Why don ’ t academic physicians seek needed professional help for psychological distress ?

a Department of Psychology, Stockholm University, Sweden b Centre of Gender Medicine, Karolinska Institutet, Stockholm, Sweden c Department of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden d School of Community & Global Health, Claremont Graduate University, USA e Keck School of Medicine, University of Southern California, USA f Department of Psychiatry and Psychology, Padua University Hospital, Italy g Department of Medicine, Cardiac Unit, Karolinska Institutet, Stockholm, Sweden h Karolinska University Hospital, Stockholm, Sweden


Introduction
Physicians face a heavy burden of work stressors, whose contribution to psychological distress is increasingly apparent [1][2][3].There is mounting evidence that physicians are at risk for burnout, depression, and suicide [1][2][3][4][5][6].Moreover, data from Switzerland indicate that the prevalence of physician burnout has increased in the last decade [6].Physicians reportedly seek help to a lesser degree and later in the course of disease than do other groups [7].Physicians appear to be reluctant to seek help for psychological health problems due to concerns about confidentiality [8].A recent study suggests that this concern is especially widespread among physicians in the surgical specialties [9].Male physicians are also reported to less frequently consult colleagues regarding their own health compared to female physicians [7].On the other hand, it has been demonstrated that professional help, through e.g., counselling in conjunction with job interventions such as reduction in work hours may ameliorate physician burnout [10].The academic clinical setting is a particularly challenging and competitive work environment for physicians, involving patient care, teaching, and research [11].This work environment is associated with adverse health effects [5,12,13].We recently reported that degrading experiences/ harassment at work are associated with suicidal thoughts among university hospital physicians [5,12].We also found a significant relation between role conflict and suicidal thoughts among this group of physicians.This was related to the multiple and often clashing demands of medical research, teaching, and administration in addition to clinical duties [12].Another salient finding from our study is that certain health-related behaviours, namely self-diagnosis and self-treatment are associated with recent suicidal thoughts among academic physicians [5]."Health and Organization among University Hospital Physicians" (HOUPE) [5,12,14] is a research project on work-related health, organisational culture, career, and working conditions in university hospitals in European countries.Subsequent to the baseline cross-sectional observational study (HOUPE I), interventions and follow-up are ongoing.Among the Swedish physicians participating in HOUPE I, over half had high levels of exhaustion; nearly 40% showed signs of disengagement from work, and over 10% had recent suicidal thoughts [14].To our knowledge, however, there are no published studies among university hospital physicians concerning help-seeking for psychological distress.Our aim with the present study was to examine help-seeking for burnout or depression among university hospital physicians participating in the HOUPE I study and who showed signs of distress, including suicidal ideation.We thereby sought to determine who is at risk for not getting needed health services, to find factors that might hinder help-seeking, but also factors that may have been triggers.The broader conceptual framework [11] for this study is to examine the complex relations among gender, work organisation, career development, certain health-related behaviours and work/family conflict as these may impact upon the mental health of academic physicians.In this paper we particularly focus upon potentially modifiable stressors, as well as aiming to identify groups of distressed academic physicians who are at risk for not seeking needed help.We hypothesise that those at risk for not having sought needed help will include male physicians, surgeons, those with taxing working conditions including role conflict, and physicians who go to work sick ("presenteeism") and those who self-diagnose and self-treat.

Methods
This is a cross-sectional study of a subpopulation of physicians with recent suicidal thoughts and/or depression/ burnout participating in the baseline (Phase I) HOUPE study.The present analyses include Italy and Sweden, since these were the only two participating countries for which all the queries pertaining to mental health were included in the questionnaires.

Participants and study design of the HOUPE I, as carried out in Italy and Sweden
Physicians were eligible to participate in the HOUPE study if they were permanently employed as physicians and actively working at the respective university hospitals, and were not on specialisation training.All eligible physicians were invited to participate on a voluntary basis in the HOUPE study.Thus, physicians working at all the departments at the respective university hospitals were included.In Sweden the study was carried out at the Karolinska University Hospital, Stockholm, with approval from the Regional Ethics Board (Stockholm) on 8 December 2004, number 04-913/2.All eligible physicians received written information about the survey, often supplemented by short oral presentations.The former included study description with an endorsement by the director of the Karolinska University Hospital, the chairperson of the local medical association, and the project manager.The HOUPE study was described to the physicians as a research programme focusing upon work-related health, organisational culture, career paths and working conditions in university hospitals in a number of European countries, with the link: www.houpe.nogiven for further information.All eligible physicians received a personal password and log-on information to the web-based questionnaire.The joint data collection via the Internet was organised centrally for the Nordic countries at www.houpe.no,hosted by Department of Research and Development at St Olavs University Hospital in Trondheim, Norway.Only authorised personnel could monitor the data collection process or designate new log-on information to participants who had lost their personal access code.All participants used their personal logon information to anonymously enter their responses to the questionnaire.Four electronic reminders were sent as follow-up to those who had not completed the questionnaire.A hardcopy of the questionnaire was also sent to provide an alternative for those who may have been reluctant to respond electronically.The Italian study took place at the Padua University Hospital, approved by The Ethics Board of Padua University Hospital, on 5 September 2005, protocol number 1039P.All the eligible Italian physicians received a letter shortly thereafter describing the study endorsed by the hospital director and by the Dean of the Medical Faculty.The Italian physicians received only a hardcopy of the questionnaire, which was sent by internal mail and an addressed return envelope without information about the sender.Informal reminders were made to the directors of the various units.Altogether 1380 physicians from Sweden were eligible for the HOUPE I study and were contacted.Of these, 841 (60.9%) physicians participated.Data were available to us

Original article
Swiss Med Wkly.2012;142:w13626 Swiss Medical Weekly • PDF of the online version • www.smw.chfrom the hospital administration concerning the gender and age of the eligible physicians.The male physicians had a lower participation rate than the females (χ 2 = 17.2, p <0.001).The Swedish physicians aged 50 or more also had a lower participation rate than those under age 50 (χ 2 = 23.2,p <0.001).There were 900 Italian physicians who were eligible and invited to participate in the HOUPE I study.Altogether 367 (40.8%) physicians participated from the Italian center.Data were also available to us from the hospital administration concerning the gender and age of the eligible physicians.The difference in participation rates between the male and female physicians did not attain statistical significance for the Italian centre (χ 2 = 2.98).Italian physicians aged 50 or more also had a lower participation rate than those under age 50 (χ 2 = 10.9, p <0.001).All the data reported in this paper are from the responses to the Phase I HOUPE questionnaire.Confidentiality was guaranteed; in particular, it was emphasised that individual data could not be identified in any way.Data collection for Phase I of the HOUPE study took place from 2005 and 2006.Further details about the HOUPE I study as carried out in Sweden and Italy are presented in references by Fridner et al. [5,12].

Criteria for the subpopulation of physicians included in the present analyses
The present paper includes all physicians from Italy or Sweden participating in the HOUPE I study who showed one or more indices of current psychological ill-health, according to their answers to the questionnaire.The indices assessed psychological distress, burnout, and suicidal thoughts.Psychological distress was gauged as a score above 4 on the General Health Questionnaire (GHQ-12) [15].Burnout was assessed by the "Mini Oldenburg Burnout Inventory" (MOLBI) via two additive scales with five items each for disengagement [16] (for HOUPE I Sweden and Italy, Cronbach's α = 0.76) and for exhaustion [16] (for HOUPE I Sweden and Italy, Cronbach's α = 0.79).A score above the upper quartile on the MOLBI scales was used as the cutpoint for exhaustion (>15) or disengagement (>12).All the queries from the GHQ and MOLBI referred explicitly to the past few weeks.Complete data for the GHQ together with both MOLBI scales were available for 1101 (91.1%) of the 1208 participating physicians from Italy and Sweden.Recent suicidal thoughts were assessed via dichotomous (yes/no) questions concerning the past 12 months.The queries were: "having thoughts about taking your life" and "thought about specific ways to take your life" [17].We combined the two queries to create a composite dichotomous variable.An affirmative response to either of the questions was considered to indicate the presence of recent suicidal thoughts [12].Data on recent suicidal thoughts were available for 1163 (96.3%) of the 1208 participating physicians from Italy and Sweden.

Survey content for the present analyses
The outcome variable was having sought professional help for depression or burnout.This was assessed from the an-swer to the dichotomous (yes/no) query: "Have you ever sought professional help for depression or burnout?" [11].We examined four types of independent factors: work environmental factors, work-home interface, health-related behaviours and demographic factors.The Questionnaire for Psychological and Social Factors at Work (QPS-Nordic) [18] was used to assess recent degrading experiences or harassment at work (dichotomous question (yes/no)) and role conflict according to the 3-item scale (Having to do things one feels should be done differently, being given assignments without adequate resources to complete them and receiving incompatible requests from two or more people).From the Physician Career Path Questionnaire (PCPQ) [11] concerning academic career and specialisation and working conditions, we included information about specialty area (surgical versus non-surgical), number of work hours per week, as well as dichotomous items (yes/ no) about currently working night shifts/taking on-call duty at night, being currently involved in medical research, and being assigned extra tasks.We evaluated the work-home interface and health-related behaviours as follows.The work-home interface was assessed from the QPS Nordic, by querying as to whether the demands of work often interfere with family life.Two items from the PCPQ ascertained whether the physicians diagnosed and treated themselves for symptoms for which they would have referred a patient to a specialist (dichotomous item (yes/no)) and whether they went to work with an illness for which they would recommend a patient to stay at home.In addition to gender and country, items concerning age, marital status (married/living with a partner versus living alone (single, divorced or widowed)), and number of children were included as socio-demographic variables.

Statistical analysis
Statistical analysis was performed using SPSS software (SPSS version 19, SPSS Inc., Chicago, Illinois).The sociodemographic and other characteristics of the group of physicians with psychological distress were assessed by numerical count and percentage.Multiple logistic regression was used to identify factors significantly associated with not having sought professional help for depression or burnout.The socio-demographic factors: age, civil status, number of children, gender and country were included in all the regression analyses.The regression analysis was performed among two subgroups likely in need of help: 1) those with recent suicidal thoughts and 2) those with GHQ above four or with MOLBI exhaustion or disengagement scores above the upper quartile.All significant adjusted odds ratios (OR) and 95% confidence intervals (CI) with regard to the outcome variable are reported.Only cases with complete data were included in each regression analysis; the missing data are indicated in table 1.

Results
As described above in the methods section and in Fridner et al. [5,12], altogether in the HOUPE study there were 841 participating physicians from Sweden and 367 from Italy.Of these 1208 physicians, 516 physicians (42.7%) showed one or more indices of current psychological illhealth, and were thus included in the present analyses.The descriptive data for these 516 physicians are shown in table 1. Therein, it is seen that of the 516 physicians with one or more indices of current psychological ill-health, a total of 404 stated that they had not ever sought professional help for depression or burnout.This amounts to 78.3% of the distressed physicians.In tables 2 and 3, we report the statistically significant as well as the non-significant associations between the independent variables and the outcome.Of the 155 physicians who reported recent suicidal thoughts, 106 had not sought help (table 2).Within this group, university hospital physicians who were currently involved in research and males were significantly more likely not to have sought help.Physicians with suicidal thoughts who had recent degrading experiences or faced harassment at work were significantly more likely to have sought help.Those who selfdiagnosed and treated themselves for symptoms for which they would have referred a patient to a specialist were also significantly more likely to have sought help.As shown in table 3, of the 461 physicians with a GHQ score above four or MOLBI score above the upper quartile for exhaustion or disengagement, 365 also had not ever

Discussion
The focus of this study is university hospital physicians who are in need of professional psychological help, based upon their answers to the HOUPE questionnaire.The most striking finding is that such a large percentage of these physicians with recent suicidal thoughts or other indicators of psychological distress have never sought help.We identified three work-related characteristics significantly related to not seeking needed help: 1) being currently involved in medical research, 2) being a surgeon, and 3) taking night call.Significant associations were not found with help seeking with regard to work role conflict, workfamily conflict, long work hours or sickness presence.A surprising finding was that self-diagnosis and treatment showed a positive association with having sought needed help for burnout or depression.Recent degrading experiences/harassment showed an even stronger association with having sought help.As hypothesised, male gender was associated with a lower likelihood of having sought needed help for university hospital physicians with recent suicidal thoughts, as well as for those with signs of burnout or depression.Italian physicians with signs of burnout or depression were significantly less likely than their Swedish colleagues to have sought help.
There is very little known about help-seeking among psychologically distressed physicians.Besides our findings that most of these physicians in need have never sought help, the only other studies (to our knowledge) which address this question are those of Shanafelt et al. [9] among surgeons in the U.S. and of Tyssen et al. [19] in a longitudinal study of medical students in Norway.A concordantly low percentage (26%) of those with recent suicidal thoughts had sought professional help within the last year in the study by Shanafelt et al. [9].Among the Norwegian medical students followed-up into their first years of postgraduate training, two-thirds of those who stated that they had had psychological health problems during the last year, also stated that they had sought help [19].However, in the latter study, psychological health problems were self-rated and the prevalence (17.1%) much lower than in our study, perhaps due to denial or lack of knowledge among these very young physicians.Among the university hospital physicians in our study, being currently involved in medical research showed the strongest association with not seeking help.Scientific activity is a key to career advancement in academic medicine, and requires the physician to continuously strive and compete to reach "the top": to publish, to attract funding and to maintain networks, inter alia.Getting help for psychological distress requires a very different mindset, since it entails recognition of vulnerability.This could be perceived as a threat to one's competitive edge, especially in light of the potential stigma surrounding mental illness among health professionals [20].In addition, the demands of research upon the physicians' time and energy may hinder their seeking needed help for mental distress.
The male physicians in this study sought help significantly less often than did their female colleagues.Other empirical data indicate that male physicians consult their colleagues significantly less often than females [7].Men in general re-portedly seek help, especially for psychological problems, less often than females [21,22].Traditional male gender norms emphasise self-reliance, emotional control, being successful and avoiding "feminine" behaviour [21].Helpseeking behaviour in those terms might be considered a weakness and threat to masculinity [22].Country (Italy) was the other sociodemographic factor significantly associated with not having sought needed help.
Our previous findings indicate that the Swedish physicians in the HOUPE study showed significantly higher burnout and GHQ scores than did the Italian physicians [14,23].
However, here we see that when Italian physicians become burned out or depressed, they are less likely to seek help.Cultural factors could be involved, and these will be explored in the HOUPE follow-up study.
Our study also identified surgeons with signs of burnout or depression, as being a group at risk for not having sought needed help.Burnout has been cited as an important problem among surgeons [24].There are no other published data (to our knowledge) concerning help-seeking for burnout or depression among surgeons.The study of Shanafelt et al. [9] indicates that a very small percentage of surgeons with suicidal thoughts seek help.Taken together, these findings suggest that more outreach is warranted to facilitate psychologically-distressed surgeons getting the needed professional help.Academic physicians with signs of burnout or depression who took night call were also less likely to seek help.This is a new finding with potential policy implications concerning work schedules.There are some data indicating that night work is associated with burnout and other untoward psychological outcomes among certain physician profiles (anaesthesiologists, intensive care specialists, as well as surgeons) [25,26].Thus, it is plausible that night work not only acts as a trigger of psychological distress, but also represents a barrier to getting help when this distress occurs.
Overall, it is likely that stigma is a barrier for distressed physicians to seek psychological help.There is a recognised stigma in seeking help for mental ill-health among physicians.Rather than so doing, physicians often state that they are taking time off for somatic health problems, when in fact the reasons are of a mental health nature [27].It has been found that anxiety about confidentiality and embarrassment are higher for mental problems than for physical problems.These concerns are present even among mental health professionals themselves [28].Notably, 86% of psychiatrists in a study from Birmingham, U.K. reported that they would be reluctant to disclose their own mental ill-health issues to colleagues or professional organisations [29].Moreover, choices regarding disclosure and treatment would be affected by concerns about confidentiality and career implications, rather than quality of care.On the other hand, and contrary to our expectations, physicians with suicidal thoughts or other signs of psychological distress who self-diagnose and self-treat were more likely to have sought help, compared to those who do not selfdiagnose and self-treat.Data from Adams et al. [27] suggest that physicians self-diagnose and self-treat predominantly for somatic problems.Notably, antidepressants are seldom self-prescribed [27].It would seem that these phys-icians realise that mental health concerns are outside their domain and recognise their limitations.Also, this finding may reflect feelings of responsibility towards their patients.In order words, a possible interpretation is that health professionals suffering from mental distress would tend to be conscientious to make sure that they are competent to provide care [20].Academic physicians with suicidal thoughts or other signs of psychological distress, who had faced harassment or degrading experiences at work, were also significantly more likely to have sought help.Harassment has been found to be strongly associated with suicide risk and other untoward mental health outcomes among physicians [5,12,30].It would seem that these experiences are so devastating that the distressed physicians cannot but acknowledge their vulnerability.

Strengths and limitations
Our outcome variable was having once sought professional help for burnout or depression.However, the academic physicians included in the present study were those with current/recent indicators of psychological ill-health.Therefore it is not possible to know how many of these distressed physicians who received help in the past are also getting the help which they currently need.On the other hand, since the queries from GHQ and MOLBI refer to the last few weeks, it might not always be expected that the psychological distress is of sufficient duration to warrant help seeking.
A potential source of bias could be from self-report of outcome and work stressors and behaviours within the same questionnaire.We attempted to counteract this by not explicitly stating that having sought professional help for depression or burnout was being assessed as an outcome variable.
Since only actively working physicians were eligible, there is a possible selection of those who were relatively healthy.This could create a bias towards the null [31] such that the associations observed in the present study may be underestimated.On the other hand, because the present study is cross-sectional, inferences about the temporal nature of observed associations must be made cautiously.There may also be unmeasured confounders that affect the results.The response rate for the Swedish physicians in the baseline HOUPE I study was within the acceptable range.However, it was low for the Italian physicians.It is likely that this was the first time that physicians in Italy had received a questionnaire of this type, and unfamiliarity may have contributed to the low response rate [5,12].Our samples are fairly large and heterogeneous, including a range of specialists within the academic medical setting.Generalisability is enhanced since two countries in different parts of Europe are included.Academic medicine is highly prestigious and a trendsetter for medical practice.Therefore, these results are likely to be relevant to broader segments of physicians and to other health providers.

Conclusions
Very few of these university hospital physicians with signs of mental distress have ever sought help from a mental health professional.We have identified groups of physicians most likely not to have sought needed help: male physicians, those from Italy, surgeons, physicians currently involved in research, and those who take night call.
Outreach is warranted particularly to these groups of physicians in need.These findings have implications for the physicians themselves as well as for patient care, clinical research, and the education of future physicians.More study, preferably of an interventional design, is warranted concerning help seeking among these academic physicians in need.

Table 1 :
Characteristics of the physicians included in the present analyses: University hospital physicians with one or more indicators of current psychological ill-health* and who participated in the HOUPE I study.Within this group, three work-related factors showed a significant direct association with not having sought help: being currently involved in medical research, taking night call and being in the surgical specialties.The strongest association was for being involved in research.Male physicians and those from Italy were also significantly more likely not to have ever sought help.On the other hand, those physicians who had recent degrading experiences or faced harassment at work were significantly more likely to have sought help.Physicians who self-diagnosed and treated themselves for symptoms for which they would have referred a patient to a specialist were also significantly more likely to have sought help.
* GHQ >4 or MOLBI Exhaustion >15 or MOLBI Disengagement >12 or suicidal thoughts HOUPE = Health and Organization among University Hospital Physicians in Europe GHQ = General Health Questionnaire, MOLBI = Mini Oldenburg Burnout Inventory Original article Swiss Med Wkly.2012;142:w13626 Swiss Medical Weekly • PDF of the online version • www.smw.chsought help for depression or burnout.

Table 2 :
Physicians with recent suicidal thoughts in the HOUPE study: Adjusted odds ratios for not having sought help for burnout or depression.
Multiple logistic regression with adjustment for socio-demographic factors: age, number of children, marital status, gender and country *Adjusted for all of the other sociodemographic factors HOUPE = Health and Organization among University Hospital Physicians in Europe, OR = odds ratio, CI = confidence intervals, NS = statistically non-significant (p <0.05)

Table 3 :
Physicians with GHQ >4 or MOLBI exhaustion scores >15* or MOLBI disengagement >12* in the HOUPE study: Significant adjusted odds ratios for not having sought help for burnout or depression.
Multiple logistic regression with adjustment for socio-demographic factors: age, number of children, marital status, gender and country GHQ >4 or scores above the upper quartile* for MOLBI exhaustion >15 or disengagement >12 *Adjusted for all of the other sociodemographic factors HOUPE = Health and Organization among University Hospital Physicians in Europe, GHQ = General Health Questionnaire, MOLBI = Mini Oldenburg Burnout Inventory [15], OR = odds ratio, CI = confidence interval, NS = statistically non-significant (p <0.05) Original article Swiss Med Wkly.2012;142:w13626 Swiss Medical Weekly • PDF of the online version • www.smw.ch