How are compassion fatigue, burnout, and compassion satisfaction affected by quality of working life? Findings from a survey of mental health staff in Italy

While mental health professionals experience similar organizational stressors to other human service workers, the intrinsic characteristics of mental health work constitute further stress factors. Mental health professionals face additional emotional strain by the very nature of their work with distressed individuals, often over extended periods.

Some critical risk factors seem to be particularly applicable to mental health care and, among these, the most important is the relationship that mental health professionals have with clients who are experiencing psychological trauma [1]. Through exposure to clients’ reenactments and accounts of traumatic experiences, helping professionals are themselves vulnerable to a variety of effects and symptoms, including depression, anxiety, intrusive imagery, numbing and avoidance phenomena, cognitive shifts, as well as social and relational problems [2‐6]. Such negative consequences have been described in terms of Compassion Fatigue, which is defined as the practitioner’s reduced capacity to be empathic or bear the suffering of clients [2, 3, 7]. On the other hand, the positive outcome from working with challenging clients has been defined as Compassion Satisfaction. Compassion Satisfaction, which reflects the rewards of caring for others, has been identified as a possible element that counterbalances the risks of Compassion Fatigue [8]. However, Compassion Fatigue and Compassion Satisfaction seem to be negatively correlated, i.e. higher levels of Compassion Fatigue may overwhelm the professional’s sense of efficacy and prevent him/her from experiencing Compassion Satisfaction [8, 9].

Burnout is defined as “a state of physical, emotional, and mental exhaustion caused by long-term involvement in emotionally demanding situations” [10]. Unlike Compassion Fatigue, Burnout is not directly related to exposure to traumatic material [11]. Compassion Fatigue differs from Burnout in that the latter evolves gradually and is a result of emotional exhaustion. In contrast, Compassion Fatigue can emerge suddenly. Thus, Burnout does not appear to capture the effects of trauma as an occupational stressor. A study of social workers involved with clients who had some exposure to the events of September 11, 2001, confirmed that exposure to clients traumatized by the World Trade Center disaster increased Compassion Fatigue, but not Burnout. Furthermore, both Compassion Fatigue and Burnout appeared to be associated with psychological problems [11].

These aspects were selected as key variables in our study as they may be a potential threat to staff wellbeing and, consequently, care quality. Among the detrimental effects of Compassion Fatigue, Burnout and lack of Compassion Satisfaction there is evidence of high turnover rates, negative attitudes towards patients, lack of communication, and clinical errors [3, 11, 12].

Possible risk and protective factors for Compassion Fatigue, Burnout, and Compassion Satisfaction have been identified, including: socio-demographic characteristics, such as age, marital status, and education [13, 14]; coping styles [2, 3, 15]; exposure to trauma [13, 16, 17]; personal history of trauma [18, 19]; job-related factors, including workload [17], length of experience [19], and attendance at training [13]; and, finally, work environment factors [3, 17].

Although the literature on work-related stress (especially on Burnout) in the healthcare sector is substantial, to date there is a limited amount of literature examining the three dimensions of Compassion Fatigue, Burnout and Compassion Satisfaction together and including the wide range of mental health professional categories; instead most studies focus on only one or two professional categories at a time.

This paper reports the findings of a study that adopted an original approach in framing the three dimensions of Compassion Fatigue, Burnout, and Compassion Satisfaction within a comprehensive model of quality of working life. We defined quality of working life as an individual’s broad job-related experience. We applied a multidimensional approach to the study of quality of working life, similar to that developed by the European Foundation for the Improvement of Living and Working Conditions (EUROFOUND) [20‐22]. Within this framework, two constitutive areas are recognized as characterizing quality of working life: the intrinsic quality of work and the quality of work-life balance [23‐26]. The intrinsic quality of work derives from the relationships between the worker’s needs and the intrinsic characteristics of the work. In this present study, six dimensions (economic, ergonomic, complexity, autonomy, control, and symbolic dimension) for the analysis of the intrinsic quality of work were included, which in turn were further subdivided into 11 indicators. Quality of work-life balance is about reconciling working life and life outside work. This is an essential condition both for encouraging entry into the labour market and for enabling people to remain at work [20]. Work-life balance was analysed in this study through two indicators (i.e., impact of work on life and impact of life on work), in order to capture bi-directional interconnections between work and life.

Against this background, the overarching aim of this study was to investigate whether and to what extent the quality of working life as perceived by mental health staff affects levels of Compassion Fatigue, Burnout, and Compassion Satisfaction.

The total number of staff invited to complete the survey was 461, of whom 416 responded to both questionnaires. Among them, 16 returns were excluded due to the exclusion criteria related to the amount of missing values, thus 400 surveys were retained for analysis, representing a response rate of 87%. Table 2 shows participants’ socio-demographic and occupational characteristics.

Overall, CF showed a mean of 14.4 (SD = 4.1), with scores of completed questionnaires ranging between 10 and 36; BO showed a mean of 22.9 (SD = 4.5) and a range of 10–39, while the mean value of CS was 32.7 (SD = 7.1), with scores ranging between 10 and 50. CF showed a first quartile of 11, a median score of 13 and a third quartile of 16; the quartiles for BO were 20, 23 and 25; finally, the CS scores had a lower quartile of 28, a median of 33 and an upper quartile of 38.

Tables 3 and 4 present the bivariate associations of the continuous and the categorical indicators of quality of working life respectively with the three outcome measures. Concerning CF, ergonomic problems and the two variables on work-life balance showed the strongest correlations (with higher values of such variables corresponding to increasing levels of CF, p-value < 0.001 in all cases). As for BO, the strongest associations were found with perceived quality of meetings and trust (in negative) and with ergonomic problems (in positive); other important associations were found with lower levels of organizational commitment and of autonomy, a higher impact of work on life (p-value < 0.001 in all the above-mentioned cases), and with perceiving more risks for the future (p-value = 0.001) and having a less positive opinion of current salary (p-value = 0.010). Finally, looking at CS, the most relevant correlations were found with higher levels of perceived quality of meetings and of organizational commitment, with important positive associations also observed with the indicators of autonomy and trust (p-value < 0.001 in all cases).

The results of the multivariate analysis are shown in Table 5. Variables that turned out significant in simple linear regressions were: for CF, ergonomic problems, job complexity, trust, perceived quality of meetings, impact of work on life and of life on work, and perceived risks for the future; in the case of BO, ergonomic problems, trust, autonomy, participation, perceived quality of meetings, organizational commitment, impact of work on life, perceived risks for the future and opinion on current salary; finally, for CS, ergonomic problems, trust, autonomy, participation, perceived quality of meetings, organizational commitment, impact of work on life, perceived need of training and risks for the future. After controlling for the other covariates, the regression models with CF and BO as outcomes showed that ergonomic problems were highly related to both (p-value equal to 0.008 in the former case and 0.002 in the latter one), with impact of work on life also emerging as a predictor of negative outcomes for CF (p-value = 0.009) and BO (p-value = 0.007). Finally, impact of life on work turned out to be a further predictor of CF (p-value = 0.011), while perceiving more risks for the future (p-value = 0.001) and lower levels of trust (p-value = 0.004) were associated with BO in multivariate analysis. As for CS, a highly significant association with perceived quality of meetings was found (p-value = 0.007), with need to attend training (p-value = 0.035) and perceiving no risks for the future (p-value = 0.012) also predicting higher CS levels. Fractional logit models (results not shown) led to similar p-values with respect to linear regressions, with direction of associations (except for participation in the CS regression, but with parameters close to 0 and non-significant in both cases) and significant results confirmed, further strengthening our findings. Moreover, organizational commitment switched from marginally non-significant to marginally significant (p-value = 0.049) in the regression related to CS, and perceived risks for the future turned out as significant in the regression on CF (p-value = 0.021). The comparison of the best subset regressions using socio-demographics and occupational variables with the final regressions highlighted that most of the predictive power of our models was accounted for by quality of working life indicators. In fact, adjusted R² increased from 0.01 to 0.15 in the case of CF, from 0.06 to 0.26 for BO, and from 0.06 to 0.18 in the regression having CS as outcome, showing that socio-demographics and occupational variables played a minor role, while variables related to job-related experience strongly increased the predictive power of our regression models.

Finally, when looking at the single items included in the indicators of quality of working life that turned out to be significant in the regressions, the strongest association was found with the item of time pressures as part of the ergonomic problems variable. More specifically, an increase in the BO score was found for people who perceived higher levels of time pressures, with a correlation coefficient of 0.32.

This study analysed in depth a core set of indicators of quality of working life and their role in affecting levels of Compassion Fatigue, Burnout, and Compassion Satisfaction among mental health staff.

Beginning with Compassion Fatigue and Burnout, one of the strongest associations was with ergonomic problems. In our models the experience of ergonomic problems significantly increased the levels of both outcomes. The presence of additional strains in the environment seems to have a negative impact on the ability to exert one’s role as a helper, thus exacerbating levels of both Compassion Fatigue and Burnout. This is consistent with previous research; Abu-Bader [14] found that higher satisfaction, lower burnout and lower turnover were a function of adequate working conditions. Interestingly, among the ergonomic problems investigated in this study, those that seemed to have a stronger impact, especially on Burnout, were time pressures. Time pressures are likely to generate a feeling of being overloaded as well as reducing the possibility of debriefing and taking time off, thus creating a vicious circle where the risk of Burnout can seriously increase [7].

Additionally, increased levels of Compassion Fatigue and Burnout were predicted by higher levels of impact of work on life. This finding indicates that work-life balance is essential in preventing work-related negative outcomes, as various authors assert [7, 39]. The importance of reconciliation of work and private life for the quality of work and employment was acknowledged in the Europe 2020 strategy [40]. Our finding that the impact of work on life was related to stronger negative consequences than impact of life on work (which was significant only in relation to Compassion Fatigue) seems to show that participants were more concerned about work interference with their personal life than the reverse (for a review of studies on the interactions between personal life - especially family relationships - and quality of work life, see Colichi and colleagues [41]).

As for Burnout only, our results showed that trust predicted lower levels of such outcome. This result is in line with research demonstrating that receipt of support and feelings of being trusted by managers and co-workers are critical in protecting against Burnout [42‐44].

When looking at factors influencing Compassion Satisfaction, we found three significant explanatory factors, i.e. perceived need of training, perceived quality of meetings, and perceived risks for the future. These factors share similar features in that they reflect aspects of workers’ motivation and engagement, and differentiate from the others analysed as they seem potentially able to make a difference in a positive way. Personal and predisposing factors may be involved. Indeed, staff who are more motivated and engaged per se are more likely to perceive meetings as more useful and to invest in developing their skills and knowledge, consequently deriving more enjoyment from taking part in meetings and being keener to receive training. Moreover, a higher motivation could lead to be less concerned about future risks (e.g., losing job, having inadequate work competence, being unable to maintain one’s family, being unable to afford a substantial item, receiving an inadequate pension). However, due to the cross-sectional nature of our study, it is also possible to see motivation as an effect rather than a cause: in particular, since team meetings and training serve several important functions (as they potentially enable professionals to improve their skills, exert control over their work routines and give each other support), their quality could increase staff motivation, while feeling insecure about one’s future might affect staff motivation and investment in work. Therefore, although we may not draw any conclusion about the quality of either meetings or training perceived by our participants, both meetings and training should be considered by managers and team leaders as a way to improve Compassion Satisfaction.

Interestingly, perceived risks and uncertainties were associated with Burnout too, further suggesting the negative effects of job insecurity on staff wellbeing [20‐22, 45]. The experience of insecurity for the future may be regarded as a ‘catalyst’ element, i.e. it is significant in affecting both positive (Compassion Satisfaction) and negative (Burnout) outcomes. Thus, additional attention should be paid to this factor as its spectrum of action seems broader than that of other variables. Indeed, all other significant variables in the final regressions were statistically associated either with Compassion Fatigue and/or Burnout or with Compassion Satisfaction, but not with both. For example, ergonomic problems seemed to play an important role for negative but not necessarily for positive outcomes (despite being associated with Compassion Satisfaction in bivariate analysis, no significant association with such outcome was found in the regression). It may be argued that an adequate working environment is necessary in order to prevent negative consequences such as Compassion Fatigue and Burnout, but it is not a crucial factor in increasing levels of Compassion Satisfaction.

The strengths and limitations of this study should be acknowledged. The main strength of the study is the relatively large number of participants and the inclusion of multiple centres and of different mental health professionals working in varied settings, such as inpatient psychiatric units, outpatient, day, and residential services. In addition, the response rate was high. A further strength of the study is the inclusion of a large set of variables measuring socio-demographics, occupational characteristics and a core set of indicators of quality of working life. The latter made it possible to capture the subjective experience and perceptions of participants about their quality of working life as a whole. However, some limitations of the study should be taken into account. First, the cross-sectional nature of the survey did not allow the determination of causality. Moreover, as only centres from two regions in Northern Italy were included, the study results do not necessarily generalize to the whole country. Furthermore, because this study emphasized the role of organizational and environmental factors, other possibly relevant variables were not included in the survey. For example, individual coping strategies with work-related stresses were not considered, although extensive literature demonstrates their crucial role in tackling difficult situations at work [46‐48]. Finally, this study analysed outcomes related to staff’s perception in mental health settings; future studies could integrate such results by considering different organisational settings and/or by including clients’ measures of effectiveness of care as main outcomes.

These findings are potentially useful for health managers and team leaders in identifying factors affecting the ability to be compassionate and the risk of Burnout. It is widely recognized that lack of Satisfaction, Compassion Fatigue and Burnout have adverse consequences on work performance and service quality. Compassion Fatigue is likely to result in problems such as misjudgements, clinical errors and poor treatment planning, all serious issues for effective care [3, 11, 12]. Therefore, in order to provide adequate mental health services, managers need to provide their employees with adequate ergonomic conditions, paying special attention to time pressures. Building trustful relationships with management and within teams is also crucial. Training and meetings are other important targets for potential improvement, although how meetings and the need for training are perceived might depend on the motivation of members of staff themselves. Additionally, the issue of insecurity for the future should be taken into serious consideration as it can affect both Burnout and Compassion Satisfaction. Lastly, to positively influence their clients’ wellbeing, managers, administrators, and staff themselves need to consider strategies to reduce possible work-life conflicts.