are denied full humanity and become ‘functionally equivalent to objects’.20 Harris and Fiske were the first researchers to report neural evidence supporting the phenomenon of dehumanisation: stereotypes judged as belonging to the ‘low-low’ social group (eg, homeless people) were not associated with medial prefrontal cortex (mPFC) activation—an index of social cognition—but rather were associated with neural patterns consistent with fear and disgust. All other social groups were associated with mPFC activation.

The pattern of results observed might help explain the pathway to serious abuses of care—essentially, social groups that prompt disgust and fear responses may become akin to objects, thereby losing human properties. Under extreme circumstances where an individual is dehumanised, unsurprisingly, abuses occur. There is growing recognition that healthcare professionals can ‘slip out’ of compassionate responding, particularly when working with patient populations who elicit strong emotions (eg, people with a learning disability) and value judgments (eg, people who have offended). The SCM, along with neurocognitive data, provide support for the idea that compassion operates within a value-based decisional framework that, depending on the resulting judgement, may lead to uncompassionate care.

Service efficiencies

In addition to the chronic understaffing and excessive working hours that plight modern hospitals, staff are increasingly under pressure to meet performance targets and achieve service efficiencies. ‘Doing more with less’ has entered common parlance in conversations about the NHS; the pressure to ‘do it quicker’ is also ever–present. In clinical practice, this means health professionals spend only minimal time with their patients and that itself may become dominated by a defensive focus on meeting targets. Clinicians in most modern healthcare settings simply do not have enough time to get to know their patients adequately, so they instead plough on under the ‘tyranny of targets’. Healthcare professionals are warned of the perils of overidentifying with their patients, but what about the converse and what are the implications of this for compassionate healthcare? Emerging evidence indicates that perceived similarity to a ‘victim’ may be a key mechanism in modulating the experience of compassion.21–24 The impact of interpersonal synchrony on affiliation (or ‘liking’) has recently been demonstrated. In a series of experiments, Hove and Risen22 reported that degree of interpersonal synchrony predicted how much a participant liked the experimenter (independent of task difficulty or baseline reports of ‘likeability’). While the precise mechanism for these effects is unknown, Hove and Risen suggest that since synchrony is associated with mutuality and sharing relationships, it is possible people infer closeness when they observe synchrony. Additionally, the authors suggest interpersonal synchrony may interfere with the ability to discriminate self versus other produced action at the neural level, thus ‘blurring’ the self–other distinction and leading to higher levels of perceived affiliation.

While Hove and Risen’s findings22 relate to the effects of interpersonal synchrony on self–other similarity and subsequent affiliation, it is possible that perceived similarity to those around us may prompt other ‘costly’ prosocial emotions, such as compassion. Indeed, Valdesolo and DeSteno23 showed that perceived self–other similarity facilitated compassionate responses towards a ‘victim’ in a series of experiments evaluating the impact of motor synchrony manipulations on compassion and subsequent helping. Prior engagement in motor synchrony with a ‘victim’ led individuals to perceive the ‘victim’ as more similar to them. Moreover, participants reported more compassion for synchronised ‘victims’ and chose to help synchronised ‘victims’ more frequently and for longer periods of time compared with unsynchronised ‘victims’ following the motor synchronisation task. Further analysis revealed that the association between synchrony and compassion was partially accounted for by perceived similarity; thus, an important mechanism for triggering compassion may be how similar an individual feels they are to a ‘victim’. Unsurprisingly, increased compassion was directly related to increased helping. While synchrony was also positively related to how much participants liked the ‘victim’—replicating Hove and Risen’s findings22—‘liking’ did not explain the relation between synchrony and compassion.

Oveis and colleagues24 provided further support for the role of perceived self–other similarity on compassion. They found that trait compassion covaried with perceived self–other similarity, and this relationship was moderated by group strength (weaker groups (eg, homeless people) vs stronger groups (eg, politicians)); that is, individuals reporting higher levels of trait compassion rated themselves as more similar to more vulnerable groups (eg, elderly people), independent of the in-group status of the target group or the individual, and positive emotionality. Furthermore, participants randomly assigned to a compassion-induction condition showed higher ratings of perceived self–other similarity to more vulnerable others across two modalities of judgment: abstract labels of groups (eg, athletes and homeless people) and concrete representations of unfamiliar individuals (ie, still photographs). Overall, results of these studies indicate that perceived self-other similarity enhances compassion. Of particular interest is the finding that induced compassion can lead to a globally heightened tendency to view oneself as similar to others but in particular to those in need.


In its short 70-year history, compassionate care in the NHS has been lacking. ‘Scandals’ relating to poor patient care are frequently exposed, often citing a lack of compassion at their core. Resulting healthcare policies mandating compassion by health professionals have achieved very little in terms of real change for patients. Healthcare ‘scandals’ reoccur, and they inevitably do so under working conditions characterised by inadequate staffing, long working shift patterns and pressure to achieve service efficiencies. This paper reviewed the psychological research on compassion and applied it to the modern healthcare setting with the aim of creating a better understanding of how these working conditions undermine compassionate care. The psychological research revealed a complex picture: a compassion system that is adaptive and highly attuned to its environment. A review of the evidence suggests a range of psychological factors modulate the experience and display of compassionate behaviour, including motivational factors (eg, the high ‘cost’ associated with compassion), individual psychological characteristics (eg, emotional regulation abilities) and sociocognitive processes (eg, perceived similarity with a ‘victim’)—each impacted by the organisational context within which a clinician practices.

Given the psychological complexity of compassion, a simplistic instruction to ‘be more compassionate’ is unlikely to be effective. Conversely, attending to the array of psychological variables that are implicated in compassion can offer valuable insights into how we can promote compassionate care in our hospitals. For example, attending to variables that might enhance coping appraisals (eg, improving emotion regulation skills) may lead to an increased capacity to experience compassion, even within environments characterised by high levels of stress, such as those encountered in healthcare organisations. Indeed, the benefits of mindfulness approaches for guarding against staff burnout in health services have been well documented.25 Furthermore, taking opportunities to explore potential similarities between staff and those they care for may help nudge compassion. While the studies reported in this paper focused on motor synchrony as a proxy measure for similarity, it is possible that highlighting similarities between staff and patient may be associated with an increase in compassionate care; for example, most healthcare professionals will be able to recall times when they too have felt helpless, frightened or alone. Seeing a little bit of ourselves in those who need our help may be a key mechanism in cultivating compassion. However, perhaps the most valuable insight the psychological evidence base has to offer so far is that the human compassion system adapts to environmental conditions. As such, we must expect compassion to fluctuate; we must expect in a chronically understaffed, exhausting and efficiency-obsessed system of healthcare the ‘collapse of compassion’.12 Under these conditions, simply mandating compassion is fruitless because fundamentally, the problem lies in the way our modern healthcare settings are organised. Until government policies on compassionate care become more psychologically informed, and the organisational factors so detrimental to compassionate care are tackled, the trickiest of all the six Cs of nursing will likely make front-page news once again.



  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Other content recommended for you