lable at ScienceDirect

European Journal of Oncology Nursing xxx (2014) 1e6

Contents lists avai

European Journal of Oncology Nursing

journal homepage: www.elsevier .com/locate/ejon

Recovery experience and burnout in cancer workers in Queensland

Michael G. Poulsen a, c, *, Anne A. Poulsen b, d, Asaduzzaman Khan b, d,Emma E. Poulsen b, e, Shanchita R. Khan b, f

a Division of Cancer Services, Princess Alexandra Hospital, Australiab The University of Queensland, Brisbane, QLD, Australiac University of Queensland, Radiation Oncology Mater Centre, Australiad School of Health & Rehabilitation Science, The University of Queensland, Australiae School of Psychology, The University of Queensland, Australiaf Institute of Health and Biomedical Innovation, Queensland University of Technology, Australia

Keywords:BurnoutHealth professionalsOncologyRecovery experience

* Corresponding author. Radiation Oncology MateSouth Brisbane, QLD 4101, Australia. Tel.: +61 7 38403

E-mail addresses: michael.poulsen@health.qld.poulsen@uq.edu.au (A.A. Poulsen), a.khan2@uq.epouslen@health.qld.gov.au (E.E. Poulsen), shanchita.k

http://dx.doi.org/10.1016/j.ejon.2014.08.0031462-3889/Crown Copyright © 2014 Published by Els

Please cite this article in press as: Poulsen, MOncology Nursing (2014), http://dx.doi.org/1

a b s t r a c t

Purpose: Two key recovery experiences mediating the relationship between work demands and well-being are psychological detachment and relaxation over leisure time. The process of recovery fromwork-related stress plays an important role in maintaining well-being, but is poorly understood in cancerworkers. The aim of this exploratory study was to examine the relationships of burnout, psychologicalwell-being and work engagement with the recovery experiences of psychological detachment andrelaxation in oncology staff.Methods: A cross sectional survey of 573 cancer workers in Queensland was conducted (response rate56%). Oncology nurses (n ¼ 211) represented the largest professional group. Staff completed surveyscontaining demographics and psychosocial questionnaires measuring burnout, psychological distress,work engagement and recovery experience. Multiple regression analyses were performed to identifyexplanatory variables which were independently associated with Recovery Experience Score (RES).Results: There was a negative association between the RES and burnout (p ¼ 0.002) as well as psy-chological distress (p < 0.0001), but not work engagement. Age >25 years was negatively correlated withRES as was having a post graduate qualification, being married or divorced, having carer commitments.Participating in strenuous exercise was associated with high recovery (p ¼ 0.015).Conclusions: The two recovery experiences of psychological detachment and relaxation had a strongnegative association to burnout and psychological well-being, but not work engagement. Furtherresearch needs to be undertaken to better understand if improving recovery experience reduces burnoutand improves the well-being of cancer workers.

Crown Copyright © 2014 Published by Elsevier Ltd. All rights reserved.

Introduction well being and are positively associated with high desire to leave

Cancer workers are exposed to a variety of work-relatedstressors including dealing with a clinical caseload that isemotionally taxing. It is well recognised that this may contribute toburnout and the literature suggests that approximately one third ofcancer workers exhibit symptoms of burnout, core features ofwhich are emotional exhaustion and disengagement (Girgis et al.,2009; Poulsen et al., 2011; Shanafelt et al., 2006). Burnout andwork stress are negatively correlated with employees' health and

r Centre, 31 Raymond Tce,255; fax: +61 7 38403298.gov.au (M.G. Poulsen), a.du.au (A. Khan), emma_han@qut.edu.au (S.R. Khan).

evier Ltd. All rights reserved.

.G., et al., Recovery experienc0.1016/j.ejon.2014.08.003

the organisation (Coffeng et al., 2012).According to the Conservation of Resources Theory (Hobfoll et al.,

2003), recovery is seen as a process to restore resources such asself-esteem or vigour that may be depleted in unfavourable workenvironments. Theoretically, workers with high daily recovery feelless fatigue and greater readiness to face new demands than thoseworkers with poor recovery.

The concept of recovery evolved from the effort-recuperationmodel (Meijiman and Mulder, 1998) and is considered to be theneed to recuperate and wind down after the effort invested inwork-related activity. Recovery refers to the process during whichthe individual's functioning returns to the pre-stressor level(Sonnentag and Natter, 2004). When fatigue builds up, there is asense of urgency to take a break and this is described as the need forrecovery (Sonnentag and Zijlstra, 2006). This is an emotional state

e and burnout in cancer workers in Queensland, European Journal of

M.G. Poulsen et al. / European Journal of Oncology Nursing xxx (2014) 1e62

characterized by reluctance to continue the present demands. Highneeds for recovery are associated with burnout and are consideredto be an important precursor for developing health problems(Sluiter et al., 1999).

The Job Demands-Resources (JD-R) model (Bakker andDemerouti, 2007) explains how low resource availability within ademanding work environment contributes to high recovery needs(Sonnentag and Zijlstra, 2006) Recovery experience is an importantmediator between demand-resource imbalance and wellbeing. Ithas been postulated that insufficient recovery contributes to poorwell-being and health problems (Geurts and Sonnentag, 2006).

Physical activity after work has been shown to be an importantfactor in recovery (Rook and Zijlstra, 2006) as well as improvinglevels of subjective well being (Poulsen et al., 2012). Relaxation isanother factor which contributes to the recovery process bydiverting attention away from work and reversing the negativeconsequences of work-related stress (Hahn et al., 2011b). Havinginsufficient time for relaxation increases the need for recoverywhich in turn leads to emotional exhaustion and sleep disturbance(Sonnentag and Fritz, 2007).

Daily recovery has been viewed as internal (i.e. happening atwork) or external (i.e. occurring after work). Internal recovery mayoccur as a short break during work hours e.g. a coffee break withcolleagues. External recovery before or after work, during weekendsand holidays refers to engaging in replenishing activities that helprebalance suboptimal systems and return stress-related reactions topre-stressor levels before the nextworking period commences e.g. avisit to the gym (Geurts and Sonnentag, 2006). When the demandsof the job increase, there is an increase in the need for recovery.

Recovery experiences have been classified as psychologicaldetachment, relaxation, mastery experiences and control over lei-sure time (Sonnentag and Fritz, 2007). Psychological detachment isdefined as the sense of being away from work both physically andmentally. People who disconnect from work in the evening and donot engage in perseverative ruminating about past stressors oranticipation of future stressors, are more likely to experience pos-itive mood and less fatigue on the following day (Sonnentag andBayer, 2005). Psychological distancing is an important buffer thatprotects workers' well-being and promotes work engagement(Sonnentag et al., 2010). Relaxation may involve activities such aslistening to music or going for a jog. Mastery experiences involvetaking on a new challenge outside of work such as learning a newlanguage or learning a new skill. Mastery and control over leisuretime have been shown to be negatively associated with emotionalexhaustion and positively associated with life satisfaction(Sonnentag and Fritz, 2007).

The purpose of this paper was to explore the relationship be-tween perceived recovery experience, burnout, psychological well-being and work engagement. It was postulated that cancer workerswho were burnt out were less likely to engage in activities thatwould lead to recovery while engaged workers were more likely toparticipate in recovery experiences. Possible explanatory factorsassociated with recovery experiences were analysed.

Methods

Sample selection and processes

Ethical clearance was obtained from the Queensland HealthState Ethics Committee to allow the line managers at fiveQueensland hospitals to be approached to distribute a survey to allcancer care workers at their hospitals. This included nursing,medical, allied health, administration, radiation therapy, physicsand research staff. Major metropolitan and regional services withinQueensland from private and public hospitals were selected.

Please cite this article in press as: Poulsen, M.G., et al., Recovery experiencOncology Nursing (2014), http://dx.doi.org/10.1016/j.ejon.2014.08.003

Eligible centres were required to have both radiation and medicaloncology services on site. Information about the study was pro-vided to line managers and this was supplemented by start upmeetings to provide information to prospective participants. At thetime of the survey, 1016 oncology staff members were employed inthe five participating hospitals. Each staff member received asealed, unaddressed survey package containing a cover letter, a 10-min survey booklet, a participant information sheet and a pre-paidreturn envelope. Reminder emails were sent to the line managerstwo weeks later.

Survey forms were unidentifiable by individual or institution. Assurveys were returned to researchers they were assigned a codenumber according to date of return.

Instruments

Demographic information included age, gender, marital status,financial information and self-reported health issues. Details ofparticipants' work situations included: professional stream; yearsof experience; hours worked per week; hours spent in direct pa-tient care; metropolitan or regional workplace; and public or pri-vate work sector.

Recovery experience was measured using the instrument devel-oped by Sonnentag and Fritz (2007). This required respondents toindicate on a 5-point Likert scale how much they agreed with thestatements, such as: “I forget about work”; “I get a break from thedemands of work”; and “I use time to relax and I take time forleisure”, with higher scores indicating higher recovery.

Burnout was assessed using the Oldenberg Burnout Inventory(OLBI) (Demerouti et al., 2002). The OLBI measures burnout withtwo dimensions: emotional exhaustion and disengagement, hasbeen shown to have high levels of reliability and validity. The 16-item self-report scale consists of 8 items evaluating EmotionalExhaustion, and 8 items for Disengagement. Respondents areinvited to rate statements, such as, “After my work I usually feelworn out and weary” on a 5-point Likert rating scale from0 (strongly agree) to 4 (strongly disagree).

Work engagementwasmeasuredwith the abbreviated version ofthe Utrecht Work and Wellbeing Survey (UWES-9) (Schaufeli et al.,2002). The UWES-9 is a recognised tool with moderate to highreliability and validity for the measurement of work engagement.This scale measures 17 items and has three elements of workengagement: vigour (six items, a ¼ 0.78 and 0.79 in differentsamples); dedication (five items, a¼ 0.84 and 0.89 in two samples);and absorption (six items, a ¼ 0.73 and 0.72 in independentsamples).

Psychological distress was measured using the Kessler 6-itemdistress scale (Kessler et al., 2002). It has demonstrated high in-ternal consistency and reliability (Cronbach's alpha ¼ 0.89). Theshort K6 demonstrates adequate clinical diagnostic precision (17).Respondents rate the frequency with which they experience sixsymptoms on a five-point scale (0 ¼ “None of the time” to 4 ¼ “Allof the time”). The possible range of total scores is 0e24 with higherscores indicating greater psychological distress. Burnout, workengagement scores and psychological distress were dichotomizedinto lower two thirds and upper one third for their respectivescores.

Self-reported physical health was classified into musculoskeletal,cardiovascular, cancer, respiratory and other diseases and thencategorised into: nil; ‘yes’, but not requiring treatment, ‘yes’, havingtreatment with low impact on daily life; and ‘yes’, receiving treat-ment with high impact on daily life.

Physical activity was defined as being: strenuous (heart beatsrapidly and you ‘huff and puff’); moderate (heart rate rises but notexhausting); and light (e.g. easy walking).

e and burnout in cancer workers in Queensland, European Journal of

M.G. Poulsen et al. / European Journal of Oncology Nursing xxx (2014) 1e6 3

Sample description

There were 573 patients who returned questionnaires. Fiftythree percent of the workforce was aged >35 years and 80% werefemale. Approximately half the staff were married and had chil-dren. Three quarters of the workforce were full-time workers(working 38 h ormore per week) with the remainder working part-time (less than 38 h per week). One third of the employees wereshift workers and these came predominately from the nursing andradiation therapy professional streams. Approximately half thework force had less than 5 years experience and almost threequarters worked in a metropolitan hospital.

Nursing represented the largest participant group with 211 re-spondents followed by 122 radiation therapists. Allied health andadministration returned 70 and 72 surveys respectively, while 51medical staff responded. The smallest professional groups werephysics and research, comprising 22 individuals in each group.

Statistical analysis

Potential explanatory variables included: shift work; years ofexperience; hours worked per week; hours of direct patient care;public versus private work sector; metropolitan versus regional;professional stream; gender; age; post-graduate qualifications;marital status; having children; elder- and other-carer commit-ments; self-reported physical and mental health. The dependentvariable used was the summated score for RES. Normality of thesescores was examined using ShapiroeWilk and ShapiroeFranciatests for normality, in addition to histogram and quintileequintile(QQ) plots. Univariate analyses were performed on explanatoryvariables to examine their possible association with the dependentvariable of RES. Total Burnout, Psychological Distress and WorkEngagement Scores were dichotimised into upper one third andlower two thirds for comparison of mean RES scores. P-values werederived from t test (when comparing two groups) or ANOVA (whencomparing more than two groups).

The data were examined for possible collinearity betweenexplanatory variables. Explanatory variables associated withoutcome measures at the 10% level of significance at the bivariatelevel were considered for inclusion in multiple regression model-ling. The analysis was carried out using the statistical package StataV10.1 (Stata Corp, College Station, TX, USA).

Results

Response rate

There were 573 questionnaires returned from 1016 surveysdistributed, representing a 56% response rate.

Univariate analysis

Themean RES was compared for a number of organizational andindividual explanatory variables. Two organizational variableswere associated with RES. Working 21e40 h had the highest RES(p ¼ 0.0009) compared to working less than 20 h or greater than40 h. Professional stream was also statistically significant(p ¼ 0.036) with the highest RES scores occurring in the radiationtherapists and physicists and the lowest levels occurring in themedical staff and research workers. Shift work, years of experience,hours of direct patient care, public versus private, location ofpractice and the area of oncology were not significant explanatoryvariables.

There were eight individual factors associated with RES.Workers less than 25 years of age had the highest level of RES

Please cite this article in press as: Poulsen, M.G., et al., Recovery experiencOncology Nursing (2014), http://dx.doi.org/10.1016/j.ejon.2014.08.003

followed by those older than 45 years. The lowest level of RESoccurred in those workers age 35e45 years (p ¼ 0.015). Workerswithout a postgraduate qualification experienced higher RES scoresthan those with a qualification (p ¼ 0.022). With regard to maritalstatus, those with a partner had the highest RES followed by thosewho were divorced/separated/widowed, married and single(p ¼ 0.019). Workers with no children had higher RES than thosewith children (p¼ 0.002), similar to higher RES for workers withoutcarer commitments versus those with carer commitments(p ¼ 0.003). There was also a strong correlation between RES andself-reported physical health (p ¼ 0.000), psychological distress(p ¼ 0.000) and burnout (p ¼ 0.000). Individual factors that werenot statistically significant on univariate analysis were gender, in-come and work engagement. The results of univariate analysis offactors associated with RES are summarized in Table 1.

Multiple regression analysis

Burnout was negatively correlated with RES (p ¼ 0.002). Psy-chological distress was significantly inversely associated with RES(p < 0.0001). Workers greater than 25 years of age had lower RESscores than those workers younger than 25 years of age with thelowest levels of RES occurring in the 35e45 year age group(p < 0.0001). Single workers had higher levels of RES than marriedworkers (p ¼ 0.025), as did those with no carer commitments(p ¼ 0.014). There was a strong negative correlation between RESscores and receiving treatment for respiratory disease (p < 0.0001).RES scores were strongly associated with the levels of strenuousactivity carried out by workers each week with workers exercisingmore than 4 times per week experiencing the highest scores(p ¼ 0.015), and those exercising strenuously only once per weekstill reporting high RES (p ¼ 0.043). These results are summarizedin Table 2.

Discussion

To the best of our knowledge, this is the first reported cross-sectional study of recovery experience in cancer workers. Thedata suggests that workers reporting high levels of burnout are lesslikely to participate in recovery experiences. We were not able toestablish a significant association with work engagement andperceived recovery experience. Professional stream was not inde-pendently associated with the levels of perceived recoveryexperience.

These data support the theoretical framework of the Jobs-Demand Resources Model (Schaufeli and Bakker, 2004) which as-sumes that job resources such as participating in recovery experi-ences, social support from supervisors and colleagues, andperformance feedback can individually or collectively contribute tomotivational processes that impact work engagement. High levelsof work engagement are negatively associated with burnout,although the two can co-exist (Poulsen et al., 2011). It is postulatedthat there is a strong relationship between burnout and workengagement on the one hand, and predictors (i.e., work pressure,autonomy) and outcomes (i.e., organizational commitment, mentalhealth) on the other hand (Demerouti et al., 2010).

Restoring or replenishing job and personal resources are ofcentral importance in facilitating recovery. High levels of negativestrain, occurring when job demands are high, act in a manner todiminish resources. Resource-demand imbalance will have adetrimental effect on workers unless recovery is possible(Demerouti et al., 2009).

While there has been considerable literature on the conse-quences of high demands in the workplace, there has been lessemphasis on the role of recovery during non-work periods. Daily

e and burnout in cancer workers in Queensland, European Journal of

Table 1Univariate analysis of factors associated with recovery experience score.

Factors N Mean Std. Dev. p-Valuea

Shift work 0.345Yes 197 18.27 6.32No 343 18.81 6.53

Years of experience 0.435�5 276 19.03 6.316e10 115 18.22 6.0811e15 76 18.70 6.6716 or more 102 17.93 7.12

Hours worked per week 34 17.91 6.43 0.0009�20 444 19.11 6.3321e40 88 16.33 6.74>40

Hours worked in direct patient care per week 0.467<10 70 18.86 7.3411e30 212 19.02 5.68>30 200 18.26 6.86

Sector of main job 0.98Public 525 18.59 6.47Private 42 18.62 6.37

Location of main practice 0.713Metropolitan 426 18.57 6.76Regional 134 18.81 5.39

Area of oncology 0.15Radiation oncology 265 19.17 6.19Medical oncology 95 18.29 7.11Haematology 86 17.59 6.61Palliative care 28 20.04 4.642 or more 22 20.23 5.71

Professional stream 0.036Administration 72 19.50 7.83Allied Health 70 18.71 5.85Medical 51 16.78 6.08Research 22 16.59 8.53Nursing 211 18.15 6.52Radiation Therapy 122 19.80 5.58Physics 22 19.73 4.33

Gender 0.101Male 113 19.45 6.21Female 460 18.34 6.47

Age (years) 0.015<25 69 20.61 5.6625e35 193 18.46 6.0235e45 140 17.57 6.78>45 169 18.66 6.70

Postgraduate qualifications 0.022No 322 19.13 6.73Yes 247 17.89 5.91

Marital status 0.019Single/never married 132 19.29 5.94Married 298 17.75 6.46Partner/cohabiting 95 19.61 6.27Divorced/separated/widowed 46 19.50 7.29

Children 0.002No 274 19.46 5.95Yes 297 17.75 6.76

Other carer commitments 0.003No 510 18.84 6.42Yes 57 16.14 5.95

Income per year ($) 0.305<40,000 69 18.29 6.5640,000 to under 60,000 181 19.39 6.5360,000 to under 80,000 151 18.58 6.41

80,000 to under 100,000 92 17.74 6.48>100,000 59 18.14 6.06

Self-reported physical health 0.000

Table 1 (continued )

Factors N Mean Std. Dev. p-Valuea

Poor/fair 70 14.83 7.01Good 211 17.91 6.35Very good 222 20.13 6.04Excellent 69 19.62 5.73

Psychological distress score 0.000Lower 2/3rd 362 20.12 5.96Upper 1/3rd 209 15.86 6.41

Total burnout score 0.000Lower 2/3rd 339 19.78 6.05Upper 1/3rd 198 16.61 6.69

Total work engagement score 0.955Lower 2/3rd 360 18.52 6.41Upper 1/3rd 192 18.49 6.57

a p-Values are derived from t-test or ANOVA.

M.G. Poulsen et al. / European Journal of Oncology Nursing xxx (2014) 1e64

Please cite this article in press as: Poulsen, M.G., et al., Recovery experiencOncology Nursing (2014), http://dx.doi.org/10.1016/j.ejon.2014.08.003

recovery is crucial for the maintenance of wellbeing and job per-formance. Conceptually, there is some similarity in the symptomsof fatigue, psychological distress and the expression of an urgentneed for recovery. In a large Dutch population study it was notedthat these entities referring to resource drain and the need forresource restoration often co-exist (Jansen et al., 2002).

So how does this model help us understand what is happeningto cancer workers and how can the model assist us in reducingburnout and increasing work engagement? Typically, job demandsfor cancer workers are difficult to modify. It is not realistic orpractical to turn cancer patients away because you are overloadedwith work. Neither is it appropriate to place undue delays in

Table 2Multiple regression analysis of factors associated with recovery experience score.

Factors Coeff (SE) p-Value

Total burnout �0.130 (0.042) 0.002

Psychological distress (K6) �0.561 (0.067) <0.0001

Age (<25 as ref)25e35 �2.594 (0.691) <0.000135e45 �3.248 (0.786) <0.0001>45 �2.336 (0.830) 0.005

Postgraduate qualification (no as ref)Yes �1.016 (0.428) 0.018

Marital status (single/never married as ref)Married �1.318 (0.584) 0.025Partner/cohabiting 0.107 (0.656) 0.871Divorced/separated/widowed �0.391 (0.936) 0.676

Other carer commitments (no as ref)Yes �1.762 (0.718) 0.014

Respiratory disease (no as ref)Yes, not requiring treatment 0.163 (0.854) 0.849Yes, having treatment �4.538 (0.851) <0.0001

Strenuous exercise (rarely as ref)Once/week 1.087 (0.537) 0.0432e3 times/week 0.695 (0.535) 0.195�4 times/week 1.619 (0.665) 0.015

Coeff e Regression coefficient; SE e Standard error of the regression coefficient.R-square¼ 0.40 suggests that 40% of the total variability in wellbeing was explainedby the fitted model.F(15,455)¼ 19.82, p < 0.0001 suggests that the fittedmodel is as a whole significant.Variables significant at 5% level of significance at bivariate level were considered inthe initial base model. The list includes: age, having children, hours worked perweek, professional stream, postgraduate qualification, marital status, other carercommitments, self-reported physical health, psychological stress, total burnout andrespiratory disease.Only significant factors are presented in the table.

e and burnout in cancer workers in Queensland, European Journal of

M.G. Poulsen et al. / European Journal of Oncology Nursing xxx (2014) 1e6 5

treatment. The greatest opportunities for reducing burnout are toimprove job and personal resources. Additionally, developingawareness of the need for recovery when there is an imbalancebetween demands and resources is another strategy that offerspossibilities to increase worker wellbeing and thereby to impacthealth and work engagement. Organisations at an operational levelcan address resource-demand imbalance by ensuring that workershave the appropriate workplace supports in place and that theyreceive timely and constructive performance feedback, as well ashaving opportunities to extend themselves and acquire newknowledge. In the era of advanced communication and being al-ways available by phone or email, it has become harder for workersto detach themselves fromwork during leisure time. Organisationsneed to assist workers in “switching off” from work during leisuretime. In some circumstances the boundaries between work andnon-work domains become blurred and workers themselves needto attempt to segment their work and non-work roles. Employerscan also discourage contact timewith employees when they are notat work such as restricting emails to work time. While organisa-tions can encourage workers to prioritise recovery time, there is aneed for workers to increase their own awareness of need for re-covery and to implement strategies at work and after work.Workers might be able to learn relaxation techniques or to joinother co-workers in lunchtime opportunities for physical exercise.

It is noteworthy that the findings of this study indicated a strongcorrelation between strenuous exercise and RES. Of all the signifi-cant factors identified in the multivariable analysis, strenuous ex-ercise was the only factor that was potentially modifiable. Theliterature would support the notion that exercise helps workersunwind from work (Rook and Zijlstra, 2006; Sonnentag, 2001) andresults in lower work stress (Norris et al., 1990), reduced absen-teeism (Lechner and de Vries, 1997) and improved job satisfaction(Rosenfeld et al., 1990). Interestingly, while participating indemanding physical activities may restore resource depletion, ifthis is viewed as a negative experience, then the effect is mitigatedor lost (Hahn et al., 2011a).

It was noteworthy that workers with no children, singleworkers, and workers with no caring commitments were eachfactors that were associated with high RES. This finding suggeststhat high RES scores were reported by those cancer care workerswho had sufficient spare time to engage in recovery experiencesduring the out-of-work hours. Although there were some differ-ences in perceived RES amongst the different professional streamson the bivariate analysis, these ceased to be significant on multi-variate analysis.

It was surprising to see that there was not a positive correlationbetween RES and worker engagement. This contrasts with previousresearch by Sonnentag (2003) who examined work-related out-comes of recovery during leisure time in 147 employees over 5days. This showed that daily recoverywas positively related to dailywork engagement during the subsequent workday. Our failure toidentify this relationship is more likely to be a methodologicalfailure rather than a lack of interaction between recovery and workengagement, as we explored general perceptions of these experi-ences over an indeterminate time period.

Another factor that is important in the recovery process isrelaxation (Hahn et al., 2011a). Relaxation is a process characterisedby decreased sympathetic activationwhich leads to a reduced heartrate, muscle tensions and has been associated with positive affect(Fredrickson and Levenson, 1998). The positive emotions generatedfrom relaxation experiences are also essential to the recoveryprocess, as the quality of the time-out activity is regarded as beingan important ingredient of restorative benefits. Relaxation divertsthe attention away from the stresses of work. Having insufficienttime for relaxation increases the perceived need for recovery which

Please cite this article in press as: Poulsen, M.G., et al., Recovery experiencOncology Nursing (2014), http://dx.doi.org/10.1016/j.ejon.2014.08.003

is associated with burnout and sleep disturbance (Sonnentag andFritz, 2007). Impaired recovery through sleep may also impair re-covery from burnout independently of the influence of depression.Workers with high levels of fatigue typically have fewer hours sleepand require extra effort to carry out their work than colleagues withlow levels of fatigue (Demerouti et al., 2009). Information on goodsleep hygiene, incorporating “wind-down” relaxation activities isrecommended as an additional component of recovery-orientedinterventions (Sonnenschein et al., 2007).

This study does have some limitations. The response rate of 56%was adequate but there was the potential for bias from the non-responders. There were also some limitations in the accuracy ofself-reported physical health as workers may have overestimatedor underestimated their true level of physical health. The effect ofsleep patterns and the types of relaxation that workers engage inwas not included in the survey suggesting that there may be manyother factors which should have been included in the model. The R-square for the multivariate model was 0.4 suggesting that 40% ofthe total variability in recovery was explained by the fitted model.Cross-sectional studies such as these are not able to show causality-for instance do burnout workers fail to participate in recovery ac-tivities because of exhaustion or are workers who are participatingin recovery activities more protected from burnout. In this regardstudies such as these can only be exploratory rather aetiological.

The relationship of RES to burnout needs to be prospectivelyassessed in an interventional study where burnout data arecollected pre and post Recovery Intervention in a randomizedcontrolled or wait listed control study.

Conclusions

There is a strong negative association between recovery expe-rience and both burnout and psychological distress in cancerworkers, but not work engagement. High recovery experience wasmore likely to occur in workers less than 25 years old, singleworkers and those without carer commitments. Participating instrenuous physical activity was also correlated with high RES. It isrecommended that efforts to reduce burnout in the workplace willbe enhanced by educational strategies and organisational effortsaimed at promoting recovery rather than looking at factors that areless readily modifiable such as workloads and exposure to stressfulexperiences. Increasing worker awareness and access to informa-tion and resources to improve recovery opportunities has the po-tential to act alongside other organisational and culturalimprovements to form the foundation pieces for workplace in-terventions aimed at reducing burnout.

Competing interests

Dr Anne Poulsen is Director of Work Life Balance SolutionsQueensland Pty Ltd. The business in involved in workshops toimprove the well-being of workers and corporate groups. She hasnot received any personal payments for this research.

References

Bakker, A.B., Demerouti, E., 2007. The jobs demands-resources model: state of theart. Journal of Managerial Psychology 22, 309e328.

Coffeng, J.K., Hendrikson, I.J., Duijts, S.F., Proper, K.I., Van Mechelen, W., Boot, C.L.,2012. The development of the Be Active and Relax “Vitality in Practice” (VIP)project and design of an RCT to reduce the need for recovery in office em-ployees. BMC Public Health 12, 592.

Demerouti, E., Bakker, A.B., Vardakou, I., Kantas, A., 2002. The convergent validity oftwo burnout instruments: a multitrait-multimethod analysis. The convergentvalidity of two burnout instruments: a multitrait-multimethod analysis. Euro-pean Journal of Psychological Assessment 18, 296e307.

e and burnout in cancer workers in Queensland, European Journal of

M.G. Poulsen et al. / European Journal of Oncology Nursing xxx (2014) 1e66

Demerouti, E., Bakker, A.B., Geurts, S., Taris, T.W., 2009. Daily Recovery from work-related effort during non-work time. Research in Occupational Stress and WellBeing 7, 85e123.

Demerouti, E., Mostert, K., Bakker, A.B., 2010. Burnout and work engagement: athorough investigation of the independency of both constructs. Journal ofOccupational Health Psychology 15, 209e222.

Fredrickson, B.L., Levenson, R.W., 1998. Positive emotions speed recovery from thecardiovascular sequelae of negative emotions. Cognitive Emotions 12, 191e220.

Geurts, S., Sonnentag, S., 2006. Recovery as an explanatory mechanism in therelation between acute stress reactions and chronic health impairment. Scan-danavian Journal of Work, Environment & Health 32, 482e492.

Girgis, A., Hansen, V., Goldstein, D., 2009. Are Australian oncology health pro-fessionals burning out? A view from the trenches. European Journal of Cancer45, 393e399.

Hahn, V.C., Binnewies, C., Sonnentag, S., Mojza, E.J., 2011a. Learning how to recoverfrom job stress: effects of a recovery training program on recovery, recovery-related self-efficacy, and well-being. Journal of Occupational Health Psychol-ogy 16, 202e216.

Hahn, V.C., Binnewies, C., Sonnentag, S., Mozja, E.J., 2011b. Learning how to recoverfrom job stress: effects of recovery training program on recovery, recoveryrelated self efficacy, and well-being. Journal of Occupational Health Psychology16, 202e216.

Hobfoll, S.E., Johnson, R.J., Ennis, N., Jackson, A.P., 2003. Resource loss, resource gain,and emotional outcomes in inner city women. Journal of Social Psychology 84,632e643.

Jansen, N.W.H., Kant, I.J., van den Brandt, P.A., 2002. Need for recovery in theworking population: description and associations with fatigue and psycholog-ical distress. International Journal of Behavioural Medicine 9, 322e340.

Kessler, R.C., Andrews, G., Colpe, L.J., Hiripi, E., Mroczek, D.K., Normand, S.L., et al.,2002. Short screening scales to monitor population prevalence and trends innon-specific psychological distress. Psychological Medicine 32, 959e976.

Lechner, L., de Vries, H., 1997. Effects of employer fitness program on reducedabsenteeism. Journal of Occupational and Environmental Medicine 39,827e831.

Meijiman, T.F., Mulder, G., 1998. In: Drenth, P.J., Thierry, K. (Eds.), PsychologicalAspects of Workload. Handbook of Work and Organisational Psychology. Psy-chology Press.

Norris, R., Carroll, D., Cochrane, R., 1990. The effects of aerobic and anaerobictraining on fitness, blood pressure and psychological stress and well-being.Journal of Psychosomatic Research 34, 367e375.

Poulsen, M.G., Poulsen, A., Khan, A., Poulsen, E., Khan, S., 2012. Factors associatedwith subjective well-being in cancer workers. Journal of Medical Imaging andRadiation Oncology 56, 347e353.

Please cite this article in press as: Poulsen, M.G., et al., Recovery experiencOncology Nursing (2014), http://dx.doi.org/10.1016/j.ejon.2014.08.003

Poulsen, M.G., Poulsen, A.A., Khan, A., Poulsen, E.E., Khan, S.R., 2011. Workengagement in cancer workers in Queensland- the flip side of burnout. Journalof Medical Imaging and Radiation Oncology 55, 425e432.

Rook, J.W., Zijlstra, F.R.H., 2006. The contribution of various types of activities to re-covery. European Journal ofWork and Organisational Psychology 2006, 218e240.

Rosenfeld, O., Tenebaum, G., Ruskin, H., 1990. Behavioural modifications followingphysical activity program in the Israeli pharmaceutical industry. Journal ofScience and Medicine in Sport 22, 93e96.

Schaufeli, W.B., Bakker, A.B., 2004. Job demands, job resources and their relation-ship with burnout and engagement: a multi-sample study. Journal of Organ-isational Behaviour 25, 293e315.

Schaufeli, W.B., Salanova, M., Gonz�alez-Rom�a, V., Bakker, A.B., 2002. The mea-surement of engagement and burnout: a confirmative analytic approach.Journal of Happiness Studies 3, 71e92.

Shanafelt, T., Chung, H., White, H., Lyckholm, L.J., 2006. Shaping your career tomaximize personal satisfaction in the practice of oncology. Journal of ClinicalOncology 24, 4020e4026.

Sluiter, J.K., van der Beek, A.J., Frings-Dressen, M.H.W., 1999. The influence of workcharacteristics on the need for recovery and experience health: a study of coachdrivers. Ergonomics 42, 573e583.

Sonnenschein, M., Sorbi, M.J., van Doornen, L.J., Schaufeli, W.B., Maas, C.J., 2007.Evidence that impaired sleep recovery may complicate burnout improvementindependently of depressive mood. Journal of Psychosomatic Research 62,487e494.

Sonnentag, S., 2001. Work, recovery activities, and individual well being. Journal ofOccupational Health Psychology 6, 196e210.

Sonnentag, S., 2003. Recovery, work engagement, and proactive behavior: a newlook at the interface between nonwork and work. Journal of Applied Psychology88, 518e528.

Sonnentag, S., Fritz, C., 2007. The recovery experience questionnaire: developmentand validation of a measure for assessing recuperation and unwinding fromwork. Journal of Occupational Health Psychology 12, 204e221.

Sonnentag, S., Bayer, U.V., 2005. Switching off mentally: predictors and conse-quences of psychological detachment from work during off-job time. Journal ofOccupational Health Psychology 10, 393e414.

Sonnentag, S., Binnewies, C., Mojza, E.J., 2010. Staying well and engaged whendemands are high: the role of psychological detachment. Journal of AppliedPsychology 95, 965e976.

Sonnentag, S., Natter, E., 2004. Flight attendants' daily recovery from work: is thereno place like home? International Journal of Stress Management 11, 366e391.

Sonnentag, S., Zijlstra, F.R.H., 2006. Job characteristics and off-job activities aspredictors of need for recovery, well-being and fatigue. Journal of AppliedPsychology 91, 330e350.

e and burnout in cancer workers in Queensland, European Journal of