Burnout in Mental Health Professionals: The Roles of Psychological Flexibility, Awareness, Courage, and Love

Introduction

The burntout phenomenon has been well-studied in psychology and is receiving increasing attention from contextual-behavioral approaches. In the dominant model of burnout, which has garnered much empirical support (Maslach et al., 2001), burnout comprises three dimensions: emotional exhaustion (EE; being emotionally overextended and depleted), depersonalization (D; feeling negative, callous, and excessively detached from clients and customers), and lack of personal accomplishment (PA; feeling incompetent and lacking experiences of success and achievement). Consistent with contextual-behavioral emphases, burnout is considered to be a syndrome involving these three dimensions, mutually influenced by contextual and individual variables (Maslach & Goldberg, 1998). Although burnout is a problem across many professions (Maslach et al., 2001), it may be a significant problem for those in medical and mental health professions who have continuous contact with human suffering (Grau et al., 2009; Ortega-Ruiz & López-Rios, 2004; Romani & Ashkar, 2014). In these contexts, depersonalization specifically may have important negative consequences for professionals and their clients.

A recent meta-analysis on the effectiveness of interventions to reduce burnout suggested that various interventions result in significant though modest reductions in emotional exhaustion but do not affect depersonalization or personal accomplishment (Maricuoiu et al., 2016). Interestingly, only one intervention, focused on teaching “soft” interpersonal skills, successfully improved feelings of personal accomplishment (Mackenzie & Peragine, 2003). A systematic review of burnout interventions with mental health workers similarly found no effectiveness of interventions on personal accomplishment (Dreison et al., 2018).

Contextual-behavioral interventions focusing on psychological flexibility—the primary target of Acceptance and Commitment Therapy (ACT; Hayes et al., 2009)—have promise for reducing burnout. Psychological flexibility (PF) refers to one’s ability to mindfully stay in contact with the present moment and act according to one’s values, even when one is experiencing unpleasant thoughts, emotions, and bodily sensations (Bond et al., 2011). PF sits on the positive end of a continuum with experiential avoidance, or the tendency to act to reduce unpleasant thoughts, emotions, and bodily sensations, even when doing so causes harm in the long run (Hayes et al., 1996). PF shows significant associations with all three dimensions of burnout in multiple samples (Losa et al., 2010; Noone & Hastings, 2011). In these studies, increased PF is related to decreased emotional exhaustion, decreased depersonalization, and increased personal accomplishment, with the strongest relations found for emotional exhaustion. An ACT intervention reduced emotional exhaustion and depersonalization in a sample of government workers, and improvements in PA mediated decreases in emotional exhaustion but not depersonalization (Lloyd et al., 2013). Likewise, PF mediated burnout reduction in a mindfulness-based intervention for nurses (Duarte & Pinto-Gouveia, 2017). This study did not break burnout into its components.

Given their continuous contact with human suffering as part of their professional role, a contextual-behavioral approach to understanding and intervening on burnout may be particularly useful for mental health professionals. A simple functional model of burnout for mental health professionals would suggest that emotional exhaustion will be high and predicted from chronic contact with suffering, especially for those who are low in PF and prone to experiential avoidance (Losa et al., 2010; Noone & Hastings, 2011; Ortiz-Fune, 2018). Consistent with available data suggesting that depersonalization is a defensive reaction to emotional exhaustion (Ashforth & Lee, 1990; Sochos et al., 2012) and that depersonalization is often expressed as decreased reciprocity in interpersonal interactions (González-Morales et al., 2012; Rose et al., 2010), emotional exhaustion may lead to depersonalized clinical interactions to avoid contact with suffering (i.e., a disconnection from clients). Such depersonalization may be negatively reinforced in the short-term but lead to decreased personal accomplishment (i.e., less successful interactions with clients), and increased emotional exhaustion, producing a burnout cycle.

While the potential roles of experiential avoidance and psychological flexibility in this process have been previously explored, we suggest that the interpersonal contextual model of awareness, courage, and love (ACL) from Functional Analytic Psychotherapy (FAP; Kuczynski et al., in press; Maitland et al., 2017) also may be helpful. According to this model, awareness refers to the need for emotional clarity and acceptance in interpersonal interactions, courage refers to being able to engage in authentic and vulnerable self-disclosures and to ask for what one needs in interactions, and love refers to being able to respond to others’ courageous self-disclosures and requests in ways that convey safety, express emotional validation, and sensitively provide what was requested. In FAP, ACL skills are taught to therapists to improve their relational interactions with clients, in the service of establishing authentic and healing interactions (e.g., Kanter et al., 2013).

In short, our aim is to understand burnout with an integrative contextual-behavioral model. In this model, we assume an overlap between PF and ACL interpersonal skills. Although PF has been defined differently by different authors (Hayes et al., 2009; Luciano, 2016; Strosahl et al., 2012), there seems to be agreement on two essential pillars: mindful awareness and acceptance of private experiences and the ability to engage in valued actions. The first pillar overlaps with the awareness concept from the ACL model. Likewise, courage and love skills may be seen as specifications of valued actions (vulnerable behaviors) that may be particularly important in the context of interpersonal interactions when one person is experiencing professional burnout.

Specifically, we would expect emotional exhaustion to relate significantly with lower levels of awareness (and perhaps be more fully accounted for by psychological flexibility), but given the presumed interpersonal nature of depersonalization and its effects on personal accomplishment, we would expect depersonalization and lack of personal accomplishment to be more significantly and uniquely associated with lower levels of courage and love, as a diminished capacity for courage and love would be experienced as depersonalization and directly impact clinical interactions. Given the general lack of efficacy of burnout interventions with respect to these components of burnout in particular, establishing such relations is important in suggesting that FAP interventions targeting increased ACL (e.g., Kanter et al., 2018; Kohlenberg et al., 2015), either alone or integrated with ACT interventions, may be helpful in addressing the complete syndrome of burnout in mental health professionals and others at risk. In fact, a combined ACT and FAP burnout intervention that did not incorporate ACL demonstrated improvements in emotional exhaustion but not depersonalization (Macías et al., 2019), while a single-subject investigation of a FAP burnout intervention that focused primarily on ACL with six mental health professionals documented large reductions in both emotional exhaustion and depersonalization for five of them (as well as large improvements in personal accomplishment; Reyes-Ortega et al., 2019).

In the current study, we explored these processes in a cross-sectional survey of Spanish mental health professionals. We were interested in (1) replicating previous findings that psychological flexibility is associated with all facets of burnout and emotional exhaustion most strongly, and (2) identifying the unique predictive power of ACL in explaining dimensions of burnout above and beyond psychological flexibility. We hypothesized that courage and love in particular would significantly predict depersonalization and personal accomplishment over and above psychological flexibility. To our knowledge, this is the first exploration of how these contextual-behavioral processes collectively and uniquely predict burnout in mental health professionals. Furthermore, previous research has employed the general form of the widely used measure of flexibility, the Acceptance and Action Questionnaire (AAQ). We instead employed AAQ work-specific adaptation (Bond et al., 2013) which may offer a more precise understanding of these processes in context.

Method

Participants

Participants were 269 Spanish mental health professionals who completed an online survey. The majority of participants were female (81%), psychologists (84.8%), working 30 or more hours weekly (83.5%), and working in the Spanish Public Health System (53.9%). A detailed description of participants’ demographic characteristics is presented in Table 1.

Measures

Burnout. The Spanish version (Gil-Monte, 2005) of the Maslach Human Services Survey Burnout Inventory (MBI; Maslach et al., 1996) is a 22-item self-report measure comprising three dimensions as discussed above: EE (9 items; e.g., “I feel emotionally drained from my work”), D (5 items; e.g., “I treat some clients as if they were impersonal objects”), and PA (8 items; e.g., “I have accomplished many worthwhile things in this job”). We used a 5-point scale from 1 (never) to 5 (everyday). A score for each subscale was computed by taking the sum across all relevant items such that higher scores reflected greater emotional exhaustion, depersonalization, and feelings of personal accomplishment. In the current sample, the internal consistencies for each scale were: EE = .88, PA = .81, and D = .75.

Work-related psychological flexibility. The Spanish version (Ruiz & Odrionzola-González, 2014) of the Work-Related Acceptance and Action Questionnaire (WAAQ) measures psychological flexibility (engaging in goal-directed actions in the presence of unpleasant emotions) in a workplace context, with a total of 7 items (e.g., “I can still work very effectively even if I am nervous about something”). Each item is assessed on a 7-point scale from 1 (never true) to 7 (always true). The final score is obtained by the total sum of items with higher scores reflecting higher levels of workplace psychological flexibility. Internal consistency is .87.

Awareness, courage and love. The Spanish-language Social Interaction Scale, developed by Reyes Ortega and Kanter (2017) as a quick assessment of awareness, courage, and love repertoires as described in FAP’s model, was used. The scale includes 13 self-report items across four subscales; three of the subscales were included in the current analysis: Awareness (A; 3 items), Courage (C; 3 items), and Love (L; 3 items). A assessed the degree of self and others’ awareness in interactions, C assessed the ability to engage in emotional self-disclosure and ask for what one needs in relationships, and L assessed the ability to respond to others’ courage with safe, validating, and giving responses. We do not consider the 4^th subscale (Accepting Love) to play a role in our burnout model. Mental health workers’ responsiveness represents better the role they play with clients than the skill of accepting love from others. Each item is responded to on a 7-point scale from 1 (totally false) to 7 (totally true). In the current sample, the internal consistencies for each scale were: A = .68, C = .64, and L = .74. English versions of the nine items used in the current study are presented in Table 2.

Procedure

The survey was created using the Google Form tool. Sampling was intentional, with a request to take the survey shared with different mental care associations such as the Spanish National Association of Clinical Psychologists and Residents (ANPIR). The survey first included an informed consent, followed by demographic questions and then the MBI, WAAQ, and SIS. The survey was anonymous. This procedure and the way to collect and save date are in compliment with the Spanish law of protection for personal information.

Data Analysis

To explore internal consistency of each measure in this sample, Cronbach’s alpha test was used (see Measures).

Two different procedures were used to explore the hypothesis. First, a multiple linear regression with the SIS subscales all entered in the same step to explore ACL contributions to burnout independently. Second, a hierarchical regression analysis including two blocks, with WAAQ scores entered at Block 1 and SIS subscales entered at Block 2, to explore the independent contributions of WAAQ (replicating previous research) and the unique contributions of ACL over and above WAAQ. For each procedure, three different analyses (one for each MBI subscale) were computed, checking the necessary assumptions.

Results

Means and standard deviations for each measure/subscale and simple correlations between measures are presented in Table 3. Means are moderated for the three burnout dimensions, with a higher dispersion found in D. Scores in WAAQ and SIS subscales are moderately high. The total score of the WAAQ and all ACL subscales correlated significantly with burnout’s components as measured by the MBI.

The relations between WAAQ and SIS subscales with the three MBI dimensions appeared linear. There is no multicollinearity in our data (VIF scores are all below 10, and tolerance scores above 0.2), the values of the residuals are independent (Durbin-Watson values between 1.85 and 1.92) and normally distributed, and the homoscedasticity assumption was accepted for all regression models. Cook’s distance values were all under 1, suggesting individual cases were not unduly influencing the model. Standardized β, t, F, r, and R squared values, and the significance values for the overall models and specific variables, are presented in Table 4 (regression analysis with the SIS) and Table 5 (hierarchical regression analysis with the WAAQ and the SIS).

Emotional Exhaustion

Regression analysis predicting EE from SIS produced a significant model, with 5.2% of the variance explained by the SIS. Only Love showed a significant (negative) relationship. The two hierarchical regression models were significant. For EE, the WAAQ accounted for 9.3% of variance, showing a negative and significant association. When SIS subscales were added, the amount of variance accounted for increased to 10.5%. No SIS subscales showed a significant association.

Depersonalization

The regression analysis predicting D from SIS subscales produced a significant model, with 8.1% of variance explained by SIS. Courage and Love showed negative and significant relationships with D. The two hierarchical regression models were significant. For D, WAAQ explained 5.8% of the variance, showing a negative and significant association. When SIS subscales were added, the model explained 9.9% of variance. Love showed a negative and significant association.

Personal Accomplishment

The regression analysis predicting PA from SIS produced a significant model, accounting for 18.8% of variance. All three subscales showed positive and significant relations. The two hierarchical regression models were significant. For PA, WAAQ accounted for 21.1% of variance, showing a positive and significant association. When SIS subscales were added, the model accounted for 28.7% of the variance. A, C, and L showed positive and significant associations.

Discussion

This paper sought to replicate, in a sample of Spanish mental health professionals, previous findings that psychological flexibility is associated with all three dimensions of burnout, and to explore the unique predictive power of a previously unexplored set of constructs—awareness, courage, and love—in explaining dimensions of burnout above and beyond psychological flexibility. We hypothesized that courage and love in particular would significantly predict depersonalization and personal accomplishment. Results demonstrated a strong relationship between psychological flexibility in the workplace and burnout’s three dimensions. Love was the only ACL skill relevant to emotional exhaustion, but this effect disappeared when psychological flexibility was taken into account. Instead, as predicted, love was uniquely relevant for both depersonalization and personal accomplishment above and beyond psychological flexibility. Courage was relevant to depersonalization and personal accomplishment as well, with effects strongest for personal accomplishment. Awareness was only related with burnout’s dimension of personal accomplishment, over and above the influence of psychological flexibility. In sum, it seems that being psychologically flexible (and fully aware) in workplace circumstances and having skills related with courageous and loving response classes are related with lower levels of emotional exhaustion (flexibility), depersonalization (flexibility and love), and higher feelings of personal accomplishment (flexibility, awareness, courage, and love).

The obtained relations between psychological flexibility and burnout are consistent with the literature (Lloyd et al., 2013; Losa et al., 2010; Noone & Hastings, 2011). Likewise, the results for interpersonal constructs of courage and love (Kuczynski et al., in press; Maitland et al., 2017) are consistent with literature on the importance of social relationships in workplaces. For example, variables such as support from colleagues (Sochos et al., 2012), relationships with colleagues, clients and the organization (Rose et al., 2010), and collective perception of burnout (González-Morales et al., 2012) are related with burnout and its dimensions.

In general, results support the proposed functional model of burnout in mental health professionals. Briefly, this model suggests that an experiential avoidance response in the context of continuous contact with human suffering may backfire and increase emotional exhaustion. Such avoidance may also manifest as a depersonalization of interpersonal interactions in response to feeling emotionally exhausted and be maintained by short term consequences (negative reinforcement). The overall result of this pattern may be a decrease in experiences of personal accomplishment. However, results are modest, especially for emotional exhaustion and depersonalization. These two dimensions are moderated by many aspects, some of them out of workers’ control. Personal accomplishment, on the other hand, could be more related with workers’ behavior and its consequences.

Courage and love findings add additional, potentially important considerations. Specifically, while WAAQ findings suggest that experiential avoidance in the context of high levels of contact with human suffering may backfire, producing more emotional exhaustion and depersonalization, the relation between love and depersonalization suggests that a diminished capacity to respond to others’ vulnerability—a primary task of therapists with their clients—may be an important element of this depersonalization response above and beyond experiential avoidance. Likewise, the relation between the three ACL dimensions and personal accomplishment suggests that skills like capacity for awareness of self and others, talking about one’s vulnerable feelings and asking for what one needs, providing safety, validation and what others have asked may be particularly important for maintaining a sense of personal accomplishment when working in environments characterized by high levels of contact with human suffering. These findings are consistent with recent successful interventions for burnout that used FAP (Macías et al., 2019) and ACL (Reyes-Ortega et al., 2019) and suggest the potential of these techniques to be useful additions to burnout interventions.

The significant predictions of personal accomplishment by both psychological flexibility and the ACL model are important findings. According to the literature, this dimension (and depersonalization) is harder to successfully intervene upon, with non-significant meta-analytic findings (see Maricuoiu et al., 2016). In the current study, combination of psychological flexibility and ACL explained more variance in personal accomplishment than in other burnout dimensions, suggesting that such variables may be targeted to improve effectiveness of burnout interventions. Although the overall amount of variance explained may be interpreted as small, the obtained effects for the current study are consistent in size with if not larger than the effects obtained in other studies (e.g., Dreison et al., 2018; Maricuoiu et al., 2016). That said, it is important to consider that currently unknown moderator variables may be important to identify to improve the prediction and influence of burnout.

Limitations of this study include measures used, the correlational nature of data, and characteristics of the sample. Regarding measures, multiple versions of the MBI have been employed making it difficult to relate results precisely with other studies. Likewise, we used AAQ’s workplace-specific version, offering a more valid estimate of flexibility in this setting but making it harder to compare to results of studies using general AAQ. Finally, to our knowledge, this is the first empirical publication using SIS; as such it is not a previously validated measure. Our results offer initial support for the scale’s internal consistency and predictive validity in this context.

Although the obtained findings generally support our proposed model, the data are correlational in nature and proposed causal relations between variables have not been supported. For example, the foundational causal sequence of experiential avoidance, in the context of high contact with human suffering, producing emotional exhaustion, has not been supported. It may be the case that emotional exhaustion is produced by direct contact with suffering regardless of avoidance and makes such avoidance more likely. It is likely the case that such relations are bi-directional and reciprocal. That said, contextual-behavioral approaches aimed to offer pragmatically useful models, geared towards maximizing intervention success (Hayes et al., 2012), and targeting psychological flexibility are useful given the armamentarium of available techniques. The same may be said, to a lesser but still valid degree, for FAP’s courage and love. Also, the overlap between concepts, such as PF and ACL subskills, could explain the poor additional contribution by the ACL model, over and beyond psychological flexibility. Analyzing different interventions (including or not ACL dimensions) could explain better their roles.

Finally, the recruitment procedure and sampling produced a sample primarily of female Spanish psychologists. While this is an important sample in its own right, and we have no reason to suspect that obtained results are idiosyncratic to their unique characteristics, it is important to explore this model with additional international samples and a broader array of mental health professionals.