1. Introduction
During the current COVID-19 pandemic in Spain, several professional groups have emerged as indispensable in individual and collective protection. First, health professionals have found themselves on the “front line”, an aspect that has influenced their high levels of contagion [
1]. Since the declaration of the State of Alarm in Spain [
2], both military and security forces have had to guarantee the population’s security and ensure compliance with the confinement measures. Throughout these hard months, social workers have emerged as the professional group responsible for protecting the most vulnerable people, guaranteeing their well-being.
Social workers play a central role in social policy [
3,
4,
5]. During the most challenging period of the pandemic in Spain, these professionals focused their roles and functions on promoting the well-being of the most vulnerable people, including those who could not cover their basic needs, those living on the street, people at risk of social exclusion, those living alone, those in need of social and health support, people with disabilities, and the elderly or those in need of long-term care. On the social level, they have provided the cushion to keep a significant number of the population integrated and protected; however, this has also had consequences, one of them being the possible development of burnout syndrome.
The number of infections and deaths during the first wave increased exponentially every day [
6]. By mid-September, and in the middle of the second wave, the total number of cases in Spain since the beginning of the pandemic was 1,751,884, with 30,495 deaths, according to official figures [
6]. In this state of affairs, and considering the statistics of an excess mortality of 22% [
7], the essential professionals have had to exercise their profession, not exempt from the stress of the continuous exposure to the virus and the possible contagion, accentuated by the transmission to their relatives [
8].
Through the present research, we approach burnout syndrome in social workers and see what variables influence this social phenomenon. The collection of data will allow us to get a precise picture of how the professional collective is to ease these workers’ labor situation, implement measures that will allow their improvement [
9,
10], and prevent possible psychological pathologies. Indeed, many studies have been developed on burnout and anxiety of workers because of COVID-19 [
8,
11,
12,
13,
14,
15]. In the international context, studies are emerging on the influence of the pandemic on the appearance of burnout syndrome in social workers, although from specific specialties such as geriatrics [
16,
17].
Burnout, as the behavioral manifestation of occupational stress, secondary traumatic stress (i.e., observable reactions that develop when working with people who have experienced trauma with symptoms characteristic of post-traumatic stress disorder [
18,
19]), and the vicarious trauma (i.e., personal behavioral and/or emotional changes that occur because of empathic involvement with people who have experienced trauma [
19]), is a highly prevalent work-related stress reaction experienced in helping professions [
20], including social policy professionals such as social workers [
21,
22]. One area where burnout is especially relevant is in child social work, having significant repercussions in its different dimensions on these workers [
23,
24,
25,
26] and persons with disabilities [
27]. However, sometimes, these differences are minimal [
28], and there is no worldwide universal pattern within social work that marks high rates of burnout.
Studies on burnout in social workers in Spain have increased over the last decade [
21,
29,
30,
31,
32,
33,
34]. We highlight the association between working conditions and labor dissatisfaction and stress, with workload, lack of managerial support, and service user/family abuse being predictive factors in this phenomenon [
35]. Recent studies have shown how, in professional practice, working conditions affect the three dimensions of burnout of the Maslach Burnout Inventory (MBI) [
36]: emotional exhaustion (EE), depersonalization (DP), and a reduced sense of personal accomplishment (PA), having similar effects on both intrinsic and extrinsic factors [
37].
Other studies show that burnout mainly affects those social work professionals with more limited work experience and public sector workers [
38]. Moreover, EE is a significant factor in the decision to continue or not to pursue the profession and can, on the one hand, be a cause of the abandonment of social work, while, on the other, of high rates of job satisfaction and feelings of self-efficacy [
39]. Therefore, considering previous research, it can be observed that the labor context is a determining factor in burnout [
40], and it is mainly conditioned by: (a) the group with which they work, (b) work experience, and (c) labor organization. Thus, the need to implement measures to mitigate the effects that burnout can produce on social workers and increase the resources and tools to deal with stressful situations is clear [
41].
Given the above, it is hypothesized that the stress experienced by social workers during the COVID-19 pandemic in the course of their professional work may have influenced the different dimensions of burnout, especially in EE. This is because this subscale is characterized by feelings of helplessness, a lack of enthusiasm, and low self-esteem, which could have been very present during the pandemic because of the workload and the danger of contagion. Likewise, the appearance of this dimension could trigger the development of negative attitudes and insensitivity toward the people on whom the service they provide falls, i.e., DP.
This research has two objectives. The first one is to approach burnout in social workers, contextualizing the phenomenon in the first wave of the pandemic in Spain concerning the health crisis of COVID-19. The second is to understand if there are predictive variables that influence the levels of burnout in social workers based on the three subscales of the MBI [
34]: EE, DP, and PA.
3. Results
In a first approximation of the descriptive data, high EE levels in social workers were observed, reaching 70.3%; when the medium level was also taken into consideration, the level further increased to 85.0%.
The DP scale reached levels of 48.7%—or 71.0% if the medium and high levels are combined—, percentages that are not considered irrelevant because of the implications of this dimension, such as the negative and insensitive attitude of others whom social workers help or attend [
43], even blaming them for how they feel [
44]. Regarding PA, 36.6% showed low levels, but if the low–medium level is considered, this percentage increases to 67.0%. Finally, the prevalence of burnout, i.e., the triple condition of high levels of EA and DP and low PA levels, reached 20.4% of the social work professionals.
Concerning subjective perceptions, 29.0% of the participants stated that they currently needed psychological or psychiatric treatment, while 87.5% considered that their workplaces should provide these treatments to their workers. In contrast, 70.8% of social workers believed they might need psychological or psychiatric treatment in the future due to COVID-19.
Regarding organizational aspects, 69.2% stated that the absence of PPE increased their stress and anxiety levels in the face of the first wave of the pandemic. Moreover, 82.4% were forced to telecommute, and 79.5% said they did not feel recognized by the organization they worked for during the first wave of the pandemic. See overall
Table 3.
A descriptive approximation of the variables used according to the professional category was made. A cross-table analysis was performed, considering the chi-square level of significance, from which the following data were extracted. For this purpose, a distinction was made between: (a) non-healthcare public services (primary care and specialized social services), (b) health social services, (c) third sector, and (d) other. In Spain, the difference between these services is as follows. Non-healthcare public services (primary care and specialized social services) are all the services managed by the public administrations of the State that do not form part of the health sector or the private sector, whether or not they are for profit. Health social services are all those public services managed by the public administrations of the State that are exclusive to the health sector. The third sector includes all private institutions, whether or not for profit. Other comprises any other entity not included in the above descriptions.
First, the same levels of feminization were highlighted, regardless of the institution for which the social workers worked, as being approximately 90%. In terms of age cohorts, health social services workers were older than the rest. Concerning NPPS, third sector workers were over-represented, accounting for 39.1%. As for PPSS, all social workers, regardless of their institution, showed very high values, between 84% and 100%. High values are also observed for PPSN, whose range was between 64.0% of social workers in health social services and 78.4% in the “others” category.
In relation to PPE, the highest values were found among health social services professionals, precisely those who were most at risk at the beginning of the pandemic in the absence of personal protective equipment. Regarding TL, the professionals in the non-healthcare public services and “others” fields performed the least TL, standing at 64.0% and 65.2%, respectively. On the opposite side were the health social services and third sector workers, with 86.3% and 85.9%, respectively. The ROW variable showed the scant recognition felt by social work professionals from their institutions, especially significant with non-healthcare public service professionals, whose feeling of professional recognition stood at 12.0%. The highest value was not much higher, with 23.6% for non-healthcare public service professionals.
According to the professional category, an approximation of the MBI scales showed that the highest values of EE were for the non-healthcare workers (73.9%), while the lowest value was for the health social services workers (52.0%). In the case of DP, the highest value was recorded in the “others” category. The third sector workers recorded a lower weight (43.8%), almost the same level as the health social services professionals (44.0%). In PA, social workers in non-healthcare public services got a higher percentage at the low level (the one taken as a reference), with 39.8%. In contrast, workers in the third sector showed a lower percentage of PA, with 31.3%. Finally, concerning the overall rate of burnout, it was more common in the “other” category (26.1%) and lowest in the health social services workers (16.0%). See overall
Table 4.
Subsequently, a cross-table analysis of dependent and independent variables was performed, according to Pearson’s Chi-square significance level, shown in
Appendix A.
Finally, binary logistic regressions were applied to the three subscales of the MBI, all of which showed acceptable, robust, and reliable models. Each one of them is shown below.
The biennial logistic regression of the Sub-EU presented a statistically significant model (χ2 = 47.224, p < 0.000). The model explained 23.5% (Nagelkerke’s R2) of the variance of moderately high consumption and correctly classified 71.8% of the cases. The Hosmer–Lemeshow test showed no significant differences between the observed and predicted results in the model (p = 0.546).
The variables included in the equation were: (a) teleworking and (b) subjective perception of needing psychological/psychiatric treatment because of COVID-19 (NPPS). Regarding teleworking, it showed an odds ratio (OR) of 2424 (95%CI, 1129–5205; p = 0.023). Regarding NPPS, it showed an OR of 17,962 (95%CI, 5372–60,060; p = 0.000). The results of the involvement of NPPS as a predictive variable are striking, given that a person in this situation is almost 18 times more likely to suffer EE than another social worker who does not currently show psychological or psychiatric treatment needs as a result of COVID-19. Additionally, teleworking led to an increase in workload, reflected in the greater possibility of suffering EE, up to 2.5 times more than other workers who had not undertaken professional practice.
The binary logistic regression of the DP subscale presented a statistically significant model (χ2 = 20.593, p < 0.002). The model explained 9.8% (Nagelkerke’s R2) of the variance of moderately high consumption and correctly classified 63.3% of the cases. The Hosmer–Lemeshow test showed no significant differences between the observed and predicted results in the model (p = 0.580).
The variables included in the equation were: (a) age (41–50 years) and (b) perception about whether psychological/psychiatric support is needed (NPPS). In relation to age, the 41–50-year-old cohort showed an OR of 9.255 (95%CI, 1012–84,630; p = 0.049). In relation to NPPS, it recorded an OR of 2.275 (95%CI, 1248–4148; p = 0.007). Thus, those aged 41–50 years had very high levels of depersonalization for the reference age (>60 years) and, in turn, comprised the only age cohort that showed representative values, a sign that professionals of this age range have higher predictive values than those of the other ages.
Finally, it presented a statistically significant model (χ2 = 20.938, p < 0.002). The model explained 10.6% (Nagelkerke’s R2) of medium-high consumption variance and correctly classified 69.6% of the cases. The Hosmer–Lemeshow test showed no significant differences between the observed and predicted results in the model (p = 0.416).
The variables included in the equation were: (a) subjective perception of needing psychological/psychiatric treatment because of COVID-19 (NPPS) and (b) feeling that their work has been recognized (ROW). With respect to NPPS, it presented an OR of 0.458 (95%CI, 0.241–0.870;
p = 0.017). However, since the B value was negative (–0.780), it indicates that people without NPPS show a greater lack of self-fulfillment than the others, although the values are shallow. In relation to ROW, it recorded an OR of 2.306 (95%CI, 1.212–4.385;
p = 0.011). Therefore, those who felt recognized in their work during COVID-19 showed higher values regarding a lack of self-fulfillment than the others—up to 2.3 times more. See overall
Table 5.
4. Discussion
As seen from the results, the most striking thing about the burnout values in social workers was the high percentage in the EE subscale (70.3%). These data alone show that these workers’ workload and emotional level are so high that immediate measures should be taken to reduce them. The DP subscale also showed high levels (48.7%), an aspect that should be paid attention to, given that social workers intervene in most cases and, over time, with people. Therefore, high levels of depersonalization can lead to a distancing of the individual they are dealing with, both from their sphere and from their problem situation. The work environment and context both influence DP, with work overload and the conflict of the role they play being stressful factors in subjects that show these high levels.
Concerning PA, the low levels—those about a lack of personal fulfillment—reached 36.6%. That is, they were not as high as in the previous subscales. One reason for this is the job satisfaction of social workers, which is recognized in many studies and could act as a compensating factor for the previous subscales. For example, the professional practice of social work is oriented toward promoting personal or collective well-being, which impacts job satisfaction [
45,
46,
47]. In the current health crisis, job satisfaction may have emerged as a counterbalance to the high values of the previous subscales, especially EE. For this reason, the burnout scale was 20.4%. Despite such high percentages of EE and DP, the triple condition that determines this phenomenon (high levels of EE and DP and low levels of PA) did not occur.
However, 29.0% said they needed psychological treatment because of the first wave of COVID-19, a slightly higher rate than the number of people suffering from burnout (20.4%). However, the high percentage of people who considered that they might need psychological care in the future due to COVID-19 (70.8%) stands out. Considering that we are entering the second wave of the pandemic and that 87.5% of the participants said that psychological support should be provided from workplaces, prevention actions should be implemented immediately because of the high risk of psychological and psychiatric disorders derived from professional activity [
37,
48].
Continuing with the data, from an organizational perspective, 69.2% stated that their level of stress and anxiety increased as a consequence of the lack of PPE, lower records than those obtained in recent research with health professionals [
11]. Moreover, 82.4% had to telework during the first wave of the COVID-19 pandemic, in itself a stressor [
49] given the difficulties involved in working from home, along with the need to care for other members of the family unit, especially children and the elderly [
50]. Furthermore, the stress derived from teleworking mainly affects women [
51]. Here, in a profession with a high level of feminization (approximately 90% in this study), it could be considered that it affects most of the professional group. In part, for these reasons, it can be understood that 79.5% of the social workers did not feel represented by their institution, given the lack of support to reconcile work and family as a consequence of teleworking.
The results of the logistic regressions presented interesting results in each of the subscales. Especially relevant are the EE results, where those who needed psychological support because of COVID-19 were up to 18 times more likely to register high values in this subscale than the others. This shows the relationship between psychological and EE factors, which may imply comorbidity of factors associated with professional activity, even two-way feedback between them. The other predictive variable was telework, which increased (by 2.5 times) the risk of suffering EE, thereby making it a stressor to be considered in the future.
Concerning DP, those between 41–50 years of age were up to 9.3 times more likely to suffer from this situation than the reference value (people over 60 years old), which is a prominent figure. This may be justified because people in this cohort have more work experience than young people and, having taken on more responsibility, have been able to do more fieldwork than their older peers, and even have greater motivation and involvement, as reflected in some research [
52]. These elements should be corroborated in further study. Additionally, DP affected the subjective perception of needing psychological treatment in the future by up to 2.3 times more than those who thought they would not suffer it in a possible second wave. Therefore, finding oneself again in the scenario experienced during the spring of 2020 in Spain increased the risk of suffering DP, and with it, the emotional distance with the people one works with.
Regarding the PA subscale, the data are very interesting and contrary to what one might initially think. In the first instance, the people who did not feel that they needed psychological or psychiatric treatment were those who showed the most remarkable prediction of suffering from lack of self-fulfillment. Most times, the paradox is that those who say they do not need psychological help are the people who try to handle situations without asking for help, sometimes even making a mistake before asking for help. This self-demanding and perfectionist characteristic implies an over-exertion, meaning these individuals exhaust all possibilities before asking for help by not recognizing their limitations. This behavior and emotional style [
53] can lead to high emotional and mental stress levels and conclude in a lack of PA. In contrast, people who felt recognized in their work during COVID-19 showed up to 2.3 times higher values regarding a lack of PA than the others.
These data might be unexpected, but they are more understandable if we consider that greater personal recognition is related to responsibility, assumption of duties, involvement in work, etc. In this way, people who showed these qualities in professional practice during COVID-19 have been affected in terms of personal fulfillment by not achieving the expected results in their work activity, an aspect that can lead to demotivation and lack of personal confidence about work performance [
54,
55,
56,
57] and immediate bosses, even reaching irritability, discouragement, and negative feelings about their professional practice [
58,
59,
60].
As a result of the above, the hypothesis raised at the beginning of this article can be confirmed since the levels of EE presented by social workers were excessively high, as was also the case with PD. Among the limitations of this study is the sample size due to the difficulty of establishing contact with social workers and the volume of work that prevented them from being predisposed to collaboration. Further limitations were the imbalance of the sample in terms of gender, since women represented the majority, and the cross-sectional design not allowing for the establishment of cause-effect relationships.