Bullying and sexual harassment of junior doctors in New South Wales, Australia: rate and reporting outcomes

Additional keywords: bullying, junior doctors, medical trainee, psychological distress, sexual harassment, wellbeing.

Bullying and sexual harassment within the medical profession is of concern in Australia. In 2015, problems within surgical training became public,¹ quickly expanding into an acknowledgement that problems existed across the medical profession.^2,3 An Australian cross-sectional survey found that 25% of doctors reported being bullied in the workplace in the previous 12 months.⁴ Detrimental effects to the medical workforce as a result of these behaviours include decreased job satisfaction, decreased workplace engagement,⁴ poor mental health and suicidal ideation⁵ and a potential for decreased patient safety.⁶

Junior doctors may be particularly vulnerable to bullying and sexual harassment given the hierarchical structure of the medical profession.^7–9 Leisy and Ahmad⁸ identified that the prevalence of bullying and sexual harassment among junior doctors ranged from 30% to 89% in international cross-sectional surveys. In addition, the tragic deaths of junior doctors in New South Wales (NSW) have heightened concern over workplace stress.^10–12 However, beyond prevalence, there is little information available that considers the actions taken in response to bullying and sexual harassment. It is therefore vital that bullying and sexual harassment of junior doctors in Australia be more fully elucidated in order to effectively address this problem.

In NSW and in the Australian Capital Territory (ACT), most doctors undertake 2 years of general prevocational training after graduation before specialist (vocational) training (PGY1 and PGY2). The annual NSW Junior Medical Officer (JMO) Census is a peer-led online survey of junior doctors during these prevocational years (PGY1 and PGY2). Included in the survey are five questions probing quantitative and qualitative aspects of the experience of being bullied and/or sexually harassed.

The present study had three aims: (1) to describe the rates of exposure to bullying and sexual harassment in junior doctors in PGY1 and PGY2 in NSW and the ACT, including the relationship with psychological distress; (2) to describe actions taken by junior doctors in response to bullying and sexual harassment; and (3) to explore the experiences of junior doctor of the reporting process.

The present study was a cross-sectional survey of PGY1 or PGY2 junior doctors in 2015 and 2016 across NSW and the ACT. The project was approved by the Hunter New England Local Health District Human Research Ethics Committee. Links to resources and contacts where participants could find support if distressed were included in the survey.

Doctors undertaking PGY1 or PGY2 training in NSW and the ACT were invited to participate in an online survey using a peer-led snowballing technique. First, an invitation to participate was generated as a web link to the survey, and the members of a representative group of PGY1s and PGY2s (n = 46) were then tasked with recruiting participants through their hospital, health service, email and social media networks to participate in the study. Those contacted were similarly encouraged to forward the web link to their peers. There were no exclusion criteria. The survey was administered in September 2015 and September 2016. The clinical year started in January of that year.

Data were collected as part of a larger annual survey entitled the ’NSW JMO Census. The data used in the present study included demographic variables, as well as nominal quantitative and free-text responses to a series of questions about bullying and sexual harassment (see Table 1). Data collection was conducted online via SurveyMonkey. The Health Education and Training Institute (HETI) of NSW hosted the SurveyMonkey account and was responsible for the security, collection and compilation of the data.

Table 1.  Survey questions on the annual New South Wales Junior Medical Officer (JMO) Census related to bullying and sexual harassment

Statistical analysis of quantitative data from the survey responses was conducted using SPSS version 24 (IBM Corp., Armonk, NY, USA). Data from the 2015 and 2016 surveys were considered separately due to the potential overlap of junior doctors responding to the survey twice about the same incident (i.e. as a PGY1 doctor in 2015 and then as a PGY2 doctor in 2016) because the wording of the survey questions did not allow differentiation of an incident as occurring in the first or second year of prevocational training.

Simple proportions and descriptive statistics were used. For 2 × 2 analyses, the Yates continuity-corrected Chi-squared test was used. In other cases, the Pearson Chi-square test was used.¹³ Psychological distress was measured as a dichotomous dependent variable based on respondent scores to the Kessler Psychological Distress Scale (K10).¹⁴ Low distress was classified as a K10 score ≤15, and moderate to high distress was classified as a K10 score ≥16.¹⁵ Differences were considered statistically significant if P < 0.05 (one-tailed).

Qualitative data regarding responses to bullying and/or sexual harassment from the 2015 and 2016 surveys were pooled. This was done due to the exploratory nature of the analyses, as well as the small sample size of respondents who provided qualitative data.

First, free-text responses regarding active actions in response to bullying and harassment were analysed using content analysis and coded into categories for simple descriptive quantitative analysis. Second, in order to explore the process of reporting and the perceived outcomes in greater depth, where junior doctors had decided not to take action in response to bullying and harassment, thematic analysis was undertaken on free-text responses describing these experience.¹⁶ Similarly, thematic analysis was undertaken on free-text responses describing junior doctor decisions not to report bullying and harassment. Thematic analysis was undertaken in an inductive manner, whereby themes were drawn from the data and with the specific research questions in mind. Four authors (AL, AK, DN, WL) independently coded free-text responses and compared coding decisions in an iterative manner to further refine emergent themes the coding approach. Any disagreements were resolved through discussion, and the iterative analytic process ceased when saturation was attained.

In 2015, of 393 doctors who completed the full survey, 374 (95%) completed the bullying portion. In 2016, of 448 doctors who completed the full survey, 440 (98%) completed the bullying portion. We estimate the response rate to be between 17% and 20%. This estimate is based on the 19th report by the Medical Training Review Panel¹⁷ citing 2195 doctors in PGY1 or PGY2 positions across NSW and the ACT for 2015, and assuming the same number of positions in 2016 (data currently unavailable). Most respondents were PGY1, aged 25–27 years, female, not married and without dependents (Table 2).

Table 2.  Demographics of the 2015 and 2016 survey respondents
Data are given as n (%)

In 2015, 54.3% (n = 203) of respondents reported having been bullied, and 15.5% (n = 58) reported sexual harassment. A significantly greater proportion of females (60.0%) than males (45.6%) reported having been bullied, and a significantly greater proportion of females (22.7%) than males (4.7%) reported sexual harassment. There were no significant differences in the proportion of bullying or sexual harassment across marital status or age (Tables 3, 4). A greater proportion of respondents who were exposed to bullying (n = 140; 69%) reported moderate to high psychological distress than those who were not exposed to bullying (n = 64 (37%); χ²₁₃₇₄ = 37.24, P < 0.001).

Table 3.  Results of Chi-square tests comparing demographic variables of respondents, in 2015 and 2016, who experienced bullying

Table 4.  Results of Yates continuity-corrected Chi-squared tests comparing demographic variables of respondents, in 2015 and 2016, who experienced sexual harassment

In 2016, 57.5% (n = 253) of respondents reported having been bullied, and 18.6% (n = 82) reported sexual harassment. A significantly greater proportion of females (62.2%) than males (51.3%) reported having been bullied, and a significantly greater proportion of females (29.2%) than males (4.8%) reported sexual harassment. A greater proportion of respondents who were exposed to bullying (n = 165; 65%) reported moderate to high psychological distress than those were not exposed to bullying (n = 81 (43%); χ²₁₄₄₀ = 20.92, P < 0.001).

Most respondents in both 2015 and 2016 reported occasional incidents of bullying and sexual harassment, occurring less than monthly (Table 5), with the perpetrator most frequently being a senior medical staff member (Table 6).

Table 5.  Number (%) of respondents in 2015 and 2016 who reported bullying or sexual harassment

Table 6.  Number (%) of respondents in 2015 and 2016 who reported bullying by perpetrator type
JMO, junior medical officer

Of the 486 respondents reporting bullying and/or sexual harassment across 2015 and 2016, 136 (28%) provided free-text responses regarding their response to the incident/s. Responses to bullying and/or sexual harassment were considered as either constituting action (took action within the system) or inaction (no active action, including avoidance). Of those respondents who provided qualitative data, 60% (n = 82) took some kind of action within the system. Action responses were further classified as either escalated (including escalation to Director of Training, management unit, other senior medical colleague, other manager, incident monitoring system, external organisations), peer sharing or direct action (talking directly to the perpetrator). Conversely, 40% (n = 54) of respondents did not take action within the system (Table 7).

Table 7.  Number (%) of respondents in 2015 and 2016 (combined) as a function of response type to bullying and/or sexual harassment (n = 136)

Thematic analysis further elucidated the reasons for not taking action (Fig. 1). Reasons included the normalisation of bullying and/or sexual harassment as ‘rampant! […] the culture is not to complain’, fear of reprisal for reporting:

Fig. 1.  Process of response types taken by junior doctors to bullying and sexual harassment.

I felt raising the concern was likely to impact negatively on my future career prospects

lack of knowledge about, or confidence in, the reporting process:

[I] did not know the appropriate channels to use

disengagement from the organisation:

I felt it would be more painful trying to change it than putting up with it

and feeling encouraged by other staff not to report the behaviour (for more quotes, see Table 8). One particular case of sexual harassment clearly illustrated the extent of the respondent’s lack of confidence in the reporting process:

Table 8.  Reasons provided by respondents for choosing not to escalate (themes, explanations, example quotes)
DPETs, Directors of Prevocational Education and Training

[The senior staff member] was touching me more regularly […] he managed to get me alone with him […] I did not escalate it because, honestly, no one would care. [Note, the full quote is not included here for risk of potential identification]

For respondents who did take action within the system in response to bullying and/or sexual harassment, most reported ineffective or personally harmful outcomes when reporting to senior colleagues.

Dismissive or blaming. Once again, many respondents described a workplace culture where they perceived bullying and/or sexual harassment behaviours to be normalised by senior staff:

DPET [Director of Prevocational Education and Training] was not helpful, dismissed my concerns.

Another junior doctor described their experience in reporting the behaviour to their DPET:

Instead of support, providing understanding, or even validation of how I was feeling, the DPET turned it back onto me and suggested […] that I am being ‘too sensitive’.

Will not trust process in future. For some respondents, their experience of the reporting process provided further stress and led to a distrust of using the reporting process again in the future:

[JMO manager and senior staff] were terrible and I will never report bullying again.

In one case, after speaking to their DPET informally about being sexually harassed, the respondent was further bullied by the perpetrator as a result:

…he told people I’d made a complaint and bullied me in other ways about it.

Overall, the process was a demoralising and discouraging experience for the respondent:

I now feel even more discouraged in making a complaint than I did before.

No effective solution offered. Despite reporting the behaviour through appropriate channels, it was common for respondents to report that senior staff, such as the DPET or Director of Medical Services (DMS), offered no solution:

…discussed with DPET and JMO manager who never followed up

…approached DPET – little feedback or response given.

Even when escalation went as far as the accreditation authority and medical indemnity insurers, respondents reported that:

…no one was interested in hearing about it, there was no support offered.

When solutions were offered by senior staff, respondents were generally dissatisfied or perceived the solutions as unsuitable:

The JMO manager, DPET & Deputy DMS were notified […] I was offered a change of hospital, which I felt was inappropriate considering I was the victim.

Logistical barriers to effective outcome. Other junior doctors described logistical barriers to effective reporting of bullying or harassment. A respondent described reporting bullying from the JMO unit manager as:

…extremely difficult, because that’s who you’re supposed to go to for issues like this, and the director of medical services had little to no time for a discussion about this matter.

Respondents described rare instances in which they felt that the response taken to bullying or harassment behaviours was effective in stopping the behaviour. Direct action (i.e. speaking directly to the perpetrator) was considered an effective response by several respondents:

I confronted issues with perpetrator himself, who apologised and backed off.

I explained to perpetrator that behaviour was inappropriate, and they stopped.

Supportive management staff were also cited in several comments as an example of an effective response strategy:

Spoke to my consultant who managed the issues with the non-medical [perpetrator].

Bullying was reported by over half the respondents, and sexual harassment by 15–19% of respondents. For most respondents who experienced bullying or sexual harassment, incidents occurred less than monthly. However, approximately 15% of respondents reported more frequent incidents (monthly, weekly, daily). Although definitions of bullying and harassment differ across studies, the exposure rates reported by the present study are in line with other recent international estimates.^8,18 Of note, sexual harassment rates in the present study were lower than those reported elsewhere (33% for students and residents pooled).¹⁸ Finally, females more often reported experiencing bullying and harassment than males, which, again, is in line with international findings from medical students through to junior doctors and trainees.^18–20

In response to a bullying or sexual harassment incident, 60% of respondents reported taking action of some kind. Escalation to a senior medical staff member was the most common response, yet most found this process either ineffective or harmful. Complaints were often dismissed or behaviours blamed on the sensitivity of the complainant, and/or no further action taken by the senior medical staff member after the complaint had been lodged. These experiences deterred some respondents from reporting incidents of bullying or harassment again in the future. Not all responses provided by senior staff were experienced as ineffective; however, cases of supportive and effective outcomes were rare.

Conversely, existing theoretical models may help explain why so many respondents (40%) in the present survey reported taking no action in relation to their experience of bullying or harassment or, indeed, went on further to explain why they felt that taking action would do more harm than good. Hollis²¹ built on the work of Kahn²² to propose that bullying in service organisations can be explained as a result of employee disengagement when resources are placed under stress in order to meet high demands for service:

…employees who must fend off harassment and bullying behaviour at work will make defending the self the priority over organisational objectives.²¹

Given that the junior doctors in the present study were on contracts of no more than 2 years and moving between terms every 10–12 weeks, this explanation seems particularly relevant.

There are limitations to the methodology used in the present study. First, the overall response rate as a percentage of possible respondents is low based on traditional conceptions of survey response. Second, survey respondents may represent a biased sample in that they were more motivated to respond to the survey to report particularly negative experiences. Third, in addition to the low response rate, it was not possible to be clear about how many junior doctors received the survey. Finally, the snowballing recruitment technique used introduces the possibility that not all respondents were junior doctors. Foreseeing the potential of these issues to weaken our study findings, we attempted to strengthen the study design in several ways. First, we argue that the innovative approach to response collection, starting with a large representative sample of junior doctors, improves the response rate. Second, the inclusion of data from across 2 years allowed for comparison of findings. Third, the additional focus on qualitative analysis into response processes and experiences enlarged the scope of the study beyond quantitative and statistical measures. Qualitative analyses of this kind are not as limited by low response rates in the same way as quantitative analyses. Finally, because the JMO census is a lengthy survey where bullying and harassment is one of many areas of focus, the sample may be more resistant to selection bias with regard to this particular topic.

We believe that the qualitative findings of the present study help explain why junior doctors often choose not to report bullying and sexual harassment. Junior doctors may feel a more compelling need to preserve their own selves over following official policy (e.g. by tolerating the behaviour until their rotation changes). The corollary of this is that a focus on interventions at the level of the junior doctor is unlikely to demonstrate an improvement in the current culture within medicine. In fact, recent calls to implement systems that improve the resilience of junior doctors in withstanding workplace stress¹¹ could be seen as unethical if not implemented as part of a broader systemic suite of interventions.

The findings suggest a need for new approaches to the problem, such as better education and training for staff who support, work with or supervise junior doctors. Such changes have already started to be implemented in Australia. External and independent programs of support have been set up for medical students and vocational trainees in some College programs and for junior doctors through the NSW Health JMO support line. But are these changes enough? The efficacy of these new initiatives needs be monitored and studied further. In addition, the role of those with governance responsibilities within healthcare organisations to more effectively address issues of this nature needs more extensive examination. Finally, the results of the present study highlight that ongoing conversations about further innovative measures to address the problem of bullying and harassment of junior doctors continue to be an important and timely priority for the medical training community.

The authors declare no competing interests.