Enhancing retention of occupational therapists working in mental health: Relationships between wellbeing at work and turnover intention

Research Article

Enhancing retention of occupational therapists working inmental health: Relationships between wellbeing at workand turnover intention

Justin Newton Scanlan,1 Pamela Meredith2 and Anne A. Poulsen2

1Discipline of Occupational Therapy, Faculty of Health Sciences, The University of Sydney and Mental Health Services,Sydney and South Western Sydney Local Health Districts, Sydney, New South Wales, and 2School of Health andRehabilitation Sciences, The University of Queensland, St Lucia, Queensland, Australia

Background/aim: Occupational therapists working inmental health who experience burnout, low work engage-ment or poor job satisfaction are at risk of poor wellbeingat work and may be more likely to leave their jobs. Theaim of this project was to explore factors associated withwellbeing at work and turnover intention in a sample ofoccupational therapists working in mental health.Methods: One hundred and three occupational therapistsworking in mental health in Queensland completed a sur-vey exploring work/life balance, effort invested in work,rewards received from work, wellbeing at work (job satis-faction, burnout and work engagement) and turnoverintention. Analyses were conducted to explore relation-ships between work/life balance, effort, reward, wellbeingat work and turnover intention.Results: All measures of wellbeing at work were signifi-cantly associated with turnover intention. A large propor-tion (33%) of the variance in turnover intention waspredicted by job satisfaction. Perceptions of both work/lifebalance and effort invested in work, as well as perceivedrewards in terms of recognition, prestige and personal sat-isfaction were significantly associated with work-relatedwellbeing scores.Conclusions and significance of the study: Resultsfrom this study deepen the understanding of factors asso-ciated with wellbeing at work and turnover intention foroccupational therapists in mental health. This knowledge

will support the development of interventions aimed atreducing turnover intention and enhancing retention ofoccupational therapists in the mental health workforce.

KEY WORDS burnout, job satisfaction, psychiatry, workengagement, workforce.

Introduction

Ultimately, every service user and manager wants the

same thing: for staff to be productive and committed to

the organisation and to the provision of quality services.

This is especially true in health services, where employ-

ees are the organisation’s most valuable resource.

Unlike some other industries where employee satisfac-

tion and wellbeing may not be directly linked with effi-

ciency or quality (Courtney, Whelan, Majoor, Ibrahim &

Day, 2006), maintaining employee wellbeing is critical

in health services. Health service employees who are

burnt out (i.e., physically and emotionally depleted due

to chronic occupational stress; Demerouti, Bakker, Var-

dakou & Kantas, 2003), or dissatisfied, are more likely

to make errors and less likely to be responsive to indi-

vidual consumer needs (Aiken et al., 2012). In extreme

circumstances, burnt out employees may even act

malevolently (Neveu, 2008).

In addition, staff turnover can be seen as having

adverse and costly outcomes for health services. In

terms of monetary costs, health services must expend

resources to advertise, screen, interview, appoint, induct

and train replacement employees (Jones, 2004; O’Brien-

Pallas et al., 2006). There are also costs in terms of client

care, with loss of organisational knowledge, compro-

mised long-term continuity of care and, potentially,

interruption of the implementation of evidence-based

practices (O’Brien-Pallas et al.; Woltmann et al., 2008).

Finally, where staff turnover causes extended vacancies,

there is cost to the wellbeing of other staff (O’Brien-Pal-

Justin Newton Scanlan PhD, MHM, BOccThy; Lecturer.Pamela Meredith PhD, BA (Hons), BSc, BOccThy; SeniorLecturer. Anne A. Poulsen PhD, BOccThy (Hons); SeniorResearch Officer.

Correspondence: Justin Newton Scanlan, Room J120, C43J,Faculty of Health Sciences, The University of Sydney, POBox 170, Lidcombe, NSW 1825, Australia. Email: [email protected]

Accepted for publication 23 June 2013.

© 2013Occupational TherapyAustralia

Australian Occupational Therapy Journal (2013) 60, 395–403 doi: 10.1111/1440-1630.12074

las et al.). When there are vacancies within a team, other

staff members often carry additional workloads. This

can lead to increased stress, dissatisfaction, and further

staff turnover, which can have an enduring effect on

the stability of teams.

The risk of burnout and poor retention in occupa-

tional therapy is an issue for the entire profession (Balo-

gun, Titiloye, Balogun, Oyeyemi & Katz, 2002; Gupta,

Paterson, Lysaght & von Zweck, 2012; Painter, Akroyd,

Elliot & Adams, 2003). Occupational therapists in men-

tal health settings have been identified as particularly

susceptible to burnout (Painter et al.). Some aspects of

working in mental health settings (e.g., working with

distressed individuals, perceived lack of consumer pro-

gress) and high workloads can contribute to burnout for

these practitioners (Bassett & Lloyd, 2001; Painter et al.).In addition to burnout, two other aspects of wellbeing

at work are job satisfaction and work engagement. Job

satisfaction is defined as a global sense of satisfaction

with one’s work (Nagy, 2002) and has been reported to

be the most significant predictor of lower turnover

intention in other employee populations (Lambert,

Hogan & Barton, 2001). Work engagement is often con-

sidered the theoretical opposite of burnout: while burn-

out is defined in negative terms, work engagement is

seen when employees are motivated, enthusiastic and

energised by their work (Schaufeli & Bakker, 2004).

Work engagement can be seen as “health at work”

rather than simply the absence of “sickness at work”

(burnout). Examining these interdependent constructs

has potential to provide greater understanding of well-

being at work from a salutogenic (health-related) per-

spective that moves beyond a unidirectional focus on

poor health. Given the importance occupational thera-

pists place on meaningful engagement and a holistic

view of health, it is surprising that work engagement

has been largely overlooked in previous research into

wellbeing at work for occupational therapists. A focus

on these three key elements of wellbeing at work (high

levels of job satisfaction, low levels of burnout and high

levels of work engagement) may provide new insights

to enhance retention of occupational therapists.

Although burnout and factors associated with job sat-

isfaction for occupational therapists have been explored

in previous studies (Balogun et al., 2002; Gupta et al.,2012; Hayes, Bull, Hargreaves & Shakespeare, 2008;

Moore, Cruickshank & Haas, 2006; Painter et al., 2003;Scanlan, Still, Stewart & Croaker, 2010), the relation-

ships between wellbeing at work and turnover intention

have received little attention. As mental health practice

is considered especially emotionally demanding (Sør-

gaard, Ryan, Hill, Dawson, & the OSCAR group, 2007),

occupational therapists in these settings represent one

population that may be at particular risk for poor well-

being at work.

A wide range of factors contribute to wellbeing at

work. These can be conceptualised as factors directly

related to the job itself, such as perceived benefits and

rewards, and person-related factors, such as appraisals

of external non-work demands. Each of these will be

discussed below in relation to theoretical models.

According to the model of Effort-Reward Imbalance

(Siegrist, 2002), employees who perceive that they invest

a high degree of effort in their work, but receive little in

return, are more likely to experience poorer wellbeing

at work than those with better perceived balance

between effort and reward (Kouvonen et al., 2006). Per-ceptions of effort and reward are useful measures of

job-related factors that may contribute to, or diminish,

wellbeing at work.

One person-related factor that may influence wellbe-

ing at work is an individual’s satisfaction with their

work/life balance. Work/life balance is the degree to

which employees consider that they can achieve a satis-

fying mix of work and non-work activities (Bauld,

Brough & Timms, 2009). Stressors in either work or

non-work domains may contribute to conflict between

job-related and personal commitments and this in turn

can impact perceptions of work/life balance and well-

being. Not surprisingly, work/life balance has been

associated with both job and family satisfaction

(Brough, O’Driscoll & Kalliath, 2005).

The present study was established to explore factors

associated with wellbeing at work, and the associations

between wellbeing at work and turnover intention, in a

group of occupational therapists in working mental

health. The study is guided by three research questions:

1. Are demographic variables related to differences in

measures of wellbeing at work (job satisfaction, burn-

out and work engagement) or turnover intention?

2. What are the relationships between job-related fac-

tors (effort and reward), person-related factors

(work/life balance) and wellbeing at work?

3. How much of the variance in turnover intention is

explained by measures of wellbeing at work?

A detailed understanding of how different elements

of wellbeing at work are associated with lower turnover

intention will provide better opportunities to address

retention issues for occupational therapists in mental

health. This is a critical issue for the profession, both

now and into the future.

Method

The study was approved by the University of Queens-

land’s Behavioural and Social Sciences Ethics Research

Committee.

Participants

During 2010, all occupational therapists registered in

Queensland (N = 2349) were invited to complete a sur-

vey exploring factors related to work and wellbeing. A

total of 951 usable surveys were returned (overall

response rate 40.5%). This article focuses on data pro-

© 2013 Occupational Therapy Australia

396 J. N. SCANLAN ET AL.

vided by 103 respondents who identified mental health

as their primary clinical area. Data regarding the prac-

tice areas for the overall sample were not available, so a

specific response rate for occupational therapists in

mental health practice could not be determined. Demo-

graphics for the sample are summarised in Table 1.

Instrumentation

Work/life balance

Work/life balance was measured by four items (Bauld

et al., 2009) rated on a scale of 0 (completely disagree)

to 6 (completely agree). Reversed items were recoded so

that higher scores represented better work/life balance

and an “overall” work/life balance score was deter-

mined by averaging ratings across the four items. This

scale has acceptable psychometric qualities and was val-

idated in a study of 3,351 employees (Bauld et al.). Inthe present study, Cronbach’s alpha was 0.95.

Effort-reward imbalance scale - abridged

This scale (Kouvonen et al., 2006) consists of four state-

ments rated on a 5-point scale, from 0 (very little) to 4

(very much). Effort is represented by one item: “How

much do you feel you invest in your job in terms of

skill and energy?” Reward is measured by three items

including self-appraisals of monetary reward, recogni-

tion and prestige and personal satisfaction. In analyses

presented in this article, individual reward items were

treated as separate variables to allow for the examina-

tion of the unique relationships between different types

of rewards and wellbeing at work.

Job satisfaction

Job satisfaction was measured by one item (“How satis-

fied are you with your job as a whole?”) rated from 0

(completely unsatisfied) to 10 (completely satisfied). Sin-

gle item measures of job satisfaction are used exten-

sively in large-scale surveys and are considered to have

a number of advantages, including reduced respondent

burden, over instruments that attempt to construct job

satisfaction from a number of component elements

(Nagy, 2002; Wanous, Reichers & Hudy, 1997).

Oldenburg burnout inventory (OLBI)

The 16-item OLBI (Demerouti & Bakker, 2008; Demero-

uti et al., 2003) measures two elements of burnout: dis-

engagement and exhaustion using a scale of 1 (strongly

disagree) to 4 (strongly agree). Higher scores represent

higher levels of disengagement or exhaustion. Although

the Maslach Burnout Inventory (Maslach, Jackson &

Leiter, 1996) is a more commonly used measure of

burnout, the OLBI is considered psychometrically supe-

rior as it contains both positively and negatively

worded items. Internal consistency is good, the factor

structure is robust and test-retest reliability is satisfac-

tory (Demerouti & Bakker, 2008; Demerouti et al.; Hal-

TABLE 1: Selected demographic characteristics of the sample,

N = 103

Domain Category n %

Gender Female 94 91.3

Male 7 6.8

Missing 2 1.9

Age <25 years 16 15.5

25–35 years 44 42.7

35–45 years 19 18.4

>45 years 22 21.4

Missing 2 1.9

Years working as

occupational

therapist

<5 years 36 35.0

6–10 years 25 24.3

11–15 years 17 16.5

16–20 years 6 5.8

21–25 years 5 4.9

>25 years 14 13.6

Work pattern Full-time 74 71.8

Part-time 23 22.3

Other 5 4.9

Missing 1 1.0

Work context Team/group practice 87 84.5

Sole practitioner 11 10.7

Other 5 4.9

Work sector Public sector 90 87.4

Private sector 11 10.7

Other 2 1.9

Work location Metropolitan 81 78.6

Regional/Rural/Remote 22 21.4

Postgraduate

qualifications

No 72 69.9

Yes 29 28.2

Missing 2 1.9

Marital status Single/never married 17 16.5

Married 50 48.5

Partner/cohabiting 22 21.4

Divorced/separated 11 10.7

Widowed 1 1.0

Missing 2 1.9

Children No 55 53.4

Yes 46 44.7

Missing 2 1.9

Income from

occupational

therapy practice

<$20,000 8 7.8

$20,000 to under $40,000 13 12.6

$40,000 to under $60,000 34 33.0

$60,000 to under $80,000 36 35.0

$80,000 to under $100,000 9 8.7

Missing 3 2.9

Income enough to

cover household

expenses

No 47 45.6

Yes 52 50.5

Missing 4 3.9

© 2013 Occupational Therapy Australia

TURNOVER INTENTION AND WELLBEING AT WORK 397

besleben & Demerouti, 2005). In the present study,

Cronbach’s alphas for the disengagement and exhaus-

tion subscales were 0.72 and 0.70, respectively.

Utrecht work engagement scale (UWES-17)

The 17 items in this instrument (Schaufeli & Bakker,

2004) are scored on a scale of 0 (never) to 6 (always)

and used to determine Vigour, Dedication and Absorp-

tion subscale scores and Total Work Engagement. The

UWES-17 has been used in many countries and indus-

tries, including a variety of healthcare settings. Internal

reliability is good and factor structure demonstrates sta-

bility across industries and countries (Schaufeli & Bak-

ker; Shimazu et al., 2008). In the present study,

Cronbach’s alphas for Vigour, Dedication and Absorp-

tion were 0.73, 0.85 and 0.68, respectively, and 0.88 for

Total Work Engagement.

Turnover intention

Turnover intention was measured by the question:

“How often have you seriously thought about leaving

your job in the past six months?” with response options

from 0 (never) to 4 (a great deal). Although turnover

intention does not invariably lead to actual turnover,

behavioural intention is considered a useful determi-

nant of actual behaviour (Tett & Meyer, 1993) and has

been used as a proxy measure for attrition in health

workforce studies (Flinkman, Leino-Kilpi & Salanter€a,

2010).

Analysis

All analyses were completed using SPSS Statistics (Ver-

sion 17.0, SPSS Inc., Chicago, IL, USA). Prior to analysis,

all variables were checked for normality (using the

criteria of P < 0.05 for Kolmogorov–Smirnov or Shap-

iro–Wilk test statistics). All relationships between vari-

ables were also checked for deviation from linearity,

and multicollinearity was examined in variables used in

regression analyses. Missing data resulted in slightly

diminished numbers in some analyses.

To determine if demographic variables were related

to differences in measures of wellbeing at work, a series

of between-group analyses were completed. Indepen-

dent-samples t-tests were used for dichotomous demo-

graphic variables (excluding “other” where necessary)

and one-way analyses of variance for other demo-

graphic variables.

To address the second research question, relation-

ships between measures of work/life balance, effort,

rewards, job satisfaction, burnout and work engagement

were examined. Initially, Pearson’s correlation coeffi-

cients were calculated. Following this, a series of step-

wise multiple regression analyses were conducted. Each

model contained one of the measures of wellbeing at

work as the dependent variable, and work/life balance,

effort, and the three measures of rewards as indepen-

dent variables.

A similar strategy was used to examine the final

research question. Correlations between measures of

wellbeing at work and turnover intention were calcu-

lated and a single stepwise multiple regression analysis

(with turnover intention as the dependent variable) was

conducted.

Results

Preliminary data checking revealed that data met

assumptions for parametric tests. All variables demon-

strated satisfactory normality in distribution. The major-

ity of variable pairs did not deviate from linearity. The

only variable pair that demonstrated significant devia-

tion from linearity was exhaustion and rewards: income

and job benefits (F3,98 = 2.79, P = 0.05). Although many

of the variables were significantly correlated, multicol-

linearity did not appear to influence the ability to inter-

pret results from regression analyses.

Between-group differences according todemographic variables

A number of differences were present in between-group

analyses based on demographic variables. Individuals

who were employed full-time reported higher turnover

intention (t95 = 2.05, P = 0.04) and disengagement

(t95 = 2.07, P = 0.04) than individuals employed part-

time. Individuals working in the private sector reported

lower disengagement than those in the public sector

(t99 = �2.35, P = 0.02). Respondents with children

reported lower exhaustion (t99 = �2.00, P = 0.05) and

higher absorption in their work (t97 = 2.02, P = 0.05)

compared with those without children. Individuals who

reported “other carer responsibilities” also reported

higher absorption in their work than respondents with-

out other carer responsibilities (t97 = 2.18, P = 0.03).

Given the unexpected results from between-group

analyses for those respondents with and without chil-

dren or other carer responsibilities, additional post hoc

analyses were completed to determine if these results

were influenced by work pattern (i.e., it was hypothes-

ised that a larger proportion of respondents with carer

responsibilities may work part-time and that this may

have an influence on results). These additional analyses

demonstrated that differences between those in caring

roles and those not in caring roles persisted (although

were no longer statistically significant), even when anal-

ysing part-time and full-time groups separately.

Relationships between work/life balance,effort, reward and wellbeing at work

Correlation coefficients (Table 2) revealed significant

relationships between work/life balance, effort, reward

and wellbeing at work. Regression analyses (Table 3)

revealed important relationships between: rewards in

terms of recognition and prestige and job satisfaction;

work/life balance, job satisfaction and exhaustion;

© 2013 Occupational Therapy Australia

398 J. N. SCANLAN ET AL.

rewards in terms of personal satisfaction, disengage-

ment and all measures of work engagement and effort

and all measures of work engagement.

Relationships between measures ofwellbeing at work and turnover intention

All measures of wellbeing at work, except for absorp-

tion, were significantly associated with turnover inten-

tion (Table 2). Results from the final regression model

(Table 4) explained a total of 40% of the variance in

turnover intention, with a large proportion of this (33%)

predicted by job satisfaction.

Discussion

Overall, this study has highlighted several important

relationships between aspects of wellbeing at work

(most especially job satisfaction) and lower turnover

intention. This is important knowledge that has poten-

tial to guide future efforts to improve occupational ther-

apist retention in the mental health workforce.

This study was guided by three research questions.

The central research question driving this study was

what elements of wellbeing at work are most associated

with lower turnover intention. This will be explored

first, followed by discussion of the other two research

questions: namely, what job-related and person-related

factors are related to wellbeing at work, and which

demographic factors are related to aspects of wellbeing

at work.

Consistent with previous research with a national

sample (Lambert et al., 2001) and nurses in the United

Kingdom (Shields & Ward, 2001), job satisfaction had

the strongest association with lower turnover intention,

predicting 33% of the variance. Job satisfaction was

most strongly associated with two factors: (1) rewards

in terms of recognition and prestige and (2) work/life

balance. This suggests that efforts to increase job satis-

faction through enhancing recognition and prestige and

supporting improved work/life balance would be

worthwhile and may support the retention of occupa-

tional therapists in the mental health workforce.

TABLE 2: Summary statistics and correlations with work/life balance, effort, reward and turnover intention

Summary statistics Correlation with

Mean (SD)

Possible

range

(Min–Max)

Work/life

balance

Effort:

skill and

energy

Reward:

income

and job

benefits

Reward:

recognition

and prestige

Reward:

personal

satisfaction

Turnover

Intention

Work/life Balance 3.56 (1.50) 0–6 –

Effort: skill and

energy

3.18 (0.71) 0–4 �0.20* –

Reward: income and

job benefits

2.19 (0.85) 0–4 0.12 0.02 –

Reward: recognition

and prestige

1.77 (1.05) 0–4 0.07 0.15 0.54*** –

Reward: personal

satisfaction

2.64 (0.92) 0–4 0.05 0.39*** 0.27** 0.46*** –

Job satisfaction 6.92 (1.67) 0–10 0.29** �0.06 0.32** 0.48*** 0.44*** �0.58***

Disengagement

(burnout)

2.20 (0.38) 1–4 0.01 �0.24* �0.16 �0.34*** �0.58*** 0.50***

Exhaustion

(burnout)

2.48 (0.39) 1–4 �0.58** 0.22* �0.27** �0.20* �0.20* 0.30**

Vigour (work

engagement)

3.42 (0.65) 0–6 0.29** 0.36*** 0.13 0.36*** 0.53*** �0.29**

Dedication (work

engagement)

4.01 (0.82) 0–6 0.04 0.43*** 0.08 0.28** 0.63*** �0.39***

Absorption (work

engagement)

3.46 (0.70) 0–6 �0.12 0.40*** 0.11 0.27** 0.46*** �0.05

Total work

engagement

3.61 (0.62) 0–6 0.08 0.46*** 0.12 0.35*** 0.63*** �0.28**

*P < 0.05; **P < 0.01; ***P < 0.001.

© 2013 Occupational Therapy Australia

TURNOVER INTENTION AND WELLBEING AT WORK 399

More effective recognition of both achievements and

contributions of individual occupational therapists may

prove a useful intervention to enhance job satisfaction

(Moore et al., 2006). Formal recognition programs (e.g.,

employee of the month) have been criticised as out-

dated and ineffective, and it has been suggested that

effective ways to enhance employees’ sense of recogni-

tion is to provide timely, genuine and personal expres-

sions of appreciation (Nelson, 2004). Drawing on a

range of previous literature, Hansen, Smith and Hansen

(2002) argued that job satisfaction and dissatisfaction

are two distinct constructs rather than opposite ends of

the same continuum. They argued that tangible rewards

(e.g., pay or awards) act to decrease dissatisfaction, but

that recognition (identifying, honouring and celebrating

achievements and innovation) is the key to enhancing

satisfaction. The most important elements of recognition

have been identified as praise (public and private; ver-

bal and written) and management styles that support

and involve employees (Nelson, 2004). Support and

involvement of employees are components of “high lea-

der-member exchange” forms of leadership which have

been associated with higher job satisfaction and lower

burnout (Thomas & Lankau, 2009). Implementing these

practical strategies to enhance employees’ sense of rec-

ognition would appear to be achievable in almost all

settings.

Increasing the recognition and prestige of occupa-

tional therapy in mental health more generally may be

a more challenging prospect. Occupational therapists in

mental health have frequently reported that they per-

ceive that their role is poorly understood and their con-

tribution is under-appreciated (Bassett & Lloyd, 2001;

Hayes et al., 2008; Lloyd, King & Bassett, 2002). Such

perceptions may be grounded in co-workers’ lack of

understanding of the occupational therapy role, but are

likely to be perpetuated and exacerbated by low profes-

sional efficacy and resilience of individual therapists

(Ashby, Ryan, Gray & James, 2013). Ashby et al. (2013)outlined a range of factors that contribute to profes-

sional efficacy and resilience in mental health occupa-

tional therapists. These included formal and informal

TABLE 3: Results from stepwise multiple regression analyses

for measures of wellbeing at work

Model Beta†

Adj. R2

change F value P value

Dependent variable: Job satisfaction

Reward: recognition

and prestige

0.33 0.22

Work/life balance 0.20 0.06

Reward: personal

satisfaction

0.35 0.05

Effort: skills and

energy

�0.21 0.03

Overall model 0.36 15.10 <0.001

Dependent variable: Disengagement

Reward: personal

satisfaction

�0.58 0.33

Overall model 0.33 49.91 <0.001

Dependent variable: Exhaustion

Work/life balance �0.55 0.33

Reward: income

and job benefits

�0.20 0.03

Overall model 0.36 29.11 <0.001

Dependent variable: Vigour

Reward: Personal

satisfaction

0.41 0.27

Work/life balance 0.32 0.06

Effort: skills and

energy

0.25 0.05

Overall model 0.38 21.13 <0.001

Dependent variable: Dedication

Reward: personal

satisfaction

0.54 0.39

Effort: skills and

energy

0.21 0.03

Overall model 0.42 36.54 <0.001

Dependent variable: Absorption

Reward: personal

satisfaction

0.35 0.20

Effort: skills and

energy

0.26 0.05

Overall model 0.25 17.46 <0.001

Dependent variable: Work Engagement

Reward: personal

satisfaction

0.56 0.38

Effort: skills and

energy

0.25 0.05

Overall model 0.43 38.02 <0.001

Adj R2 = Adjusted R2.

†Standardised coefficient for variable in final model.

TABLE 4: Results from regression analysis for turnover

intention

Model Beta† Adj. R2 change F value P value

Dependent variable: Turnover intention

Job satisfaction �0.41 0.33

Disengagement 0.37 0.04

Absorption 0.23 0.03

Overall model 0.40 22.89 <0.001

Adj R2 = Adjusted R2.

†Standardised coefficient for variable in final model.

© 2013 Occupational Therapy Australia

400 J. N. SCANLAN ET AL.

support from occupational therapy colleagues, opportu-

nities to use occupationally focussed interventions, the

ability to clearly articulate the occupational therapy role

to colleagues and managers, and the capacity to asser-

tively manage expectations that other forms of knowl-

edge (e.g., cognitive behavioural therapy) should be

integrated into practice. Supporting occupational thera-

pists to identify and enhance these elements of practice

may have a positive impact on therapists’ perceptions

of their own competence and how valued and respected

they feel.

While these results are important, it must be remem-

bered that 60% of the variance in turnover intention was

not predicted by the measures of wellbeing at work used

in this study. Turnover intention is influenced by a broad

range of factors, only some of which are related to well-

being at work (O’Brien-Pallas et al., 2006). Occupational

therapists are often young, mobile and have good oppor-

tunities for change in positions. A previous study

reported that the most common reasons for occupational

therapists leaving mental health positions were: desire

for a different type of work; lifestyle reasons; career

development opportunities; dysfunctional teams and

unsatisfactory management (Scanlan et al., 2010). Only

the last two of these factors would be considered “nega-

tive” factors associated with poorer wellbeing at work.

The second research question explored the relation-

ship between job-related and person-related factors and

indices of wellbeing at work. Results suggested that

four variables: work/life balance; rewards in terms of

recognition and prestige; rewards in terms of personal

satisfaction and effort, were particularly important pre-

dictors of wellbeing at work. These are discussed in

more detail below.

Although participants in this study reported generally

positive work/life balance, poorer work/life balance

was associated with higher levels of exhaustion. Impor-

tantly, work/life balance is a matter of perception rather

than an absolute. For example, Bauld et al. (2009) foundthat employees who were happy with their working

hours experienced higher job satisfaction than those

who would have preferred to work fewer hours, regard-

less of the actual number of hours worked. For those

perceiving work/life imbalance, a genuine commitment

to ‘family-friendly’ policies including flexitime, no

unpredictable overtime and holding meetings during

regular daytime hours, and other flexible work practices

may enhance perceptions of work/life balance and

improve wellbeing at work (Brough et al., 2005).Rewards in terms of recognition and prestige,

rewards in terms of personal satisfaction, and effort

invested in one’s work in terms of skill and energy,

were all significantly associated with wellbeing at work.

Rewards that provided recognition and prestige demon-

strated the strongest associations with job satisfaction.

Rewards in terms of personal satisfaction were associ-

ated with variables indicating positive wellbeing at

work. Effort, however, had a mixed pattern of relation-

ships. On the one hand, high effort was associated with

high exhaustion (representing poor wellbeing at work),

but on the other hand, high effort was associated with

low disengagement and high work engagement (repre-

senting positive wellbeing at work). Notwithstanding

the finding that high effort may be associated with

higher exhaustion; these results suggest the possibility

that high effort, particularly in the context of high

rewards, may support greater wellbeing at work in

terms of work engagement. This is consistent with pre-

vious findings that occupational therapists in mental

health frequently reported “challenging work” (suggest-

ing a reasonable degree of effort required) as a key

attraction of their position, and “seeking a new chal-

lenge” as a reason for leaving positions (Hayes et al.,2008; Scanlan et al., 2010). This is also consistent with

other research that demonstrated the positive impact of

working in “innovative roles” on retention for nurses

and allied health professionals (Collins, Jones, McDon-

nell, Jones & Cameron, 2000).

Overall, each variable studied (with the exception of

rewards in terms of income and job benefits) demon-

strated significant relationships with measures of well-

being at work which were in turn related to lower

turnover intention. Although the final regression analy-

sis for turnover intention focused attention on job satis-

faction as the most important element of wellbeing at

work, the significant correlations between all measures

of wellbeing at work suggest that improvements in any

area are likely to contribute to better overall wellbeing

and potentially enhance retention.

Finally, exploration of the first research question,

“Are demographic variables related to differences in

measures of wellbeing at work (job satisfaction, burnout

and work engagement) or turnover intention?” uncov-

ered some additional interesting results. Possibly, the

most intriguing of these was the finding that individu-

als with more commitments, in terms of carer responsi-

bilities, could experience greater absorption in their

work and less exhaustion. Theories surrounding “multi-

ple roles” (Barnett & Hyde, 2001) provide some expla-

nation for these relationships. Contrary to traditional

understandings of multiple-role conflict, Barnett and

Hyde proposed that holding multiple roles may actually

contribute to wellbeing through several processes

including: buffering (successes in one role may buffer

against the negative effects of challenges in another

role); increased social networks; more opportunities for

success and affirming experiences and broader experi-

ences and a wider frame of reference. Similar results

were reported from a recent study: the presence of a

partner or children in the home supported work-home-

segmentation and contributed to employees’ wellbeing

(Hahn & Dormann, 2013).

Full-time workers reported higher turnover intention

and disengagement compared with part-time workers.

© 2013 Occupational Therapy Australia

TURNOVER INTENTION AND WELLBEING AT WORK 401

It is possible that full-time employees may have more

opportunities to change positions and therefore may

have greater turnover intention which could also lead

to greater disengagement. Another perspective might be

that part-time employees have a greater appreciation of

the opportunity to work part-time and therefore are less

likely to consider leaving and feel less disengaged from

their work.

Participants in the private sector also reported lower

disengagement (M = 1.95) than participants in the pub-

lic sector (M = 2.23). Previous studies have reported

mean ratings of disengagement between 2.21 and 2.38

(Demerouti & Bakker, 2008; Demerouti et al., 2003).

Therefore, the difference between private and public

sector respondents appears related to the lower than

average disengagement reported by private sector

respondents. Although more detailed data on the con-

text of work was not available, it may be that respon-

dents from the private sector worked in smaller

services or in private practice, experienced greater flex-

ibility in their work, and/or had the opportunity to

work more intensively with their clients. Private sector

working environments may also contribute to greater

perceived autonomy and job control which may also

support enhanced wellbeing at work (Crawford, LePine

& Rich, 2010).

While interesting and potentially important, these dif-

ferences according to demographic variables should be

interpreted with caution. The differences found were

generally modest and in the context of multiple compar-

isons; hence, there is an increased risk of Type 1 errors

(i.e., concluding that there is a difference where there is

none). These findings should be replicated in other

studies prior to being used to inform decision making.

Although results from this study are consistent with

those from previous longitudinal studies, the cross-sec-

tional nature of this study prevents conclusions about

causation. For example, while it seems likely that

reduced satisfaction with one’s job might lead to

increased turnover intention, it may be that a desire

to leave one’s job (for example, due to a desire to move

to a more desirable location) may lead to reduced job

satisfaction over time. The non-random nature of this

sample may also limit the generalisability of findings,

although the use of a complete sampling frame (all

occupational therapists registered in Queensland) is a

strength of the study.

Future work in this area should focus on the design

and evaluation of targeted interventions to enhance

occupational therapists’ job satisfaction (as well as other

forms of wellbeing at work) and examine longitudinal

effects on retention. Although evidence is sparse, it is

likely that an experienced and engaged occupational

therapy workforce in mental health will support high

quality service provision and promote optimal out-

comes for consumers of mental health services.

Conclusion

Although some level of turnover is unavoidable (and

may even be quite healthy), this study adds valuable

information to the developing body of literature sur-

rounding retention and wellbeing at work of occupa-

tional therapists in mental health. Results suggest a

range of directions that are likely to promote greater

wellbeing at work and, in turn, enhance the retention of

occupational therapists in the mental health workforce.

These include improving recognition of individual occu-

pational therapists and of the profession as a whole,

supporting satisfactory work/life balance, and ensuring

that therapists have access to work that they find chal-

lenging and personally rewarding.

References

Aiken, L. H., Sermeus, W., Van den Heede, K., Sloane, D.

M., Busse, R., McKee, M. et al. (2012). Patient safety, sat-

isfaction, and quality of hospital care: Cross sectional sur-

veys of nurses and patients in 12 countries in Europe

and the United States. British Medical Journal, 344, doi:10.

1136/bmj.e1717

Ashby, S. E., Ryan, S., Gray, M. & James, C. (2013). Factors

that influence the professional resilience of occupational

therapists in mental health practice. Australian Occupa-

tional Therapy Journal, 60, 110–119.Balogun, J. A., Titiloye, V., Balogun, A., Oyeyemi, A. &

Katz, J. (2002). Prevalence and determinants of burnout

among physical and occupational therapists. Journal of

Allied Health, 31, 131–139.Barnett, R. C. & Hyde, J. S. (2001). Women, men, work, and

family: An expansionist theory. American Psychologist, 56,

781–796.Bassett, H. & Lloyd, C. (2001). Occupational therapy in

mental health: Managing stress and burnout. British Jour-

nal of Occupational Therapy, 64, 406–411.Bauld, R., Brough, P. & Timms, C. M. (2009). Working to

live or living to work? The impact of time demands and

preferred working hours on work and family outcomes.

In: P. H. Langford, N. J. Reynolds & J. E. Kehoe (Eds.),

Proceedings of the 8th Industrial and Organisational Psychol-

ogy Conference (pp. 12–16). Sydney, Australia: Australian

Psychological Society Ltd.

Brough, P., O’Driscoll, M. P. & Kalliath, T. J. (2005). The

ability of ‘family friendly’ organizational resources to

predict work-family conflict and job and family satisfac-

tion. Stress and Health, 21, 223–234.Collins, K., Jones, M. L., McDonnell, A., Jones, R. & Cam-

eron, A. (2000). Do new roles contribute to job satisfac-

tion and retention of staff in nursing and professions

allied to medicine? Journal of Nursing Management, 8,

3–12.Courtney, M., Whelan, A. K., Majoor, J., Ibrahim, J. E. &

Day, G. E. (2006). Work design in health care. In: Mary G

Harris and Associates (Ed.), Managing health services: Con-

cepts and practice (2nd ed., pp. 245–269). Sydney: Mosby.

© 2013 Occupational Therapy Australia

402 J. N. SCANLAN ET AL.

Crawford, E. R., LePine, J. A. & Rich, B. L. (2010). Linking

job demands and resources to employee engagement and

burnout: A theoretical extension and meta-analytic test.

Journal of Applied Psychology, 95, 834–848.Demerouti, E. & Bakker, A. B. (2008). The Oldenburg

Burnout Inventory: A good alternative to measure burn-

out and engagement. In: J. R. B. Halbesleben (Ed.),

Stress and burnout in health care (pp. 65–78). Hauppage,

NY: Nova Sciences.

Demerouti, E., Bakker, A. B., Vardakou, I. & Kantas, A.

(2003). The convergent validity of two burnout instru-

ments: A multitrait-multimethod analysis. European Jour-

nal of Psychological Assessment, 19, 12–23.Flinkman, M., Leino-Kilpi, H. & Salanter€a, S. (2010).

Nurses’ intention to leave the profession: Integrative

review. Journal of Advanced Nursing, 66, 1422–1434.Gupta, S., Paterson, M. L., Lysaght, R. M. & von Zweck, C.

M. (2012). Experiences of burnout and coping strategies

utilized by occupational therapists. Canadian Journal of

Occupational Therapy, 79, 86–95.Hahn, V. C. & Dormann, C. (2013). The role of partners

and children for employees’ psychological detachment

from work and well-being. Journal of Applied Psychology,

98, 26–36.Halbesleben, J. R. B. & Demerouti, E. (2005). The construct

validity of an alternative measure of burnout: Investigat-

ing the English translation of the Oldenburg Burnout

Inventory. Work & Stress, 19, 208–220.Hansen, F., Smith, M. & Hansen, R. B. (2002). Rewards and

recognition in employee motivation. Compensation and

Benefits Review, 34, 64–72.Hayes, R., Bull, B., Hargreaves, K. & Shakespeare, K.

(2008). A survey of recruitment and retention issues for

occupational therapists working clinically in mental

health. Australian Occupational Therapy Journal, 55, 12–22.Jones, C. B. (2004). The cost of nurse turnover: Part 1: An

economic perspective. Journal of Nursing Administration,

34, 562–570.Kouvonen, A., Kivim€aki, M., Virtanen, M., Heponiemi, T.,

Elovainio, M., Pentti, J. et al. (2006). Effort-reward imbal-

ance at work and the co-occurrence of lifestyle risk fac-

tors: Cross-sectional survey in a sample of 36,127 public

sector employees. BMC Public Health, 6, doi:10.1186/

1471-2458-6-24.

Lambert, E. G., Hogan, N. L. & Barton, S. M. (2001). The

impact of job satisfaction on turnover intent: A test of a

structural measurement model using a national sample of

workers. The Social Science Journal, 38, 233–250.Lloyd, C., King, R. & Bassett, H. (2002). A survey of Aus-

tralian mental health occupational therapists. British Jour-

nal of Occupational Therapy, 65, 88–96.Maslach, C., Jackson, S. E. & Leiter, M. P. (1996). Maslach

Burnout Inventory Manual (3rd ed.). Consulting Psycholo-

gists Press: Palo Alto, CA.

Moore, K., Cruickshank, M. & Haas, M. (2006). Job satisfac-

tion in occupational therapy: A qualitative investigation

in urban Australia. Australian Occupational Therapy Jour-

nal, 53, 18–26.

Nagy, M. (2002). Using a single-item approach to measure

facet job satisfaction. Journal of Occupational and Organiza-

tional Psychology, 75, 77–86.Nelson, B. (2004). Formal recognition programs do not

work. Industrial and Commercial Training, 36, 243–246.Neveu, J.-P. (2008). Burnout and consequences: A review of

health professional maltreatment of the patient. In: J. R.

B. Halbesleben (Ed.), Handbook of stress and burnout in

health care (pp. 157–172). New York, NY: Nova Science.

O’Brien-Pallas, L., Griffin, P., Shamian, J., Buchan, J., Duf-

field, C., Hughes, F. et al. (2006). The impact of nurse

turnover on patient, nurse and system outcomes: A pilot

study and focus for a multicenter international study.

Policy, Politics, & Nursing Practice, 7, 169–179.Painter, J., Akroyd, D., Elliot, S. & Adams, R. D. (2003).

Burnout among occupational therapists. Occupational

Therapy in Health Care, 17, 63–78.Scanlan, J. N., Still, M., Stewart, K. & Croaker, J. (2010).

Recruitment and retention issues for occupational thera-

pists in mental health: Balancing the pull and the push.

Australian Occupational Therapy Journal, 51, 102–110.Schaufeli, W. B. & Bakker, A. B. (2004). UWES: Utrecht

Work Engagement Scale Premilinary Manual. Version

1.1. Retrieved 4 December, 2012, from http://www.wil-

marschaufeli.nl/publications/Schaufeli/Test%20Manu-

als/Test_manual_UWES_English.pdf

Shields, M. A. & Ward, M. (2001). Improving nurse reten-

tion in the National Health Service in England: The

impact of job satisfaction on intentions to quit. Journal of

Health Economics, 20, 677–701.Shimazu, A., Schaufeli, W. B., Kosugi, S., Suzuki, A., Nash-

iwa, H., Kato, A. et al. (2008). Work engagement in Japan:

Validation of the Japanese version of the Utrecht Work

Engagement Scale. Applied Psychology: An International

Review, 57, 510–523.Siegrist, J. (2002). Effort reward imbalance at work and

health. In: P. L. Perrew�e & D. C. Ganster (Eds.), Historical

and current perspectives on stress and health (pp. 261–291).Oxford: Elsevier Science.

Sørgaard, K. W., Ryan, P., Hill, R. & Dawson, I., & the

OSCAR group. (2007). Sources of stress and burnout in

acute psychiatric care: Inpatient vs. community staff.

Social Psychiatry and Psychiatric Epidemiology, 42, 794–802.Tett, R. P. & Meyer, J. P. (1993). Job satisfaction, organiza-

tional committment, turnover intention, and turnover:

Path analysis based on meta-analytic findings. Personnel

Psychology, 46, 259–293.Thomas, C. H. & Lankau, M. J. (2009). Preventing burnout:

The effects of LMX and mentoring on socialisation, role

stress, and burnout. Human Resource Management, 48,

417–432.Wanous, J. P., Reichers, A. E. & Hudy, M. J. (1997). Overall

job satisfaction: How good are single-item measures?

Journal of Applied Psychology, 82, 247–252.Woltmann, E. M., Whitley, R., McHugo, G. J., Brunette, M.,

Torrey, W. C., Coots, L. et al. (2008). The role of staff turn-

over in the implementation of evidence-based practices in

mental health care. Psychiatric Services, 59, 732–737.

© 2013 Occupational Therapy Australia

TURNOVER INTENTION AND WELLBEING AT WORK 403

Copyright of Australian Occupational Therapy Journal is the property of Wiley-Blackwell andits content may not be copied or emailed to multiple sites or posted to a listserv without thecopyright holder's express written permission. However, users may print, download, or emailarticles for individual use.