HEALTH POLICY AND SYSTEMS

Job-Related Stress and Sickness Absence Among Belgian Nurses: A Prospective Study Jeroen Trybou, PhD1 , Sofie Germonpre, MSc2 , Heidi Janssens, MSc3 , Annalisa Casini, PhD4 , Lutgart Braeckman, PhD5 , Dirk De Bacquer, PhD6 , & Els Clays, PhD7 1 Post-doctoral research fellow, Department of Public Health, Ghent University, Belgium 2 Lecturer, Department of Healthcare, HU Brussel, Brussels, Belgium 3 Doctoral researcher, Department of Public Health, Ghent University, Ghent, Belgium 4 Post-doctoral researcher, Department of Epidemiology and Health PromotionSchool of Public Health, Free University of Brussels, Brussels, Belgium 5 Professor, Department of Public Health, Ghent University, Ghent, Belgium 6 Professor, Department of Public Health, Ghent University, Ghent, Belgium 7 Assistant professor, Department of Public Health, Ghent University, Ghent, Belgium

Key words Job-related stress, sickness absence, demand-control support, effort-reward-imbalance, overcommitment Correspondence Jeroen Trybou, De Pintelaan 185, 9000 Ghent, Belgium. E-mail: [email protected] Accepted: January 17, 2014 doi: 10.1111/jnu.12075

Abstract Purpose: The purpose of this study was to investigate the influence of job stress on sickness absence of nurses and determine the predictive power of the Demand-Control-Support (DCS) model, the Effort-Reward ImbalanceOvercommitment (ERI-OC) model, and a combination of both. Design: A survey was conducted to measure job stress in a sample of 527 Belgian nurses, followed by prospective data collection of sickness absence (longterm, short-term, and multiple episodes). Findings: Perceptions of job strain and ERI increased the odds for long-term (adjusted odds ratio [OR] = 2.26; 99% confidence interval [CI; 1.27–4.04]) and multiple episodes of sickness absence (adjusted OR = 1.64; 95% CI [1.01– 2.65]). Iso-strain and ERI-OC increased the odds for long-term (OR = 1.75; 95% CI [0.98–3.11]), multiple episode (adjusted OR = 1.93; 95% CI [1.14– 3.26]), and short-term (adjusted OR = 1.69; 95% CI [1.03–2.76]) sickness absence. Conclusions: The combined model of DCS and ERI-OC predicts the odds for long-term and short-term sickness absence and multiple episodes. Clinical Relevance: This study has implications for human resources management in nursing organizations. Nursing administrators are advised to monitor and balance nurses’ job demands and efforts. They should recognize the importance of social support, job control, job rewards, and overcommitment in order to reduce the job stress of nurses.

Nursing is characterized by high levels of job-related stress (McVicar, 2003). In recent decades, nursing job demands have increased remarkably and continue to increase due to a growing aging population, the rapid evolution of medical technologies, and higher patient expectations (Simoens, Villeneuve, & Hurst, 2005). In this challenging environment, healthcare organizations struggle to find and retain an adequate number of nurses (World Health Organization, 2011). This issue is a global research and policy priority (Van den Heede & Aiken, 2013). Specifically, high levels of sickness absence have 292

contributed to the nursing shortage (Buchan & Aiken, 2008). Several studies have demonstrated that the way work and jobs are organized has an influence on employee health (e.g., Weyers, Peter, Boggild, Jeppesen, & Siegrist, 2006) and sickness absence (e.g., Schreuder, Roelen, Koopmans, Moen, & Groothoff, 2010). Different job characteristics such as the level of job demands, the degree of control (e.g., Verhaeghe, Mak, Van Maele, Kornitzer, & De Backer, 2003), and the (im)balance between effort and reward (e.g., Lavoie-Tremblay, O’Brien-Pallas, Gelinas, Desforges, & Marchionni, 2008) have received Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

Trybou et al.

attention in previous nursing studies. Since healthcare organizations struggle with a shortage of nursing staff, these insights are of major importance to reduce absenteeism. Moreover, the absence of nurses causes greater pressure of work for the remaining nurses who are present at work. In turn, this could result in a vicious circle of absenteeism and a chronic shortage of nursing staff. Sickness absence is not only associated with additional costs but could also have consequences for patients. Previous research has revealed that sickness absence can result in patient dissatisfaction and lower quality of care (Aiken, Clarke, Sloane, Sochalski, & Silber, 2002). This prospective study investigated the relationship between work characteristics of nurses and sickness absence. To define work characteristics, we built on two well-established but competing theoretical models: the Job Demand-Control-Support (DCS) model (Karasek, 1979) and the Effort-Reward ImbalanceOvercommitment (ERI-OC) model (Siegrist, Siegrist, & Weber, 1986). Unlike most previous research building on these models, this study combined both models. Both models have received attention in previous empirical research. Overall, it can be concluded that scientific evidence exists in support of the DCS model ( de Lange, Taris, Kompenier, Houtman, & Bongers, 2003) and the ERI-OC model (van Vegchel, de Jonge, Bosma, & Schaufeli, 2005) on a large range of health outcomes and well-being measures. However, a central issue in contemporary research studies is whether these two models can be considered complementary, and thus advantageous to combine, or overlapping, and therefore mutually exclusive. This has led to a call for additional research combining both models. This study is one of the first to examine the combined impact of these work characteristics on sickness absence in nursing using in a single model using a prospective design.

Conceptual Framework Two theoretical models explaining job stress and the related sickness absence have received major attention in past research: the DCS model, developed by Karasek (1979), and the ERI-OC model, conceived by Siegrist et al. (1986). Both models build on the psychosocial work environment to analyze the relationship between working conditions and health outcomes.

Job-Related Stress Among Nurses

as work load, time pressure, and role conflict. Job control represents the extent to which the individual employee has control over his or her tasks and the way work is performed (being able to make task-related decisions and to apply specific skills). Accordingly, different combinations of job demand and job control result in different types of work environments. In particular, jobs characterized by high demands and low control are stressful. This situation is referred to as high job strain (Karasek, 1979). The basic model was extended with “social support at work” and, accordingly, this theoretical framework is referred to as the extended DCS model (Johnson & Hall, 1988). Social support at work refers to the assistance provided by supervisor and colleagues, which can lower the perceived stress induced by high job demands and low job control. Accordingly, jobs with high job strain (jobs with high demands and low control) combined with low social support bear the greatest risks for maladjustment and illnesses. This situation is referred to as iso-strain (van Der Doef & Maes, 1999). Figure 1a provides an overview of the theoretical model.

The Effort-Reward Imbalance-Overcommitment Model The basic ERI model emphasizes the perceived imbalance in effort and reward by nurses. Effort represents job demands or obligations imposed by the employer. Reward consists of wage, esteem, job security, and career opportunities. The model is based on the assumption that high effort in combination with low reward leads to continuous tension and a stressful imbalance (van Vegchel et al., 2005). In addition to these characteristics, overcommitment was incorporated as a third dimension in the extended ERI model. This is a personality characteristic based on the cognitive, emotional, and motivational elements of behavior and reflects an exorbitant ambition in combination with the need to be approved and esteemed (Siegrist, 1998). Accordingly, Siegrist (2002) argued that an imbalance between high effort and low reward increases the risk for poor health and sickness absence (the extrinsic ERI hypothesis) and assumes that an extrinsic ERI in combination with a high level of overcommitment leads to the highest risk for poor health and sickness absence (the interaction hypothesis). A high level of overcommitment is also supposed to increase the risk for poor health (the intrinsic overcommitment hypothesis). Figure 1b provides an overview of the model.

The Job Demand-Control-Support Model The basic Job Demand-Control model considers psychological job demands and job control as essential job characteristics that affect employee well-being. Job demands refer to stressors in the work environment, such Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

A Conceptual Comparison Between DCS and ERI-OC In general, both models integrate sociological and psychological theories to conceptualize the relationship 293

Job-Related Stress Among Nurses

Trybou et al.

Job Demand Control Support -model High Job Demands

Low Job Control -

+

Effort Reward Imbalance -model

DC

Social Support

High Effort

-

DCS

Low Reward -

Extrinsic

Overcommitment Interac on

Job Stress

Job Stress

DC= Demand-Control DCS= Demand-Control-Support

Figure 1. a: Job Demand-Control-Support model. b: Effort-Reward Imbalance model.

between psychosocial working conditions and health outcomes or behavior. Although there are overlapping features between them, notable important differences can also be identified. Nonetheless, theoretical and empirical arguments in support of both models exist, and accordingly they can be considered equivalent. Firstly, while the focus of the DCS model is exclusively on workplace characteristics, the components of the ERIOC model (salaries, career opportunities, and job security) are also linked to labor market factors and the work role in a macro-social and macro-economic perspective (Rydstedt, Devereux, & Sverke, 2007). Secondly, it also can be argued that ERI-OC offers a broader approach in comparison with DCS because it includes both a stress dimension in relation to health (perceived ERI) and a motivational dimension in relation to active behavior and learning (overcommitment). This implies that the subjective meaning of the work experience is taken into account in the development process of job stress (Calnan, Wainwright, & Almond, 2000). Since nurses are professional employees, partly driven by professional motives, the way they interpret and react to an organizational context could be very different from that of nonprofessional employees (Trybou, Gemmel, Pauwels, Hennick, & Claeys, 2013), making overcommitment theoretically a potentially important component. Thirdly, while the DCS model is focused on the critical need of task control, the ERI-OC model is mainly focused on the broader concept of reciprocity and fairness in the social exchange process (Siegrist, 1996). This is based on the assumption that organizational members tend to reciprocate beneficial (detrimental) treatment they experience with positive (negative) work-related behavior (Gouldner, 1960). These three elements mentioned above highlight the difference between both theoretical models and show that the focus of the ERI-OC model is broader than that of the DCS model. On one hand, this broader focus increases our holistic understanding of job stress. Therefore, 294

this can be considered an advantage of the ERI-OC model when compared with the DCS model. However, on the other hand, this induces a less delineated conceptualization of job stress (compared with the clear focus of the DCS model on job characteristics), which could be detrimental in explaining sickness absence.

Methods Study Design and Sampling The data used in this study were collected as part of a larger Belgian cohort study (the BELSTRESS III study) investigating risk factors for sickness absence (Clays, Kittel, Godin, De Bacquer, & De Backer, 2009). Data on sickness absence were collected prospectively through the records of the employer of the participants over 12 months. The study was approved by the medical ethics committee of the University Hospital of Ghent and the Faculty of Medicine of the Universite´ Libre de Bruxelles.

Measurement The questionnaire was composed of existing instruments with sound psychometric properties. The questionnaire included standardized measures for the characteristics of the work environment (based on the DCS and ERI-OC models), individual characteristics (gender, age, and educational level), and health indicators (overweight, smoking, drinking behavior, and suffering from chronic illnesses). Perceptions of job stress by nurses were measured using validated instruments of the DCS (Karasek et al., 1998) and the ERI-OC (Siegrist, 1996) models. Adequate levels of reliability and construct validity were demonstrated in past research, and the use of both measures of work-related stress is therefore justified. Previous studies clearly distinguished job demands, job control, and social support by factor analysis, and Cronbach’s α of the Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

Job-Related Stress Among Nurses

Trybou et al.

DCS instruments was satisfactory (Pelfrene et al., 2001). Similarly, previous studies confirmed the factorial structure of effort, reward, and overcommitment. Cronbach’s α of the ERI instruments was sufficiently high (Siegrist et al., 2004). In the DCS model, psychological job demands (Cronbach’s α = 0.76) were measured by five items that are related to mental work load, organizational constraints on task completion, and conflicting demands. Job control (Cronbach’s α = 0.69) consists of the subscales “skill discretion” (six items) and “decision authority” (three items). Social support (Cronbach’s α = 0.85) was also composed of two subscales, namely “social support from colleagues” (four items) and “social support from supervisor” (four items). Using the ERI-OC model, effort (Cronbach’s α = 0.79) was measured by the sum score of five items, comprising time pressure, interruptions, responsibility, working overtime, and increasing demands. Reward (Cronbach’s α = 0.75) consists of the sum score of 11 items with three subscales: esteem (five items), wage (two items), and job security and career opportunities (four items). Overcommitment (Cronbach’s α = 0.82) was composed of the sum score of six items. Prospective data of sickness absence were collected over 12 months. The registered data were obtained from the human resources departments. Because the recording of sickness absence in Belgium is strictly ruled by law and requires medical certification, we may assume that the sickness absence registration is highly accurate. In our study, the outcome variables duration (short-term and long-term) and frequency of sickness absence were used. Short-term sickness absence corresponds to a minimum of one episode of 1 to 7 \days of sickness absence during a full calendar year. Long-term sickness absence corresponds to a minimum of one episode of 15 or more \days of sickness absence. Finally, multiple episodes of sickness absence correspond to two or more episodes of sickness absence during a full calendar year (regardless of the length of the absence). Gender, age, educational level, smoking, overweight, alcohol consumption, and suffering from chronic illnesses are known risk factors for sickness absence and were therefore considered as possible confounding variables (Allebeck & Mastekaasa, 2004). In our sample, smoking, overweight (body mass index > 25 kg/m2 ), excessive alcohol consumption (for men an average of four glasses a day; for women an average of two glasses a day), and suffering from chronic illnesses (e.g., chronic bronchitis, diabetes) were not predicting sickness absence. In addition, these variables were not related to the psychosocial stressors of the ERI-OC and DCS models and were therefore not retained as confounders. In accordance with contemporary research studies, besides the crude odds Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

ratios (ORs), the ORs while adjusting for basic confounding sociodemographic characteristics were also calculated. Gender (male or female), age (continuous variable), and educational level (continued secondary school or university/college) were retained as control variables to calculate the adjusted OR. The ERI score was calculated following the established method of Siegrist et al. (2004) using the formula e/(r∗c), with e as the effort sum score, r as the reward sum score, and c as the correction factor for different numbers of items in the nominator and the denominator, in this case 11/5. To define job strain, a ratio of demands and control was calculated according to the procedure used in recent studies: the sum score measuring demand is divided by the sum score measuring control, with higher values indicating higher control (Peter et al., 2002). Specifically, to calculate the job strain score, the formula d/(jc∗c) was applied, with d as the sum score for demand, jc as the job control sum score, and c as the correction factor, which here is 9/5. The cut-off for the binary variables ERI and job strain were determined by the median of the frequencies for the ERI score and the job strain score. In this sample of nursing professionals, the cut-off was 1.10 for ERI and 0.82 for job strain. Subsequently, the variables ERI-OC and iso-strain (job strain combined with low social support) were created. In line with previous research (e.g., Clays et al., 2009), dichotomous variables (low vs. high) were created for social support and overcommitment based on the median values.

Data Analysis Chi-square tests were conducted to assess significant differences in sociodemographic variables (gender, age, and educational level). The relation between job stress and sickness absence was examined by means of logistic regression analysis. The ORs and corresponding 95% confidence intervals (CIs) were calculated. Firstly, the association between the different stress models and sickness absence was studied. This was investigated using both the ERI-OC model and the DCS model individually and subsequently by combining both models. Secondly, the relationship between the several components of both stress models and absenteeism was examined. This analysis was performed for the basic models (ERI and demand-control [DC]) and the extended stress models (ERI-OC and DCS). Four exposure categories were made: (a) a category not exposed to any stressor according to the stress models (the reference group), (b) a group of nursing professionals exposed to only job or iso-strain, (c) a category exposed to only ERI/ERI-OC, and (d) a group exposed to both stressors (combined model). Figure 2 provides an overview. Besides studying 295

Job-Related Stress Among Nurses

Trybou et al.

Figure 2. Empirical model.

the crude associations, the relationships were evaluated after adjustment for potential confounding variables (gender, age, and educational level). All analyses were conducted using SPSS 20.0 (SPSS Inc., Chicago, IL, USA).

Results Participants This study focuses on 527 nurses (86.0% women), who were employed in a hospital (65.1%), a general service center for persons with disabilities (27.5%), and a nursing home (7.4%). The participants belonged to randomly chosen healthcare organizations in Belgium. The response rate was 25.5%. Nonrespondents did not differ from respondents with respect to gender or age. Unfortunately, we were not able to test whether nonrespondents differed from respondents with respect to the level of sickness absence. The mean age of the sample was 43.5 years. Table 1 shows an overview of the sample characteristics.

Descriptive Statistics Table 2 summarizes the frequency of the work characteristics and sickness absence in the study population. 296

High ERI combined with overcommitment was experienced by approximately one third of the nursing professionals (n = 179). Similarly, about one third of the participants (n = 176) perceived iso-strain (DC score higher than 0.82 combined with low social support). The combination of a DC score of more than 0.82 and ERI higher than 1.10 was perceived by 35.4% (n = 174) and 19.8% (n = 96) of the nurses also experienced overcommitment and low social support. Regarding sickness absence, only 42.5% (n = 220) of the nurses were absent a single day due to sickness; 43.6% (n = 226) were absent for a short term (1– 7 \days); and 18.9% (n = 98) were absent for a long period (>15 \days); 28.8% (n = 149) were absent multiple times (at least two episodes). No significant differences in gender, age, and educational level were apparent in short-term, long-term, and multiple episodes of sickness absence.

ERI and Job Strain in Relation to Sickness Absence The basic stress indicator of an ERI (extrinsic hypothesis) was not associated with any type of sickness absence (short-term, long-term, and multiple episodes). In Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

Job-Related Stress Among Nurses

Trybou et al.

Table 1. Means, Standard Deviations, Cronbach’s α, and Correlation of Variables

Job demand (JD) Job control (JC) Effort (Ef) Reward (Re) Social support (SS) Overcommitment (OC) ∗

Mean

SD

JD

JC

Ef

Re

SS

OC

31.94 67.97 14.98 29.39 22.61 15.57

6.31 9.15 2.69 4.80 3.74 3.50

0.76 −0.068 0.645∗∗ −0.361∗∗ −0.274∗∗ 0.520∗∗

0.69 0.098∗∗ 0.339∗ 0.343∗∗ −0.061

0.85 −0.253∗∗ −0.130∗∗ 0.554∗∗

0.75 0.709∗∗ −0.391∗∗

0.79 −0.258∗∗

0.82

p < .05; ∗∗ p < .01. Italics = Cronbach’s α.

ERI With Overcommitment and Iso-strain in Relation to Sickness Absence

Table 2. Demographic and Nursing Job Characteristics

Male Female Age (yr) Educational level Continued secondary school University/college Organization Hospital Disability center Nursing home Sickness absence Short-term Long-term Multiple episodes Effort-reward imbalancea Effort-reward imbalancea and overcommitment High job strainb Iso-strainb Effort-reward imbalancea and high job strainb Effort-reward imbalancea and overcommitment and iso-strainb a b

n

%

74 453 SD: 6.9

14 86 Mean: 43.5

126 400

24 76

343 145 39

65.1 27.5 7.4

226 98 149 250 179

43.6 18.9 28.8 49.9 35.8

259 176 174

50.3 34.6 35.4

96

19,.8

Effort-reward imbalance score > 1.10. Demand-control score > 0.82.

contrast, after adjusting for gender, age, and educational level, nurses with high job strain (high demands combined with low control) had a 2.30 times higher odds for long-term sickness absence (95% CI = 1.12–4.70). When both components of ERI and the DC model are combined, nurses with high levels of job stress had a 2.26 times higher odds for long-term sickness absence (99% CI = 1.26–4.04). In addition, a significant positive association was revealed between the combination of high job strain and ERI and multiple episodes of sickness absence (OR = 1.63; 95% CI = 1.01–2.65). This implies higher odds for sickness absence compared with the isolated stress models. Table 3 gives an overview of these results. Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

Similar to the null result of the isolated effects of an ERI, ERI-OC (interaction hypothesis) also demonstrated no significant association with any type of sickness absence (short-term, long-term, and multiple episodes of sickness absence). The same result was found for iso-strain (high demands combined with low control and low social support). However, when ERI-OC and iso-strain were combined, a borderline significant OR was revealed for long-term sickness absence (OR = 1.75; 95% CI = 0.98–3.11). Nurses who experienced ERI-OC and iso-strain had a 1.93 times higher odds for multiple episodes of sickness absence (95% CI = 1.14–3.26) and a 1.69 times higher odds for short-term sickness absence (95% CI = 1.03–2.76). Table 4 gives an overview of these results.

Discussion This paper provides an overview of psychological work characteristics (measured by the DCS and ERI-OC models) that could lead to sickness absence in nursing. The analyses support the hypothesized model only partially. Our results confirm previous evidence demonstrating that the way nursing work and jobs are organized has an influence on sickness absence (Schreuder et al., 2010). Our study adds to this knowledge by demonstrating the supplementary value of combining the DCS and ERIOC models. The combined models predict long-term and multiple episodes of sickness absence significantly. The OR is increased when compared with the application of the isolated models. This is in accordance with the studies of Griep, Rotenberg, and Chor (2010) and Schreuder et al. (2010), which concluded that combining the ERI-OC and DCS models predicts sickness absence to a greater degree. Remarkably, the isolated ERI-OC model was not significantly associated with any of the sickness absence parameters under study (short-term, long-term, and multiple episodes). This finding contrasts with several previous 297

Job-Related Stress Among Nurses

Trybou et al.

Table 3. Results Effort-Reward Imbalance (ERI) and Job Strain Adjusted odds ratioa

Crude odds ratio Short-term SA

No ERI

ERI

No ERI

ERI

No job strain Job strain Long-term SA No job strain Job strain Multiple SA No job strain Job strain

1 1.40 (0.80–2.43) No ERI 1 2.21 (1.09–4.49)∗ No ERI 1 1.35 (0.73–2.51)

1.36 (0.78–2.37) 1.10 (0.71–1.69) ERI 1.60 (0.75–3.41) 2.29 (1.29–4.07)∗∗ ERI 1.57 (0.85–2.89) 1.62 (1.01–2.62)∗

1 1.33 (0.76–2.35) No ERI 1 2.30 (1.12–4.70)∗ No ERI 1 1.26 (0.67–2.36)

1.31 (0.75–2.30) 1.11 (0.72–1.72) ERI 1.63 (0.76–3.48) 2.26 (1.27–4.04)∗∗ ERI 1.53 (0.82–2.84) 1.64 (1.01–2.65)∗

Note. SA = sickness absence. a Adjusted for age, gender, and educational level. ∗ p < .05; ∗∗ p < .01. Table 4. Results of Effort-Reward Imbalance-Overcommitment (ERI-OC) and Iso-Strain Adjusted odds ratioa

Crude odds ratio Short-term SA

No ERI-OC

ERI-OC

No ERI-OC

ERI-OC

No iso-strain Iso-strain Long-term SA No iso-strain Iso-strain Multiple SA No iso-strain Iso-strain

1 1.34 (0.77–2.31) No ERI-OC 1 1.48 (0.76–2.88) No ERI-OC 1 1.34 (0.74–2.45)

1.41 (0.68–1.91) 1.58 (0.97–2.56) ERI-OC 0.99 (0.50–1.96) 1.77 (1.00–3.16) ERI-OC 1.51 (0.87–2.63) 1.81 (1.08–3.03)b

1 1.39 (0.79–2.43) No ERI-OC 1 1.55 (0.79–3.03) No ERI-OC 1 1.34 (0.73–2.47)

1.19 (0.71–2.01) 1.69 (1.03–2.76)b ERI-OC 0.97 (0.49–1.94) 1.74 (0.98–3.11) ERI-OC 1.58 (0.90–2.76) 1.93 (1.14–3.26)b

Note. SA = sickness absence. a Adjusted for age, gender, and educational level. b p < .05.

studies that found significant associations with sickness absence frequency (Schreuder et al., 2010) and long-term sickness absence (Derycke, Vlerick, Van de Ven, Rots, & Clays, 2013). However, our results are similar to the findings of Griep et al. (2010), whose study did not reveal any positive association. Our study therefore confirms the limited predictive value of the ERI-OC model and highlights the importance of studying job stress and sickness absence in greater depth by extending the ERI-OC model with other work characteristics. Similarly, the limited predictive value of the DC(S) model was demonstrated in our study. Job strain (low demand and low control) and iso-strain (low demand, low control, and low support) were only associated with longterm sickness absence during follow-up. This finding contrasts with several previous studies that found significant associations of job strain and iso-strain with sickness absence frequency (Schreuder et al., 2010). However, the limited value of applying the DCS model in an isolated way is illustrated by our study. 298

We contend that job stress and sickness absence are determined by a multitude of working conditions. Therefore, we may conclude that combining or extending the existing stress models is an important avenue for further research. When considering our results of the sickness absence analyses, it is important to note that short-term sickness absence was only explained significantly by the combined and extended model ERI + overcommitment and isostrain. The isolated ERI, ERI-OC, DC (job strain), and DCS (iso-strain) and the combination of DC (job strain) and ERI-OC showed no significantly higher odds of shortterm sickness absence. As demonstrated in the study of Aronsson and Gustafsson (2005), this could be explained in part by the high sense of responsibility characterizing nurses. It can be argued that these professionals experience pressures to go to work, even if they are sick, because patients depend on their care. Similarly, since nursing is characterized by a chronic shortage and therefore nursing managers experience difficulties to replace Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

Job-Related Stress Among Nurses

Trybou et al.

them, sickness absence causes greater pressure of work for nurses who are present at work. Collegial obligations toward coworkers could also increase the pressure to go to work. This in turn could also contribute to prolonged sickness absence and multiple episodes and explains why we did not find statistically significant higher odds for short-term sickness absence. In addition, the nonsignificant findings of short-term sickness absence when both job strain (high demands and low control) and ERI-OC are present contrasts with the significant findings of long-term sickness absence and multiple episodes of absenteeism. It can be argued that these finding are in line with the conclusions of Bakker, Demerouti, de Boer, and Schaufeli (2003). Indeed, while sickness absence duration can be seen as an indicator of health problems, sickness absence frequency could be related to motivational and behavioral features of the nurses. The latter is then considered to be a way to withdraw from adverse work conditions. Therefore, longterm sickness absence is considered as “involuntary absenteeism,” resulting from the inability (i.e., due to illness) rather than the unwillingness to work (Steel, 2003). In contrast, absence frequency is considered as an indicator of “voluntary absenteeism” and can be interpreted as a coping behavior (Schaufeli, Bakker, & Van Rhenen, 2009). From this we can assume that both coping behavior as well as associated health problems can be the reason for absenteeism (Schaufeli et al., 2009). Although the results of this study are significant, it has several limitations. First, the participation rate was relatively low, which may have resulted in a selection bias. However, when comparing the group of participants with the total population of nurses, it appeared that respondents did not differ from nonrespondents in terms of gender or age. Nevertheless, we acknowledge that it is possible that nonrespondents had higher levels of stress and sickness absence. Unfortunately, we were not able to measure this aspect. Even if this was the case, there is no reason to assume that the results presented here would be overestimated as a result of this. Second, due to our limited sample size (527 nurses), we were not able to include both sociodemographic characteristics (gender, age, and educational level) and health indicators (overweight, smoking, drinking behavior, and suffering from chronic illnesses). However, these health indicators were not significant predictors of absenteeism in our sample. The impact of omitting these variables is therefore limited. Finally, the studied constructs job demands and job effort correlated highly. Indeed, when nurses experience high job demands, they can be expected to

Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

perceive high job effort. In addition, overcommitment correlated highly with job demands and job effort. One could argue that ambitious nurses who have a high need for being approved and esteemed are more sensitive to perceptions of high job demands and high job effort. However, we showed that although both models (DCS and ERI-OC) are partly overlapping, they are advantageous to combine. Bundled, they increase insight into the relationship between work characteristics of nursing and sickness absence.

Implications of the Study Our study shows that a considerable number of nurses experience high levels of job stress, which increases sickness absence. Subsequently, the absence of nurses causes greater pressure of work for the remaining nurses who are present, possibly resulting in a vicious circle of absenteeism and a chronic staff shortage. Since the way nursing work and jobs are organized clearly has an influence on employee health and sickness absence, this can be considered to be a management and policy priority for head nurses, nurse leaders, and nurse executives. They should recognize the importance of social support (e.g., a friendly work environment), job reward (e.g., job appreciation by the head nurse), overcommitment (e.g., preserving a healthy work-life balance), and an adequate level of job control (e.g., adequate decision latitude) in order to reduce job stress and prevent sickness absence. In practice, nursing leaders and managers are advised to monitor and balance job demands and efforts experienced by nurses (e.g., by enabling nurses to give feedback on these issues). Since a multitude of work characteristics can contribute to perceived job stress and absenteeism of nursing professionals, a holistic approach concentrating on all identified aspects is advisable.

Conclusions The present findings increase insight into the importance of several work characteristics of nursing to sickness absence. Since the nursing shortage is considered a global research priority, this can be considered an important finding. Furthermore, our study shows that the two dominant models of job stress (DCS and ERI-OC) are complementary and thus advantageous to combine. The combined model of DCS and ERI-OC predicts the odds for long-term, short-term, and multiple episodes of sickness absence significantly.

299

Job-Related Stress Among Nurses

Clinical Resources

r r

Stress at the workplace: http://www.who.int/ occupational health/topics/stressatwp/en/ Protecting workers’ health: work organization and stress: http://www.who.int/occupational health/ publications/stress/en/

References Aiken, L. H., Clarke, S. P., Sloane, D. M., Sochalski, J., & Silber, J. H. (2002). Hospital nurse staffing and patient mortality, nurse burnout, and job dissatisfaction. Journal of the American Medical Association, 288, 1987–1993. Allebeck, P., & Mastekaasa, A. (2004). Risk factors for sick leave—General studies. Scandinavian Journal of Public Health, 32, 49–108. Aronsson, G., & Gustafsson, K. (2005). Sickness presenteeism: Prevalence, attendance-pressure factors, and an outline of a model for research. Journal of Occupational and Environmental Medicine, 47, 958–966. Bakker, A. B., Demerouti, E., de Boer, E., & Schaufeli, W. B. (2003). Job demands and job resources as predictors of absence duration and frequency. Journal of Vocational Behavior, 62, 341–356. Buchan, J., & Aiken, L. (2008). Solving nursing shortages. Journal of Clinical Nursing, 17, 3262–3268. Calnan, M., Wainwright, D., & Almond, S. (2000). Job strain, effort-reward imbalance and mental distress: A study of occupations in general medical practice. Work & Stress, 14, 297–311. Clays, E., Kittel, F., Godin, I., De Bacquer, D., & De Backer, G. (2009). Measures of work-family conflict predict sickness absence from work. Journal of Occupational and Environmental Medicine, 51, 879–886. de Lange, A. H., Taris, T. W., Kompenier, M. A. J., Houtman, I. L. D., & Bongers, P. M. (2003). The very best of the millennium: Longitudinal research and the demand-control-support model. Journal of Occupational Health Psychology, 8(4), 282–305. Derycke, H., Vlerick, P., Van de Ven, B., Rots, I., & Clays, E. (2013). The impact of effort-reward imbalance and learning motivation on teachers’ sickness absence. Stress & Health, 29(1), 14–21. Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American sociological Review, 25, 161–178. Griep, H. R., Rotenberg, L., & Chor, D. (2010). Beyond simple approaches to studying the association between work characteristics and absenteeism: Combining the DCS and ERI models. Work & Stress, 24(2), 179–195. 300

Trybou et al.

Johnson, J. V., & Hall, E. M. (1988). Job strain, work place social support, and cardiovascular disease: A cross-sectional study of a random sample of the Swedish working population. American Journal of Public Health, 78, 1336–1342. Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain: Implications for job-redesign. Administration Science Quarterly, 24, 285–307. Karasek, R. A., Brisson, C., Kawakami, N., Houtman, I., Bongers, P., & Amick, B. (1998). The Job Content Questionnaire (JCQ): An instrument for internationally comparative assessments of psychosocial job characteristics. Journal of Occupational Health Psychology, 3(4), 322– 355. Lavoie-Tremblay, M., O’Brien-Pallas, L., Gelinas, C., Desforges, N., & Marchionni, C. (2008). Addressing the turnover issue among new nurses from a generational viewpoint. Journal of Nursing Management, 16(6), 724– 733. McVicar, A. (2003). Workplace stress in nursing: a literature review. Journal of Advanced Nursing, 44(6), 633–642. Pelfrene, E., Vlerick, P., Mak, R.P., De Smet, P., Kornitzer, M., & De Backer, G. (2001). Scale reliability and validity of the Karasek “Job Demand-Control-Support” model in the Belstress study. Work & Stress, 15(4), 297–313. Peter, R., Siegrist, J., Hallqvist, J., Reuterwall, C., Theorell, T., & the SHEEP Study Group. (2002). Psychosocial work environment and myocardial infarction: Improving risk estimation by combining two complementary job stress models in the SHEEP study. Journal of Epidemiology & Community Health, 56, 294–300. Rydstedt, L. W., Devereux, J., & Sverke, M. (2007). Comparing and combining the demand-control-support model and the effort reward imbalance model to predict long-term mental strain. European Journal of Work and Organizational Psychology, 16(3), 261–278. Schaufeli, W. B., Bakker, A. B., & Van Rhenen, W. (2009). How changes in job demands and resources predict burnout, work engagement, and sickness absenteeism. Journal of Organizational Behavior, 30, 893–917. Schreuder, J. A. H., Roelen, C. A. M., Koopmans, P. C., Moen, B. E., & Groothoff, J. W. (2010). Effort-reward imbalance is associated with the frequency of sickness absence among female hospital nurses: A cross-sectional study. International Journal of Nursing Studies, 47, 569–576. Siegrist, J. (1996). Adverse health effects of high-effort/low reward conditions. Journal of Occupational Health Psychology, 1, 27–41. Siegrist, J. (1998). Adverse health effects of effort-reward imbalance at work: Theory, empirical support, and implications for prevention. In C. L. Cooper (Ed.), Theories of organizational stress (pp. 190–240). Oxford, UK: Oxford University Press. Siegrist, J. (2002). Effort-reward imbalance at work and health. In P. L. Perrewe´ & D. C. Ganster (Eds.), Historical Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

Trybou et al.

and current perspectives on stress and health (pp. 261–291). Oxford, UK: Elsevier Science. Siegrist, J., Siegrist, K., & Weber, I. (1986). Sociological concepts in the etiology of chronic disease: The case of ischemic heart disease. Social Science & Medicine, 22, 247–253. Siegrist, J., Starke, D., Chandola, T., Godin, I., Marmot, M., Niedhammer, I., & Peter, R. (2004). The measurement of effort-reward imbalance at work: European comparisons. Social Science & Medicine, 58, 1483–1499. Simoens, S., Villeneuve, M., & Hurst, J. (2005). Tackling nurse shortage in OECD countries. Report No. 19, OECD health working papers. Retrieved November 26, 2013, from http://www.oecd.org/dataoecd/11/10/34571365.pdf Steel, R. P. (2003). Methodological and operational issues in the construction of absence variables. Human Resource Management Review, 13, 243–251. Trybou, J., Gemmel, P., Pauwels, Y., Hennick, C., & Clays, E. (2013). The impact of organizational support and leader-member exchange on the work-related behaviour of nursing professionals: The moderating effect of professional and organizational identification. Journal of Advanced Nursing, 70(2), 373–382.

Journal of Nursing Scholarship, 2014; 46:4, 292–301.  C 2014 Sigma Theta Tau International

Job-Related Stress Among Nurses

Van den Heede, K., & Aiken, L. (2013). Nursing workforce a global priority area for health policy and health services research: A special issue. International Journal of Nursing Studies, 50(2), 141–142. Van Der Doef, M., & Maes, S. (1999). The Job Demand-Control (-Support) Model and psychological well-being: A review of 20 years of empirical research. Work & Stress, 13(2), 87–114. van Vegchel, N., de Jonge, J., Bosma, H., & Schaufeli, W. (2005). Reviewing the effort-reward imbalance model: Drawing up the balance of 45 empirical studies. Social Science & Medicine, 60, 1117–1131. Verhaeghe, R., Mak, R., Van Maele, G., Kornitzer, M., & De Backer, G. (2003). Job stress among middle-aged healthcare workers and its relation to sickness absence. Stress and Health, 19, 265–274. Weyers, S., Peter, R., Boggild, H., Jeppesen, H.J., & Siegrist, J. (2006). Psychosocial work stress is associated with poor self-rated health in Danish nurses: A test of the effort-reward imbalance model. Scandinavian Journal of Caring Science, 20, 26–34. World Health Organization. (2011). Enabling solutions, ensuring healthcare. Geneva, Switzerland: Author.

301