Introduction

There is increasing concern about the amount and impact of mental health problems, such as depression, among doctors.1-4 The exact prevalence of depression within this population has been difficult to determine but a recent meta-analysis estimated that 29% of junior doctors had significant symptoms of depression.5 Rates of suicide within doctors are significantly higher than the general population, with evidence suggesting that female doctors are at twice the risk of dying by suicide compared with the general population, while male doctors have 1.4 times the risk.6 While some of this increased risk almost certainly relates to doctor's access to and knowledge about high lethality means of suicide, it is also likely that doctor's high rates of suicide are linked to the difficulty that many doctors have in asking for help. Doctors, particularly those in training, appear to be less likely than the general population to seek help for mental illness.7, 8

Several barriers and facilitators of help-seeking for mental illness in doctors have been investigated. Barriers include concerns about career implications,9-13 stigma,9-11, 13-15 ambivalence,9, 10, 13, 16 confidentiality concerns,9, 10, 13, 17 time constraints11, 13, 16, 18 and limited knowledge of services.9, 10, 13, 16 In contrast, proposed facilitators of early help-seeking include good organisational and practical support, confidential and flexible services,9, 13 openness between colleagues, emotional support from peers, good relationship with a general practitioner (GP),13 positive experience with mental health services and sharing sessions with others experiencing similar problems.19 While many of the identified barriers to doctors seeking help are potentially modifiable, at present there is little understanding about the relative importance of these factors. This information is crucial in order to guide the design of interventions to promote mental health help-seeking by medical professionals.

In addition, particular groups of doctors may be at increased risk of not seeking help for mental illness. Generally, female doctors make up a larger proportion of those accessing Employee Assistance Provider (EAP) support services.19-21 Other potential factors that may impact help-seeking include stage of training and medical specialty, although there is limited evidence available examining these. Preliminary evidence suggests that senior doctors may be more likely to seek help compared with junior doctors21, 22 and that GP may be more likely to seek help compared with other specialties.23 There is also some evidence that doctors working in regional, rural or remote communities may have different rates of help-seeking for mental illness compared to those working in metropolitan areas, due to differing availability of services and levels of stigma.18 There is some evidence that, within the male general population in Australia, seeking help from a professional for mental illness is less likely in non-metropolitan areas than in metropolitan areas.24 Finally, although ethnicity has been found to be associated with help-seeking in the general population,25-27 there have yet to be any published studies exploring this factor within a medical population.

In order to address these gaps and to provide information to guide local interventions, this study aimed to: (i) determine rates of help-seeking for depression among Australian doctors; (ii) compare the relative impact of various perceived barriers to seeking help for depression among Australian doctors; and (iii) identify if particular groups of Australian doctors are less likely to have sought professional help for depression.

Methods Sampling strategy

This study used data from the Australian National Mental Health Survey of Doctors and Medical Students. The methodology of this survey has been published previously.28 In brief, 42 942 doctors in Australia were randomly invited to participate in an online- or paper-based survey between February and April 2013. They were initially sent a letter to inform them that they would receive a questionnaire pack. Participation was voluntary and anonymous. Reminder letters were sent to all potential participants. A total of 12 252 doctors completed the survey, a response rate of approximately 27%. The demographic profile, such as age, gender, state/territory and indigenous status, of participants was similar to Australian doctors as described in the most recent national Census.29 However, in terms of work location, doctors from regional/rural/remote areas were over-represented.30

The population of interest for the present study were doctors who reported having ever felt seriously depressed, even if not diagnosed. Retired doctors were excluded from analyses in the present study. Respondents who failed to identify their stage of career were included (7.2%; n = 299), as were doctors who remained part of the workforce but currently not working (0.4%; n = 16).

Measures Demographics

Demographic variables of interest included gender, location of primary medical degree (Australia or overseas), location of primary place of work (metropolitan or non-metropolitan), stage of training (junior doctors – intern, pre-vocational trainee and vocational trainee or senior doctors – consultants) and specialty (GP, psychiatry, medicine, surgery, anaesthesiology, obstetrics and gynaecology/paediatrics, emergency department/intensive care unit/pain management, oncology, imaging/pathology, administrative/education or others).

Rates of help-seeking

Rates of help-seeking for depression were calculated by identifying doctors who answered ‘Yes’ to both questions, ‘Have you ever felt seriously depressed, even if not diagnosed?’ and ‘Did you seek personal support or professional treatment for depression?’

Patterns of help-seeking

Participants who were identified as having sought help or treatment for depression were further asked from where they sought personal support or professional treatment. Rates of ‘personal’ help-seeking were measured by identifying those who reported accessing one or more forms of personal support, including friend, family member, spouse/partner, work colleagues, library, university services, Internet and peer support program. Rates of ‘professional’ help-seeking were measured by identifying doctors who reported accessing one or more forms of professional support, including a GP, psychiatrist, psychologist/counsellor, indigenous support worker, EAP, telephone helpline, doctors' health advisory service and workplace support. Participants were instructed to mark all sources of help that applied to them.

Barriers to help-seeking

There were 19 items in the questionnaire that measured potential barriers to help-seeking: impact on registration and right to practice; concerns about career development/progression; fear of lack of confidentiality/privacy; impact on colleagues; lack of confidence in professional treatment; reliance on self – do not want help; difficulty identifying symptoms of mental illness; do not want to burden others; do not believe it will help; embarrassment; lack of time; fear of unwanted intervention; fear or stress about help seeking or the source of help; cost; stigmatising attitudes to mental illness; lack of knowledge about mental health services; impact on patients; income loss and lack of locum cover.28 Participants were asked to mark all items that apply to them and their decisions on whether to seek help or not.

Statistical analysis

All analyses were conducted using the IBM spss version 24, (Armonk, New York, USA). Descriptive statistics were used to explore sample characteristics, rates of help-seeking, patterns of help-seeking and barriers to help-seeking. Logistic regression analyses were used to examine the association between professional help-seeking as the dependent variable and a small group of predetermined independent variables. After a review of the available literature, five variables from the survey were selected for inclusion in the analysis (gender, location of primary medical degree – Australia or overseas, location of primary place of work – metropolitan or non-metropolitan, stage of training, specialty). Univariate logistic regression was first performed to establish the unadjusted associations between each variable and the likelihood of seeking professional help for depression. In the subsequent multivariate logistic regression, predictor variables were added in blocks sequentially to determine the unique variance contributed by each new variable, before a final model was determined.

Ethical approval for the original study was obtained from Monash University Human Research Ethics Office and Committee (CF12/2295–2012001228). The use of this data and the analysis conducted for the present study were also approved by University of New South Wales Human Research Ethics Committee (HC190896).

Results Demographics and rates of help-seeking for depression

Of the 12 252 doctors who completed the survey, 4154 (33.9%) indicated that they had ever felt seriously depressed. This group of 4154 doctors constituted the sample for the present study. The proportion of male and female doctors in this group were almost equal, with slightly more females (55.7%; n = 2314) than males. Almost 80% of doctors completed their undergraduate training in Australia and almost two-thirds (59.8%; n = 2483) worked in metropolitan areas (Table 1). The majority of the sample were senior doctors (62.5%) and the most dominant specialty was GP (36.0%) with a wide spread of other specialties included in the analysis. Of the 4154 doctors who reported ever having felt seriously depressed, 2695 (64.9%; 95% confidence interval (CI): 63.4–66.4) reported some form of help-seeking for mental health problems. A total of 2493 (60.0%; 95% CI: 58.5–61.5) doctors in the sample reported having sought help from a professional.

Table 1. Demographics of Australian doctors who have ever felt seriously depressed and rates of help-seeking for depression Variable n (%) Gender Male 1809 (43.5) Female 2314 (55.7) Age (years) 22–25 140 (3.4) 26–30 469 (11.3) 31–40 960 (23.1) 41–50 981 (23.6) 51–60 1066 (25.7) ≥61 508 (12.2) Location of primary medical degree Local (Australia) 3293 (79.3) Overseas 826 (19.9) Location of primary place of work Metropolitan 2483 (59.8) Non-metropolitan 1635 (39.4) Stage of training Junior doctors 1257 (30.3) Senior doctors 2598 (62.5) Specialty GP 1494 (36.0) Psychiatry 238 (5.7) Medicine 426 (10.3) Surgery 174 (4.2) Anaesthesiology 273 (6.6) Obstetrics and gynaecology/paediatrics 301 (7.2) ED/ICU/pain management 243 (5.8) Oncology 85 (2.0) Imaging/pathology 143 (3.4) Administrative/education 97 (2.3) Others 25 (0.6) Help-seeking for depression Overall 2695 (64.9) Personal support† 2418 (58.2) Professional treatment† 2493 (60.0) ED, emergency department; GP, general practitioner; ICU, intensive care unit. Patterns of help-seeking for depression

As demonstrated in Figure 1, among those doctors who sought help from personal supports, a friend, spouse/partner or family member were the most frequently reported sources. Only one-quarter (n = 616) of doctors who sought personal support for depression felt able to gain this support from a work colleague. Meanwhile, for those who sought professional treatment, GP were the most frequently reported source of help, followed by psychologists/counsellors and psychiatrists. Notably, only a small proportion of doctors reported seeking help from services offered at the workplace, such as EAP and Workplace Support (Fig. 2).

Patterns of help-seeking (personal support) among doctors who have ever felt seriously depressed (n = 2418).

Patterns of help-seeking (professional treatment) among doctors who have ever felt seriously depressed (n = 2493). DHAS, doctors' health advisory service; EAP, Employee Assistance Provider; GP, general practitioner.

Barriers to professional help-seeking for depression

The most commonly reported barrier to professional help-seeking for depression among depressed doctors was privacy and confidentiality concerns, with more than half of the sample endorsing this concern (Table 2). Almost half of the sample cited ‘embarrassment’ and ‘impact on registration and right to practice’ as barriers to professional help-seeking.

Table 2. Barriers to help-seeking among doctors who have ever felt seriously depressed Barrier Number endorsing each barrier (% of total sample = 4154) Fear of lack of confidentiality/privacy 2572 (61.9) Embarrassment 1863 (44.8) Impact on registration and right to practice 1710 (41.2) Lack of time 1471 (35.4) Concerns about career development/progression 1381 (33.2) Reliance on self, do not want help 1379 (33.2) Stigmatising attitudes to mental illness 1368 (32.9) Impact on colleagues 1229 (29.6) Fear of unwanted intervention 1200 (28.9) Do not want to burden others 1145 (27.6) Lack of confidence in professional treatment 1067 (25.7) Do not believe it will help 868 (20.9) Income loss 740 (17.8) Fear or stress about help-seeking or the source of help 684 (16.5) Impact on patients 519 (12.5) Lack of locum cover 410 (9.9) Cost 379 (9.1) Difficulty identifying symptoms of mental illness 193 (4.6) Lack of knowledge about mental health services 190 (4.6) Predictors of professional help-seeking for depression

The regression model to estimate the impact of several factors on the likelihood of doctors seeking professional help for depression examined five predetermined variables. The full model containing all predictors entered simultaneously (gender, location of primary medical degree, location of primary place of work, stage of training, specialty) was able to differentiate between doctors who reported and did not report professional help-seeking for depression (χ2 (14) = 117.7; P < 0.001). However, primary place of work did not account for any significant amount of variance in the likelihood of professional help-seeking and was subsequently excluded from the final model (χ2 (13) = 117.4; P < 0.001). Table 3 presents the final model demonstrating the associations between each factor and the likelihood of professional help-seeking for depression. Females were more likely to seek professional help for depression than males (odds ratio (OR) = 1.74; 95% CI: 1.50–2.01; P < 0.001), Australian-trained doctors were more likely to seek professional help for depression than overseas-trained doctors (OR = 1.34; 95% CI: 1.12–1.59; P = 0.001) and senior doctors were more likely to seek professional help for depression than junior doctors (OR = 1.35; 95% CI: 1.14–1.61; P = 0.001). Furthermore, psychiatrists were more likely to seek professional help than GP (OR = 1.57; 95% CI: 1.15–2.13; P = 0.004) whereas surgeons (OR = 0.52; 95% CI: 0.37–0.72; P < 0.001) and pathologists/radiologists (OR = 0.70; 95% CI: 0.490–0.99; P = 0.043) were less likely to seek professional help than GP.

Table 3. Logistic regression predicting likelihood of seeking professional help for depression Variable n % seeking professional help OR OR 95% CI Adjusted OR Adjusted OR 95% CI Significance in the final model (P) Gender Male 1809 52.8 Reference Reference Reference Reference <0.001 Female 2314 65.6 1.702 1.50–1.93 1.736 1.50–2.01 Location of primary medical degree Overseas 826 54.2 Reference Reference Reference Reference Local 3293 61.6 1.351 1.16–1.58 1.338 1.12–1.59 0.001 Stage of training Junior doctors 1257 55.4 Reference Reference Reference Reference 0.001 Senior doctors 2598 61.9 1.303 1.14–1.49 1.354 1.14–1.61 Specialty GP 1494 62.7 Reference Reference Reference Reference Psychiatry 238 71.0 1.456 1.08–1.96 1.565 1.15–2.13 0.004 Medicine 426 59.4 0.869 0.70–1.08 0.984 0.78–1.24 0.890 Surgery 174 43.1 0.450 0.33–0.62 0.518 0.37–0.72 <0.001 Anaesthesiology 273 61.2 0.937 0.72–1.22 1.066 0.81–1.40 0.648 ObGyn/paediatrics 301 60.1 0.897 0.70–1.16 0.953 0.73–1.24 0.718 ED/ICU/pain 243 57.2 0.795 0.60–1.05 0.990 0.74–1.32 0.943 Oncology 85 58.8 0.849 0.54–1.33 0.922 0.59–1.46 0.729 Imaging/pathology 143 52.4 0.656 0.47–0.93 0.695 0.49–0.99 0.043 Administration/education 97 68.0 1.266 0.82–1.96 1.315 0.83–2.09 0.244 Others 25 68.0 1.263 0.54–2.95 1.677 0.64–4.43 0.296 CI, confidence interval; ED, emergency department; GP, general practitioner; ICU, intensive care unit; ObGyn, obstetrics/gynaecology; OR, odds ratio. Discussion

To our knowledge, this is the first study to describe help-seeking for depression in Australian doctors across all stages of training and specialty areas. Reassuringly, the present study found most doctors who had experience significant depression were able to seek professional help, with most choosing to use their GP, psychologist or psychiatrist rather than mental health services related to their workplace. However, rates of help seeking remained concerning low for some groups of doctors, with male doctors, overseas-trained doctors, junior doctors, surgeons and pathologists/radiologists being less likely than their counterparts to seek professional help for depression.

The rate of professional help-seeking for depression in the present study (60%) was notably higher than rates reported among doctors in previously published studies (24%),11, 15, 16, 25 and higher than the general Australian population (35%).8 This is a positive finding that may indicate some benefits from work done on encouraging help-seeking among Australian doctors over the past decade. Alternatively, the prevalence of help-seeking may have been inflated by the risk of sampling bias in a survey that was specifically about doctors' mental health and through selecting a sample that were able to recognise that they had experienced depression.

Knowing where doctors seek help for depression could aid in directing resources and modifying existing, underutilised systems. A majority of the doctors in this sample who had sought professional help for depression had received treatment from GP, followed by psychologists/counsellors and psychiatrists. This indicates that when doctors seek help for depression, they have a tendency to follow traditional medical pathways to specialist services. Conversely, relatively few doctors had accessed services offered by the workplace or related to the medical profession, such as the Doctors' Health Advisory Service, Employee Assistance Program or Workplace Support. The barriers to professional help-seeking for depression reported by doctors in the present study were in line with their preferred sources of professional support. Specifically, the most commonly reported barriers – concerns regarding privacy and confidentiality, embarrassment, impact on registration and right to practice and career development or progression – may lead to doctors avoiding mental health services within their workplace or related to the medical profession. This may be partially due to mandatory reporting requirements in most Australian states regarding doctors with impairment.31 Together, these findings suggest that strategies to promote help-seeking for depression in doctors may be more successful if they harness doctors' preference for seeking professional treatment from GP, psychologists and psychiatrists, and they address doctors' concerns about privacy and impact on their career. One additional potential avenue for addressing the mental health of doctors is e-mental health, given the evidence for Internet and mobile-based interventions in reducing symptoms of depression28and their capacity to overcome barriers to help-seeking, such as embarrassment and privacy concerns.

This study also identified several groups of doctors at greater risk of untreated depression, given their lower rates of professional help-seeking for depression. First, male doctors were significantly less likely to seek professional help for depression than their female counterparts. This finding was congruent with findings in the Australian general population, in which males are less likely than females to use mental health services.32 Possible reasons for this include gender socialisation,33 or men experiencing a lower level of confidence in professional mental healthcare than women.34, 35 Second, our findings showed that doctors trained in Australia were 1.3 times more likely to seek professional help than overseas-trained doctors. This may be due to initiatives by Australian medical schools to raise awareness of mental health and support services, such as the AMSA Mental Health Campaign.36, 37 Conversely, culture may play a vital role. Perceived stigma of mental illness within the general population has been found to be more prevalent in developing countries.38 As approximately 32% of doctors in Australia in 2016 having obtained their undergraduate training overseas,39 the influence of culture on likelihood to seek help for mental health problems represents a critical factor to consider in future education campaigns. Third, and consistent with previous research,21, 22 senior doctors were more likely to seek professional help than junior doctors. In addition to concerns regarding career development, being time-poor may also be a barrier to help-seeking for junior doctors, due to the demands of study and exams. Regardless of the underlying cause, this finding is critical as mental disorders are more likely to occur in junior doctors; hence, the very group that are most likely to need help appear less able to ask for it. Last, the present study found that some specialities had particularly low rates of help-seeking for depression,