Edinburgh Research Explorer An exploration of the factors influencing career choice in mental health

Aims and objectives: To identify the factors that are associated with considering a career in mental health. Background: The mental health specialty is facing a recruitment crisis in the United Kingdom but there is limited evidence about which factors encourage and discourage people from considering a career in mental health. Design: Quantitative, observational, online survey using a multiple ordinal logistic regression model to identify if there were any significant predictors of the extent to which participants would consider a career in mental health. The design and write up of the study were guided by the STROBE checklist. Method: We gathered the views of 231 participants (female = 188, 81.7%)


| INTRODUC TI ON
There is anecdotal evidence about the different ways in which NHS organisations have attempted to improve staff recruitment, including introducing values-based recruitment and interviewing (NHS Employers, 2014, 2015, offering structured work experience to young people (Sheffield Teaching Hospitals NHS Foundation Trust, 2019), targeting recruitment at particular under-represented groups (NHS Employers, 2016), and using internet media, such as YouTube to promote NHS routes and career opportunities for young people (Health Education England, n/d).
There is, however, very limited empirical evidence, both about the factors that influence recruitment to mental health services, and of the effectiveness of interventions to increase recruitment. Barriball et al. (2015) conducted a comprehensive review which focused on the recruitment and retention of the health workforce in Europe. The authors reviewed published literature and practice examples from countries across the world and identified four main approaches that have been used: changes in the way education is provided, for example the use of more practice-based coaching and mentoring; using monetary incentives; implementing structural changes to provide greater support to staff, such as child friendly policies; and changes in the regulatory framework that impact on areas such as management and employment arrangements. The authors concluded, however, that: "The primary studies which we reviewed, in which one or more interventions were directly observed do not include reports of systematic evaluations and none which is able to conclude to a relation of causality between an intervention and some observed effects" (p88).
A later review by McKenzie et al. (2017) of literature that was relevant to the recruitment and retention of staff working in support services for people with a learning disability and/or autism identified a number of factors as being likely to influence recruitment into the caring professions more widely. These included rewards, such as pay; relationships factors, such as supervision; organisational factors, such as values and culture; and features of the role, such as impact on work-life balance. The authors noted, however, that there has been limited research into whether interventions which address these factors actually influence recruitment.
Of the limited intervention research available that is directly relevant to mental health, Sutton, Patrick, Maybery, and Eaton (2016) report on an evaluation of a five-day "vacation school" which offered allied health and nursing students experience of employment and career opportunities in a rural area in Australia. Significant, positive changes in student attitudes towards living and working in the rural area were found, with the latter being sustained at 6-month follow-up (Sutton, Maybery, & Patrick, 2015).

Relevance to clinical practice:
The results can inform more targeted recruitment strategies and development of suitable career pathways for those interested in a career in mental health.

K E Y W O R D S
facilitators, mental health, nursing: staff: barriers, recruitment What does this paper contribute to the wider global clinical community?
• Many countries experience difficulties with attracting staff to mental health services, but there is limited research into the factors that influence recruitment.
• Our results suggest that students and graduates of psychology and social science degrees may be an important recruitment pool.
• Recruitment messages need to provide clear and targeted messages that provide key information about the profession, highlight the helping nature and effectively address concerns about the perceived negative impact on staff well-being.
• Opportunities that offer alternative routes into mental health careers need to be of short enough duration and financially viable in order to attract applicants.
In another Australian study, Foreman, McMillan, and Wheeler (2017) evaluated the impact of a scholarship, which paid 50% of programme costs to obtain a mental health practice qualification. The original numbers were small, with 19 students being offered a scholarship, of whom 17 (89%) commenced study. All students completed the study baseline questionnaire, however, this number dropped to seven and two, respectively, at the 12-and 24-month timepoints. The completing students reported a commitment to working in the field of mental health, while studying and after graduating, however, no longer-term follow-up data was available to determine if these commitments were realised.
Figures from the UK also suggest that financial considerations are influencing recruitment to health programmes, with applicants and entrants to nursing, midwifery and allied health profession preregistration programmes in England having fallen two years in a row since the bursary for such students was replaced by student loans in 2017 (Buchan, Charlesworth, Gershlick, & Seccombe, 2019). By contrast, in Scotland and Wales, where the bursary was retained, numbers have increased. In Scotland, where there has been a focus on recruiting to mental health and learning disability nursing programmes, there was a further increase in student intake for 2019/2020.
The lack of an evidence base about the factors influencing recruitment to mental health services and the effectiveness of recruitment interventions represents the first challenge to increasing the mental health workforce, as it becomes difficult to draw robust conclusions about the extent to which past and current recruitment strategies have been successful or otherwise. The second challenge is where to recruit from in order to meet the identified targets (Health Education England, 2017). One large workforce pool from which mental health staff could be recruited, is students and graduates of psychology and social studies degrees, as research suggests that these people may constitute a large, well-educated group with interests that align with caring professions such as nursing (Durcan, Stubbs, Appleton, & Bell, 2017;The Psychological Professions Network, 2018).
The lack of a robust evidence base about what attracts individuals to, and discourages them from, undertaking a career in mental health, however, makes it difficult for NHS organisations to identify the most effective recruitment strategies to increase the mental health workforce in their particular area. The present study, therefore, aimed to identify the factors that were facilitators and barriers to taking up a career in mental health, with a particular focus on students/graduates from psychology/social studies degrees.

| Design
The first author was granted ethical approval for the research on behalf of the research team by Northumbria University Department of Psychology Ethics committee (REf: 15326). All participants provided consent for their data to be used in the study. The study used an observational design, with information from participants being obtained from an online survey. The design and write up of the study were guided by the "Strengthening the Reporting of Observational Studies in Epidemiology" (STROBE) checklist (see supplementary File S1).

| Sampling
The study primarily adopted a purposive sampling technique, with the aim of recruiting groups who had been identified in previous literature as being likely to have an interest in a career in mental health (Durcan et al., 2017;The Psychological Professions Network, 2018).
These included those studying on, or graduates of, a psychology or social studies degree, those studying psychology in 6th form at school or college and those who currently worked or had previously worked in a caring profession, who were known to the research team. In all cases, communication about the study was initially made via email, which contained information about, and a link to, the online survey. Emails were sent to 31 universities and 39 colleges/schools in the North of England. Project information and survey link were also posted on the psychology Facebook groups and LinkedIn group at the researchers' University. Of the schools/colleges approached, five responded to say they did not participate in research. It is unknown how many participants took part from each of the remaining organisations, as this information was not collected to protect the anonymity of respondents. The sample size was determined based on the time and resource constraints of the study.

| Participants
Participants were students studying on or graduates of relevant programmes (e.g. psychology), individuals currently working in or who had previously worked in care-related work and young adults at school/college who were studying a relevant topic (e.g. psychology). Inclusion criteria were that the participant was aged 18 years or older, or in the case of school and college students, aged 15-17 years and with parental consent to participate. In total, 231 participated. The participant characteristics are provided in the results section.

| Procedure
Potential participants and their parents (where relevant) were given information about the study and a link to the online survey, which provided more detailed information and a button to confirm consent. Participants provided some demographic information and were asked to complete the online questionnaire. This was developed for the study, as no relevant pre-existing measure of the areas of inter- Participants were asked to identify the extent to which they would consider undertaking a career in mental health (rated on a five-point scale from 5 = "definitely yes" to 1 = "definitely no"), which (if any) professional roles in mental health services they would be interested in (with the options of "no," "possibly," and "yes") and why, as well as their knowledge about these roles and about a career in mental health in general (rated on a five-point scale, from 1 = "nothing at all" to 5 = "a great deal").
They were then asked to identify the factors that would encourage and discourage them from considering a career in mental health; the extent to which they agreed that information about a range of factors, such as good pay, promotion prospects and supervision would help them decide whether to take up a career in mental health (rated on a seven point scale from 1 = "completely disagree" to 7 = "completely agree"); the importance of a range of factors (such as low work-related stress and good pay) to them in their job (rated on a five-point scale from 1 = "very unimportant" to 5 = "very important"); and the extent to which they thought a career in mental health would offer each one (rated on a five-point scale from 1 = "not at all" to 5 = "completely").
Finally, they were asked in an open text question to identify ideas they felt would improve recruitment to mental health services.
These responses were categorised according to common themes.
Responses were anonymous. The project had a focus on specific roles within the NHS, which were identified as being a priority for services in the North of England, including mental health nursing and roles open to nurses such as High Intensity Psychological Therapist.

| Statistical procedure
In order to identify if there were any significant predictors of the extent to which people would consider a career in mental health, a multiple ordinal logistic regression model was fit. These models can be used to predict an ordered categorical outcome from several predictors, mutually adjusting for the effects of other predictors in the model. Like multinomial logistic regression, it relies on a logistic link function; however, unlike multinomial logistic regression, it models the ordering of the categories for the outcome variable and thus uses relevant information about the ordinal nature of the variable.
Odds ratios derived from the model represent the odds of transitioning between adjacent categories (e.g., from category 1 to 2,3 or category 1,2 to 3). It is assumed that the coefficient that describes the relations between each pair of outcome variable levels is the same, meaning that a single coefficient can be used to describe the effect for each predictor. This is the proportional odds assumption and is useful for providing parsimony in otherwise complex models.
Using this model, we examined the effect of the above-described demographic predictors as well as experience and knowledge of mental health careers on the self-reported likelihood of choosing a career in mental health. All relevant predictors were included in our model based on our a priori hypotheses regarding which variables would predict self-reported likelihood of choosing a career in mental health. Nonsignificant predictors were not trimmed from the model but were retained to ensure that all predictors were mutually adjusted for, in order that we could identify the effects of each predictor over and above the effects of the others.

| Participant characteristics
In total, 231 people participated. Table 1 provides a detailed breakdown of participant characteristics and of the information included as predictors of considering a career in mental health. The majority of participants were female, white and from the UK. The ages ranged from 16-65 (M = 22.7, SD = 8.9). Most were students at university or college, 159 of whom were studying psychology at some level. All but 3 (who were retired, unemployed and working as a volunteer respectively), of the remaining participants were in full or part-time employment. Thirtyeight participants (16.5%) reported having a mental health difficulty.

| Undertaking a career in mental health
The majority of participants reported that they would definitely (n = 101, 44.1%) or probably (n = 62, 27.1%) consider undertaking a career in mental health. Only six (2.6%) reported that they would definitely not consider a career in mental health.

| Predictors of the extent to which individuals were considering a career in mental health
Several significant predictors were found (see Tables 1 and 2). The following were associated with a significantly greater likelihood of considering a career in mental health: being female, having a mental health condition and greater knowledge of mental health. Having had experience of working with people with mental health difficulties was significantly negatively associated with likelihood of considering a career in mental health. As compared to university students, school students were less likely to consider a career in mental health; however, this likely reflects the predominance of psychology students in the sample. Age, qualification and occupational group were not significant predictors.

| Specific mental health professions
Participants were asked to indicate whether they would consider undertaking a career in specific roles within mental health and their knowledge about the particular role (see Table 3), and about careers in mental health overall. Over half of participants (51.4%) reported that they would consider a career as a mental health nurse and a similar percentage (55.2%) reported they would consider a career as a High Intensity Psychological Therapist, a profession that requires an existing professional qualification, such as mental health nursing.
Clinical psychology was the profession that the highest number of participants considered as a career choice. The mean score for overall knowledge about careers in mental health was 2.9 (SD = 0.91), indicating only a moderate level of knowledge. Table 4 illustrates the range, mean and standard deviation in relation to the level of importance of a range of factors in influencing their career choice and the extent to which they felt each one would be achieved in a career in mental health. The factors that were identified as being most important to participants in terms of their career choices in general were the ability to help others and receiving appropriate training required for the role. These were also the factors that were seen as being most likely to be achieved in a career in mental health.

| Facilitators and barriers to a career in mental health
Participants were asked to indicate the personal factors that would encourage them to undertake a career in mental health. The most common positive reasons were that the career choice would allow participants to help others (n = 143, 67.4%) and would be consistent with participants interest in other people (n = 32, 15%). Other things that would encourage participants to consider mental health as a career were if barriers to access were reduced, for example more information was available, easier access to training (n = 25, 12%); and if the pay and other associated benefits were better (n = 21, 10%).
The most common factors that would discourage participants from considering a career in mental health were the stress or other negative personal impact, such as poor work-life balance, that they felt the profession would have on them (n = 132, 65.3%); organisational factors, such as lack of knowledge about the available roles and working patterns (n = 39, 19.3%) and training issues, such as the length of time and expense that training would involve (n = 34, 16.8%). Note: The odds ratios represent the increased odds of being in a higher versus lower set of categories along the ordinal variable (i.e. between categories 1 versus 2,3,4,5; 1,2 versus 3,4,5; 1,2,3 versus 4,5; 1,2,3,4 versus 5) for each unit increase in the predictor. An odds ratio > 1 suggests that the predictor increases the likelihood of being higher on the ordinal variable whereas an odds ratio < 1 suggests that it decreases the likelihood. Good job security 1-5 4.4 0.9 1-5 3.5 0.9

TA B L E 4
Level of importance of each factor in influencing career choice and extent to which participants felt this would be achieved in a career in mental health

| Information that would most help participants decide about a career in mental health
The types of information that had the highest mean rating in terms of helping participants decide whether to take up a career in mental health were in relation to the available roles (M = 6.

| Suggestions to improve recruitment in mental health
Participants' suggestions were categorised into five themes as illustrated in Table 5. The most common suggestions related to the provision of greater information about careers in mental health and highlighting the intrinsic rewards, in particular the opportunities the work offered to help others.

| D ISCUSS I ON
The project aimed to identify those factors which would encourage and discourage potential employees from considering a career in mental health. By design, most participants were currently studying, or graduates from, a degree that was likely to indicate a potential interest in a career in mental health (Durcan et al., 2017;The Psychological Professions Network, 2018  Provision of more career information/awareness raising about mental health issues 91 (52) "Having teachers talk to us about it in school and university." "Maybe an in-depth session outlining the careers available." "I think as mental health awareness is increased people will talk about it more, and therefore will want to help." Greater highlighting of the rewards (e.g. ability to help others)

(24)
"Show examples of how these people have helped others." "Showing that having that career would actually benefit someone else." "Have it talked about more as a rewarding career." Provision of opportunities for experience 17 (10) "Work experience." "More placement opportunities." "More opportunities to gain experience." Developing clearer and more accessible career pathways 15 (8) "Less financially draining (e.g. not having to pay for courses when getting almost no chance to work during them-doesn't leave opportunity for anyone not middle to upper class.)" "Job and career opportunities." "Easier routes to the higher up jobs or at least more clear routes." Pay/benefits 10 (6) "Higher wage brackets for roles such as nurses and social workers." "Like good wage, making the working place more attractive." "Higher pay. Greater team and managerial support. Better work-life balance." opportunities for young people (Health Education England, n/d) and resources for the long-term unemployed (NHS When considering the key information to include in such communications, our results suggest that the caring and helping nature of the profession should be emphasised, this being the key factor that attracted our participants and also a significant factor associated with increased likelihood of considering a career in mental health. A second important message is to highlight the ways in which mental health services tackle the stressful nature of the work, for example through staff support, supervision and mentoring. This would help address the main factor that discouraged people from considering this career choice, that is concern about the impact on their own personal well-being and work-life balance. These are realistic concerns, as working in mental health services can be stressful and result in staff burnout (Sadiku, 2016); however, many individuals may not be aware of the mitigation strategies that exist. Providing opportunities for direct contact with those being supported and the staff supporting them, such as service visits (McConkey, McAuley, Simpson, & Collins, 2007) and opportunities for placements/voluntary work/shadowing to potential employees (Owen & Standen, 2007;Sutten et al., 2015), have been identified in other caring contexts as important ways of recruiting staff with the right values, with realistic expectations, and who are committed to the area of work, as well as increasing opportunities to recruit from a more diverse workforce (Alink, Euser, Bakermans-kranenburg, & van IJzendoorn, 2014;Werner, 2011 Psychology, 2019). Many psychology graduates may also take approximately five years after graduation before beginning on a traditional career pathway (Coulthard, 2017).
This indicates that there may be opportunities for early interventions to encourage psychology graduates to undertake training in careers other than clinical psychology. Conversion programmes for mental health or learning disability nursing offer opportunities for such graduates to embark on a mental health career pathway, however, our results suggest that such programmes would need to be of relatively short duration and not incur additional costs for students in order to maximise their appeal.
Such developments and the wider promotion of relevant existing training opportunities to psychology students and graduates may help reduce the focus on clinical psychology as their main career of choice.
Obtaining the views of students and graduates from relevant degrees was a priority because of the likelihood that these groups may be particularly attracted to a career in mental health. As a result, however, the study had the limitation that most participants were white, from the UK, female, and had already embarked on or completed a relevant degree. This meant that it was not possible to fully explore the potential to encourage school or college students into a career in mental health at an early stage, for example, by diverting some prospective undergraduates from psychology degrees into a wider range of mental health career-related degrees, such as nursing.
Similarly, the limited diversity in terms of participants' gender, age and ethnicity reflects a key issue across the workforce and highlights the need to undertake a more targeted approach to understanding barriers and opportunities for men, and those from older and different ethnic groups. Future research with a more diverse participant group may further increase our understanding of how best to attract under-represented groups into mental health services. A second limitation was that there was very little research, specific to recruitment in mental health services, which could be utilised to inform the design of the study. Those factors that were identified as influencing career choice in other areas of health and support services (e.g. Barriball et al., 2015;Mckenzie et al., 2017) were, however, also rated as important to the participants in the present study.

| CON CLUS ION
In order to meet the unprecedented ambitions for growth in the mental health workforce over the next 10 years, the NHS needs to maximise supply of new recruits into training for mental health careers. There are many potential sources from which staff could be recruited to mental health nursing that are currently being underutilised, in particular psychology and social studies entrants and graduates. This largely appears to be due to the limited promotion of such careers to these groups.

| RELE VAN CE TO PR AC TICE
Our research suggests that recruitment messages need to highlight the helping nature of careers in mental health and effectively address concerns about the perceived negative impact on staff wellbeing. Clear and targeted information about the roles, access routes and required qualifications needs to be provided at different stages of the supply pipeline. New opportunities that offer alternative routes need to be of short enough duration and financially viable in order to attract applicants. Interventions that aim to increase recruitment need to be evaluated in order to develop a more robust evidence base for what is effective.

ACK N OWLED G EM ENTS
This work was supported by funding from Health Education England (HEE).

CO N FLI C T O F I NTE R E S T
The study was funded by Health Education England (HEE). CB and AC are employed by HEE.

AUTH O R CO NTR I B UTI O N S
KMcK, AM, KM, CB and AC contributed to the study conception and design. Material preparation and data collection were conducted by MC and RM. Analysis was conducted by AM and KMcK. The first draft of the manuscript was written by KMcK, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.