J Occup Rehabil (2011) 21:17–22 DOI 10.1007/s10926-010-9249-1
Sickness Presenteeism Among Swedish Police Officers Constanze Leineweber • Hugo Westerlund • Jan Hagberg • Pia Svedberg • Marita Luokkala Kristina Alexanderson
Published online: 9 June 2010 Ó Springer Science+Business Media, LLC 2010
Abstract Introduction The aim was to describe the prevalence of sickness presenteeism (SP) and to explore possible associations with work characteristics among Swedish police officers. Methods Questionnaire data from 11,793 police officers were analysed. Relative risks (RR) and 95% confidence intervals (CI) for SP were calculated with modified Poisson regression. SP was defined as having gone to work on two or more occasions during the past 12 months despite judging that one’s health would have motivated sickness absence. Results Of the police officers, 47% reported SP. All studied work environment factors were significantly associated with SP. The strongest association was found for stress (RR = 1.46; 95% CI: 1.41– 1.52). Low support from colleagues and low control had higher impact on the risk estimates for SP among older subjects. Adjustment for self-rated health lowered the RRs, however, estimates remained statistically significant. The results indicated that SP was most affected by work environment among subjects with good self-rated health. Conclusions SP was high among police officers. Work environment factors seem to be associated with SP, particularly among subjects with good general health. Keywords Work
Police officers Sickness presence
C. Leineweber (&) H. Westerlund M. Luokkala Stress Research Institute, Stockholm University, 106 91 Stockholm, Sweden e-mail: [email protected]
H. Westerlund J. Hagberg P. Svedberg K. Alexanderson Division of Insurance Medicine, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden
Introduction Sickness presenteeism (SP) is estimated to cost companies more than $ 150 billion a year in the US,  significantly more than sickness absence [2–4]. From the employer’s perspective, SP can have negative consequences for productivity, even if the individual does not feel that he or she should have been sickness absent, and a variety of health problems have been shown to be associated with decreased productivity at work . For the individual, SP may have long-term health consequences and it may also affect colleagues and clients, e.g. through increased risks for contagion, stress or accidents . On the other hand, unless competent replacement personnel are brought in, SP may be preferable to SA in terms of productivity. It has also been argued that longterm SA may have detrimental effects on health [7–10] and that SP may thus in many cases be preferable also for the individual. Few studies have explicitly focused on the possible detrimental effects of SP on health, and longitudinal studies about health consequences of SP are scarce. A recent prospective Danish study found that participants who had gone to work ill more than six times in the year prior to baseline had a 74% higher risk of becoming sick-listed for more than 2 months, even after controlling for several potential confounders ; and in recent Swedish studies SP was found to increase the risk for future sickness absence  and poor general health . However, the evidence for a causal link between SP and negative health effects is still very weak. Data from the Swedish Work Environment Surveys (SWES) indicates that SP increased sharply between 1997 and 2000, in parallel with a dramatic rise in sickness absence . In 1997, 37% of the respondents in SWES
stated that they had gone to work at least twice during the last 12 months despite being so ill that they should have reported in sick. In 2000, the corresponding figure was 53% , in 2007 that rate had decreased slightly to 48% . Several causes or determinants of SP have been identified, both ‘work related’ and ‘personally related demands for presence’ . Among work related demands for presence, time pressure [13, 14], good relationship to colleagues , job stress , high workload , and job insecurity  could be identified. Employees in human services organisations have been found to have higher rates of SP than others, and especially high rates have been found among teachers and health care employee [17, 18]. According to the 2007 SWES, SP is most common in occupations with relatively large individual responsibility, where the work requires that personnel are in place, and where temporary replacements are difficult to find, e.g. primary school teachers, childminders, preschool teachers, and health care personnel . The police force is an important institution in all democratic societies and SP among police officers could affect not only the individual, but also colleagues, clients and citizens. To have a healthy Police Force is of vital importance for every democratic state. At the same time, police officers have an extremely strenuous psychosocial work environment and employees more often have disorders of a psychosocial character than employees in other occupations [19, 20]. The most common complaint is ‘too much to do’, but also relational problems at the workplace are common . Despite the fact that police work involves some very specific occupational exposures, such as shift work, representing ‘law and order’, violent situations, and threats, the association of polices’ work environment and their health is quite unexplored. Some studies have investigated sickness absence among police officers. A recent systematic literature review could identify twentyone studies on sickness absence in the police published in peer-reviewed scientific journals . The main finding of this review was that there was a tendency for higher sickness absence among female police officers, but that more studies in this area are needed to draw scientific conclusions. Studies on SP are even more scarce. We have not found a single published paper describing SP among police officers. A systematic review from 2004 of all studies of SP in general could not identify more than eight relevant studies, several of them based on physicians/nurses . Since then, several studies on SP have been published, however, none of these focusing on the special situation of police officers. The aim of the present study was to describe the prevalence of SP and to explore possible associated work characteristics among Swedish police officers.
J Occup Rehabil (2011) 21:17–22
Methods Between October 8th and 28th in 2007 Statistics Sweden, commissioned by the National Police Board, conducted a web-based, anonymous survey among all employees of the Swedish police force . The aim of the survey was to explore how the employees experience their work and comprised questions covering several areas, e.g. health, support, work stress, control, physical work environment, and leadership. A total of 17,892 persons answered the survey (response rate: 74%) . In the present study the analyses were restricted to respondents working as police officers (n = 11,793). The study was approved by the Regional Ethical Review Board in Stockholm. Sickness Presence Sickness presence (SP) was measured by a single question, which also has been used in nearly identical formulation in other studies: ‘‘How many times during the past 12 months did you go to work even though you should have been off sick due to the state of your health?’’ [12, 17] Response categories were ‘‘I have been sick, but stayed home each time’’, ‘‘Once’’, ‘‘2 to 4 times’’, ‘‘5 to 9 times’’, ‘‘10 times or more’’, and ‘‘I have not been sick during the past 12 months’’. Participants were defined as SP if they stated that they had gone to work twice or more during the past 12 months despite judging that their health would have motivated sickness absence. Demographics Information on sex, age, and seniority (supervisor/superior vs. subordinated) were obtained from the questionnaire. The question about age had 10 response options (\25, 26– 29, 30–34, 35–39, 40–45, 46–49, 50–54, 55–59, 60–64 and 65 years or older) and was converted into a semi-continuous variable by replacing the category number with the mean age for the specific age group (e.g. 25–29 was replaced by 27.5 years). Self-Rated Health Self-rated health was measured by one single question ‘‘How would you describe your general health status?’’ with 5 response alternatives (very good to very poor). For the analyses, self-rated health was dichotomised into good (very good or good) and suboptimal (fair to very poor) in accordance with epidemiological praxis [24, 25]. Variables Regarding Work All variables about work were measured by a scale reaching from 1 (strongly disagree) to 10 (strongly agree)
J Occup Rehabil (2011) 21:17–22
with a higher number indicating a better work environment. Answers to negatively formulated questions, e.g. ‘I often feel under stress while performing my work’, were reversed. Composite measures were constructed including all items available within a specific category. For analyses, answers regarding the different work characteristics were added and the sum dichotomised with the worst 30% indicating a poor work environment. Ergonomics were measured by two items: (1) The equipment/facilities I use at work are adapted to my needs; (2) Other ergonomic conditions at my workplace or in my individual work area are good (Cronbach’s alpha 0.86). Support from the closest supervisor was measured by three items: (1) My immediate supervisor gives me the encouragement I need; (2) When the workload is heavy, my immediate supervisor makes sure that my burden is lessened; (3) I am confident enough to express critical opinions without fear of reprisals from my immediate supervisor (Cronbach’s alpha 0.84). Support by colleagues was measured by three items: (1) The atmosphere in my immediate work team is good; (2) My colleagues give me the encouragement I need; (3) I am confident enough to express critical opinions without fear of reprisals from my colleagues (Cronbach’s alpha 0.77). Leadership was measured by 5 items: (1) My immediate supervisor is good at leading and assigning tasks; (2) My immediate supervisor has the ability to motivate the work team to achieve the set goals; (3) My immediate supervisor has confidence in and relies on co-workers; (4) If problems arises at my workplace, my immediate supervisor takes the initiative to solve the difficulties; and (5) My immediate supervisor encourages an open dialogue at our workplace (Cronbach’s alpha 0.94). Control was measured by three items: (1) I am allowed to make my own decisions about how to perform my work; (2) I have enough authority to be able to effectively take responsibility on the job; (3) I actively participate in and influence the planning of my work (Cronbach’s alpha 0.91). Stress was measured by one item: (1) I often feel under stress while performing my work. Statistics To test differences in prevalence of SP between groups, the Chi-square test was used. To test the effect of work environment on SP, relative risks (RR) and 95% confidence intervals (CI) were calculated using modified Poisson regression . Age, sex, seniority (i.e. subordinate or supervisor), and self-rated health was controlled for separately in several steps. When a significant (P \ .05) interaction term between the work environment factor and the control variable was found we adjusted in a further step
for this interaction. The logarithm of the interaction terms was added to the logarithm of the corresponding work environment factors and the sums were exponentiated in order to get interaction adjusted RR’s. Interaction adjusted confidence intervals were calculated from variances of the sums above. See e.g. Hosmer & Lemeshow  for further details. In a last step, we adjusted for all work environment variables, age, and sex simultaneously. SPSS 17.0 was used for the analyses.
Results Most of the participants were male (75.8%) between 35 and 54 years of age (54.6%). Twenty-two percent were supervisors. Self-rated health was quite good, but 15.8% stated their self-rated health as fair and 2.5% as poor or very poor (Table 1). Background characteristics (age groups, sex, seniority, and self-rated health) and their univariate relationships with SP are shown in Table 1. Sickness presence was common, 46.5% had gone to work even though they thought that they should have been off sick due to the state of their health at least twice during the past 12 months. Significant differences in SP were found for sex, age, seniority, and self-rated health. Men, participants over 54 years of age, and supervisors stated more often that they had not been ill at all. Women and subordinates reported most often extreme SP (10 times or more) as well as participants above 35 years of age. Lower self-rated health was clearly associated with a higher number of periods of SP. Risk ratios for SP and different work environment factors are shown in Table 2. In the unadjusted model, all studied work environment factors (poor support from colleagues and supervisors, poor ergonomics, poor leadership, low control, and high stress) were statistically significantly associated with high SP. Adjustment for sex, age, and seniority did not change the results substantially for any of the work environment indices. However, age showed a significant interaction with support from colleagues, and with control. Taking these interactions into account increased the risk estimates significantly among older participants. When we adjusted for self-rated health, the risk estimates were considerably attenuated, but remained statistically significant. All work environment variables showed a significant interaction with self-rated health, indicating a stronger association between work environment and SP among participants rating their health as very good. After adjustment for the interaction term, poor leadership was no longer statistically significantly associated with SP. After mutual adjustment control, ergonomics and stress remained weakly but statistically significantly associated with SP.
J Occup Rehabil (2011) 21:17–22
Table 1 Basic characteristics for Swedish police officers in the study and the relationship between these characteristics and sickness presenteeism Sickness presenteeism during the last 12 months (Row %) N (col. %)
Not ill at all (%)
Sickness absent each time when ill (%)
2–4 times (%)
5–9 times (%)
10 times or more (%)
Age group \35 years
9,194 (78.0) 2,595 (22.0)
Seniority Subordinate Supervisor
Table 2 Relative risks and 95% confidence intervals for sickness presenteeism for different work factors
Poor support from colleagues
Adjusting for interaction Poor support from supervisor
Adjusting for interaction Poor ergonomics
Adjusting for interaction Poor leadership Adjusting for interaction
Adjusting for interaction Low control Adjusting for interaction
M2 M1 ? age
M3 M1 ? sex
M4 M1 ? seniority
M5 M1 ? SRH
1.19 (1.15–1.24) NS
1.18 (1.14–1.23) NS
1.18 (1.14–1.23) NS
1.12 (1.08–1.16) 1.02 (0.97–1.07)
Model 1 shows the unadjusted univariate (simple) associations between each independent variable and sickness presenteeism, Model 2 adjusts for only age, Model 3 adjusts for only sex, Model 4 controls for only seniority, and Model 5 adjusts for only self-rated general health. NS indicates that the interaction was not significant
Discussion The aim of the present study was to describe the prevalence of sickness presenteeism (SP) and to explore possible associations with work characteristics among Swedish police officers. SP was high among the police officers, but comparable to
the rates found in other Scandinavian studies of SP in the general population [10, 12]. Forty-seven percent of the police officers reported that they had gone to work ill at least twice during the past 12 months. Among the total Swedish work force SP decreased from 51% in 2001 to 48% in 2007, which is nearly identical to the rate we found in our study .
J Occup Rehabil (2011) 21:17–22
Our results indicate that work environment factors are associated with SP, particularly among officers with good general health. In line with findings by Hansen & Andersen , stress and poor social support from colleagues and supervisors were associated with an increased risk of SP. Results from that cross-sectional study  also indicated that insufficient time and resources increased the risk for SP by around 30% after adjustment for several other possible risk factors, including other work environment factors and health. Employees who are stressed might decide to go to work while ill because they know that their work is left undone when being sickness absent , or that their colleagues will become excessively burdened. Another reason to go to work while ill might be the fact that the first day of sick leave is an unpaid day followed by days with reduced wage. However, stress may explain less of the SP among respondents with suboptimal health, which is indicated by the fact that controlling for the interaction term between stress and self-rated health attenuated the association significantly. In addition, the association with poor social support was partly explained by poor self-rated health, which indicates that the social support had less importance for the SP among subjects who rated their general health status as poor. High control, and hence the possibility to adapt work tasks and work pace to one’s health condition, could be hypothesized to increase the risk for SP . However, this seems not to have been the case in the present study. As in previous studies [10, 12], our results showed that low control at work was associated with higher SP. Aronsson  gives several possible explanations, one is that people with high control are healthier and therefore have less SP just because they have fewer ‘‘chances’’ of being ill at work; a hypothesis which is worth further studies and partly supported by our findings. The association between control and SP was somewhat attenuated when we adjusted for self-rated health and decreased further with adjustment for the interaction between control and self-rated health, indicating that control has less impact on SP among respondents rating their health as poor. Few studies have explored the association between aspects of leadership and SP. A recent cross-sectional study found a dose–response relationship between integrity, i.e. an honest, sincere, just, and trustworthy leadership and SP . Having a manager who seldom showed integrity was associated to an increased risk for SP among both men and women. However, in that study, leadership seemed to have less importance for SP than working conditions and general life satisfaction, as the association disappeared after controlling for these variables. In our study the association between leadership and SP did decrease when we adjusted for self-rated health, as it did for all other work
environment factors, and the association became non-significant when controlling for the interaction term, indicating that good leadership may not be able to protect against SP among workers with suboptimal health. The results indicate that police officers who judged their equipment or ergonomic conditions poorly adapted to their needs more often were SP. We would expect a gender-effect, as the equipment often is adjusted to the male norm, but the risk estimates did not change when controlling for age, sex, or seniority. The risk estimates decreased when controlling for self-rated health but the change was not statistically significant. However, a statistically significant drop in the risk was found when adjusting for the interaction between self-rated health and ergonomic conditions. The phenomenon of ‘being on double risk’  is possible for all the work environment factors, as all factors we studied are known to increase the risk for ill-health and subjects with high morbidity have a higher risk for SP than healthy persons. As this study is cross-sectional, we cannot know for sure what is cause, and what is effect. The causal relationship as well as possible health outcomes of SP should be investigated in further studies. The main strength of the study is the very large study population, and that all were included, that is, all police officers in a country, an occupational group with many occupational hazards and high demands on their health. Additionally, the response rate was high, the questionnaire very comprehensive, and the number of non-participants small. Furthermore, the anonymous, web-based questionnaires were administrated by Statistics Sweden, not by the Police Authority, providing a lower risk of biased responses from fear of the employers seeing the answers. However, the anonymity made it impossible to conduct detailed attrition analyses. Another limitation is that the scales used to measure work environment are not validated. Despite its limitations, this study fills a knowledge gap, especially in view of the important societal function of the police force and the very limited amount of scientific studies on these aspects.
Conclusions Our results indicate that work environment factors are associated with SP among police officers. This seems to be particularly true among subjects with good general health, as the risk ratios were substantially attenuated when adjusting for the interaction between self-rated health and work environment factors. Support from colleagues and control over work tasks interacted with age, indicating that poor support from colleagues and low control are of greater importance for SP among older employees.
22 Acknowledgments This study was financially supported by the Swedish Council for Working Life and Social Research and the National Police Board. All authors are independent from their funders. Conflict of interest statement All authors declare that the answer to the questions on your competing interest form are all No and therefore have nothing to declare.
References 1. Stewart WF, Ricci JA, Chee E, Morganstein D. Lost productive work time costs from health conditions in the United States: results from the American Productivity Audit. J Occup Environ Med. 2003;45(12):1234–46. 2. Hemp P. Presenteeism: at work–but out of it. Harv Bus Rev. 2004;82(10):49–58, 155. 3. Mattke S, Balakrishnan A, Bergamo G, Newberry SJ. A review of methods to measure health-related productivity loss. Am J Manag Care. 2007;13(4):211–7. 4. Loeppke R, Taitel M, Richling D, Parry T, Kessler RC, Hymel P, et al. Health and productivity as a business strategy. J Occup Environ Med. 2007;49(7):712–21. 5. Schultz AB, Edington DW. Employee health and presenteeism: a systematic review. J Occup Rehabil. 2007;17:547–79. 6. Vingard E, Alexanderson K, Norlund A. Swedish council on technology assessment in health care (SBU). Chapter 9. Consequences of being on sick leave. Scand J Public Health Suppl. 2004;63:207–15. 7. Kivima¨ki M, Head J, Ferrie J, Hemingway H, Shipley M, Vahtera J, et al. Working while ill as a risk factor for serious coronary events: the Whitehall II prospective cohort study. Am J Public Health. 2005;95:98–102. 8. Bergstrom G, Bodin L, Hagberg J, Aronsson G, Josephson M. Sickness presenteeism today, sickness absenteeism tomorrow? A prospective study on sickness presenteeism and future sickness absenteeism. J Occup Environ Med. 2009;51(6):629–38. 9. Bergstrom G, Bodin L, Hagberg J, Lindh T, Aronsson G, Josephson M. Does sickness presenteeism have an impact on future general health? Int Arch Occup Environ Health. 2009; 82(10):1179–90. 10. Hansen CD, Andersen JH. Sick at work—a risk factor for longterm sickness absence at a later date? J Epidemiol Community Health. 2009;63(5):397–402. 11. SCB. Andel av de sysselsatta enl. arbetsmiljo¨underso¨kningen fo¨r vald arbetsmiljo¨fra˚ga efter ko¨n, a˚lder och socioekonomisk in˚ r 1997–2007 (Percentage of employed acc. to the delning SEI. A Swedish Work Environment Survey for selected question by sex, age and socio-economic classification SEI. Year 1997–2007). ¨ rebro: Arbetsmiljo¨underso¨kning (Work Environment Survey). O SCB (Statistics Sweden); 2009. 12. Aronsson G, Gustafsson K. Sickness presenteeism: prevalence, attendance-pressure factors, and an outline of a model for research. J Occup Environ Med. 2005;47(9):958–66.
J Occup Rehabil (2011) 21:17–22 13. Hansen CD, Andersen JH. Going ill to work–what personal circumstances, attitudes and work-related factors are associated with sickness presenteeism? Soc Sci Med. 2008;67(6):956–64. 14. Caveley N, Cunningham JB, MacGregor JN. Sickness presenteeism, sickness absenteeism, and health following restructuring in a public service organization. J Manag Stud. 2007;44 (304–319). 15. Elstad JI, Vabo M. Job stress, sickness absence and sickness presenteeism in Nordic elderly care. Scand J Public Health. 2008;36(5):467–74. 16. de Vroome E. Eurofound: prevalence of sickness absence and ‘presenteeism’ European foundation for the improvement of living and working conditions, 2006. 17. Aronsson G, Gustafsson K, Dallner M. Sick but yet at work. An empirical study of sickness presenteeism. J Epidemiol Community Health. 2000;54:502–9. 18. Vingard E, Alexanderson K, Norlund A. Swedish council on technology assessment in health care (SBU). Chapter 10. Sickness presence. Scand J Public Health Suppl. 2004;63:216–21. 19. Arbetsmiljo¨verket. Poliser (Polices). Korta arbetsskadefakta (Short facts about work injury). Solna: Arbetsmiljo¨verket (Swedish Work Environment Authority), 2009. 20. Vilhelmsson B. Poliser (Polices). Korta sifferfakta (Short facts in numbers). Solna: Arbetsmiljo¨verket (Swedish Work Environment Authority), 2003. 21. Ko¨rlin J, Alexanderson K, Svedberg P. Sickness absence among women and men in the police: a systematic literature review. Scand J Public Health. 2008;37(3):310–9. 22. SCB (Statistics Sweden). SCB:s analysmodell med No¨jd-Medarbetar-Index. Medarbetarna om sin arbetssituation. (Statistics Sweden analytic model with satisfied-employee-index. The employees about their work situation). In: Polisen (The Police), editor, 2007. 23. Polisens a˚rsredovisning (Annual report of the Swedish Police, in Swedish). In: Rikspolisstyrelsen (The National Police Board), editor, 2007. 24. Manor O, Matthews S, Power C. Dichotomous or categorical response? Analysing self-rated health and lifetime social class. Int J Epidemiol. 2000;29(1):149–57. 25. Svedberg P, Bardage C, Sandin S, Pedersen NL. A prospective study of health, life-style and psychosocial predictors of selfrated health. Eur J Epidemiol. 2006;21(10):767–76. 26. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(7):702–6. 27. Hosmer D, Lemeshow S. Applied logistic regression. New York: Wiley; 2000. 28. Johansson G, Lundberg I. Adjustment latitude and attendance requirements as determinants of sickness absence or attendance. Empirical tests of the illness flexibility model. Soc Sci Med. 2004;58(10):1857–68. 29. Nyberg A, Westerlund H, Magnusson Hanson L, Theorell T. Managerial leadership is associated with self-reported sickness absence and sickness presenteeism among Swedish men and women. Scand J Public Health. 2008;36:803–11.