Mental health in the veterinary profession: an individual or organisational focus?

Abstract

The veterinary profession is experiencing a shortage of veterinarians, with attrition recognised as a substantial contributor. Research has also indicated increased levels of mental ill health and alarming suicide rates in practitioners. The primary aim of this study was to investigate the impact of eleven modifiable workplace factors on mental health outcomes, job appreciation and intention to leave the veterinary profession. The second aim was to ascertain whether workplace factors influence mental health outcomes after controlling for individual resilience. An online survey was completed by 73 practising Australian veterinarians. Unfavourable workplace factors correlated with adverse outcomes including depression, stress, reduced job appreciation and increased likelihood of leaving both the role and the profession. Workplace factors remained linked with the outcomes of job appreciation, depression and stress whilst controlling for practitioner resilience. Job appreciation was a significant predictor of intention to leave both the current role and the profession. Via multiple linear regression, two categories were identified as associated with improved psychological outcomes and job appreciation. These were workplace factors that represent breaks from workload and control or decision latitude in the workplace. Whilst resilience represents a key area for intervention, workplace factors potentially represent an easier‐to‐modify area for intervention.

‘When I grow up, I want to be a vet!’ is a phrase that can be heard in classrooms around the world. Yet, in a 2018 veterinary workforce survey, 60% of Australian veterinary practices were looking for a veterinarian, and 24% of those practices had been advertising for over a year. ¹ Adding to this, the COVID‐19 pandemic has seen a 20%–30% increase in demand for veterinary services. ² In 2021, veterinarians were added to the Australian priority skilled migrant list in an attempt to ease the shortage.

In a recent Australian media report the national shortage of veterinarians was discussed, describing the profession as ‘demanding and exhausting’. ³ With seven Australian universities turning out approximately 500–550 veterinary graduates annually, the shortage appears to stem from retention problems in the profession, rather than the issue of attracting new veterinarians. The 2018 workforce survey supports this suggestion identifying that 20% of Australian veterinarians were considering leaving the profession in the following year, a figure that had increased by over 50% from the 2016 survey. ¹ Likewise, a 2020 study of UK vets identified that 43.7% of respondents were likely or very likely to leave their roles within 2 years, the three most commonly cited reasons for leaving included; negative mental impact, negative client interactions, and inadequate salaries. ⁴

High attrition rates and other negative sequelae have sparked a plethora of research into negative veterinary practitioner experiences including burnout, compassion fatigue, drug or alcohol addiction, mental ill health (depression, anxiety, psychological distress), suicide, stress, and workplace injuries. ⁵ , ⁶ , ⁷ , ⁸ , ⁹ Concerns have been raised after research identified poor psychological health in one‐third of Australian veterinarians, with younger and/or early‐career veterinarians at greater risk. ¹⁰ Similarly, negative associations between resilience and overall health, mental illness and mental health have been found in Canadian veterinarians. ¹¹

The veterinary profession has an increased rate of suicide (relative to the general population) including a fourfold suicide rate for female veterinarians, a rate that is twice other healthcare professions. ⁵ , ¹² In Australia, up to 80% of veterinarians who die by suicide have done so using readily accessed animal euthanasia drugs. ¹³ In a 2020 study, 26% of surveyed Canadian veterinarians reported having contemplated suicide in the preceding 12 months, a rate that far exceeds that seen in the general Canadian population. ¹⁴

The mental health of veterinary professionals is impacted by external (to practice management) factors, including; personality traits, gender, socioeconomic status of clientele, client interactions, performing convenience euthanasia, time since graduation, type of (veterinary) practice, unexpected outcomes, medical mistakes and responsibility for patient care. ⁶ , ⁷ , ⁹ , ¹⁰ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ However, strong associations are also found with workplace antecedents including; demands of practice, long hours, overload (volume of work), workplace pressure (time per client, number of consults, time in surgery, number of phone calls), social isolation, not enough holidays, and work/home‐life balance. ⁶ , ⁸ , ⁹ , ¹⁰ , ¹⁵ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³

Prolonged job stressors can lead to the psychological syndrome known as burnout. ²⁴ Three key elements of burnout syndrome have been acknowledged including, fatigue and exhaustion, cynicism and a feeling of ineptitude. ²⁵ Six sources of burnout have been identified: work overload, lack of control, insufficient reward (intrinsic and/or extrinsic), conflicting values, feeling at odds with colleagues (breakdown of community), and a feeling of unfairness. ²⁵ Potentially linked with vulnerability to poorer mental health, a high rate of burnout (55%) in Australian early career veterinarians has been found. ²⁶ While burnout rates have been found to decrease with time in practice, they still remain above population norms after 20 years in the profession. ²⁶ Burnout has also been recognised as a significant predictor of intention to leave the profession. ¹⁹

Veterinarians who left the profession have indicated that the decision to leave was influenced by both personal and work experiences. ¹⁵ A recent retrospective study surveyed veterinary practice managers, asking why they thought previous employees had left. Responses largely focussed on external factors, such as family, location and work‐life balance. ⁴ By contrast veterinarians answering the same survey identified poor management as a key reason for leaving/intending to leave with ‘the team’ identified as a key reason to stay in the practice. This employee/employer ‘disconnect’, was recognised as a possible failure of management to understand the current workforce. ⁴ Medical professions identify similar importance in adapting working conditions to suit the needs and values of their employees, acknowledging that there are generational differences in these needs and values. ²⁷ Research examining work attitudes of Generation Y employees suggest they may have greater expectations for work‐life balance and freedom of work values with incongruency between individual and organisation values increasing intentions to leave. ²⁸ Thus, veterinary employees, and particularly younger graduates', concerns may not be being heard, resulting in practitioners choosing to leave the profession rather than advocate for organisational change.

Despite the aforementioned findings, discussion and recommendations drawn from extant research often refer to building individual resilience, providing coping skills, training (to deal with work‐related distress, anxiety and depression), breaking down barriers to seeking help, and learning anti‐stress skills. ⁸ , ⁹ , ¹⁰ , ²⁰ , ²⁶ A paradoxical focus on individual solutions despite evidence of organisational and situational factors having a significant and substantial influence, is consistent with the broader literature around workplace burnout. ²⁹ This scenario has been likened to asking the blackened coughing canaries coming out of the coalmine what they had done wrong to become so unwell. ³⁰ After identifying this individually focussed response in a multitude of workplaces and professions, researchers suggest this cognitive bias likely favours the simplicity of individual change over organisational or even cultural change. ²⁹

Job appreciation for Dutch veterinarians correlated significantly with the physical and emotional workload, aspects of practice management and economic concerns. ³¹ In a 2017 study, German practitioners were found to be less satisfied with their work and other facets of life than similar subgroups of the general population. ³² , ³³ Amongst these German veterinary practitioners; those working long hours and paid low incomes were found to experience less work satisfaction, whilst employees (i.e., not sole traders/owners) experienced less life satisfaction. ³² Importantly, satisfaction with time worked was recognised as having a direct influence on veterinarians' life satisfaction. ³³

In a 2007 survey of Australian small animal practices, 55% of veterinarians were found to be working over 40 h per week, with 94% working more than their rostered hours and 63% taking no sick days per year. ²² The same study identified that increased working hours, staying back late, total number of consultations, surgical hours, number of client phone calls and Outside of Office Hours calls were all significantly correlated to workplace stress, poor psychological health and decreased job satisfaction. More than 50% of veterinarians interviewed in 2012, who had experienced suicidal acts or thoughts, identified workplace concerns including hours and volume of work as having an impact. ⁹

Our study proposes that workplace factors that contribute to workload represent modifiable workplace norms that are potentially contributing to impaired wellbeing and driving veterinary staff out of the workforce. Based on the aforementioned literature, eleven easily modifiable workplace conditions specific to the veterinary profession were identified. These included average total weekly working hours, ²⁰ , ²¹ , ²² , ³³ consultations per day, ⁹ , ²⁰ , ²² , ²³ daily hours in surgery, ²⁰ , ²² working beyond rostered hours, ⁶ , ²⁰ , ²² , ²³ taking regular breaks, ⁹ , ²⁰ sacrificed lunch breaks, ⁹ , ¹⁵ , ²⁰ allocated time to catch up, ²⁰ , ²³ walk‐in appointments, ²² ability to take holidays, ²⁰ ability to take sick leave, ²² and control over the structure of the work day. ⁶

From the eleven identified workplace antecedents this study aimed to assess those that associate most strongly with depression, anxiety, stress and job appreciation. Additionally, the study investigated practitioner resilience and what role it may play in the association between workplace factors and mental ill‐health and/or job appreciation. Finally, the study examined the predictive influence of the eleven workplace factors and job appreciation in practitioners' intention to leave their role and the profession.

Three hypotheses were proposed:

Hypothesis 1

Unfavourable workplace factors will be positively associated with depression, anxiety, and stress and negatively associated with job appreciation.

Hypothesis 2

Unfavourable workplace factors will remain associated with depression, anxiety, stress and job appreciation when practitioner resilience is controlled for.

Hypothesis 3

Workplace factors and job appreciation will predict a practitioner's intention to leave.

Materials and methods

Participants

One hundred and five veterinarians currently practising in Australia initially responded to the survey with a final sample of 73 responses informing the current study once incomplete responses were removed. Respondents were primarily female (76.7%, n = 56; 20.5%, n = 15 male and 2.7%, n = 2 non‐binary/prefer not to say), ages ranged from 24 to 61 years (M = 36.96 years, SD = 9.27) and the majority worked in small animal practice (69.9%, n = 51) with 15.1% (n = 11) indicating ‘mixed practice’ and ‘other speciality fields’ respectively. Length of time in the profession ranged from <1 to 37 years (M = 11.74, SD = 10.11), 74.0% were employees (n = 54) and 26.0% (n = 19) were business owners or locum/contract workers.

Measures

Consisting of 76 items, the survey contained personal (age, gender, years in practice) and work‐related (practice type, salary, employer/self‐employed or employee status) demographic questions, items regarding the current experience of the 11 workplace factors and three validated psychometric scales measuring resilience, depression/anxiety and job appreciation. Participants were also asked about the likelihood of leaving their current role and/or the profession (five‐point Likert scale from Very Unlikely to Very Likely) if the profession had met their expectations and if they would recommend the profession to others (Yes/No answers). Answering the survey took less than 15 min.

Workplace factors

Possible workplace factors were identified via a search of the literature, of these 11 were selected as being the most modifiable within a veterinary practice workplace drawing on the experience of the first author (a practising veterinarian for 17 years). Four questions asked for numeric responses including average total weekly working hours, average consultations per day, average hours spent daily in surgery and average number of walk‐in appointments per day. Seven questions utilised a 5‐point Likert scale response format (e.g., 1 = Almost always through to 5 = Almost never) and included working beyond rostered hours, taking breaks from work, working through lunch breaks, having time allocated to catch up, the ability to take holiday leave, the ability to take sick leave and having control over daily work structure. To calculate a total workplace factor score, numeric responses were converted into a score 1–5 (using SPSS visual binning with 4 cuts, binning was done to allow summation with factors utilising a Likert response format) and questions with negative wording were reverse scored prior to summation. Higher total (possible range 11–55) scores indicate workplaces with greater workload pressure.

Resilience

An individual's resilience was measured via the Connor‐Davidson Brief Resilience Scale (CD‐RISC‐10). The 10‐item scale has demonstrated excellent psychometric properties and high internal consistency (α = 0.85), ³⁴ with a similar score achieved in the current study (α = 0.86). Participants were asked how each item had related to them during the previous month. For example, ‘under pressure I stay focussed’, with Likert‐type responses (0 = Not true at all, through to 4 = True nearly all the time). Total scores range from 0–40 with higher scores indicating greater resilience.

Depression, anxiety and stress

Practitioner depression, anxiety and stress were measured via the short form of the Depression Anxiety and Stress Scale (DASS). Participants were asked how each statement has applied to them during the preceding week, for example, ‘I found myself getting agitated’. The scale scored 0–3 points per question (0 = Never to 3 = Almost always), higher scores indicate greater levels of depression, anxiety and stress. DASS is a widely used 21‐item scale (7 items per three subscales) that has demonstrated good internal consistency and concurrent validity (depression α = 0.97, anxiety α = 0.92, stress α = 0.95). ³⁵ Reliability for our study was slightly lower but still acceptable, α = 0.924 for depression, α = 0.799 for anxiety and α = 0.870 for stress.

Job appreciation

Job Appreciation (JA) was assessed through a 17‐item measure developed for use in the veterinary context with good reliability (α = 0.86). ³¹ Each statement is answered via a Likert scale (1 = Totally disagree to 5 = Totally agree) for example, ‘I am proud of this practice’. Two statements regarding a practitioner's experiences during pregnancy included a ‘not applicable’ option. An average score is calculated with higher scores indicating greater job appreciation. Internal consistency in our study was high, α = 0.92 and α = 0.90, for those with and without pregnancy experience respectively.

Likelihood of leaving current role or profession

To provide an insight into an individual's thoughts on leaving their current role and the profession, participants were asked to rate how likely they were to leave their current role in the next 12 months, and the profession in the next 12 months or 5 years (1 = Very likely to 5 = Very unlikely). Scores were subsequently reversed so that higher scores indicated a greater likelihood of leaving. Scores from the two ratings of likelihood of leaving the profession (12 months and 5 years) were summed with possible scores ranging from 2–10, again higher scores indicating an increased likelihood of leaving the profession.

Procedure

The study used an online cross‐sectional survey, with responders remaining anonymous and non‐identifiable. Human ethics approval was given by Central Queensland University (2021‐045). Participants were recruited via Australian veterinary social media groups and snowball sampling via email to professional networks. Participants had to be currently registered to practice in Australia. Recruitment occurred between 7 July and 8 April 2021.

Statistical analysis

Using SPSS software, the raw data was analysed, with missing data identified as well as one outlier (working hours) that was corrected via Windsorizing and replaced with the nearest non‐outlying value. ³⁷ Other substantial variations in the sample were not considered to be outliers as they represent real‐world differences in the range of practice types surveyed rather than outliers (e.g., time in surgery and walk‐in appointments). Data were graphed to check for normality, and some non‐normal results were identified (e.g., depression, anxiety, likelihood of leaving role). However, the central limit theorem was applied due to a sample size of over 30. ³⁷ Reliability and validity of psychometric scales were checked. To facilitate summing across continuous (e.g., average hours) and discrete variables continuous scores were sorted into low, medium and high levels via visual binning with two cuts. Similarly, the total workplace factor score was binned into low, moderate and high categories to allow MANCOVA analysis (Hypothesis 2). For all testing, significance was recognised as P = <0.05 and where relevant Bonferroni correction was implemented to help control elevated error rates inherent in repeated testing. Correlation coefficients were calculated using Spearman's test for categorical responses (e.g., gender, type of practice) and Pearson's test for continuous and ordinal variables.

Results

Workplace factors

When considering unfavourable workplace factors, 45.2% (n = 33) of respondents reported almost always or often working back late beyond a rostered shift, with 61.6% (n = 45) sometimes or almost never taking regular breaks from their work. Ratings on the workload factor showed that 52.1% (n = 38) of vets sometimes to almost always lose their lunch/mid‐shift break, 69.8% (n = 51) had sometimes or almost never time allocated to catch up and 61.7% (n = 45) of veterinarians felt that they were given little to no control over the structure of their workday. There were 27.4% (n = 20) of practitioners who felt it would be unlikely they could take a holiday in the next 12 months, whilst 24.6% (n = 18) felt they would be unlikely to be able to take sick leave at short notice. Responses to the likelihood of leaving revealed that 38.4% (n = 28) of veterinarians felt they were likely or very likely to leave their current role in the next 12 months and 30.1% (n = 22) were likely or very likely to leave the profession in the next 5 years. Importantly, 41.1% (n = 30) of veterinarians felt that the profession had not lived up to their expectations and 49.3% (n = 36) would not recommend the profession to others.

Significant correlations between the eleven unfavourable workplace factors were also identified and are presented in Table 1. As can be seen in the table strong positive correlations were found between factors involving staying back late, ability to take regular breaks, sacrificed lunch breaks and control over the daily structure.

TABLE 1.

Correlation matrix for workplace factors

	1	2	3	4	5	6	7	8	9	10	11
1. Total working hours	1
2. Number of consultations	−0.074	1
3. Time in surgery	0.198	0.131	1
4. Staying back late	−0.191	−0.234*	−0.194	1
5. Taking regular breaks	0.292*	0.134	0.254*		1
6. Sacrificed lunch breaks	−0.249*	−0.096	−0.150	0.600**	−0.587**	1
7. Allocated catch‐up time	0.259*	0.096	0.182	−0.187	0.392**	−0.402**	1
8. Number of walk‐in appointments	0.231*	0.016	0.009	−0.069	0.226	−0.154	0.209	1
9. Ability to take holidays	0.424**	0.108	0.115	−0.319**	0.344**	−0.245*	0.199	0.199	1
10. Ability to take sick leave	0.133	0.355**	0.020	−0.282*	0.237*	−0.183	0.299*	0.103	0.376**	1
11. Control over daily structure	0.145	0.381**	0.195	−0.451**	0.274*	−0.374**	0.336**	0.084	0.206	0.278*	1

^*P < 0.05;

^**P < 0.001.

Workplace factors and mental health

The mean score for practitioner resilience was 27.04 (SD = 6.29), lower than the reported general population average of 32.1. ³⁸ Mean response stress scores were 18.19 (SD = 8.68), consistent with extremely severe stress. Depression scores indicated a mean of 12.22 (SD = 9.88), a score consistent with severe depression, whilst anxiety scores resulted in a mean of 8.99 (SD = 7.58), consistent with severe anxiety. ³⁶

Significant correlations were identified between nine of the eleven workplace factors and depression, anxiety, stress and job appreciation as can be seen in Table 2. The strongest significant correlations were noted for staying back late, taking regular breaks, sacrificing lunch breaks, allocating catch‐up time and control over daily structure. Despite finding no significant correlations between total working hours or walk‐in appointments with the psychometric scales these factors were still included in the total workplace factor score.

TABLE 2.

Descriptive statistics and correlations for workplace factors

	Mean	SE	Depression	Anxiety	Stress	Job appreciation
1. Total working hours	40.72	1.35	0.062	0.108	0.107	−0.056
2. Number of consultations	14.15	0.957	0.240*	0.012	0.036	−0.283*
3. Time in surgery	3.55	0.731	0.240*	0.117	0.274*	−0.093
4. Staying back late	2.71	0.165	−0.187	−0.158	−0.332**
5. Taking regular breaks	3.55	0.178	0.134	0.211	0.241*	−0.301**
6. Sacrificed lunch breaks	2.73	0.166	−0.153	−0.220	−0.423**	0.259*
7. Allocated catch‐up time	3.77	0.157	0.343**	0.071	0.339**	−0.411**
8. Number of walk‐in appointments	3.69	0.791	0.032	−0.071	0.090	0.024
9. Ability to take holidays	2.30	0.162	0.246*	0.032	0.182	−0.199
10. Ability to take sick leave	2.23	0.168	0.137	0.129	0.019	−0.400**
11. Control over daily structure	3.67	0.161	0.361**	0.237*	0.443**	−0.467**

n = 73.

^*P < 0.05;

^**P < 0.00.

Correlations between demographic details and psychometric scales are presented in Table 3. Significant positive correlations were identified between self‐employment and job appreciation and resilience along with a significant negative correlation with DASS scores. Resilience did not correlate with age or time in practice. A significant negative correlation was also seen between anxiety and age or years in practice along with a negative correlation between age and stress. The only correlation identified with annual salary was a significant positive correlation with resilience scores.

TABLE 3.

Correlations between demographics and psychological scales

	Depression	Anxiety	Stress	JA	Resilience
Age	−0.176	−0.318**	−0.234*	0.155	0.100
Years in practice	−0.155	−0.331**	−0.202	0.177	0.119
Gender b	0.000	0.103	0.182	−0.057	‐
Employer (2)/Employee (1) a , b	−0.367**	−0.354**	−0.357**	‐	0.347**
Salary	−0.150	−0.218	−0.126	−0.047	0.276*
Type of practices b	−0.043	−0.024	−0.029	0.153	0.139

^a Variable coded as 1 (employee) or 2 (employer/locum).

^b Spearman correlation.

^*P < 0.05;

^**P < 0.001.

When looking at the correlation between demographics, workplace factor scores and the likelihood of leaving a role or profession a significant negative correlation was found between age and years in the profession with the likelihood of leaving a role (r = −0.236, r = −0.292, P < 0.05 respectively). Significant negative correlations were also found between self‐employment and the likelihood of leaving the role and the profession (rs = −0.477, rs = −0.503, P < 0.001 respectively). Further investigation found a significant negative correlation between self‐employment and workplace factor (11) not having control over daily structure (rs = −0.455, P < 0.001). This workplace factor also had a strong significant correlation with likelihood of leaving one's role and profession (rs = 0.440, P < 0.001; rs = 0.319, P < 0.01).

Presented in Table 4 are the correlations between individual workplace factors and the psychometric scales. A very strong positive correlation was identified between job appreciation and likelihood of leaving both the current role and the profession. Resilience showed a stronger association with likelihood of leaving the profession than likelihood of leaving the role.

TABLE 4.

Correlation matrix for scale scores

	Workplace factor score	Depression	Anxiety	Stress	Job appreciation	Resilience	Likelihood of leaving role	Likelihood of leaving profession
Workplace factor score	1
Depression	0.311**	1
Anxiety	0.177	0.580**	1
Stress	0.362**	0.622**	0.628**	1
Job appreciation	−0.482**	−0.504**	−0.286*		1
Resilience	−0.133	−0.490**	−0.476**	−0.427**	0.410**	1
Likelihood of leaving role	0.423**	0.386**	0.156	0.324**	−0.750**	−0.262*	1
Likelihood of leaving profession	0.290*	0.409**	0.351**	0.390**	−0.579**	−0.457**	0.709**	1

^*P < 0.05;

^**P < 0.001.

Controlling for resilience

A one‐way MANCOVA was conducted on four dependent variables; depression, anxiety, stress and JA after controlling for practitioner resilience. Homogeneity of variance–covariance matrices was assumed (Box's M = 29.50 F[20,15282.72] = 1.35, P = 0.136). Using Wilks' lambda, the combined DVs were significantly different by workplace factor levels (Wilk's Λ = 0.722, F[8,132] = 2.92, P = 0.005, partial η2 = 0.150), after controlling for resilience. No significant interaction was found (Wilk's Λ = 0.95, F[8,128] = 0.41, P = 0.91, partial η2 = 0.03). A significant Levene's test was recognised for depression (F[2,70] = 10.36, P < 0.001), suggesting that results for depression should be interpreted with some caution. A univariate ANOVA was used to investigate the impact of workplace factor level on the individual psychometric tests. The main effects for workplace factor level were significant on JA (F[2,69] = 8.73, P < 0.001, partial η2 = 0.20), depression (F[2,69] = 4.71, P = 0.012, partial η2 = 0.12) and stress (F[2,69] = 4.18, P = 0.019, partial η2 = 0.11). Pair‐wise comparisons with a Bonferroni adjustment indicated a significant difference between low and moderate workplace factor levels for depression and stress, whilst for job appreciation, a significant difference was seen between low and moderate, as well as low and high workplace factor levels. This results in an estimated marginal means pattern demonstrating a significant increase in depression and stress between low and moderate groups, then a plateau from moderate to high groups. Similarly, a significant decrease in job appreciation from low to moderate groups was observed with a smaller relative decline seen from moderate to high groups.

Influences on intention to leave

Separate multiple linear regression equations were used to predict the likelihood of leaving one's current role or profession based on workplace factors and job appreciation scores. The predictor variables included workplace factors and job appreciation scores and the dependent variables likelihood of leaving one's role and likelihood of leaving the profession. Assumption testing demonstrated appropriateness for regression calculation, Durban Watson = 1.92 and 2.16 respectively, VIF = 1.30, Cook's distance all <1, Mahalanobis scores <15.2.

For the likelihood of leaving one's role, a significant regression equation was found (F[2,70] = 45.86, P < 0.001) with 57% of total variance accounted for (R² = 0.57). Of all the variables loaded on Job Appreciation proved to be a significant predictor (β = −0.71, P < 0.001). A significant regression equation was found to predict the likelihood of leaving the profession (F[2,70] = 17.65, P < 0.001) with 34% of total variance accounted for (R² = 0.34). As for intent to leave their current role only Job Appreciation loaded as a significant predictor (β = −0.57, P < 0.001) of intent to leave the profession.

Discussion

The first hypothesis in the current study, that unfavourable workplace factors would correlate positively with depression, anxiety, and stress and negatively with job appreciation was supported for nine of the eleven workplace factors examined. The second hypothesis, that unfavourable workplace factors would still provide an adverse influence whilst controlling for individual resilience was supported for job appreciation, stress and depression. The final hypothesis was also partially supported with job appreciation being a strong predictor for both the intent to leave their current role and the profession.

Contrary to the findings in other studies, ⁸ , ²⁰ , ²³ total working hours and the number of walk‐in consultations proved to have no significant correlation with wellbeing potentially linking with findings that the impact of the actual work was over and above the total hours worked for our study participants. ²² Two themes were identified when correlating workplace factors with mental health outcomes and job appreciation. The first theme being strong and significant correlations with factors that represent downtime from work including not taking regular breaks or lunch breaks, having no allocated catch‐up time and not leaving work on time. In addition, all these factors correlated significantly with each other suggesting that workplaces with unfavourable conditions were more likely to have multiple unfavourable conditions and that these had a cumulative effect on the individual. Little extant research exists in this space. However, a 2009 survey of Queensland veterinarians reported ‘not taking rest breaks’ as a source of extreme stress. ²⁰ The study also recognised not enough holidays, time per patient, seeing too many patients and long working hours as stressors. Associations between work stress and staying back late, number of consultations and surgeries have also been identified with researchers recommending practices allocate 30 min of catch‐up time at the end of each day, limit surgical loads, book complex cases earlier in the day, extend consultations from 15 to 20 min and ensure complex cases get longer time slots. ²² Interestingly, while the Veterinary Services Award (at the time of writing) states that all veterinary support staff must have a minimum 30 min meal break between 4 and 5 h of work as well as 10 min paid rest break after every 4 h, there are no minimum requirements for employee veterinarians. ³⁹ In conjunction with extant research the findings here suggest that veterinary workplaces should apply these standards to all staff. Whilst not all practices will be able to control all the identified workplace factors, striving to reduce as many as possible is likely to be of benefit. What the present study does highlight is that little has changed in the decade since the 2009 Queensland study ²⁰ and furthermore, that a level of focus needs to be placed upon the workplace in order to target both veterinary wellness and attrition.

The second theme, control, emerged from the relationship between depression, anxiety, stress and job appreciation. Lack of control is one of the six contributors to burnout, the absence of autonomy and decision making resulting in an inability to shape the workplace to match one's professional values. ²⁵ High decision latitude has been extensively researched in the organisational space, particularly as part of the Job Demand‐Control‐Support (DCS) model of work strain. ⁴⁰ The model proposes that employees working in demanding roles experience greater stress if they cannot decide how they will complete their work. Within the model control acts as a buffer, improving well‐being, job satisfaction and reducing illness. The addition of control can even improve staff engagement, a drive to try new things and build their skills. ⁴⁰ Within the veterinary field, decision latitude has been shown to be protective of mental health and influence how work‐related stressors contribute to veterinary suicide risk. ⁶ , ⁴¹ Relatedly employee veterinarians have been found to be more stressed than practice principals and partners whilst greater job appreciation has been recorded in Australian veterinarians/employers. ²⁶ , ³¹ The present study supports these findings with self‐employment associated with reduced depression, anxiety and stress and increased job appreciation. Importantly, as self‐employment has no association with workplace factor scores, the results support the DCS model with workplace control acting as a buffer rather than conscious management of boundaries resulting in improved outcomes.

Prior research has proposed that the association seen in the veterinary profession between age and mental health is due to improved resilience with time. ¹⁰ Whilst the present study did find a correlation between greater mental health and age (excluding depression) there was no association between age and resilience. These findings add additional support to the suggestion that control may be the buffer protecting mental health in older physicians. The authors propose that whilst employers express frustration at the movement of veterinary staff to locum and contract roles, this move may be acting as a protective factor for those staff. Improving decision latitude in the workplace could serve to reduce the movement of veterinarians towards locum roles whilst improving both wellbeing and attrition and meeting the freedom of work values that Generation Y desire. ²⁷ Similar conclusions have been reached in UK‐based research where increased participation in decision making was linked with improved mental health in veterinarians. ⁴² Future research examining differences between practice owners and locums may be valuable, particularly potential effects of self‐employment versus practice management and how this relates to the buffering aspect of decision latitude. Importantly workplace control may encompass more than one single factor and further research into how and what this control should look like is needed.

Resilience has been identified as a mediator of the response to stressors and advocated as a key area for intervention in veterinary practice. ¹¹ While this is supported here through an association with positive psychological outcomes, unfavourable workplace factors alone continued to correlate with scores for depression, stress and job appreciation. This suggests that even the most resilient practitioner will be negatively influenced by an unfavourable workplace. Further, these findings support the proposition that alongside promoting resilience to practitioners and undergraduates, teaching the importance of healthy workplaces would provide considerable benefit. Workplace factors are fundamentally more modifiable than developing personal resilience and their importance can be communicated succinctly.

In summary, both perceived workplace control and factors that prevent downtime from work have the strongest influence on mental wellbeing and job appreciation. Workplace factors continue to influence wellbeing and job appreciation despite individual practitioner resilience with job appreciation in turn being a strong predictor of practitioner intention to leave. The current findings suggest that the profession risks losing practitioners who have little appreciation for their current role. Given the ease of use, it may be worth including The Job Appreciation ³¹ scale in business evaluation and feedback exercises to track this important aspect of employee experience. Importantly, the findings of this study suggest that instead of a sole focus on the individual, simultaneous focus should be placed on the unfavourable pressures or expectations practices place on their staff and those that practitioners place on themselves.

Limitations of this study include the relatively small, younger, sample and use of a self‐report survey with the potential for response bias. The results represent a cross‐section of the profession during a complex period of time that included a global pandemic that impacted workplaces and workload. Potentially because social media channels were used for appealing to subjects, a non‐representative and younger sample was recruited. However, with younger and new‐to‐profession veterinarians more likely to leave, the current sample captures a valuable target in the profession. Veterinary staff often voice their concerns that poor remuneration is a cause for attrition, despite many studies (including the present one) showing little to no association. It would be valuable to investigate employee awareness regarding the impact of workplace factors, or if practitioners incorrectly attribute other factors as the cause of their distress. If awareness does exist, further direction for study may include investigating the factors that influence a practitioner's ability to self‐advocate in the workplace.

Conclusion

Mental wellbeing and attrition continue to be an area of concern for the veterinary profession. Unfavourable workplace factors are associated with reduced mental wellbeing and a lower appreciation for the job. Lower job appreciation is, in turn, a direct predictor of leaving the job role and leaving the profession. While resilience does correlate with improved outcomes, when resilience is accounted for statistically, unfavourable workplace factors continue to associate with poor outcomes. This indicates that workplace interventions targeted solely at individual resilience may not be effective alone. Modifiable workplace factors potentially represent more easily targeted aspects of a veterinarian's workplace than resiliency training/interventions. Veterinary practitioners also need to be aware of the cognitive bias that results in placing blame on the individual (the blackened canary) and look to the workplace to adapt to these modifiable factors that will improve attrition and mental wellbeing. As employers and employees, veterinarians need to collectively ask the profession to do better. Focus needs to be on ways to improve at the workplace level, and not consistently expected to be shouldered by individuals.

Conflicts of interest and sources of funding

The authors declare no conflicts of interest or sources of funding for the work presented here.

Hilton, KR. , Burke, KJ. and Signal, T. , Mental health in the veterinary profession: an individual or organisational focus? Aust Vet J. 2023;101:41–48. 10.1111/avj.13215