Integral to holistic “big picture” nursing care is an empathy that strives for social justice. Social empathy (SE), a construct developed to guide social workers, names this ability.1,2 Social empathy is “the ability to more deeply understand people by perceiving or experiencing their life situations and as a result gain insight into structural inequalities and disparities.”3(p266) Social empathy requires more than technical skills or even interpersonal empathy or other-focus; it also requires a perspective that appreciates the impact of social determinants and seeks action to address them. Given the potential benefit of valuing and instilling SE within nurses, this study measured SE among nurses and explored factors that potentially could contribute to it.
SOCIAL EMPATHY: CONCEPTUAL AND EMPIRICAL BACKGROUND
Because SE was recently developed as a construct, empirically generated evidence about it is limited. Thus, this background will provide a theoretical description about SE, review pertinent research findings pertaining to SE, and summarize relevant evidence about the impact of clinician empathy on patient health outcomes.
The components of SE are several.1,2 Although most researchers differentiate affective from cognitive empathy, Segal and colleagues1 conceptualized SE as comprised of 3 aspects of interpersonal empathy plus the ability to have a big picture perspective. Interpersonal empathy is comprised of one's affective response (AR), the unconscious mirroring of another's experience. This is a neurologically mediated resonance allowing one to feel the sensations and emotions of the witnessed other. Once one has had this AR, then a self-other awareness (SOA) is needed to appreciate that there is a difference between one's experience and the other's experience; this allows a consciousness of the mirroring. This thinking process about one's empathy is the cognitive (COG) component of SE. It involves imagining or perceiving how the other's experience would be if it were to happen to me.
When these components of interpersonal empathy—mixed with emotional regulation—are then complemented by macro-perspective taking (MPT) and a contextual understanding of the systemic barriers to others' well-being, SE exists.2 These components of SE, according to Segal et al, are what prompt social justice. To reiterate, SE involves not just pity, sympathy, or even empathy—feeling another's suffering, but also a desire to relieve that suffering at its source. Empathy serves as a building block for SE.4 When one has empathic insights into discrimination, injustice, or inequality, one is then able to take action that promotes social justice.
While most of the published studies about SE primarily focus on examining the psychometric properties of a scale to measure SE,1,4 a few studies do provide some glimpses into the nature of SE. Supporting construct validity, interpersonal and social empathy are directly correlated; this finding supported the theoretical contention that SE is unlikely without interpersonal empathy.1,5 Segal et al6 examined survey data from 312 undergraduate and graduate students to learn whether SE differed between Latinos (ie, ethnic identification with a culture that values collectivism and community), Caucasians, and “others.” Indeed, Latinos had higher SE than did the other ethnic groups. Higher SE manifested in such ways as Latinos being more likely to help a stranger and more likely to attribute social problems to public policies. A study of 173 social workers determined that aspects of SE can impact burnout, compassion satisfaction, and secondary traumatic stress.5 That is, SOA contributed to increased compassion satisfaction and decreased secondary trauma stress, while AR only contributed to increased compassion. These researchers concluded that SE can be harnessed to help health care professionals cope with the emotional demands of their work, a finding particularly relevant for pandemic fatigued nurses.
Given the nascence of SE as a studied phenomenon, it is beneficial to briefly review how clinician empathy is associated with health outcomes. Two recent meta-analyses documented how client perception of psychotherapist empathy is correlated with perception of therapeutic alliance (r = 0.5; calculation based on 53 studies),7 and predictive of treatment outcome (r = 0.28; based on 82 samples).8 Two systematic reviews of nurse or physician empathy strongly suggested that clinician empathy contributes to decreased psychological distress among patients.9,10 Lelorain et al9 observed that perception of clinician empathy was often linked with greater patient satisfaction and decreased psychological distress. Similarly, Yu and Kirk10 found that nurse empathy was often inversely related to psychological distress. Both systematic reviews, however, observed that clinician empathy is not usually associated with patient quality of life, pain management, or several other patient outcomes.
As the above systematic reviews observed, few studies provide evidence about patient responses to nurse empathy. A classic study from a quarter century ago measured 70 hospital patients' perceptions of nurse empathy and observed that it was inversely related to distress, including anxiety, depression, and anger.11 It was also important to learn from these findings that patients' assessment of nurse empathy was only moderately correlated with their nurses' self-reported perceptions. Since the 2008 review conducted by Yu and Kirk,10 additional studies have investigated the impact of nurse empathy. For example, a study of 30 Iranian patients who had received care from male nurses also found these nurses' self-reported empathy explained 18.5% of the variance in patient satisfaction.12
Fascinating, yet less prevalent, are studies that document how receiving empathy not only makes patients feel good about their health care professionals and comply with their recommendations, but may also improve physical well-being. For instance, patients who scored their physician as highly empathic were significantly more likely to have optimal diabetes control (ie, A1c and low-density lipoprotein) than those who had low empathic scoring physicians.13 Similarly, 719 patients beginning to have cold symptoms, who perceived a caring and empathic attitude in their physician, reported less severity and duration of their cold, and interleukin-8 and neutrophils decreased significantly.14
Similarly, one study conducted in China documented patients' physical health outcomes resulting from nurse empathy.15 Yang et al15 measured 365 surgical lung cancer patients' immune response in relation to a primary nurse's empathy. Empathy was assessed among the 30 nurses in this study using a self-report questionnaire; these nurses were then categorized a possessing low, medium, or high empathy. After 7 to 10 days of receiving care, blood samples were drawn from the patients. Nurse empathy was observed to significantly correlate with B-cell (r = 0.16) and natural killer-cell (r = 0.84) numbers, but not CD3+, CD4+, or CD8+ cells. Thus, nurse empathy and patient immune function, in part, were directly related. If this evidence is transferrable to SE, it suggests that nurse SE might promote patient satisfaction, treatment adherence, emotional well-being, and immune function. In summary, the evidence about the psychological and physical outcomes of empathy—including SE—suggests that it is a vital attribute for nurses to possess.
The aim of this study was to measure SE among registered nurses (RNs) and to explore how selected demographic, personal, and work-related factors are associated with SE. The following research questions were addressed. Among RNs employed in a large health care system: (1) How socially empathic are they? (2) How are work-related (ie, burnout, satisfaction with work environment, years employed as an RN, level of nursing education, role, hours worked per week, and current work role and setting), personal factors (ie, self-compassion, spirituality/religiosity, and advocacy), and demographics (ie, age, gender, ethnicity, and history of receiving government subsidies) associated with SE?
This quantitative, cross-sectional, observational study used online survey methods to generate descriptive and correlational evidence.
Setting, sample, and recruitment
The population from which the sample was drawn was the 3663 RNs employed in a nonprofit, faith-based health care system in the northwestern United States. This health care system had as a core value to be of service to the poor and vulnerable. Inclusion criteria included being able to write and understand English.
The RNs were recruited via an email distributed through the organizations' email system. Embedded in the invitational email was a secure link to access the survey, available in REDCap. One follow-up email was sent to all potential study participants at midpoint during the 4-week survey period (in November, 2018).
Power analysis indicated that to achieve 80% power to detect a moderate effect size (f2 = 0.15) in the outcome variable of SE in a linear regression model with 20 variables and an adjusted α level of .0025 to account for the number of hypotheses tested, a sample size of 285 was required.
The survey became available to the participants after they read an introductory letter, which informed them that by responding they were indicating their consent. The online survey was comprised of several self-report questionnaires; together, there were 60 items. Participants could proceed at their own pace and access the survey via a personal code as often as needed. Participating RNs were encouraged to complete the survey in its entirety; however, progression of the online survey was not limited to having answered each question. The survey was comprised of 5 self-report scales that were presented in the following order.
Demographic and work-related items
These items, developed by the researchers, inquired about gender, age, ethnicity, education, work setting and role, number of years and hours per week worked, and an item assessing whether the participant or their family member received government assistance (eg, housing allowance, student loan, and food allowance) in the past. Another item inquired as to whether their institution supported them in “providing empathic and compassionate care.”
Interpersonal and Social Empathy Index
The Interpersonal and Social Empathy Index (ISEI) measures SE—that is, the merge of interpersonal empathy and the contextual understanding of systemic barriers and MPT.1 The ISEI is comprised of 4 subscales: macro-perspective taking (MPT; 5 items); cognitive empathy (COG; 4 items); self-other-awareness (SOA; 3 items); and affective response (AR; 3 items). Response options include 6-point Likert-type scales that range from never (1) to always (6). Higher scores indicate higher levels of interpersonal and social empathy. The ISEI was developed and tested among social workers. Both exploratory and confirmatory factor analyses confirmed the structural validity of the instrument. In this study, the internal reliability for the ISEI total matched that of the instrument's developers (α = .85). Cronbach's α values for the subscales were as follows: AR = 0.79; SOA = 0.66; COG = 0.77; MPT = 0.64.
Social Issues Advocacy Scale
The Social Issues Advocacy Scale (SIAS) measures awareness and behaviors related to social justice with 4 subscales (ie political awareness, social issues awareness, affective response, and confronting discrimination).16 Likert-type response options range from 1 (strongly disagree) to 5 (strongly agree). Evidence for structural and discriminant validity exists. In this study, internal consistency for the SIAS was 0.88.
Self-Compassion Scale–Short Form
The Self-Compassion Scale–Short Form (SCS-SF) measures compassion directed toward oneself, especially in situations causing personal suffering.17 This validated 12-item version comprises 6 subscales that measure Self-Kindness, Self-Judgement, Common Humanity, Isolation, Mindfulness, and Over-Identification. Likert 5-point scale response options range from almost never to almost always. Because internal reliability for these subscales was weak, Raes et al17 recommended that the SCS-SF total should only be used. In this study, internal reliability for the SCS-SF total was 0.58.
Single-item measures of emotional exhaustion and depersonalization
Conceptually, burnout includes emotional exhaustion (EE) and depersonalization (D), as well as a reduced sense of personal accomplishment. Given the gold standard for measuring burnout (Maslach Burnout Inventory [MBI]) contains 22 items and is costly to use, West et al18 determined that 1 item measuring EE and 1 item measuring D could provide a psychometrically satisfactory burnout scale. Concurrent validity was supported when comparing data collected with these 2 items and the MBI. Thus, this scale only includes: “I feel burned out from my work” and “I have become more callous toward people since I took this job.” Response options include never (0) to daily (6). In this study, the 2 items correlated (r = 0.74).
Brief Trust/Mistrust in God Scale
The Trust/Mistrust in God Scale contains 3 items that assess trust in God (TIG) and 3 items that focus on mistrust in God (MIG).19 Response options allow respondents to indicate how true for them an item is. The scale assumes a “God or higher power”; therefore, in this study the response Do not believe in a higher power was added. Analyses were conducted assuming the TIG and MIG to be continuous variables, with a scale ranging from 0 (Do not believe in a higher power) to 5 (Very much). Support exists for the structural convergent validity, as well as acceptable internal reliability and test-retest reliability. In this study Cronbach's α values of 0.98 (TIG) and 0.90 (MIG) were observed.
Institutional review board (IRB) approval was obtained from the organization where data were collected (IRB# STUDY2018000306); the investigators' university IRB deferred their approval to this site (IRB#5180094). Standard practices for respecting the dignity of human subjects were implemented. That is, a waiver of consent statement was included in the introductory materials that appeared before the online survey items. Digital identifiers were removed from the dataset before analyses. Data were managed and transmitted securely.
SPSS version 26 was used to analyze these data. Data were initially examined for missing values and normality. Although missing data meant dropping 3 cases, no imputations or transformations were required. Depending on whether the variable was categorical or continuous, frequencies and percentages and/or measures of central tendency were computed for all variables to allow description. To initially determine what explanatory study variables were associated with SE, analyses of variance and Pearson correlations were computed. Variables significantly associated with SE were included then in models analyzed using multiple linear regression. The standard for determining significance was P < .05.
The sample was comprised of 614 RNs who responded to the online survey (17% response rate). Participants were predominantly female, White, and worked as a staff RN; although a third identified as medical-surgical nurses, they represented a variety of units/settings. Their mean age was 44.6 (SD 12.44) years. A majority (85%) agreed that their employer supported their provision of compassionate care. See Table 1 for further details.
TABLE 1. -
Categorical Demographic Variables and Their Distributions Among 614 Participating RNs
|Prefer not to respond
|Degree in nursing
|ICU (any type)
|Obstetrics and gynecology
|Location of health care institution
|Worked per week
|<36 h (part-time)
|36-40 h (full-time)
|Perceived support to provide compassionate care
Abbreviations: ICU, intensive care unit; NP/CNS, nurse practitioner/clinical nurse specialist.
Nurse social empathy
Measured using the ISEI, SE appeared high in this sample. That is, with the exception of the AR subscale with which participants responded to with higher endorsement, ISEI total and subscale scores hovered around 80% of the maximum scores possible (Table 2).
TABLE 2. -
Descriptive Statistics for Major Study Variables (N = 614)
|ISEI subscale MPT
|ISEI subscale COG
|ISEI subscale SOA
|ISEI subscale AR
|Social issues advocacy
|Trust in God
|Mistrust in God
|Years in nursing
Abbreviations: AR, affective response; COG, cognitive; ISEI, Interpersonal and Social Empathy Index; MPT, macro-perspective taking; SOA, self-other awareness.
a117 (19.9%) reported 0, indicating no belief in a higher power.
b135 (23.4%) reported 0, indicating no belief in a higher power.
Association of personal, work-related, demographic factors with social empathy
Initially, multiple linear regression calculated how much variance in SE was independently explained by the set of personal factors, the work-related factors, and the demographic factors. The personal factors of social issues advocacy (B = 0.35, t = 11.19, P < .001) and self-compassion (B = 0.22, t = 3.46, P < .001) were both significantly associated with SE (R2 = 0.22; P < .001). Neither trust nor mistrust in God/higher power was associated with the ISEI. It is worth noting, surprisingly, that TIG was inversely weakly correlated with self-compassion (r = −0.11, P = .007) and burnout (r = −0.18, P < .0001); mistrust in God was not correlated with major study variables.
Next, work-related variables were also examined to see which were associated with SE. Of all these variables, only burnout and current nursing position significantly explained total ISEI scores (R2 = 0.05, P ≤ .001). That is, an increase in burnout was negatively associated with SE (B = −0.38, t = −.32, P = .001), and holding a management position (in contrast to being a staff RN) was associated with increased SE (B = 3.83, t = 2.64, P = .009). During regression analysis examining what demographic variables contributed to SE, only gender was identified as significantly associated (B = −2.83, t = 2.44, P = .02, R2 = 0.034). Indeed, females had more SE (mean = 72, SE = 0.34) than males (mean = 68.9, SE = 1.09).
Next, demographic, work-related, and personal factors in concert were subjected to multiple linear regression to assess their association with SE. The results of the likelihood ratio test indicated that the demographic variables and personal variables and work-related variables together explained significantly more variability in the ISEI compared with the previously presented models that used only one of these factors. The following variables showed significant associations with the total ISEI: age, social issues advocacy, self-compassion, burnout, amount of time worked, nursing educational level, and current role (Table 3). More specifically, younger nurses, those more inclined to advocate, those more self-compassionate, those less burned out, those working part-time (in contrast only with those working overtime), and those with an associate degree (vs those with a baccalaureate or graduate degree in nursing) were more socially empathic.
TABLE 3. -
Multiple Linear Regression Results: Demographic Variables, Personal Variables, and Work-Related Variables Explaining Social Empathy (N = 614)
2 = 0.288
|Outcome: ISEI Total
|Male vs female
|African American vs Caucasian
|Hispanic vs Caucasian
|Asian vs Caucasian
|Other vs Caucasian
|Do not want to respond vs Caucasian
|Yes vs none
|Prefer not to respond vs none
|Trust/mistrust in God
|Years worked as RN
|Full-time vs part-time
|Overtime vs part-time
|Baccalaureate vs associate degree
|Graduate degree vs associate degree
|Current nursing position
|Management vs staff nurse
|Nurse practitioner/clinical nurse specialist vs staff nurse
|Other vs staff nurse
|Work settings: Urban vs rural
|Employer supports compassionate care? No vs yes
Abbreviation: ISEI, Interpersonal and Social Empathy Index.
These study findings provide evidence about SE, a construct hitherto unexplored among nurses. Not only was SE found to be seemingly quite high in its various dimensions, but several factors were found to be associated with it. Factors that most strongly contributed to SE included age, social issues advocacy, self-compassion, burnout, amount of time worked, nursing educational level, and current role. In initial analyses, gender also showed itself to be a contributing factor, as women were more socially empathic than men. Likewise, a history of receiving any form of government assistance and perception of employer support for compassionate caring were found to be associated with higher SE. Surprisingly, theistic beliefs (or lack thereof) were not associated with SE. The implications of these findings for nursing practice, administration, education, and research are worth exploring so that SE can increasingly manifest in nursing care, and a holistic “big picture” approach can be fostered.
In this sample, SE—the ability to empathize in a manner that leads to social justice—was substantial. Unfortunately, comparison with other nurse samples is impossible, given nascence of the phenomenon in nursing. Nevertheless, these findings documenting intensity of SE may provide nurse administrators with an initial benchmark or realistic goal for an aggregate of nurses. It must be noted, however, that the sample was recruited within a religious and socially conscious health care system (eg, preferentially provided care to the poor). It is unknown whether socially conscious or empathic nurses are attracted to work in faith-based health care systems, or whether the organizational ethos of a religious organization influences how socially empathic a nurse is. Future research can investigate what organizational factors, such as guiding philosophical framework, contribute to promoting SE within nurses.
Expectedly, nurses who were more self-compassionate and less burned out possessed more SE. Feeling burned out from work (EE) and becoming callous toward people (D) were predictably negatively associated with SE. These findings provide reinforcement for similar findings observed by other researchers that link burnout to moral distress, compassion fatigue, decreased patient safety, and so forth.20,21 Hunt et al's22 synthesis of 5 studies (n = 562 nurses within the samples) linking empathy and burnout led to the conclusion that this association is more complex; that is, in 3 studies empathy and burnout were directly associated, and in 2 studies indirectly. These scholars conjectured that the underlying factor explaining whether empathy led to burnout was whether the nurse was able to self-regulate and curb emotional arousal. Thus, our finding of a direct relationship between self-compassion and SE is helpful. It suggests that nurses who can be compassionate toward themselves naturally are compassionate toward others. Findings from a study of 280 Portuguese nurses indicated self-compassion may protect against burnout caused by an unregulated affective empathy23 may lend insight. Perhaps self-compassion, which inherently involves emotional regulation, can increase SE and minimize burnout.
There is strong evidence from a meta-analysis of 18 clinical trials that empathy (especially cognitive empathy) can be taught.24 These findings suggest that nurse educators might introduce self-compassion as a technique for developing SE. Nurse administrators can strive to create work milieus that foster self-compassion and decrease the risk for burnout. Because burnout is prevalent among nurses, results raise a red flag: Given burnout is negatively associated with SE, does burnout manifest in nurse inability to mirror and connect with patients to affect their well-being? This study adds to the evidence that suggests yes.25 Thus, nursing education must include experiential training about self-compassion and burnout; such training likely needs reinforcement in work settings (eg, opportunities for mindfulness while at work, retreats promoting self-care). The finding that part-time nurses were more socially empathic than those who worked overtime also suggests that self-care, a balanced life, and nurse wellness may explain SE.
It was likewise unsurprising that the proclivity to advocate and engage in social issues (eg, to vote, to discuss or monitor current legislation, confront discrimination) explained SE. Indeed, while not all nurses with considerable SE may express it in social advocacy, it is hard to imagine that those who do measure high on the SIAS being low on SE. Conceptually, it is arguable that social issues advocacy is fueled by SE1; hence, the association we observed.
This study is the first to our knowledge to measure SE using the ISEI among nurses. Our findings indicate the scale, especially when considered in total, has qualities indicating reliability and validity. The ISEI's internal reliability was strong, and evidence for its convergent and hypothesis testing validity is suggested. That is, the ISEI was correlated in expected directions with the SIAS, SCS-SF, and burnout. Its efficacy for future nursing research is also enhanced by its ability to generate an acceptable distribution of scores, avoiding floor or ceiling effects.
Limitations constrain the application of findings from any study, including this one. It must be noted that study participants were recruited from one health care system in one geographical area using convenience sampling during one point in time. Thus, generalizability is limited. The internal reliability of the SCS-SF was less than 0.60; likewise, of course, the inner reliability of the burnout measure is limited given it contained only 2 items.
Social empathy was strong in this sample of nurses working in a context perceived to be supportive of compassionate care. Thus, these data provide a benchmark for SE among nurses. Although some factors found to explain SE are unchangeable (eg, age and gender), others are. These include potentially intervention-amenable contributors such as burnout, self-compassion, and interest in advocating for social issues. These findings, therefore, not only highlight a construct that should be of interest to nurses, but provide hope that SE can be nurtured.
1. Segal EA, Cimino AN, Gerdes KE, Harmon JK, Wagaman MA. A confirmatory factor analysis of the Interpersonal and Social Empathy Index. J Soc Soc Work Res. 2013;4(3):131–153.
2. Segal EA, Wagaman MA. Social Empathy as a framework for teaching social justice. J Soc Work Educ. 2017;53(2):201–211.
3. Segal EA. Social empathy: a model built on empathy, contextual understanding, and social responsibility that promotes social justice. J Soc Serv Res. 2011;37(3):266–277.
4. Gerdes KE, Lietz CA, Segal EA. Measuring empathy in the 21st century: development of an empathy index rooted in social cognitive neuroscience and social justice. Soc Work Res. 2011;35(2):83–93.
5. Wagaman MA, Geiger JM, Shockley C, Segal EA. The role of empathy in burnout, compassion satisfaction, and secondary traumatic stress among social workers. Soc Work. 2015;60(3):201–209.
6. Segal EA, Gerdes KE, Mullins J, Wagaman MA, Androff D. Social empathy attitudes: Do Latino students have more? J Hum Behav Soc Environ. 2011;21(4):438–454.
7. Nienhuis JB, Owen J, Valentine JC, et al. Therapeutic alliance, empathy, and genuineness in individual adult psychotherapy: a meta-analytic review. Psychother Res. 2018;28(4):593–605.
8. Elliott R, Bohart AC, Watson JC, Murphy D. Therapist empathy and client outcome: an updated meta-analysis. Psychotherapy (Chic). 2018;55(4):399–410.
9. Lelorain S, Brédart A, Dolbeault S, Sultan S. A systematic review of the associations between empathy measures and patient outcomes in cancer care. Psychooncology. 2012;21(12):1255–1264.
10. Yu J, Kirk M. Measurement of empathy in nursing research: systematic review. J Adv Nurs. 2008;64(5):440–454.
11. Olson JK. Relationships between nurse-expressed empathy, patient-perceived empathy and patient distress. Image J Nurs Sch. 1995;27(4):317–322.
12. Goodarzi N, Azma K, Tavakolian E, Peyvand P. Association of nurses' self-reported empathy and mu suppression with patients' satisfaction. J Caring Sci. 2015;4(3):197–205.
13. Hojat M, Maio V, Markham FW, Louis DZ, Gonnella JS. Physician empathy and diabetes outcomes. J Gen Intern Med. 2019;34(10):1966.
14. Rakel D, Barrett B, Zhang Z, et al. Perception of empathy in the therapeutic encounter: effects on the common cold. Patient Educ Couns. 2011;85(3):390–397.
15. Yang N, Xiao H, Cao Y, Li S, Yan H, Wang Y. Influence of oncology nurses' empathy on lung cancer patients' cellular immunity. Psychol Res Behav Manag. 2018;11:279–287.
16. Nilsson JE, Marszalek JM, Linnemeyer RM, Bahner AD, Misialek LH. Development and assessment of the Social Issues Advocacy Scale. Educ Psychol Meas. 2011;71(1):258–275.
17. Raes F, Pommier E, Neff KD, Van Gucht D. Construction and factorial validation of a short form of the Self-Compassion Scale. Clin Psychol Psychother. 2011;18(3):250–255.
18. West CP, Dyrbye LN, Satele DV, Sloan JA, Shanafelt TD. Concurrent validity of single-item measures of emotional exhaustion and depersonalization in burnout assessment. J Gen Intern Med. 2012;27(11):1445–1452.
19. Rosmarin DH, Pirutinsky S, Pargament KI. A brief measure of core religious beliefs for use in psychiatric settings. Int J Psychiatry Med. 2011;41(3):253–261.
20. Karakachian A, Colbert A. Nurses' moral distress, burnout, and intentions to leave: an integrative review. J Forensic Nurs. 2019;15(3):133–142.
21. Zhang YY, Han WL, Qin W, et al. Extent of compassion satisfaction, compassion fatigue and burnout in nursing: a meta-analysis. J Nurs Manag. 2018;26(7):810–819.
22. Hunt PA, Denieffe S, Gooney M. Burnout and its relationship to empathy in nursing: a review of the literature. J Res Nurs. 2017;22(1/2):7–22.
23. Duarte J, Pinto-Gouveia J, Cruz B. Relationships between nurses' empathy, self-compassion and dimensions of professional quality of life: a cross-sectional study. Int J Nurs Stud. 2016;60:1–11.
24. Teding van Berkhout E, Malouff JM. The efficacy of empathy training: a meta-analysis of randomized controlled trials. J Couns Psychol. 2016;63(1):32–41.
25. Copanitsanou P, Fotos N, Brokalaki H. Effects of work environment on patient and nurse outcomes. Br J Nurs. 2017;26(3):172–176.