Therapist self-compassion and compassion fatigue: the mediating role of resilience

: The demands placed on professionals working with psychological problems can result in a risk of burnout (Yang & Hayes, 2020). The ob-jective of this study was to analyze the possible role of resilience as a mediator variable in the relationship between self-compassion and compassion fatigue, burnout, and compassion satisfaction. A descriptive cross-sectional study design was used. Sixty-five clinical psychologists (50 women and 15 men) aged between 23 and 71 years ( M = 33.8, SD = 10.8) participated in the study. The therapists completed the Self-Compassion Scale, the 14-Item Resilience Scale, and the Professional Quality of Life Scale. Separate simple mediation models were tested to examine the extent to which each of the Resilience scale variables attenuated the relationship between Self-compassion and Professional Quality of Life. Serial multiple mediator models were performed to explore whether Resilience variables acted interactively as mediators in the association between Self-compassion and Professional Quality of Life. Simple mediation analyses showed that Personal Competence was a significant partial mediator in the relationship be-tween Self-Compassion and Compassion Satisfaction and Burnout. Personal Competence and Acceptance of Self and Life were not significant mediators of the relationship between Self-Compassion and Secondary Traumatic Stress.


Introduction
Establishing helping relationships with the people with whom one works in different contexts involves the development of skills and capacities linked to the management of our empathic ability and the development of a helping alliance (Friedlander et al., 2009).In the case of therapists, one of the characteristics that enables them to be more effective is having a sincere interest in the person with whom they work (Cormiern & Cormier, 1994).Still, persistent exposure to patient suffering and inadequate measures to promote self-care sometimes generate negative consequences (Gimeno, 2021).The demands placed on professionals working with psychological problems can result in a risk of burnout (Yang & Hayes, 2020).According to Simionato and Simpson (2018), half of the psychotherapists surveyed claimed to suffer burnout at moderate to high levels.The stress experienced by therapists can have a negative impact on their ability to put themselves in the client's shoes and has been linked to a range of adverse outcomes (Lambert & Barley, 2001).The very nature of the therapist-client relationship can result in indirect adverse stress reactions such as compassion fatigue and burnout (Cieslak et al., 2014).
Compassion fatigue is a secondary form of stress in the therapeutic helping relationship.It occurs when the emotional capacity of the health professional to cope with the empathic commitment to the patient's suffering is overwhelmed (Figley, 1995).Compassion fatigue reduces our ability or our interest in bearing the suffering of others, according to the theoretical model for the study of compassion fatigue developed by Stamm (Stamm, 2009).It can be defined as the negative aspect of professional quality of life and is divided into two dimensions: (1) burnout and (2) secondary trauma, vicarious trauma, or secondary traumatic stress, which refers to negative feelings driven by work-related fear and trauma.Moreover, there is also a benefit derived from helping people who are suffering: compassion satisfaction.This phenomenon has been defined as the feeling of achievement and pleasure derived from doing one's work well and effectively (Radey & Figley, 2007;Stamm, 2010).Stamm (2010) asserts that compassion satisfaction does not prevent the development of compassion fatigue.However, he affirms a relationship between both variables since compassion satisfaction would increase the ability to withstand secondary traumatic stress.
Therapist self-care strategies have proven to be a practical resource for developing the ability to manage one's own emotions and impact those of others (Boellinghaus et al., 2013).Self-compassion has consequently appeared as the emerging construct in the literature on therapist self-care.Self-compassion is a multidimensional state that concerns being kind, warm, and understanding when one is suffering or makes mistakes.Self-compassion encompasses multiple dimensions that can be developed and learned: self-kindness, common humanity, and mindfulness (Neff, 2003).Preliminary theory and evidence suggest that relating to oneself with compassion is a promising means of promoting self-care, professional well-being, and improved coping with stress among health professionals (Finlay-Jones et al., 2015;Kemper et al., 2015).Previous studies conclude that more compassionate and self-compassionate professionals experience fewer symptoms of compassion fatigue and greater compassion satisfaction (Mantelou & Karakasidou, 2019;Patsiopoulos & Buchanan, 2011).However, those who are more selfcritical are more likely to experience compassion fatigue (Beaumont, Durkin, Hollins Martin, & Carson, 2015).
Some studies have identified resilience as a mediating factor between self-compassion and psychological outcomes (Fang-Fang et al., 2022;Pérez-Aranda et al., 2021).For Wagnild (2009), resilience is defined as a personality characteristic that moderates the negative effect of stress and promotes adaptation.Resilience has been identified as an essential protective factor against traumatic events and the effects of adversity on mental and psychological health.Specifically, previous research has shown that resilience may have a crucial mitigating role in reducing the devastating impact of compassion fatigue and burnout among disaster mental health providers (Burnett & Wahl, 2015).
Given that the above research shows evidence supporting self-compassion in addressing psychological issues, it would be interesting to more deeply understand the potential influence of self-compassion in enhancing the quality of life of psychotherapists.For this reason, the objective of this study was to analyze the possible role of resilience as a mediator variable in the relationship between self-compassion and compassion fatigue, burnout, and compassion satisfaction.Based on the literature review, the following hypotheses were formulated.First, we hypothesized that selfcompassion might negatively influence compassion fatigue and burnout and positively influence compassion satisfaction (Hypothesis 1).We also hypothesized that resilience might mediate the relationship between self-compassion and compassion fatigue , burnout, and compassion satisfaction (Hypothesis 2).

Method Design
A descriptive cross-sectional study design was used.

Participants
The original sample comprised 120 clinical psychologists.The following inclusion criteria apply: to have a degree or degree in psychology and experience as a therapist.Exclusion criteria included not having signed the informed con-sent.One person refused to continue collaborating with the study, so that case was excluded from the sample.In another case, a clear acquiescence bias in responses to the scales was detected; for reasons of prudence, this case was also excluded.No missing data were allowed; therefore, 53 subjects were excluded.The final sample comprised 65 clinical psychologists (50 women and 15 men) aged between 23 and 71 years (M = 33.8,SD = 10.8).All of them were actively working at the time of the study.The mean number of years of professional experience was 6.62 (SD = 7.65).Almost twothirds assumed a cognitive-behavioral theoretical approach (63%).The sociodemographic information and outcome measures are presented in Tables 1 and 2, respectively.

Procedure
The participants were recruited through snowball sampling via social media.Moreover, strategic centers, such as the Official Association of Psychologists of Andalusia, were contacted.On the first page of the survey, an information sheet explaining the objectives and nature of the study was presented.The participants provided informed consent before completing the form.The study was approved by Ethics Committee of the Loyola University.

Measures
Demographic information.The participants were asked to provide information about their sex, age, and years of work experience.They were also asked to indicate the psychological model according to their clinical practice.
Self-Compassion.The Self-Compassion Scale (SCS; Neff, 2003;Spanish version, Garcia-Campayo et al., 2014) was used to measure self-compassion.The SCS is a 26-item scale with Likert response items ranging from 1 (Almost never) to 5 (Almost always).The scale consists of six subscales; three are positive (Mindfulness, Common Humanity, and Self-Kindness), three are negative (Over-Identification, Isolation, and Self-Judgment), and a Total Self-Compassion Score.In the current study, only the Total Self-Compassion Score was used in the analyses.A higher Total Self-Compassion Score was interpreted positively.The Total Self-Compassion Score showed good internal consistency (McDonald's ω = .90).
Resilience.The 14-Item Resilience Scale (RS-14; Wagnild, 2009;Spanish version, (Sánchez-Teruel & López-Bello, 2014) was used to measure resilience.The scale is a brief version of the original 25-item scale (Wagnild & Young, 1993).The scale rates the items on a 7-point Likert Scale ranging from 1 (Strongly disagree) to 7 (Strongly Agree), with higher scores indicating greater levels of resilience.The scale consists of two subscales: Personal Competence (11 items) and Acceptance of Self and Life (3 items).The internal consistency was adequate for Personal Competence (ω = .72)and poor for Acceptance of Self and Life (ω = .46).
Compassion Satisfaction, Burnout, and Secondary Traumatic Stress.The Professional Quality of Life Scale (PROQOL; Stamm, 2010) is a 30-item self-report questionnaire designed to measure compassion fatigue, work satisfaction, and burnout in helping professionals.Participants were asked to rate the items about their experience as a therapist on a 5-point Likert scale from 1 (Never) to 5 (Very Often) limited to the last 30 days.The scale contains three subscales: Compassion Satisfaction (10 items), Burnout (10 items), and Secondary Traumatic Stress (10 items).Higher scores represent higher levels of compassion satisfaction, burnout, and secondary traumatic stress.The internal consistency was adequate for Compassion Satisfaction (ω = .78)and poor for Burnout (ω = .51)and Secondary Traumatic Stress (ω = .67).

Data Analyses
The data distribution was explored to select parametric or non-parametric analysis.The normal distribution was checked by inspecting visual Q-Q plots; the Z-statistic for skewness and kurtosis was also calculated.Given the sample size, a cutoff of Z > |2.58| for a non-normal distribution was used in conjunction with examining histograms (Mayers, 2013).To check if age or years of experience needed to be controlled for in later analyses, their correlations with the outcome measures were examined.The sex variable was also explored as a potential covariable.Zero-order intercorrelations between all outcome measures were computed to substantiate the consideration of possible indirect influences of Self-compassion on Professional Quality of Life.All significantly correlated variables were retained to be included in the mediation analyses.A correlation coefficient of .10 is considered a small effect, .30 is considered a medium effect, and .50 is considered a large effect (Cohen et al., 2002).Separate simple mediation models were tested to examine the extent to which each of the Resilience scale variables attenuated the relationship between Self-compassion and Professional Quality of Life.All analyses were completed using biascorrected bootstrapping to minimize Type II error (Shrout & Bolger, 2002).Bootstrapping was used to establish the statistical significance of all total, direct, and indirect effects.If the upper and lower bounds of the bias-corrected 95% confidence intervals do not contain zero, the effect is considered significant.The mediation analyses were conducted using the PROCESS macro for IBM SPSS (Hayes, 2017).Finally, we performed serial multiple mediator models to explore whether Resilience variables acted interactively as mediators in the association between Self-compassion and Professional Quality of Life.We used 10,000 bootstrap samples for both simple and multiple mediator models.

Results
Age, years of work experience, Self-Awareness, Compassion Satisfaction, and Secondary Traumatic Stress variables were non-normally distributed.Thus, non-parametric analyses were conducted.A Mann-Whitney test indicated that the Total Self-Compassion Score was higher for women (Mdn = 22.0) than for men (Mdn = 20.4),U = 223, p = .024,r = -.39.The sex variable was therefore included as a covariate in the subsequent mediation analyses.Only the years of work experience variable was correlated with an outcome measure variable (Secondary Traumatic Stress; ρ = -.35,p = .005).
The intercorrelations among outcome measures are presented in Table 3.As expected, the Self-Compassion and Professional Quality of Life measures were significantly correlated.The Total Self-Compassion Score and the Compassion Satisfaction subscale were strongly positively correlated (ρ = .47,p < .001).In the opposite sense, the Total Self-Compassion Score was moderately negatively correlated both with the Burnout (ρ = -.31,p = .012)and Secondary Traumatic Stress (ρ = -.35,p = .001)subscales.However, the correlation between the Self-Compassion and Resilience measures was only significant for the Self-Management subscale (ρ = .43,p < .001).All correlations between Resilience and Professional Quality of Life measures were significant (from ρs = -.62 to ρs = .43),except the correlation between Secondary Traumatic Stress and the Acceptance of Self and Life subscales (ρ = -.22,p = .085).
Results indicated that whereas the direct effect of Self-Compassion on Personal Competence was significant (Figure 1, a paths; B = 1.00; 95% CI [.55, 1.46]), the effect on Acceptance of Self and Life was not significant (B = .21;95% CI [-.02, 0.44]).Personal Competence was a significant partial mediator (Figure 1a) in the relationship between Self-Compassion and Compassion Satisfaction (indirect effect = .40;95% CI [ .15, .68]).That is, Self-Compassion was positively related to Personal Competence, which was positively related to Compassion Satisfaction.In addition, Personal Competence was also a significant partial mediator (Figure 1c) in the relationship between Self-Compassion and Burnout (indirect effect = -.22;95% CI [-.48, -.02]).Similar to the previous analysis, Self-Compassion was positively related to Personal Competence, which was negatively related to Burnout.

Serial Mediation Analyses
The Resilience variables (Personal Competence and Acceptance of Self and Life) were introduced as mediators in the relationship between Self-Compassion and Professional Quality of Life (only for the Compassion Satisfaction and Burnout variables subscales).As with simple mediation analyses, the effect is considered significant based on 95% confidence intervals that do not include zero.
As can be seen in Figure 2, the total effect of Self-Compassion on Compassion Satisfaction (Figure 2; path c; B = .74,95% CI [.42, 1.05]) was significantly attenuated when Personal Competence and Acceptance of Self and Life were included as mediators (Figure 2, path c'; B = .33,95% CI [.03, .63]).While Personal Competence had a unique effect on Compassion Satisfaction (B = .35,95% CI [.19, .51]),Acceptance of Self and Life did not (B =.26, 95% CI [-.05, .57]).Neither Acceptance of Self and Life alone (i.e., independent of the influence of Personal Competence) nor the combined effect of Personal Competence and Acceptance of Self and Life significantly explained between-group differences in Compassion Satisfaction (indirect effects: B = .03,95% CI [-.05, .09]and B = .03,95% CI [-.00, .10],respectively).The indirect effect of the resilience variables Personal Competence and Acceptance of Self and Life on the relationship between Self-Compassion and Burnout was also non-significant (Figure 3; B = -.06;95% CI [-.14, .01]).Serial mediation analysis (Personal Competence and Acceptance of Self and Life as mediators, Self-Compassion as the dependent variable) was rerun by introducing age as a covariate.Sex showed no significant effect in any variable of the model.The indirect effect was significant only for Personal Competence (B = .34,95% CI [-.11, .64])when sex was included as a covariate in the model.Including the covariate did not enhance the model fit.

Discussion
The present study aimed to understand how self-compassion improves quality of life in helping professionals.Specifically, the potential mediating effect of resilience on burnout, secondary traumatic stress, and compassion satisfaction was examined in a cross-sectional sample of clinical psychologists.Our findings suggest that self-compassion plays a significant role in the quality of life of therapists.This hypothesis has already been explored in previous studies (for example, Mantelou, & Karakasidou, 2019;Patsiopoulos & Buchanan, 2011), but why therapists with high self-compassion report better quality of life than others is still unclear.
Our results show that the effect of self-compassion was mainly mediated by Personal Competence but not by Acceptance of Self and Life.In addition, the mediation effect was significant only for Compassion Satisfaction and Burnout, but not for Secondary Traumatic Stress.In line with previous studies (Kemper & Khayat, 2015), our results conclude that therapists with high self-compassion face suffering with an attitude of kindness and warmth.Furthermore, they show higher levels of self-reliance, determination, and personal satisfaction than therapists with lower levels of selfcompassion.Resilience was found to yield protective effects against the effect of emotional exhaustion derived from one's work, which improves the therapist's quality of life (Voon et al., 2021).However, personal competence does not mediate between self-compassion and secondary traumatic stress.As a prior study suggested (Figley, 2002), both self-compassion and secondary traumatic stress could be mediated by other variables such as empathic ability, the therapist's traumatic memories, or disengagement (understood as the ability of psychotherapists to distance themselves between sessions from client suffering).
The present study was subject to some potential methodological weaknesses and limitations.First, the partial mediating effect of resilience on the relationships between selfcompassion and Compassion Satisfaction suggests that other factors might not have been considered.Future research should explore other potential variables which could mediate the relationship between self-compassion and therapist quality of life.For example, some authors suggest that emotion regulation could mediate the relationship between selfcompassion and psychological health by facilitating strategies to manage stressful situations derived from therapeutic work (Finlay-Jones et al., 2015).Second, in the present study, the internal consistency of some dimensions of the Resilience Scale (Self-Awareness) and the PROQOL (Burnout and Secondary Traumatic Stress) was poor.As in previous studies, these subscales appear to show weaknesses in internal consistency in other non-English speaking populations (Galiana et al., 2017).It is possible that, for this reason, our results are not as expected with these variables.
Despite its limitations, the present study provides several implications for clinical practice.To date, systematic studies of therapist self-care are scarce.Compared to all the published studies on how our patients change, we know far less about how therapists can manage discomfort and distress from working with other people's suffering (Norcross & VandenBos, 2018).Therefore, the results focused on specific strategies, such as self-compassion and resilience, that clinicians can implement to improve their psychological wellbeing and professional quality of life.Developing these

Figure 1 Figure 2
Figure 1 Indirect effects of Resilient measures on the relationships between Self-Compassion and Professional Quality of Life measures

Figure 3
Figure 3 Indirect effects of Resilient measures on the relationships between Self-Compassion and Burnout Therapist self-compassion and compassion fatigue: the mediating role of resilience 225 anales de psicología / annals of psychology, 2024, vol.40, nº 2 (may) techniques can offer therapists the necessary resources to help them adapt and improve their quality of life.Incorporating self-compassion-and resilience-based interventions into therapist training programs can help practitioners improve their professional quality of life and, ultimately, their effectiveness in clinical practice.