Around 1 million deaths by suicide are registered every year―17% of them are related to Indians and 14% happen in developed countries    .
However, the rate of suicide is not homogenous in all countries   . Physicians usually commit suicide in higher frequency compared to the general population    . Thus, the decrease of suicide mortality among physicians should be the main purpose of this working class to change this situation  . Hence, these professionals have 2.5 times higher chances of committing suicide than the active population   . Since suicidal ideation is another risk factor for suicide, prevention actions aiming to avoid these thoughts and suicide itself have great importance not only for these physicians, but also for the entire workgroup     .
Furthermore, mental health studies including Medicine students point out high levels of depression, anxiety, and Burnout Syndrome. Nevertheless, these studies about suicide and mental health risk factors among active male physicians are rare    . Additionally, expressive levels of psychic suffering were found among physicians who committed suicide    .
Therefore, more pieces of information about suicide among health professionals are necessary in most of the known sources. Such information could be used to monitor the risk of suicide, to define the concomitant risk factors for self-mutilation in medical population, or to help analyze mental health factors that contribute to physicians’ suicide   .
Thus, we aim at exemplifying the research of such theme, “Suicide among Physicians,” and its correlation with the available scientific content regarding Health Sciences in the main electronic bibliographic databases of open access to the full text content. This is done through the Brazilian Department of Health, which is named Biblioteca Virtual em Saúde-BVS/BIREME, in Brazil, and the US National Library of Medicine, also known as MEDLINE/PubMed, in the USA.
Therefore, this investigation presents the methodological development similarities of BVS/BIREME and MEDLINE/PubMed databases, as well as their reported controlled vocabulary: Descritores em Ciências da Saúde (DeCS) and Medical Subject Headings (MeSH).
This is a systematic review with metanalysis following Cochrane Handbook  and Meta-Analyzes (PRISMA)  guidelines recommendations for conduction of both the systematic review and the metanalysis (Figure 1). Previously chosen inclusion/exclusion criteria were adopted to select the sample in the MEDLINE/ PubMed and BVS/BIREME databases, since they are the object of this study.
Figure 1. Flow diagram summarizes the process of study inclusion in this review. From: Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med, 6(7): e1000097. doi:10.1371/journal.pmed1000097. Source: Developed by the authors (2016).
the research question was structured based on the PICO acronym. Each word of the PICO component means: P: MEDLINE and BVS scientific databases; I: categorization of the selection of source-data and hierarchical system available in each analyzed database; C: collection of the available scientific information from January 1, 1996 to August 26, 2016; O: guarantee of reliable collected information to support the decisions taken by the investigators and Health Sciences professionals, which enabled the detailed description of the creation of the eligible criteria adopted herein.
The research question consists in: “Have the main open access bibliographic databases, MEDLINE/PubMed in the United States, and BVS/BIREME in Brazil, presented methodological similarities in the indexed scientific content during the last 20 years as a way to ensure reliability of the recovered information?” The theme “Suicide among Physicians” was used to illustrate such methodological similarities. The following types of studies were included: Classical Article, Clinical Study, Clinical Trial, Clinical Trial, Phase I, Clinical Trial, Phase II, Clinical Trial, Phase III, Clinical Trial, Phase IV, Controlled Clinical Trial, Corrected and Republished Article, Journal Article, Meta-Analysis, Multicenter Study, Observational Study, Overall, Randomized Controlled Trial, whose main theme was “Suicide among Physicians”, in English, Spanish and Portuguese languages. Such time period was chosen due to the creation date of the PubMed health sciences scientific research portal in January of 1996. Exclusion criteria were comprised of articles that did not directly approach the theme ‘Suicide among Physicians’ or situations that resulted in such action, as well as of studies that did not have the open access availability to full text. Studies were chosen through electronic search in MEDLINE/PubMed and BVS/BIREME databases, in addition to gray literature. The R 3.3.1 statistical program was used for the statistical analysis.
Throughout the process of study selection, two reviewers worked independently and analyzed the studies to be included. When occurred disagreement between them, a third reviewer was used to make the final assessment on the inclusion or non-inclusion of the study. The entire content of chosen studies was analyzed.
Based on research strategy, 282 publications were identified at MEDLINE and 31 at BVS, resulting in 313 publications; however, after undergoing the criteria and analyses of eligibility, 16 studies remained (Table 1). The repeated articles were computed only once in the final counting.
Based on the joined analysis of all articles submitted to the test, a combined Odds Ratio value of 1.85 CI (1.83 - 1.87) was found. It shows the association of MEDLINE database with BVS database and the Odds Ratio regarding the suicide theme among medical professionals. The obtained p-value (<0.05) was statistically significant for the analyzed datum. The joined analysis of the included studies was made (Figure 2). Interpreting the figure below, each group of studies is represented with a line.
Table 1. Suicide among physicians in the MEDLINE/PubMed and BVS/BIREME databases: main findings and limitations.
Figure 2. Statistical analysis of the studies. Source: Developed by the authors (2016).
4.1. Database Comparison
The square represents the effect of studies and the line indicates the confidence interval. The size of squares represents the weight of each group of studies for the statistical analysis. The vertical line shows the absence of effect, and Fixed-Effects (FE) symbolize the result of the analysis.
Thus, the Health Sciences area and its unfolding in Collective Health have been facing important scientific challenges seeking social health and welfare, including the interdependence of knowledge. Then, the relevance of free availability of scientific information (Open Access) in bibliographic databases supports the sustainability of sciences and health professionals’ development, because it helps the development, preservation, spreading, access, and use of scientific information without costs, including health information systems  .
Health professionals have high performance expectations. Job positions in these areas are necessarily stressful, which might lead to adverse effects on health  .
Indeed, the healthy worker’s effect refers to a consistent trend of the active employee in having a lower mortality rate than the general population due to selection effects   . However, this is not applied to physicians, regarding death by suicide. In comparison with other occupations    , other university graduates   and people in general    , suicide and suicidal thoughts are significantly more common among physicians, as well as burnout and other signs of psychic suffering. Mortality rate by almost all causes of death, except suicide, is however lower for physicians than for the general audience   , which is in agreement with the healthy worker’s effect  .
In addition, precipitating factors are comprised of reduction of the risk of a recent loss or crisis and increase of risk of a job problem that might reflect the experience of a physician in dealing with death and loss, but inability of dealing with problems related with a physician’s identity. Medical self-identification is centered many times on the professional role, which can pass through job aspects and their lives at home. For someone whose job helps defining and identifying his/her personal and professional profile, crisis in a job situation can make him/her feel more threatened than for someone whose personal identity depended less on job satisfaction  .
Although speculative, medical suicide rate is believed to be higher than the reported rate, since it can be poorly codified in death certificates, sometimes deliberately. For instance, a death can be notified as an accidental overdose of a distracting prescription drug or medicine, instead of a purposeful overdosing  .
Under these circumstances, the study of Jiao Y and collaborators compared the points of view of members from a community to psychiatrists’ in Shanghai, China, regarding the posture of these populations before suicide. They found that although physicians considered suicide as a preventable worldwide public health social issue, they had stigmatized views about the suicidal patient and comprehended such patient less with lower sympathy. Even after filtering sociodemographic characteristics, the analysis remained significant  .
In a certain article, only 14.9% of the responders were satisfied or very satisfied with their jobs, whereas 52.5% were dissatisfied or very dissatisfied with their jobs. Almost a tenth of the physicians had had suicidal thoughts, although none had tried committing suicide  .
Furthermore, physicians, other university graduates and human services professionals presented a lower rate and decrease of mortality compared with the general population  . Therefore, regarding specific causes of death, physicians had a higher mortality suicide rate than the general population  .
In another study, there were 17 defined and 15 possible suicides. Among the 17 defined cases, 16 were of the male gender. 70.6% of these individuals died at the age of 40 and other 11.8% were 51 to 64 years old when they died. Only 17.6% of the subjects were aged ≤40 years when they died. Among the cases of defined suicide  , the most frequent suicide methods were hanging/asphyxia (n = 5, 29.4%), vegetal charcoal burning (n = 5, 29.4%), and drug poisoning (n = 5, 23.5%).
Thus, results showed a consistent increase of suicide mortality with a standardized mortality ratio around 1.65, mainly due to the excessive number of suicides by poisoning  . Suicide verdicts had been provided by 87% of the remaining 223 physicians and open verdicts by 13% of them  . A cohort study identified that therapists considered 42% of the physicians at risk  .
Responders showing the highest burnout rates were younger physicians that had worked for about 8.5 years, period in which they were still taking or had just concluded training with specialists  . These young physicians, with moderate experience and who needed to work during shifts, seemed more vulnerable  .
Thereby, physicians who committed suicide had higher chances of being 40 years old, of serving the community and of living in urban areas. Additionally, a study found a predominance of medical suicide among physicians serving the community  .
Another study has showed that no specific cause was highlighted for female death, only suicide. In general, mortality rate indices were quite similar for men and women. This indicates that educational level and health behavior explain differences in mortality rather than gender, with exception of specific gender- related illnesses  .
Although the suicide rate of white women was only around half of the high rate compared to their male colleagues, such rate was approximately twice higher than the normal working population  . These female physicians had a Suicide Rates Ratio (SRR) of 52.10 compared with working professionals, which was still statistically high  .
Female physicians presented higher suicide rates than the standard population in almost all age groups, and the elevations were statistically significant in some cases. The high amount of suicide rate for females also increased with age (p = 50.015)  . The suicide rate for females was higher than that for the active female population from the USA  .
Besides, there was a strong trend of suicide rate increase for male physicians with age increase  . This is important because some studies indicate that white male physicians > 45 years have higher rates of suicide, whereas those aged 45 or less present decreased ones  .
The female physicians’ suicide ratio is another remarkable finding. Women comprised around 13.2% of the total sample and 24% of the suicide death group; thus, they are represented in the suicide group (p = 0.02). This result does not reflect the suicide rate of the Brazilian population, in which males are 2.3 to 4.0 times at higher risk  .
Nevertheless, with regard to the psychiatry service, there is a quite similar amount of male and female physicians in general, and no differences were found in the degree of morbidity  . There was no difference in gender distribution in the two verdict categories  . The youngest age range of both genders presented relatively high rates  . The largest group of physicians was comprised of specialists with medical registration (residents), who were concluding their under graduation course; therefore, they were looking for consultant positions (specialist staff), and there was a significant number of new consultants  .
The main exception to the low mortality rate among female physicians was almost twice higher than mortality rate due to injuries and poisoning. This was mainly attributed to an excessive amount of suicides, mainly of female anesthetists  . It has been found that mortality rates in male specialist physicians were also lower in men from the I social class  , considered “High-Level Professionals and business workers”.
Agreeing these ideas, one of the most stressful moments for residents is the transition from being resident to being completely responsible  . Most of the responders had concluded their graduation tests, academic performance was therefore not compromised  .
Four specialties presented significantly higher suicide rates than the general medicine, as follows by order of risk: community health, anesthesiology, psychiatry, and general practice. The pattern of suicide rates by gender and time period was similar in almost all specialties, but the community health that had a male rate of 12.3 times higher than women  . Psychiatrists showed a significantly higher global mortality rate compared to all the other specialized physicians, with significant high rates of poisoning, for instance  .
The total suicide mortality rate was significantly increased. An increase of suicide frequency by poisoning was a common characteristic seen in all subgroups, especially among male members of the Association of Specialist Physicians (AMS, acronym in Portuguese) and of the Organization of Practicing Physicians (OGP). Physicians presented an increase in the number of accidents and suicides by poisoning, as well as by other methods  .
4.4. Risk Factors
Some articles approached through multivariate, comparative or descriptive analyses      the possible risk factors that could develop high levels of emotional overload and, as consequence, genesis of suicidal ideations among physicians. A study carried out in Taiwan, published in 2009, points out that the highest rates of emotional disorders among physicians occur at the age of 40 years, a moment when physicians have gone through a relative extenuating working period and, therefore, believe they are less capable of taking new personal and professional directions  .
However, Siu, Yuen and Cheung, 2012, after analyzing 226 questionnaires applied to physicians enrolled in the Hong Kong Public Doctors’ Association, observed that higher levels of physical and emotional burnout were associated with early age, less working experience, need of job shift changes, less children, and lower physical workout  . In compliance with the outcomes of this study, Dyrbye and collaborators (2014) showed that depressing symptoms, as well as general burnout, depersonalization and intense fatigue were more commonly found among Medicine students and residents compared with the general prevalence rates of these symptoms in the North-American population of the same age, which decreased after development in the professional career stages  . This result is similar to what was seen by Eneroth M and collaborators, 2014, who also pointed out as causes of suicidal ideation: job concern, frequent meetings to discuss experiences and demands, sick presenteeism and job disengagement  .
Carpenter L, Swerdlow A and Fear N, 1997, showed that adverse job conditions with long working hours and task overload are also important in this condition, since they have a direct impact on the professional’s job quality  . The amount of hours spent by a medical professional for the genesis of this situation is around 56, in a way that burnout among physicians is possibly related to a wide interaction of factors, including physical exhaustion, too many shifts, unhealthy working positions and unhealthy sleep patterns  . Additionally, the combination of extenuating working conditions and possible selection of subjects with particular personality marks might contribute to high risks of suicide among physicians    . Siu, Yuen and Cheung, 2012, reported that physical exercises and colleagues’ social support were not considered factors of burnout relief  , and the culture the physicians undergone did not have so much importance as a risk factor for the genesis of suicidal thoughts  .
In general, inappropriate handling, increase of work-related issues and stress seem to be risk factors, which can be the key to decrease deaths by suicide among physicians when dealt correctly  . This can also be done through the positive influence on the job environment, with leaders able to teach and ask correctly for such physicians, since these influences were considered to have a positive impact on the resident’s professional and personal lives  (Figure 3
Figure 3. Protection factors against suicide among physicians, based on the analyzed studies. Source: Developed by the authors (2016).
and Figure 4).
These factors were compared through the Pearson Linear Correlation Graph, in order to better clarify the relationship between these elements and suicide among physicians. The correlation coefficient found was 0.89, which indicates a strong relation between these variables (Figure 5).
4.5. Impact on Public Health and on the Individual’s Professional Health
Several analyzed studies indicate that suicide among physicians, as well as actions that resulted in this condition, was harmful not only to the career of the professional facing this situation, but also to the provision of health to the general population   .
Gold K, Sen A and Schwenk T, 2012, approach that the physician’s self-iden- tity is almost always focused on his/her professional role, which has a result in his/her working and personal life. Thus, a crisis at work could be more threatening to a physician than to those whose personal identity is less dependent on job satisfaction  .
4.6. Measures to Be Implemented―Family―Seek Support
Several measures can be taken to avoid conditions that lead to the practice of suicide among physicians, such as: initiatives to decrease workload-related stress, provide job prestige and supply good practices of job safety are viable
Figure 4. Risk factors for suicide among physicians based on the reviewed studies. Source: Developed by the authors (2016).
Figure 5. Pearson linear correlation graph presenting the correlation analysis between the risk factors and suicide among physicians. Source: Developed by the authors (2017).
choices  , since job satisfaction becomes a need  . Hence, building connection networks and promoting mutual support among physicians in the community are essential to the medical professional’s welfare  . There must be strategies focused on improving inter-professional relations, as well as encouraging physicians to seek help in the event of depression, burnout or other emotional/mental issue  .
In addition, some studies indicate an association between being single and/or divorced with suicide, which shows the role of family support in detecting and treating predispositions to such practice. We also known that compared to the general population with the same cause of death, physicians who committed suicide have a higher incidence of mental illnesses associated with working issues  . However, other studies observed that only the factor “amount of children” is related to suicide, and the physician’s age or marital status had no importance. All the same, this relevance was noticed only among resident male physicians. We assumed that family support, whether from a companion or children or other person who provides emotional support, might be important in the prevention of such problem  .
Furthermore, another factor of concern is the poor seek of support by medical professionals. The depression-related stigma and lack of knowledge about it hinder millions of people of seeking proper medical support in the correct time period, which result in stress and burnout, as well as in high mortality and morbidity rates, since it not only affects the person committing suicide, but also his/her family.
Although the mental illness-related stigma is usual, it is more frequently seen in developing and underdeveloped countries  . A study found three work- related characteristics that are also related to the fact of not seeking help for depression: being currently involved in a medical research, being a surgeon and working at night shifts. On the other hand, no significant associations between seeking help and conflict of job functions, job and family conflicts, long working hours and, even, illness were found  . For a better understanding on this subject, knowledge on how to develop a healthy lifestyle and the elimination of risk factors are necessary, which might explain the mortality patterns among physicians by suicide  .
Another study showed that physicians use higher rates of anti-psychotic, benzodiazepines and barbiturates. A physician can get these medicines more easily. Additionally, suicide attempts by overdose in these people are considered less prevalent than by fire gun or hanging. Nonetheless, among physicians who know about the toxicological effects and doses for specific medications, an overdose can represent a greater risk  . This reinforces the idea of improving welfare in all career phases as a need  . Resident male physicians mostly need support, because they have more difficulties in discussing this kind of problem openly. Thus, individual talks might be more advantageous to them  .
Thus, despite many factors that contribute to suicide among physicians, a study found that poor working conditions of many of these professionals and unusual routine to which they undergo are the greatest reasons for mental illnesses and other disorders that might result in suicide  . A possible strategy to decrease suicide among physicians would be regular screenings in the search of mental illnesses among them, which is characterized as a health policy  . In addition, meetings to approach working situations could contribute to decrease of suicidal thoughts  . Hence, a Norwegian study showed decrease of suicidal thoughts among male physicians in Norway during the last 10 years, due to the improvement of their working conditions, provided through health system reforms that were implemented in the last decade in such country   . We understand that measures like these are important not only to medical professionals, but also to the general population, because the good condition of physicians health is also important to their patients  .
Therefore, this study intends to occupy an empty space in scientific knowledge, considering that the access to scientific and technical information is present as an essential characteristic for the development of health research and that the act of sharing the available resources in both analyzed databases presents common methodologies. Such tools of open access research are seen as reliable tools to collect the available scientific content through open access to the available indexed content regarding the methodological characteristics, as well as the scope of recovery of the available information in the bibliographic databases of open access to the full content: MEDLINE/PubMed and BVS/BIREME.
Scarce recovery of the indexed articles in the analyzed databases is considered a limiting factor of this study, because of the small amount of publications on the theme, especially in the public health context for the theoretical basis on the theme “Suicide among Physicians”, which may have hindered carrying out this research. Further studies are recommended in order to extend the research universe by associating the researched theme with the medicine students’ category.