Introduction

According to the World Health Organization (WHO), depression is a major cause of disability worldwide and a lead contributor to the global burden of disease (WHO 2020). Even though an individual could present with depressive symptoms, he/she may not necessarily be diagnosed with Major Depressive Disorder (MDD). MDD requires that five or more of the following symptoms be present during the same 2-week period and represent a change from previous functioning: depressed mood most of the day, nearly every day; markedly diminished interest or pleasure in all or almost all activities nearly every day; significant weight loss when not dieting or weight gain; insomnia or hypersomnia nearly every day; psychomotor agitation or retardation nearly every day; fatigue or loss of energy nearly every day; feelings of worthlessness or excessive or inappropriate guilt; diminished ability to think or concentrate or indecisiveness as well as recurrent thoughts of death and suicidal ideation (APA 2013). Compared to non-depressed workers, those with depression are more likely to display poorer work performance (Parent-Lamarche et al. 2020) more absenteeism, and higher unemployment rates (Lerner and Henke 2008). In fact, depression has been linked to a loss of productivity amounting to $36 billion annually in the United States (Kessler et al. 2009b). As such, work stress and the subsequent mental health problems that might result are now considered work and societal concerns (Jonge and Dormann 2017).

As for the etiology of depression, many biological, psychological, and environmental factors could contribute to the onset or maintenance of the disorder. There is now accumulating evidence that work organization conditions could also contribute to worker depressive symptoms (Parent-Lamarche and Marchand 2019; Parent-Lamarche et al. 2020; Penix et al. 2019; Saijo et al. 2016). Studies examining mental health in the working population are numerous (Monahan and Swanson 2019; Saade and Marchand 2013a, b) and so are those pertaining to the helping professions (Booker et al. 2020; Favrod et al. 2018; Saijo et al. 2014). Depressive symptoms presented by individuals working in helping professions have become a major concern due to their higher depression prevalence rates compared to the general population (Kessler et al. 2005; Letvak et al. 2012). Several theoretical models have subsequently been put forth in an attempt to explain the association between work organization conditions and workers’ mental health. The most well-known models are the Demand-Control model (Karesek and Theorell 1990), the Effort–Reward Balance model (Siegrist 1996), the Multilevel model (Marchand et al. 2015) and the Job-Demands-Resources model (Demerouti et al. 2001).

The Demand-Control model (Karesek and Theorell 1990) posits that workers who are subjected to a high job strain (high work demands coupled with low job control) are at an increased risk of health problems. Demands at work are usually experienced in the form of psychological and physical stressors (e.g., high workload, exposure to dangerous substances at work), whereas job control refers to decisional latitude (a high level of skill utilization and a weak decisional authority). According to this model, social support plays an important moderating role between work demands and workers’ mental health (Karesek and Theorell 1990). More specifically, a worker enjoying a high social support at work who is faced with high job demands and low job control is less likely to feel stressed compared to someone without sufficient social support (Karesek and Theorell 1990).

The second well-known theoretical framework is the Effort–Reward model (Siegrist 1996). According to this model, the gap between work demands and the rewards gained from the work could generate a negative emotional state in workers. This model is based on the following hypothesis: effort is exerted at work in exchange for rewards. Those rewards could take several forms such as monetary compensation, work security, increased self-esteem and better career opportunities. In the long term, the lack of balance between efforts exerted at work and rewards reaped, in exchange for this effort, could increase one’s susceptibility to stress-induced disease (Siegrist 1996).

The third work-stress model worth mentioning is that of Marchand et al. (2015). This multilevel model adopts a more comprehensive view of work organization conditions and stress. Compared to the previous two, this model views workers’ mental health in the context of larger socio-environmental structures. In addition to accounting for work organization conditions, this model includes social, economic, political and cultural systems an individual has to interact with on a daily basis. Agent personality (e.g., workers’ personality traits) is also accounted for as they might impact the way an individual interprets those stressors. According to Marchand et al. (2015), the way an individual interacts with various daily structures could be a source of frustration, negatively impacting their mental health.

Lastly, the final theoretical model is the Job-Demands-Resources model (Demerouti et al. 2001). The Job-Demands-Resources model adopts a dual perspective of work organization conditions and mental health problems. In a first step, demands at work could overtax one’s resources. In a second step, a lack of resources could impede one’s ability to meet those demands. According to this model, the interaction between excessive job demands and a lack of resources is at the heart of mental health problems. Even though all the models previously presented have been vastly tested and validated (Marchand et al. 2015; Ylipaavalniemi et al. 2005), we relied on the Demand-Resources model for several reasons. Foremost among those reasons is our interest in work-related variables and their possible contribution to workers’ mental health. Even though each profession has its own characteristics that could impact worker’s mental health, one is generally able to divide those characteristics into two large categories: demands and resources (Demerouti et al. 2001). Because of its comprehensiveness, the Job-Demands-Resources model could lend itself to a number of professions without necessarily being limited to demands and resources specific to a particular job (Demerouti et al. 2001). Another reason for our choice is attributed to its emphasis on the interaction between job demands and resources (Demerouti et al. 2001). Even though demands at work are not necessarily negative, they could be perceived negatively by a worker exerting a great deal of effort in an attempt to meet those job demands (Meijman and Mulder 1998).

Examining the relationship between work variables and mental health is particularly important in helping professions. Beyond the business repercussions, a depressed individual could incur on his organization and the ones placed on him as an individual, depression is likely to affect the people he is working with. Individuals working in the helping professions play a key role in caring for other people. As such, those professionals are often at the forefront of contact with patients (Maharaj et al. 2019). Nurses suffering from depression will inevitably see the quality of care they provide erode (Gao et al. 2012; Letvak et al. 2012; Welsh 2009). The same goes for doctors, with a number of studies linking depression to poorer quality of care and increased medical errors (Brunsberg et al. 2019; Halbesleben and Rathert 2008; West et al. 2009), potentially affecting a population’s health. In a study conducted by Siebert (2004), 56% of depressed social workers reported at least one incident due to their impairment, 44% reported more than three and 61% continued working even though they were too distressed to be effective. Depressive symptoms experienced by social workers could affect their work performance which may in turn impact client safety and professional credibility (Siebert 2004).

With the rapid developments in medical sciences (Wang et al. 2014) and fast-changing technology (Lim et al. 2010; Moustaka and Constantinidis 2010; Salmond and Echevarria 2017), a number of jobs in the helping professions have become increasingly stressful (Maharaj et al. 2018; Song et al. 2017). For instance, in the nursing profession, an effort to reduce personal cost coupled with a reduction in the nursing workforce has translated into mounting work pressure (Wang et al. 2014). As such, nursing has come to be considered as a stressful and challenging occupation (Lim et al. 2010; Maharaj et al. 2018; Moustaka and Constantinidis 2010). The same can be said for doctors (Clough et al. 2017; Tomioka et al. 2011), social workers (Cho and Song 2017; Siebert 2004), psychologists (Simpson et al. 2019; Smith and Moss 2009) and many others. In such professions, depression is now being recognized as an important occupational health problem due to its high prevalence rates (Letvak et al. 2012; Tomioka et al. 2011).

Even though several studies highlighted workers’ mental health concerns in the helping professions (Asaoka et al. 2013; Gu et al. 2017; Hall et al. 2018), few included a vast array of professions. Most studies have instead focused on one profession in particular (e.g., doctors and nurses). Comparing depression prevalence rates in a wild variety of helping professions could help identify the ones associated with the highest rates. The results of this study will hopefully help prevent depressive symptoms in at-risk professions. To the best of our knowledge, no study to date has conducted a systematic review pertaining to depression rates in helping professions while attempting to link work organization conditions to this disorder. With the numerous advancements in the work environment, evaluating the role work organizations play in workers’ depression is important. This is even the more important given the pivotal role workers in the helping professions play. This statement has never been more true than in 2020 and 2021 when COVID-19 paralyzed the workforce and all eyes turned toward health care professionals. By identifying possible contributors to workers’ depression, the hope is to inform prevention and intervention strategies in the workplace.

The aim of the present study is twofold. Our first aim is to provide an overview of depression prevalence rate in a vast array of helping professions. Our second aim is to identify work organization conditions that seem to be associated with this depression risk. Workers examined are those working in the helping professions given the high prevalence rate of depression reported in this population (Kessler et al. 2009a; Letvak et al. 2012). In the current review, helping professions examined were: doctors, nurses, social workers, psychologists, psychiatrists, midwives, occupational therapists, speech pathologists, laboratory and X-ray technicians, community health workers, physical therapists, and eldercare workers.

Methods

Search procedure

A systematic review was conducted on October 7th, 2019 to pinpoint the search terms and the databases to review.

The search terms were meant to identify work variables (“work* environment” OR workplace OR job OR occupation*) experienced by helping professionals (doctor* OR physician* OR clinician* OR nurse* OR “social worker*” OR psychologist* OR psychiatrist* OR counsellor* OR counselor* OR therapist* OR psychotherapist* OR “health care worker*” OR “health care personnel” OR “health care staff” OR “health care professional*” OR “medical personnel” OR “medical worker*” OR “medical staff” OR “helping profession*” OR “helping relationship*” OR “aid relationship*”) presenting with depressive disorder or depressive symptoms (depress* OR “mood disorder*”). We limited our search to scientific articles published between 2004 and 2019.

We systematically searched the following databases: (a) PubMed, (b) PsycInfo, (c) CINAHL and (d) Web of Science. The CINAHL search yielded 9289 records, the PsycInfo search yielded 18,186 records, PubMed 25,986 and Web of Science 34,165. In total, 87,626 records were retrieved and downloaded onto Endnote than onto an Excel file. After duplicates were eliminated, we were left with 53,525 records to screen.

Inclusion criteria

To be included in the review, studies had to meet the following criteria:

  1. a)

    Peer reviewed.

  2. b)

    Published between 2004 and 2019.

  3. c)

    Written in English.

  4. d)

    Examined depression as a dependant variable.

  5. e)

    Examined depressive symptoms in individuals working in the helping professions. In the current review, such professions included doctors, nurses, social workers, psychologists, psychiatrists, midwives, occupational therapists, speech pathologists, laboratory and X-ray technicians, community health workers, physical therapists, and eldercare workers. Examined the contribution of work organization conditions to workers’ mental health.

Exclusion criteria

Studies were excluded from this review if they:

  1. a)

    Pertained to a student population including medical residents. We decided to exclude medical residents from the present review due to our decision to only include individuals having completed their studies and occupying a professional position with minimal or no supervision necessary to conduct their jobs.

  2. b)

    Did not pertain to individuals not working in helping professions in the healthcare field. Those professions included but were not limited to lawyers, financial analysts, real estate agents, management consultants, accountants, spiritual leaders, teachers, clergy, professors, firefighters, etc.

  3. c)

    Relied on a qualitative design.

  4. d)

    Were descriptive and/or did not use a predictive model. More specifically, we excluded studies having only conducted correlational analysis between work organization conditions and depression risks or having solely reported descriptive statistics. Relatedly, it is important to mention here that studies that reported on both correlational and regression analysis were included. Inversely, studies that only reported descriptive findings such as descriptive statistics or correlational ones were not.

  5. e)

    Pertained to the effects of an intervention or a treatment plan on worker’s mental health.

  6. f)

    Were not empirical (e.g., systematic review, meta-analysis, literature reviews, letters to editors, reports, commentaries, and points of view).

  7. g)

    Reported on depression following a situational and unique event (e.g., depression symptoms presented by doctors caring for 9/11 victims). We decided to exclude those studies since their findings were specific to a particular event and were not necessarily generalizable to other health care workers.

  8. h)

    Did not evaluate any work organization variable. More specifically, studies that reported on depression symptoms without examining work-related variables were excluded. Similarly, studies reporting on perceived stress or value incongruence between the worker’s values and an aspect of the work organization were not considered specific to the work organization but rather a subjective evaluation made by the worker of that work-related variable and that is unique to him. Examples of such variables include: perceived work satisfaction, perceived value incongruence, perceived stress. It is, however, important to mention that if a study reported on both a personal subjective variable such as perceived work satisfaction in addition to some work-related variables (e.g., physical demands at work), the study was retained.

  9. i)

    Reported on perceived stress as the only dependant variable without reporting on depression. The same goes for anxiety. Studies with depression scores being lumped with those of other disorders (i.e., no clear depression score could be computed) were also excluded.

Data extraction

After all records had been retrieved, screening was conducted in a series of steps (Fig. 1). The first step consisted of screening the titles and the abstracts of records retrieved. If deemed eligible, the record was retained for a second step. In this second step, we retrieved the full-text article and read it. Given the large number of retrieved records, each study was independently screened by one rater. In case of doubt about the article’s eligibility, the first and/or fourth author made the final decision. The large heterogeneity in sample sizes, methodology adopted, depression definition, measurement tool, whether the researchers evaluated depression symptoms, antidepressant use, or a clinical depression diagnosis precluded us from conducting a meta-analysis. Figure 1 illustrates the results of the search and of the screening and selection process for the inclusion of studies in our review.

Fig. 1
figure 1

PRISMA flow diagram (Moher et al. 2009)

Full size image

The following information was extracted from each study:

  1. a)

    Country where the data were collected.

  2. b)

    Profession.

  3. c)

    Depression measurement tool.

  4. d)

    Depression prevalence rate (if reported).

  5. e)

    Work setting.

  6. f)

    Sociodemographic characteristics.

  7. g)

    Number of participants.

  8. h)

    Work-related variables.

  9. i)

    Risk of bias.

Reporting the depression prevalence rate was not always easy. We decided to report this prevalence rate when it was clearly stated in the study or when computing it was straightforward. Studies that only reported a mean score on a particular depression scale with no mention of a prevalence rate were not included in this analysis. We considered studies on depression prevalence rates such as those reporting on a 2-week prevalence rate, a 12-month prevalence rate and lifetime prevalence rate. A research assistant evaluated all the studies’ risk of bias based on Checklist for assessing the quality of quantitative studies (Kmet et al. 2004). This evaluation tool was selected given its comprehensiveness and its inclusion of all criteria considered relevant in the evaluation of quantitative studies.

According to this quality assessment criteria, a study was considered to have low bias if:

  1. a)

    The question or the objective was sufficiently described.

  2. b)

    The design was evident and appropriate to answer the study question.

  3. c)

    The method of subject selection (and comparison group selection, if applicable) or source of information/input variables (e.g., for decision analysis) was described and appropriate.

  4. d)

    The subject (and comparison group, if applicable) characteristics or input variables/information (e.g., for decision analyses) were sufficiently described.

  5. e)

    The outcome and (if applicable) exposure measure(s) were well defined and robust to measurement/misclassification bias. The means of assessment were reported.

  6. f)

    The sample size was appropriate.

  7. g)

    The analysis was described and appropriate.

  8. h)

    Some estimate of variance (e.g., confidence intervals and standard errors) was reported for the main results/outcomes (i.e., those directly addressing the study question/objective upon which the conclusions are based).

  9. i)

    The authors controlled for confounding variables.

  10. j)

    The results were reported in sufficient detail.

  11. k)

    The results supported the conclusions.

Given that no intervention/treatment study were included in our review, we did not score the following three items:

  1. a)

    If random allocation to treatment group was possible, was it described?

  2. b)

    If interventional and blinding of investigators to intervention was possible, was it reported?

  3. c)

    If interventional and blinding of subjects to intervention was possible, was it reported?

Each item was scored as follows: a score of 2 was allocated to a “Yes”, a score of 1 was allocated to a “Partial yes” and a score of 0 was allocated to a “No”. In total, each study was rated using this 14-item questionnaire with the exception of the three items pertaining to intervention/treatment studies. Each study’s risk of bias was, therefore, evaluated based on 11 items. A low score indicated a high risk of bias while a high score indicated a low risk of bias. The minimum bias score was 0 and the maximum was 22.

Results

In total, this systematic review included 17,437 workers in various helping professions. The number of participants ranged between 23 and 17,437 per study (M = 1291.56, Mdn = 435). Those individuals worked in more than 29 different countries in a vast array of industries and professional categories. Among those professional categories were doctors, nurses, social workers, psychologists, psychiatrists, midwives, occupational therapists, speech pathologists, laboratory and X-ray technicians, community health workers, physical therapist, and eldercare workers. In addition, the doctors sampled had different specialities including critical care doctors, surgeons, obstetricians, trauma doctors, pediatricians, palliative doctors, gerontologists, dentists, psychiatrists, and veterinarians. To facilitate reading, those different specialities were referred to as doctors in Table 1. As for their work setting, those health care professionals were working in hospice centers, hospitals, nursing homes, long-term care facilities, home health care, private clinics, health centers, medical centers, public hospitals and community clinics and some were even deployed with the military. As mentioned earlier, the data covered in this review originated from 29 countries with Japan, the US and China being the most frequently cited (14.01% of the studies collected data in Japan, 13.08% in the US and 11.21% in China).

Table 1 Synthesis of Studies Pertaining to Depression in Health-Care Professions
Full size table

Unsurprisingly, 73.83% and 30.84% of studies included nurses and doctors in their sample. More specifically, most studies pertained to a sample of healthcare workers including doctors and nurses as well as other health care workers (e.g., occupational therapist and X-ray technician). Studies that included both doctors and nurses were computed twice (once in the doctors’ category and once in the nurses’ category). 62.62% of studies reported a depression prevalence rate (Table 2). Among those studies, depression prevalence rate (when reported) varied between 2.5 and 91.30% (M = 27.23%, Mdn = 23.28%). Given the large variations in sample size across studies, we decided to report the depression median and the mean to avoid having studies with a large number of participants skewing the results. Those depression prevalence rates need to be interpreted with caution due to the large heterogeneity in the definition of depression, measurement tool, whether the researchers evaluated depression symptoms, antidepressant use, or a clinical depression diagnosis. In addition, worth considering is the large heterogeneity displayed in sample sizes and the methodology adopted. As for the depression tools used, a large variability was also noted. The three most frequently used depression tools were the Center for Epidemiologic Studies Depression Scale (CES-D; 32.71%), the Hospital Anxiety and Depression Scale (HADS; 13.08%) and the Beck Depression Inventory (BDI; 11.21%).

Table 2 Work-Related Variables Associated with Depression in Health-Care Professions and Studies' Risk of Bias
Full size table

As for work factors’ contribution to depression, variables related to various demands at work and resources were highlighted (Table 2). In terms of demands, physical demands placed on a worker (e.g., inadequate physical resources for work (Asaoka et al. 2013)), short sleep duration due to one’s work schedule (Chaiard et al. 2019), psychological demands (Chen et al. 2016; Zhang et al. 2017), working night shifts (De Vargas and Dias 2011), being on-call at night (Balch et al. 2010), bullying (Rodwell and Demir 2012; Yildirim 2009), workplace violence (Da Silva et al. 2015; Eriksen et al. 2006), work–family conflict (Dyrbye 2014) and work–life imbalance (Compton and Frank 2011), perception of lack of justice in the workplace (e.g., low procedural justice, low relational justice (Ylipaavalniemi et al. 2005)), negative work-to-family spillover (Franche et al. 2006), rapidly rotating shifts (Hall et al. 2018), weekly paid overtime (Hall et al. 2018), low decisional authority (Franche et al. 2006; Zhang et al. 2017), low decisional latitude (Franche et al. 2006), and lack/insufficiency of autonomy (Enns et al. 2015) were all found to be positively associated with depression risk. More details are provided in Table 2.

Inversely, a number of resources have also been identified (Table 2). Examples of such variables include: job satisfaction (Tarrant and Sabo 2010), satisfaction with one’s choice of medical specialty, satisfaction with support provided in training (Berman et al. 2007), social support (Berthelsen et al. 2015; Chana et al. 2015; Saksvik-Lehouillier et al. 2016), supervisory support (Chen et al. 2016), fair leadership (Berthelsen et al. 2015), and peer support (Hsieh et al. 2016).

Lastly, risk factors frequently reported by individuals working in the helping profession and that might be more commonly noted than in other professions are: high workload (Yoshizawa et al. 2016), high emotional strain (Kubik et al. 2018; Muntaner et al. 2004) little sleep (Flo et al. 2012; Tsutsumi et al. 2011), high number of working hours (Tsutsumi et al. 2011), bullying or conflict at work (Rodwell and Martin 2013) in addition to the aforementioned work factors. One might think that the combination of those variables might interact with resources an individual has to explain their  risk of suffering from depression.

Discussion

The systematic review conducted highlights the alarmingly high depression prevalence rate in the helping professions, varying between 2.5 and 91.30%. Those findings raise concerns for the professionals themselves as well as for their patients. Depression in the helping profession is very likely to erode the quality of care offered to patients and potentially put them at risk (Brunsberg et al. 2019). The results presented are consistent with existing literature on the impact of work organization conditions on depression risk. This relationship can be explained by the Job-Demands-Resources model. According to this model, demanding jobs could exert pressure on the worker, potentially depleting them of physical and mental resources, resulting in health problems (e.g., depression). Demands at work frequently associated with depression risk in helping professionals include: physical demands, psychological demands, job insecurity, irregular work schedule, lack of sleep, lack of decision authority, and latitude. Inversely, resources frequently identified as being negatively associated with depression risk in helping professions include support (supervisor, family, and coworkers), perception of work justice, and fair leadership.

The results of our review extend previous findings by aggregating the results of previous studies. Whereas most previous studies focused on a limited number of helping professions (e.g., doctors and nurses), this systematic review included doctors, nurses as well as other professions (social workers, psychologists, psychiatrists, midwives, occupational therapists, speech pathologists, laboratory and X-ray technicians, community health workers, physical therapist, and eldercare workers and many others) working in more than 29 countries. Relatedly, the scope of this review is also worth noting. The screening of 4 different databases allowed us to retrieve 87,626 records. The number of records screened and retained is quite large. Whereas most systematic reviews had mainly reported on the helping professionals’ burnout rates (Adam et al. 2018; Rotenstein et al. 2018), or on depression’s prevalence rates (Mata et al. 2015), this study attempted to explore depression prevalence rate as well as work organization variables. To the best of our knowledge, this is the first systematic review to be this comprehensive and of this magnitude. It is also the first study to comprehensively assess prevalence rate in a large array of helping professions while simultaneously examining what work organization conditions could be linked to it.

Those results point to the important role employers could play in preventing and intervening on such mental health problems. Prevention and intervention efforts are more likely to be effective if they were to be implemented at both the organizational and individual level. In the past, most preventative interventions focused on individual characteristics. The evidence for the effectiveness of organizational intervention, although weak, is now mounting (Myette 2008). Based on this systematic review, organizations and workplaces should direct their efforts on work-related risk factors that are associated with depression (e.g., reducing the workload, reducing the emotional strain, reducing the impact of work schedule on a worker’s sleep, decreasing the number of work hours, and trying to avoid bullying and conflict at work). Standardized risk and protective factors assessment at the organizational level could be one way of going about it. An evaluation of the demands placed on a worker coupled with the resources available could also be pertinent. Given that depression is a multifactorial disorder with some of its risk factors being individually based, employers could screen for possible depression symptoms. In doing so, one might need to pay particular attention to avoid possible discrimination and stigmatization. Keeping those records confidential and possibly web-based could be an option (Myette 2008).

Limitations

This study presented with a number of limitations that are worth mentioning. The important heterogeneity in terms of depression definition, measurement tool, sample obtained, and methodology adopted complicated the comparison process. This lack of consensus in terms of number of symptoms necessary to warrant a depression disorder or a cutoff score used across studies to report on participants’ depression precludes us from deriving a consistent prevalence rate. The same problem is also noted with regards to work organization conditions that are labeled differently and hence measured differently across studies. We also decided to limit the scope of this review mainly to demands and resources that are work-related. In doing so, we excluded personality traits and predisposing variables that could explain some of the obtained results. Some articles could not be retrieved online. The results of such studies were, therefore, not reported. Our decision to exclude certain professions and students could have biased our results. In our review, we wanted to focus on individuals having graduated, with no or minimal supervision, and therefore, occupying professional roles as opposed to that of a resident or a trainee. Even though we decided to use the term “risk” to refer to depression risk, several studies were cross-sectional. Caution, therefore, needs to be taken when interpreting the results of such studies. Lastly, we only included studies written in English.

Conclusion

Future researchers are encouraged to expand the pool of studies reviewed to non-English ones as well as to medical residents and trainees who will eventually be on the job market. Identifying risk factors in residents and trainees could be worth the effort and may serve as a preventive measure before those individuals start working with vulnerable populations. Relatedly, conducting a meta-analysis could help shed further light on the impact of work organization variables on depressive symptoms. Lastly, examining the influence of non-work-related variables such as workers’ personality traits and predisposing variables on worker’s depression could also be pertinent.

Despite those limitations, our results should serve as an important reminder to pay attention to the mental health of workers in helping profession. Investing in employees’ mental health by preventing and reducing depression risk could prove to be a valuable investment from an employer’s point of view (Myette 2008) as it is likely to increase productivity and reduce absenteeism (Oliveira Santana and Barros 2019; Rost et al. 2004). The work organization variables highlighted in this review could be one way of going about it. Individuals working in the helping professions are encouraged to recognize early signs of depression in themselves as well as in their colleagues. Early detection will hopefully help prevent the development of depressive symptoms. Similarly, vulnerable employees are encouraged to carefully evaluate potential employment places based on variables identified in this review. Health care professionals already working in stressful environments should be made more aware of the association between certain job characteristics (e.g., high number of hours) on their risk of suffering from depression. Better awareness of those risk factors’ negative repercussions will hopefully serve to mitigate the risk of employee depression. Healthcare organizations are encouraged to put in place work organization conditions and human resource practices targeting the risk factors identified in this review. Supervisors and managers alike should make every effort to reduce those risk factors while amplifying the protective role resources could play in workers’ mental health (Siebert 2004). Relatedly, employees at risk should be encouraged to seek help (Siebert 2004). It is also worth reiterating that individuals working in helping professions have an ethical responsibility toward themselves and to their patients (Siebert 2004). Clients served by individuals in health care services are often in vulnerable situations. The negative repercussions from workers’ depression symptoms should be prevented before being endured by their patients.