Examining the relationship between burnout and empathy in healthcare professionals: A systematic review

Objective: Empathy and burnout are two related yet distinct constructs that are relevant to clinical healthcare staff. The nature of their relationship is uncertain and this review aimed to complete a rigorous, systematic exploration of the literature investigating the relationship between burnout and empathy in healthcare staff. Design: A systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidance. Data sources: Search terms (Burnout OR Burn-out OR “Burn out”) AND (Empathy OR Empath*) enabled identification of studies investigating burnout and empathy in healthcare staff, using five electronic data bases (MEDLINE, PsycINFO, CINAHL Plus, PubMed, and SCOPUS). Manual searching amongst reference lists of eligible articles was also completed. Review methods: Databases were searched for studies published in the English language, from inception to February 2017. Key inclusion criteria were: 1) participants who were nurses or medical professionals, 2) full written manuscript in English, 3) use of the Maslach Burnout Inventory to assess burnout and a standardized outcome measure for empathy, 4) quantitative methodology exclusively. Results: Ten eligible studies were reviewed. Of those, seven were conducted in countries where English was not the first language. Eight of the studies provided empirical support for a negative relationship between empathy and burnout. One study provided support for a positive relationship between burnout and empathy. One study reported contradictory evidence with positive and negative correlations between different subscales of the empathy and burnout measures. In general, the quality of the studies was assessed to be good. However, some of the studies failed to provide information pertaining to sample size, with the reporting of data less than adequate


Introduction
Empathy is a core element of an effective therapeutic relationship (Yu & Kirk, 2009); however it is a subtle concept that is hard to conclusively define. It is often confused with related concepts such as compassion fatigue and sympathy. Burnout is a related but distinct concept (Maslach, 2003), that needs to be distinguished from empathy. Both of these concepts have been cited in the literature as fundamental to quality of healthcare (Brockhouse, Msetfi, Cohen, & Joseph, 2011), and therefore the exact relationship between the two needs to be examined rigorously.

Burnout
Maslach and Jackson (1981) defined burnout as a psychological syndrome involving physical depletion, feelings of helplessness, negative self-concept, and negative attitudes towards work, life, and others. Their conceptualization cited burnout as an internal reaction to external stressors (Adriaenssens, De Gucht, & Maes, 2015). The Maslach Burnout Inventory ([MBI]; Maslach & Jackson, 1981) is referred to as the 'gold standard' for measuring burnout in empirical research (Bradham, 2008;Lee & Ashforth, 1990). Lee and Ashforth (1990) Maslach and Jackson's (1981) definition did not have universal agreement it is widely cited in the literature. This is cited in the literature as the most commonly used measure for assessing burnout in human services (Halbesleben & Demerouti, 2005;Lee & Ashforth, 1990). Indeed, a review of the literature demonstrated 90% of studies utilized the MBI as an outcome measure for burnout (Schaufeli & Enzmann, 1998), and it continues to be used more recently (Torres, Areste, Mora, & Soler-Gonzalez, 2015;Walocha, Tomaszewski, Wilczek-Rużyczka1, & Walocha, 2013).
In line with Maslach and Jackson's (1981) definition of burnout, the MBI measures burnout across three dimensions: emotional exhaustion (EE), depersonalization (DP), and personal accomplishment (PA).
EE is defined as a state of emotional and sometimes physical depletion. Those experiencing EE are likely to feel over-extended and unable to offer emotional support to others; Nyatanga (2014) refer to EE as being central and often the most obvious manifestation of the syndrome. DP is conceptualized as an unfeeling and impersonal response towards recipients of one's care Paris and Hoge (2009). This conceptualization has been supported in the literature as clinicians' development of negative or cynical attitudes towards service user (Baxter, 1992). Lee and Ashforth (1990) discuss how DP can be seen as a defense which serves to protect against unwanted demand, or reduce perceived threat. Therefore it has been associated with psychological strain, and escape as a way of coping. Maslach (2003) defined a reduced sense of PA as involving a negative view of oneself, particularly in relation to one's work with service users.
Whilst the MBI has good reported reliability and validity (Maslach & Jackson, 1981), it has come under some criticism in relation to the wording and scoring of items. All of the DP and EE items are worded negatively and the PA items are worded positively (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001), indicating that this uni-directional wording may have caused artificial clustering of factors (Bouman, te Brake, & Hoogstraten, 2002;Lee & Ashforth, 1990). Additionally researchers have suggested that 'exhaustion' should also include cognitive and physical aspects (Pines, Aronson, & Kafry, 1981;Shinn, 1982).
In response to these criticisms other measures have been developed to address these limitations (e.g. Halbesleben & Demerouti, 2005), however, the utilization of this measure within the empirical literature does not compare with that of the MBI (Maslach & Jackson, 1981).
Prevalence is documented to be as high as 70% worldwide amongst physicians (Lamothe, Boujut, Zenasni, & Sultan, 2014), with 30%-50% of nurses reaching clinical levels of burnout on self-report measures (Aiken, Clarke, Sloane, Sochalski, & Silber, 2002;Gelsema et al., 2006;Poncet et al., 2007). Burnout has been linked to quality of care, with an international study, Poghosyan, Clarke, Finlayson, and Aiken (2010) reporting that higher self-ratings of burnout were associated with lower self-ratings of quality of nurses own care. Similarly Maslach (2003) cites burnout as the principle reason for job attrition within nurses. Burnout is also linked with increased rates of job turnover and stressrelated absences (Potter et al., 2010), estimated to cost £450,000 a year per National Health Service (NHS) Trust in the United Kingdom (Wright, 2005). It is not surprising therefore, that burnout has been widely researched in healthcare settings.

Empathy
Empathy, like burnout, has been widely discussed within the context of medical, nursing, and other healthcare professions in relation to its role in therapeutic relationships and quality of care (Brockhouse et al., 2011;Cunico et al., 2012;Smajdor, Stöckl, & Salter, 2011). Theoretically and conceptually, empathy has seen much attention in the philosophical, psychological, and more recently, cognitive neuroscience literature, with varying definitions and conceptualizations (Decety & Lamm, 2006). It is not within the scope of this review to consider all of these definitions; instead, the reader will be guided through the clinically relevant conceptualizations of empathy, its measurement, pertinence to clinical practice, and links with burnout as a construct. Rogers (1957) termed empathy as the ability of the clinician to sense the service user's private world as if it were their own, without losing the 'as if', hypothetical quality. This sense of distancing, or appropriate level of detachment from the service user's emotion, is supported in subsequent definitions offered by Hojat et al. (2002) and Mercer and Reynolds (2002). The common factor amongst these definitions is the suggestion that empathy bridges the gap between selfexperience and that of others (Hodges & Klein, 2001). This may be important for clinicians who, through their therapeutic relationships, are required to empathize for long periods with service users experiencing intense and often negative emotions.
Within this context empathy is understood to have four key dimensions: emotive, cognitive, behavioral, and moral (Morse et al., 1992). The emotive and cognitive components relate to clinicians' abilities to experience and share in another person's feelings, and intellectually identify and understand another person's feelings from an objective stance. The behavioral dimension refers to a clinician's ability to communicate their understanding of another person's perspective. The fourth, moral dimension, was referred to by Morse et al. (1992) as an internal altruistic motivation to be empathic towards others. This dimension was not supported by a subsequent review of the literature by Decety and Jackson (2004). Despite this lack of support, the moral component could be considered relevant when reflecting on the recent exposure of failing hospital organizations in the UK (Mid Staffordshire; Southern Health). Subsequent reports (e.g. Francis, 2013) recommended the need for a change of culture within the NHS, embodying compassionate and patient centered care that is underpinned by the NHS constitution and values. These values could be seen to reflect the moral obligation of healthcare staff to work in an empathic way with service users.
The clinical relevance of the emotive, cognitive, and behavioral dimensions have been demonstrated empirically with varied emphasis (Decety & Jackson, 2004;Eisenberg & Eggum, 2009;Mercer & Reynolds, 2002). Stepien and Baernstein (2006) discussed how engagement on a solely cognitive level could lead to empathic statements appearing superficial, therefore emotional engagement is necessary to enhance the interaction, building trust within the therapeutic relationship. Here the focus is on the importance of the cognitive and emotional dimensions.
Conversely, service users have reported that a clinician's ability to firstly, understand them (cognitive dimension) and secondly, express this understanding (behavioral dimension), is a key aspect in the therapeutic relationship (Shattell, Starr, & Thomas, 2007). This emphasis on understanding, and the links with developing a meaningful relationship, are supported by Hojat et al. (2002) who highlight how developing a meaningful relationship with service users is contingent on an understanding of their cognitive and affective states. Mercer and Reynolds (2002) also considered 'understanding' to be an important facet in responding empathically.
This connection between empathy and relationship with service users has been cited in previous research. Roter et al. (1997) andSuchman, Roter, Green, andLipkin (1993) found that service users and clinicians felt greater satisfaction with an interaction when there was an increase in empathy. Improved clinical outcomes have also been linked to increased clinician empathy and a good therapeutic relationship (Burns & Nolen-Hoeksema, 1992; Elliott, Bohart, Watson, & Greenburg, 2011; Krupnick et al., 1996). Therefore empathy, irrespective of the