Burnout syndrome and weekly workload of on-call physicians: cross-sectional study

ABSTRACT CONTEXT AND OBJECTIVES: Burnout syndrome (BS) is characterized by three dimensions: emotional exhaustion, depersonalization and reduced personal fulfillment. The objectives of this study were to evaluate a possible association between BS and weekly workload, and to describe the prevalence of BS and the sociodemographic and occupational profile of on-call physicians in Maceió. DESIGN AND SETTING: Cross-sectional study in intensive care units (ICU) at public and private hospitals in Maceió. METHODS: A self-administered form was used to evaluate sociodemographic characteristics and BS through the Maslach Burnout Inventory (MBI) among 67 on-call physicians at ICUs in Maceió. Pearson’s R correlation test was used to compare workload and emotional exhaustion. For other dimensions, Spearman’s S test was used (P < 0.05). Other variables were represented by simple frequencies. The 95% confidence interval was calculated for each variable. RESULTS: Among the physicians studied, 55.22% were female and the mean age was 43.9 ± 8.95 years. The mean weekly workload on call was 43.85 ± 24.49 hours. The frequency of high scores in at least one of the three dimensions of MBI was 70.14%. CONCLUSIONS: Despite the high prevalence of BS, especially among physicians who did not practice regular physical activity, our data did not indicate any significant correlation between weekly workload and any of the three dimensions of BS in this sample. The high prevalence of BS draws attention to the importance of investigating other possible causes, in order to prevent and adequately treat it.


INTRODUCTION
Burnout syndrome (BS) is a well-defined condition characterized by emotional exhaustion, depersonalization and reduced personal fulfillment. 1 Emotional exhaustion consists of depletion of an individual's emotional resources. 2 It is considered to be the first feature of the syndrome and mainly occurs because of overload and personal conflicts in interpersonal relationships. 3 Depersonalization is characterized by emotional numbness among professionals, who start to treat clients and colleagues as objects. 4 This is a fundamental aspect of BS, given that the other features of BS can be found in depressive disorders in general. 2 Finally, reduced personal fulfillment (or feelings of incompetence) reveals negative self-evaluation associated with unhappiness and dissatisfaction with work. 4 Regarding the working environment of physicians in particular, some stressors that would raise the possibility of occurrence of BS can be highlighted: excessive demands that diminish the quality of care; long working hours; large numbers of shifts; low pay; requirement to deal with suffering and death; and constant exposure to risk, among others. In addition, the huge demands from society need to be taken into account. Society expects physicians to be infallible professionals, thereby creating pressure on them that is sometimes unsustainable. 3 The reported prevalence of BS has varied widely among studies, depending on the population assessed and the conceptual values used as the reference point. The prevalence of BS in a study among intensive care physicians in Salvador, Bahia, was 63.3%, when taking into account high scores in at least one dimension.
Thus, it was shown that there was high prevalence of the syndrome among the attending physicians studied. 4

OBJECTIVE
The objective of the present study was to evaluate a possible association between BS and the weekly workload of on-call physicians in six hospitals in Maceió, and to describe the sociodemographic and occupational profile and the prevalence of BS among those physicians. The hypothesis tested was that the correlation between BS and weekly workload among the on-call physicians in Maceió would be 0.3.

METHODS
This was a cross-sectional observational descriptive study. The data was gathered from public and private intensive care units in Maceió, Alagoas, between January and February 2011.
The research subjects included were medical graduates who were working on call regularly and were registered with the Regional Medical Council of Alagoas. Subjects were excluded from the study if they did not provide complete responses to the Maslach Burnout Inventory (MBI), on the data extraction form, and if they did not fill out the question referring to their weekly workload. The study did not make any exclusions or distinctions in relation to age groups, gender, race or social group, based on the sample size calculation. Sixty-seven on-call physicians who met the inclusion criteria and who filled out the data extraction form completely were evaluated.
The participants selected were on-call physicians who were working within this type of regimen during the study period. We obtained data about their lifestyles, work, health and stress as well as any presence of BS among these doctors (as diagnosed by the MBI). On-call physicians were chosen because of the high stress burden that is associated with shiftwork, as well as with medical work in general, due to constant involvement in suffering and death, excessive workloads, low pay and large numbers of shifts, among other factors. 3 This study was approved by the Research Ethics Committee of the Federal University of Alagoas (Universidade Federal de Alagoas, UFAL) before data-gathering started.
A consent statement was prepared for carrying out this project, since it was an investigation that dealt directly with people who voluntarily provided personal information, in response to a research protocol. The informed consent statement was presented to the subjects by the researchers prior to handing over the data extraction form, so that these individuals could be aware of all the issues involved in the investigation and could be assured of the security and confidentiality of their information. The participants were asked to fill out the form after firstly signing a consent statement. This document assured participants that the data gathered would remain confidential, assured them of the right to refuse to participate, explained that no risks were involved and Visits were made to two public and four private hospitals in Maceió, where physicians working in shifts were met, in order to invite them to participate. The information used was what the physicians themselves supplied on the data extraction form. Since the form was designed to be self-administered, the researchers did not interfere with the way in which the subjects filled it out. The forms were collected immediately after completion, and thus, there were only three forms that were not answered completely. The researchers continued to seek participants until the necessary sample size (calculated as described below) had been attained.
The data-gathering form consisted of four sets of questions: first set: general identification of the participants, with optional identification by name, with the aim of characterizing these individuals according to sex, marital status, age, expertise, length of employment, total weekly workload, work shifts, lifestyle, etc.; second set: evaluation of the level of burnout through the MBI; third set: questions about health problems, diseases and symptoms, in order to assess the overall health situation of the study population; fourth set: questions relating to stress in the workplace. The evaluation of symptoms and variables concerning personal, social, work and organizational characteristics correlated with the main symptoms presented in the BS process and its triggering variables. 6 The MBI is the most commonly used questionnaire for assessing BS in studies on this syndrome worldwide. It was first published by Maslach  The following additional data were also analyzed: gender, age, marital status and presence of children.
It was calculated that 67 participants would be required, taking a correlation between MBI scores and the weekly workload of 0.3, a significance level of 5% and a statistical power of 80%. A virtual calculator was used, which was obtained through the following site: http://www.stattools.net/StatToolsIndex. php#Sample%20size.
The correlation between the emotional exhaustion score and the weekly workload was analyzed using Pearson's R correlation test. Spearman's S correlation test was used between the weekly workload and other dimensions. Both of these tests took a significance level of 5% and statistical power of 80%, and were twotailed tests. The weekly workload was represented by mean ± standard deviation. The other variables were represented by calculating the simple frequency. The 95% confidence interval (CI) was used for each point that was estimated.

RESULTS
Sixty-seven intensive care physicians were evaluated and included in the study, as envisaged in the sample size calculation.

DISCUSSION
In this study, we sought to investigate whether any correlations existed between weekly workload and the three dimensions of BS, among intensive care physicians working in Maceió. However, no significant correlation was found in the studied sample.
While there are other instruments for measuring this syndrome, MBI is the one that is most used worldwide to independently describe each of the dimensions that characterize BS. 7,10 There is still no consensus in the literature regarding the diagnosis of the syndrome. According to Ramirez et al., only individuals who exhibit the three characteristic dimensions of professional burnout can be considered to present the syndrome. 11 However, we took the diagnosis of the syndrome to be in accordance with Grunfeld et al., who classified burnout as the presence of at least one high level, regardless of whether this consisted of emotional exhaustion, depersonalization or low levels of personal fulfillment. 8 Thus, although the Maslach questionnaire is specific for diagnosing the syndrome, the lack of standardization in evaluating the results makes them difficult to interpret, which may explain our finding that large numbers of people presented BS. New studies are being performed to find instruments that better assess the Brazilian reality. 10 In our research, we combined application of MBI with other factors of relevance for characterizing individuals and their working conditions. All the participants answered the MBI completely, but some items in the other sets of questions on the form were left unanswered or were not answered as expected, thus causing difficulty in statistical analysis and standardization. However, this did not hinder the objective of the study.
In cross-sectional studies, groups are formed only at the data analysis stage, as it is known at this point which individuals are exposed or not exposed to the variables, or which of them are healthy or sick. Cross-sectional studies examine the relationship between exposure and disease in a given population or sample, at a particular time, thus providing a picture of how the variables are correlated at that time. Therefore, this type of study does not establish causality and just points out the association between variables. 4 The doctors answered the questions during their shifts, which may have compromised the responses, because of possible difficulty in fully concentrating on the form.
The profile of the critical care physicians studied consisted of a relatively young population, with slight female predominance.
By analyzing burnout as a syndrome (rather than as independent variables), women health workers were seen to make up the most vulnerable group, and this can perhaps be explained by excessive workloads when summed their family commitments. 7 The physicians surveyed had on average graduated more than 10 years earlier and also had done more than 10 years of ICU work, and a slight majority did not have a specialist title in intensive care.
These findings contrast with what was observed in similar studies, such that the physicians' mean age, length of time since graduation and length of work in ICUs were slightly greater in the present study. 4,9,12 Having a specialist title would indicate a desire to spend more time working in the ICU, which would imply that physicians who possessed this title would have a lower frequency of BS. 4 Most of the interviewees were married and had children: the literature indicates that people who are married or are living with a stable partner and who have children experience feelings of family responsibility and greater resistance to the syndrome, through an associated ability to face emotional problems. 7 A study on oncologists found a connection between working exclusively in public institutions and lower levels of emotional exhaustion. 2 It was suggested that for these professionals, it was rewarding to work with students as part of the work in hospitals and public institutions, and that these institutions were places that offered less bureaucracy and greater creative freedom for The frequency of BS in this study, when taking into account a high score in at least one dimension, was 70.14% (47/67; 95% CI: 63.66 to 80.62). This level was higher than what was observed in other studies, using the same criteria. 2,4,9,12 Emotional exhaustion was the dimension that had the second highest frequency in our study. It is regarded as the first reaction to the stress generated by the demands of work. Once exhausted, people feel physical and emotional fatigue, with difficulty in relaxing and performing their activities. The internal resources that professionals possess for facing up to situations experienced at work, as well as their energy available for performing these activities become reduced.
The characteristics of this dimension, compared with the others, allow it to be easily accepted by professionals in expressing consistent aspects of burnout. 4 Given the psychological and physical symptoms, professionals develop depersonalization, which is characterized by cold and negative attitudes, thereby generating derogatory treatment towards the people directly involved with the work. 12 They even start to become cynical and ironic towards the recipients of their work. 9 However, depersonalization was the least prevalent dimension in the subjects studied. This can possibly be explained by the difficulty that physicians feel in accepting some of the claims associated with this aspect of the syndrome, since these involve attitudes that are generally not well accepted by society.
If these professionals feel that they are inefficient, with diminished self-confidence and a sense of failure, there is a reduction in personal fulfillment at work. 4 It is important to emphasize that this dimension is considered by some authors to be the last reaction to the stress generated by work demands. 4,8,9,12 Nonetheless, this was the dimension with the greatest frequency of high levels among these research subjects, and it did not appear to be related to salaries, weekly workload or number of years of doing the work. It was beyond the objectives of the present study to investigate possible correlations with this finding, although it was noted that there were correlations between the syndrome scores and the weekly workload, as well as between these scores and the length of time working in the ICU. This would suggest that the greater the workload was, the lower the scores for personal fulfillment would be, and consequently, the higher the frequency of high levels of this dimension would be, which also happened with shorter lengths of ICU work in years. However, the significance of these correlations could not be proven in this study.
The greater the subjects' ages and therefore the longer that they had been working, especially doing ICU work (which would indicate greater experience and appreciation of the field), the lower the frequency of burnout was, such that younger workers were more susceptible. 7 This can be explained by the fact that at the beginning of the profession, young physicians are at a time of growing awareness of realities, in which the transition from idealistic expectations to the reality of everyday practice shows that the reality does not always turn out as promised or expected. 7 Considering that BS affects about one in every two physicians, of whom one third are significantly affected and one tenth present a severe form, with irreversible characteristics, evaluation of the prevalence of this syndrome and its possible causes is of great importance. 2 13 Physicians are directly connected to this reality by being exposed to extensive working hours, regardless of the type of service, and bear a heavy burden placed on them by society, through their role of extremely high responsibility. 3 Therefore, practical efforts should be directed towards decreasing the stressors in the workplace that contribute towards developing the syndrome, which is perceived as harmful to health and leads to emotional exhaustion.
BS remains unknown among many health professionals, and greater clarification is required in relation to the syndrome, its manifestations and causes, and consequently in relation to effective ways for treating and preventing it, and for intervening in it.
Clarification is also needed in order to understand the impact that the work has on the professionals' lives and consequently on the quality of services provided for the population. 3 This was a pioneering study in the sense of providing a profile of doctors working in ICUs in Maceió and assessing the frequency of BS in this population. However, the limitations mentioned above regarding the study method and its objectives need to be taken into consideration. This study raised the hypothesis of a correlation between high weekly workload and BS, which could not be proven in this population. However, from the study results, we can raise hypotheses about other sociodemographic and work-related factors that may be associated with BS among ICU physicians. For better confirmation of this result, statistical analysis of greater complexity would be required, with better organization of the data extraction form to obtain more objective variables and responses. A more detailed statistical analysis on other variables that could be associated with the syndrome was not part of the main objective of this study, but this needs to be addressed in future studies, with the purpose of clarifying the causal factors of BS.
The high prevalence of BS draws attention to the importance of assessing other possible causal factors, in order to adequately prevent and treat it, since it may affect the quality of the services offered by physicians and the success of medical interventions. Thus, it is recommended that future studies should attempt to ascertain correlations between BS and other relevant variables. Since reduced personal fulfillment was the most frequent dimension observed in the sample, this particular dimension may need further investigation, given that it indicates that people are unhappy with their work. It may also be important to reevaluate the use of the MBI for diagnosis, since it may not be the most adequate tool for Brazilian realities and its scores (especially for personal fulfillment) may lead to excessive diagnosis.

CONCLUSION
Despite the high prevalence of BS, especially among physicians who did not practice regular physical activity, our data did not indicate any significant correlation between weekly workload and any of the three dimensions of BS among intensive care physicians working in Maceió.