Skip Navigation LinksHome > February 2012 - Volume 54 - Issue 2 > Low Workload as a Trigger of Sick Leave: Results From a Swed...
Journal of Occupational & Environmental Medicine:
doi: 10.1097/JOM.0b013e31823fdf68
Original Articles

Low Workload as a Trigger of Sick Leave: Results From a Swedish Case-Crossover Study

Hultin, Hanna PhD; Möller, Jette PhD; Alexanderson, Kristina PhD; Johansson, Gun PhD; Lindholm, Christina PhD; Lundberg, Ingvar PhD; Hallqvist, Johan PhD

Free Access
Article Outline
Collapse Box

Author Information

From the Department of Public Health Sciences, Division of Public Health Epidemiology (Drs Hultin, Möller, and Hallqvist) and Department of Clinical Neuroscience, Division of Insurance Medicine (Drs Alexanderson and Lindholm), Karolinska Institutet, Stockholm, Sweden; National Centre for Work and Rehabilitation, Department of Medicine and Health Sciences, Linköping University, Linköping, Sweden (Dr Johansson); Department of Medical Sciences, Division of Occupational and Environmental Medicine (Dr Lundberg) and Department of Public Health and Caring Sciences, Division of Preventive Medicine (Dr Hallqvist), Uppsala University, Uppsala, Sweden.

Address correspondence to: Hanna Hultin, PhD, Karolinska Institutet, Department of Public Health Sciences, Division of Public Health Epidemiology, Norrbacka 7th floor, SE-171 76 Stockholm, Sweden (hanna.hultin@ki.se).

The authors declare no competing interests.

Collapse Box

Abstract

Objectives: To investigate if exposure to an unusually low workload when ill can trigger taking sick leave.

Methods: A case-crossover design was applied to 546 sick-leave spells obtained from a cohort of 1430 employees within six Swedish workplaces. New sick-leave spells were reported from the workplaces during 3 to 12 months follow-up. Exposure was assessed in structured participant interviews at sick leave. Case and control periods from the same individual were sampled according to the matched-pair and usual-frequency approaches. Results are presented as odds ratios with surrounding 95% confidence intervals.

Results: The odds ratio of sick leave on a day with an unusually low workload was 2.57 (confidence interval, 1.07–6.16).

Conclusions: Becoming ill on a day with a lower workload than usual can trigger the decision to take sick leave.

Sick leave varies over time in the population in a way that does not coincide with the variations in population health.1,2 Moreover, sick leave also varies among individuals with the same diagnosis.1 This implies that to better understand the social phenomenon that is sick leave, we need to look into other factors which may influence the association among illness, sickness, disease, and sick leave.

Previous studies have suggested that nonmedical factors may influence the association between illness and sick leave.37 Factors in the work and home environment, which promote absence or attendance, may be of importance when an individual makes the decision to either take sick leave or to go to work despite illness.1,39 However, empirical research is scarce.

To generate hypotheses regarding possible factors promoting absence or attendance we performed a focus-group study, with 16 Swedish individuals of different sexes, ages, and occupational backgrounds, who all had experience of sick leave during the previous 2 years. The participants were instructed to discuss different aspects they considered when deciding whether to report sick or not when ill. Among others, the results suggested that a low workload may incline superiors to encourage employees to take sick leave and take the opportunity to get well. This could point at the existence of an absence culture, which is more lenient when the workload is low, in a way that would be in line with the work of Nicholson and Johns.10 However, previous empirical studies have mainly focused on the effects of a high workload or high work demands on sick leave.1114 One exception is the study of Gimeno et al,15 which found increased risks of sick leave among individuals in so called “passive jobs” characterized by a combination of low demands and low control. Another result from the focus groups study was that a low workload made it easier to make the decision to take sick leave since the employee did not have to worry about colleagues getting more work tasks. This is in line the with previous studies which indicate that the tendency to take sick leave may be associated with the extent to which the employee experiences attendance requirements at work.4,16 A low workload could also imply that the satisfaction of the employee is lowered which also may reduce the attendance motivation.7 However, previous studies of the association between work satisfaction and sick leave have been conflicting.7,1725 To the best of our knowledge it has never been investigated if becoming ill when having a low workload affects the likelihood of taking sick leave.

A previous study has indicated that aspects at work during the last days before sick leave, or expected work conditions during the first sick leave day, may act as triggers in the decision to report sick when ill.26 To be able to investigate if exposure to a low workload may trigger an individual to report sick when ill, we used a case-crossover design. It is an epidemiologic study design, which has been invented to identify and quantify the effect of intermittent factors that trigger acute outcomes.27,28 It was initially used to study trigger factors of myocardial infarction but has since then also been used to study nonmedical triggers of decision-making processes.26,29,30

The aim of this study was to test the hypothesis that exposure to a lower workload than usual when ill can trigger individuals to take sick leave.

Back to Top | Article Outline

MATERIAL AND METHODS

Study Design

This study is based on data from the Triggers of Sick Leave project, a case-crossover study nested within a cohort, carried out at six Swedish workplaces, between April 2005 and February 2007.

In a case-crossover study, the frequency of exposure during a time period in close proximity to the outcome, the case period, is compared with exposure frequencies during one or more outcome-free control time periods for the same individual.27,28,31 If low workload has a trigger effect, it should be present when ill and specifically when deciding to take sick leave. We therefore investigated two different case periods (Fig. 1); one consisting of the first day of a new sick-leave spell (*), and one extended case period consisting of the first sick-leave day and the previous workday (†). If a respondent worked part of a day and then took sick leave, that day was still defined as the first sick-leave day.

Figure 1
Figure 1
Image Tools

Four different control time periods were defined (Fig. 1): A usual frequency of exposed workdays, on the basis of a 2-week period prior to the first sick-leave day, excluding the case period (‡), and a usual frequency of exposed workdays during the 2 months prior to the first sick-leave day (§), a matched-pair control period corresponding to the second last and third last workdays before the first sick leave day (‖), and a matched-pair control period corresponding to the last workday before the first sick-leave day controlled for weekday (¶).

Back to Top | Article Outline
Study Sample

The six workplaces that participated in the project were selected to cover different geographical regions and three different occupational sectors as follows: manufacturing industry (one manufacturing plant), white-collar office work (one insurance company), and health care (four public and municipal health care facilities). Within each sector, several very different occupations were represented. For instance, the manufacturing plant employed both engineering specialists and manufacturing process operators, the health care facilities both head nurses and janitors, and the insurance company both computer specialists and insurance sales persons. The project group approached the workplaces through the executive management and the project was approved by workplace union representatives. The project was also approved by Stockholm's Regional Ethics Committee and conforms to the principles of the Helsinki Declaration.

Between 30 and 1200 people were employed at each workplace. All employees who were not currently on parental leave, sick leave for more than 30 days, or other leave of absence were eligible for participation in the cohort, and 3020 employees fulfilled these inclusion criteria. The study cohort consists of the 1430 individuals (47%) who returned a baseline postal questionnaire and a consent form sent to their home addresses.

Back to Top | Article Outline
Data Sources

We used three sources of data as follows: (1) baseline questionnaire data regarding health, private life, and work environment, (2) daily reports from the workplaces on start and end dates of all new sick-leave spells among the participants during follow-up, and (3) telephone interview data collected during the first days on sick leave, concerning exposure to potential trigger factors. The follow-up varied between 3 and 12 months at the different workplaces for administrative reasons.

New sick-leave spells were reported daily by e-mail or fax from the employers. The logistics for this was part of an already existing organization for staff management at five of six workplaces; and at one, a specific group of informants was recruited. At one workplace, daily reporting of sick-leave spells was not possible, and instead reporting was carried out on a weekly basis. Deidentified group-level data on sick-leave spells among nonparticipants were also reported to the project.

Back to Top | Article Outline
The Sick-Leave Spells

A sick-leave spell was defined as each time an employee reported sick to their workplace. Thus, all spells are in essence self-certified, although for some of the spells the participant may have received a medical certificate as well. All new spells from participants that were reported during the follow-up, regardless of length and grade, were included, except planned sick leave (ie, for planned surgery). If overlapping or extended spells were reported (ie, two sick-leave dates being reported with no return-to-work date in between), these were combined into one spell. Since the unit of analysis was sick-leave spells, some participants contributed with more than one spell during the follow-up. However, after an initial pilot period, individuals that had participated in three interviews were not contacted if on sick leave again, leaving a total of 877 sick-leave spells eligible for inclusion in the study. For 679 of these spells an interview was conducted, and in 198 (23%) spells the absentee declined or could not be reached. In 111 (13%) of the interviewed spells, the absentee did not have time or strength to complete a full interview, and a shortened version of the interview, with no exposure information, was conducted. In 22 of the spells (3%), more than 14 days had passed before the interview, and they were excluded at analysis. The total number of sick-leave spells included in the analyses was thus 546 (62% of all recorded eligible sick-leave spells). A flowchart of the data collection is presented in Fig. 2.

Figure 2
Figure 2
Image Tools
Back to Top | Article Outline
Exposure

Exposure to a lower workload than usual was assessed by a set of questions in the telephone interview at sick leave. First the respondent was asked, “Have you, at any time during the last year, experienced a low workload at work?” followed by these examples, “maybe you have had a hard time filling your work hours with work tasks because of a lower inflow of orders, being in a period between two projects, more available staff, fewer patients/clients, or having a smaller area of responsibility than usual.” If the respondent answered no to this gate question, he or she was considered as unexposed in the case periods as well as in all control periods. If the respondent answered yes to the gate question, exposure on each workday during a 2-week period prior to the first sick-leave day was recorded in a schedule. Both case periods and the two matched-pair control periods (Fig. 1: control period ‖ and ¶) were extracted from this schedule. The extended 2-day case period was considered as exposed if exposure was reported on either of the days. The 2-week usual frequency control period (Fig. 1: control period ‡) was created by adding all exposed days in the 2-week schedule, excluding the case period. Finally, the respondents were asked to estimate the number of exposed workdays, during the 2 months prior to the first sick-leave day. This was the base for the 2-month usual frequency control period (Fig. 1: control period §).

Respondents who could not answer the gate question regarding exposure during the last year (0.2%), or who could not estimate their exposure on the first sick-leave day (10.9%), were excluded from the analyses. If respondents could not estimate exposure during the last 2 weeks (0.2%) or during the last 2 months (5.5%), they were excluded from the respective analyses based on those questions.

Back to Top | Article Outline
Effect Modifiers

The following variables were extracted from the baseline questionnaire: age, sex, self-rated health, being in partner relationship, presence of children younger than 18 years, percentage of housework performed, baseline adjustment latitude, attendance requirements, and history of sick leave (definitions of these are given in Table 1). Occupational titles from the baseline questionnaire were manually coded into the 2 digit SEI codes (Socioeconomic Classification) of Statistics Sweden.32 At baseline, the respondents estimated their normal work ability on a 0 to 10 scale (0 representing no ability to work at all, and 10 representing having the best possible work ability). In the telephone interview, the respondents estimated their work ability at the time of taking sick leave, using the same scale. On the basis of this, the percentage of normal work ability at time of sick leave was calculated and categorized into three groups, 0% to 20%, 21% to 50%, and 51% to 100%. In the interview, the respondents reported when the first symptoms of the illness, which lead to their sick leave, began. From this date and the date of the first sick-leave day, a dichotomous variable was created, indicating if symptoms began before the first sick-leave day or not. They were also asked if there were any circumstances other than their stated health problem which were of importance when deciding to take sick leave. These were coded into work and/or home related.

Table 1
Table 1
Image Tools
Back to Top | Article Outline
Statistical Analyses

Data analysis of case-crossover data is analogous to that of matched case-control studies. When applying the usual frequency approach (using the usual frequency of exposure during the 2-week and 2-month periods), the odds ratios (OR) were calculated using a Mantel-Haenszel estimator with 95% confidence intervals (CI) for sparse data.27,28,33 In the matched-pair interval approach (using control information based on the matched-pair control periods), conditional logistic regression was used, with each sick-leave spell being regarded as one stratum.28,33

The odds ratios can be interpreted as estimates of the incidence rate ratios of the outcome, comparing exposed to unexposed conditions.28,34,35 Effect modification was assessed by stratifying the analysis based on the case period consisting of the first sick-leave day and 2-week usual frequency control period (control period a) by sex, age groups, socioeconomic status, occupational sector, and the percentage of normal work ability and onset of illness symptoms prior to the first sick-leave day.

The interviewers used special codes to indicate when the respondent was uncertain about whether a specific workday was exposed or not. In the presented results, all days during which such “uncertain events” were reported are considered as exposed, but alternative analyses were also made using stricter coding of such days (data not shown). An estimated sick-leave incidence rate was calculated, for participants and nonparticipants, with person-time based on calendar days, including both the days on sick leave and work-free days.

All analyses have been performed in the software package SAS 9.1 (SAS Institute, Inc, Cary, NC).

Back to Top | Article Outline

RESULTS

Of all cohort participants, 39% reported sick at least once during the follow-up. Sick-leave spells which lasted shorter than 8 days accounted for 83% of all eligible spells and 86% of all spells included in the case-crossover study. About half of the participants included in the case-crossover study worked at the manufacturing plant, but all socioeconomic categories were of approximately similar size except higher nonmanuals, which only accounted for 5% of the participants (Table 1). A large part of the self-reported health disorders at sick leave concerned minor acute infections indicating a clearly lowered work ability (Table 2). Most participants reported that no other circumstance at work or at home had affected their need for reporting sick.

Table 2
Table 2
Image Tools

In 37% (n = 203) of the included spells, the participants reported that they had been exposed to a lower workload than usual at least once during the last 12 months. During the 2 weeks prior to sick leave, 88% reported that none of their workdays had been exposed, 2% reported that all workdays had been exposed, and 9% reported a varying frequency of exposure.

The respondents had a higher tendency to take sick leave on a day when they expected to be exposed to a lower workload than usual, than on days when the workload was not lower than usual (Table 3). The odds ratio of sick leave was 2.62 (95% CI, 1.45-4.74) when using a case period consisting of the first sick-leave day and a 2-week usual frequency control period. When using an extended case period of the first sick-leave day and the workday before that and comparing it to a matched-pair control period consisting of the second and third last workdays, the odds ratio was 4.00 (95% CI, 1.50-10.67). The weekday-controlled matched pair analysis yielded an odds ratio of 2.57 (95% CI, 1.07-6.16). The effect estimate when using a 2-month usual frequency control period was in the same direction but somewhat lower (data not shown).

Table 3
Table 3
Image Tools

Table 4 shows odds ratios greater than 1 in all subgroups, however, not always statistically significant. Although there are some indications of effect modification, the stratum-specific confidence intervals are wide and consequently overlap each other.

Table 4
Table 4
Image Tools
Back to Top | Article Outline

DISCUSSION

This is the first study to show that exposure to lower workload than usual when ill may trigger the decision to take sick leave. Elevated risk estimates were found for analyses with different types of control periods. Both exposure on the first sick-leave day and exposure either on the first sick-leave day or the previous workday implied increased risks of sick leave among the respondents. This may indicate that the effect of experiencing a low workload on the decision to report sick may be more distinct when one has experienced an unusually low workload during the previous workday.

The question is what mechanism this captures? If interpreted in the framework of the previously mentioned focus-group study, the results could indicate that employees are encouraged to take sick leave when the workload is low, we will call this the encouragement mechanism. However, another mechanism might be that a low workload may also make employees feel less satisfied and therefore less motivated to go to work despite illness, the satisfaction mechanism. A third possible mechanism is that employees out of loyalty try to be flexible and schedule their sick leaves to days when they interfere least with activities at the workplace, the flexibility mechanism.

Nicholson and Johns10 has proposed that workplace absence levels and employees absence patterns are affected by a psychological contract that emerges between an individual employee and the organization, which dictates not only the accepted level of absence but also the legitimate reasons for being absent and in what situations the work tasks and the general work situation permit absence. Nicholson and Johns do not specifically consider a lower workload than usual, but using their terminology, our results could be seen as an indication that the psychological contract at the included workplaces may allow a higher level of sick leave when the workload is lower than usual. This would be an indication of the encouragement mechanism presented earlier.

Theoretical reference to the satisfaction mechanism may be found in the Process Model, formulated by Steers and Rhodes7 which focuses heavily on importance of motivation and satisfaction for explaining absence and attendance. The determinants of attendance motivation and satisfaction which Steers and Rhodes mention are relatively stable characteristics of the work environment such as job level, workgroup size, and leadership style. Perhaps as a result of this, motivation and satisfaction are frequently studied as stable individual characteristics predicting sick leave, and the results have been conflicting.7,1725 If a low workload makes an individual less satisfied and therefore less motivated to go to work despite illness, satisfaction could vary from day to day and be part of the explanation for our finding. Steers and Rhodes do not discuss whether the effect of satisfaction on sick leave is explicit, that is, whether the individual chooses not to go to work because of boredom, or whether it is implicitly affecting an individual's choices. In our study, none of the participants indicated that a low workload was one of the reasons for reporting sick. Although a self-reported measurement does not necessarily imply truthfulness, it suggests that if low workload had its' effect through a satisfaction mechanism, at least part of it was implicit to the individual.

Previous studies have shown that individuals experiencing attendance requirements, for instance because they feel that absence introduces inconveniences for colleagues, patients, or customers, may be less inclined to take sick leave.4,3638 When such attendance requirements do not exist, employees may be more inclined to take the chance to rest and get well. Attendance requirement is one concept in the Illness Flexibility Model, a theoretical model aimed at explaining the decision process of the ill individual.3,4,8 When discussing their model, Johansson and Lundberg explicitly state that most of the concepts are likely to have very short induction times.8 Exposure to lower workload than usual may be seen as lack of attendance requirements. The suggested stronger effect among respondents that reported symptoms prior to the first sick-leave day or that state that their work ability was only moderately reduced when taking sick leave, might also imply that individuals who have been ill for a while try to schedule their sick leave to days when the workload is low, which may indicate a flexibility mechanism. However, to make sound statements of such associations one needs a study with higher statistical power.

It is important to note that all three theoretical mechanisms, the encouragement mechanism, the satisfaction mechanism, and the flexibility mechanism, are interconnected, in that they all are assumed to have their effect through motivation. According to the Process Model, low satisfaction causes lowered attendance motivation and in the Illness Flexibility Model low attendance requirements lower the motivation to attend work when ill.3,4,7 Whether the found association is a result of any of these mechanisms is impossible to state from this study, but the presented results indicate that employers and human resource personnel should recognize that employees may consider their expected workload when they are ill and are deciding whether to take sick leave or not.

However, exposure to a low workload was uncommon among the study participants (88% reported not being exposed during the previous 2 weeks). This implies that the absolute effect of exposure to a low workload may be small. The exposure frequency differed somewhat between occupational sectors; exposure during the previous 2 weeks was reported by 11% of manufacturing plant employees, 7% of health care employees, and 12% of office employees. Furthermore, employees from the manufacturing plant contribute to a large part of the exposed cases and the reported ORs. This may affect generalizability; however, when stratifying the results on occupational sector, the estimated ORs were similar for the other occupational sectors, albeit with lower power. Further studies are needed to investigate effect modification by occupational sector.

Back to Top | Article Outline
Strengths and Limitations

To our knowledge, it has never been investigated before whether a low workload can trigger the decision to take sick leave when ill. The study group comprises of individuals from different occupations, workplaces, and geographical regions. The major advantage of the employed case-crossover design is that by comparing individuals to themselves, confounding from stable risk factors of sick leave is eliminated.28

Although confounding from stable risk factors is eliminated by design, confounding from other triggers could still occur in case-crossover studies. The most important potential confounder is that of illness, which can be expected to be a strong trigger of sick leave. Although drawing on the Demand-Control model of Karasek and Theorell39,40 and Csikszentmihalyi's41 Flow Theory, one might hypothesize that a low workload implies exposure to unchallenging, low-skilled work tasks which may possess a health risk to the employees, it is likely that such a relationship requires a longer induction time than one or two workdays. Furthermore, we do not have any reason to expect that illness could cause a low workload. Neither do we expect that the two exposures would share any covarying pattern, but it is important to note that we cannot estimate the effect of exposure to a lower workload than usual independent of illness in this study, since this would require a case series of sick-leave spells without illness. The presented results can be considered as estimates of the trigger effect of becoming ill when being in a period of lower workload than usual, that is, of when the two varying patterns of illness and low workload coincides.

In a previous study we have shown that exposure to psychosocial events at work, such as problems in the relationship with a superior or colleagues, may trigger sick leave.26 Furthermore, we found a triggering effect of exposure to stressful work situation on the first sick-leave day.26 We do not consider any of these as potential confounders since we do not see any reason why exposure to such events would covary with a lower workload than usual. However, it is still possible that multiple exposures may enhance the effect of each other. Unfortunately we do not have the power to adjust for multiple exposures in our analyses, but 27% of the cases that were exposed to a lower workload than usual on the first sick-leave day were also exposed to at least one work-related psychosocial event. Analyses excluding double exposed cases showed approximately same effect estimates as the reported results, but with slightly lower effect estimates (data not shown).

The matched-pair control period was used to handle the possible issue that exposure to a lower workload than usual may vary by weekday, and may come in periods lasting longer than one workday.

Like all studies based on retrospective self-reported information, it is important to consider whether the results may have been effected by recall bias. To minimize the risk that the sick-leave status, or being ill, would influence the exposure information given regarding the case and control periods differentially, neither the respondents nor the interviewers were informed of the hypothetical length of the hazard period. To minimize general memory problems we tried to make the recall period as short as possible and the median time from the first day of sick leave to the interview was 2 days. If exposure is underreported in the control periods due to a longer period of recall than the case periods, it would lead to overestimated effects. However, the similar results for the different types of control information, whether based on 2 weeks or the last two workdays before the case period, are also an argument against important recall bias. The ORs from the 2-month usual frequency control period are smaller rather than larger than the ORs from control periods requiring shorter recall. In additional analyses we coded uncertain exposure events as missing (data not shown), but the effect estimates generated by those analyses were still of same direction and similar magnitude as those reported in Table 3. Although nothing in the explicit wording of the interview questions suggested how exposure might be related to sick leave, social desirability may influence how exposure is reported, especially for the case period, such that respondents may aim to justify their absence by over or under reporting exposure to a lower workload than usual. Our general impression from a separate question on the participants' own view of the reasons for their sick leave is, however, that they foremost stressed the illness and did not reflect much on circumstances in their work environment. If this tendency was operating when collecting exposure information as well, it would lead to an underestimation of the effects.

The results are generated from a strategic sampling procedure and this together with the nonparticipation, on cohort and interview level, must be considered when interpreting the results. Descriptive comparisons of participants and nonparticipants revealed that nonparticipation was more common in the health care sector and among younger employees. Since the design imply that only cases are included (ie, individuals on sick leave), selection bias is mainly a problem if participation, and thereby case selection, is related to exposure in the case period. If individuals exposed to a lower workload than usual in the case period would be more prone to participate, this could imply overestimated ORs. However, we see no indication of this, in fact in 12.5% of the interviews the respondent reported a higher workload than usual during the first sick-leave day. In a previous study we have reported that exposure to a high workload also increased the risk of sick leave.26 To invalidate our findings, exposure to a lower workload than usual would have to be a deterrent of sick leave among the not included sick-leave spells. Since the estimated sick-leave incidence rate was 4.30 spells/1000 person days among nonparticipants compared with 2.85 spells/1000 person days among participants, we find it unlikely that such an association would be operating on a large scale among the nonparticipants. The data consist of mainly short-term sick-leave spells, and although an increased risk was found both among sick-leave spells shorter and longer than 8 days (see Table 4), it is likely that the effect of exposure to a low workload may be different for longer spells. Of the recorded eligible participant sick-leave spells, 38% were not interviewed. However, more than half of the individuals contributing to these excluded sick-leave spells contributed with other sick-leave spells and subsequent interviews during follow-up. The main reasons for declining to be interviewed were lack of time or feeling too ill; however, descriptive comparisons between sick-leave spells included in analyses and sick-leave spells in which only a short interview was conducted, did not reveal any differences regarding self-reported reasons for reporting sick or self-rated work ability at the time of reporting sick. The individuals who declined to be interviewed because of lack of time were mainly individuals that had a very short sick-leave spell and who had already returned to work by the time the interviewers contacted them.

Back to Top | Article Outline

CONCLUSION

This study shows that individuals who get ill on workdays characterized by a lower workload than usual are more likely to take sick leave. Our results indicate that nonmedical factors may have trigger effects on short-term sick leave, but the measures used are new the study needs to be repeated in samples that allow for broader generalization and for investigation of effect-modification.

Back to Top | Article Outline
ACKNOWLEDGMENTS

This work was supported by Karolinska Institutet, the Swedish Council for Working Life and Social Research, the Swedish Research Council, Stockholm County Council, the National Swedish Social Insurance Board, and the Swedish National Institute of Public Health. The researchers are independent from the funders, and the funding sources had no involvement in any step of the research process, from design to submission. The authors thank Professor Olle Lundberg for the input during the design phase of the study, and Jenny Hansson for her invaluable contribution during the data collection.

Back to Top | Article Outline

REFERENCES

1. Alexanderson K, Norlund A eds. Sickness absence: causes, consequences, and physicians' sickness certification practice. A systematic literature review by the Swedish Council on Technology Assessment in Health Care (SBU). Scand J Public Health. 2004;32:3–263.

2. Mulder M. Social Insurance in Figures 2010. Stockholm: Swedish Social Insurance Agency; 2011.

3. Johansson G. The Illness Flexibility Model and Sickness Absence [thesis]. Dep Public Health Sci. Stockholm: Karolinska Institutet; 2007.

4. 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;58:1857–1868.

5. Kristensen TS. Sickness absence and work strain among Danish slaughterhouse workers: an analysis of absence from work regarded as coping behaviour. Soc Sci Med. 1991;32:15–27.

6. Nicholson N. Absence behavior and attendance motivation—a conceptual syntesis. J Manage Stud. 1977;14:231–252.
7. Steers RM, Rhodes SR. Major influences on employee attendance—Process Model. J Appl Psychol. 1978;63:391–407.

8. Johansson G, Lundberg I. Components of the illness flexibility model as explanations of socioeconomic differences in sickness absence. Int J Health Serv. 2009;39:123–138.

9. Kristensen TS. Sygefavaer som coping (Sickness absence as coping) (in Danish). Socialmedicinsk tidskrift. 1995;1:12–17.

10. Nicholson N, Johns G. The Absence culture and the psychological contract—who's in control of absence. Acad Manage Rev. 1985;10:397–407.

11. Boedeker W. Associations between workload and diseases rarely occurring in sickness absence data. J Occup Environ Med. 2001;43:1081–1088.

12. Duijts SFA, Kant J, Swaen GMH, van den Brandt PA, Zeegers MPA. A meta-analysis of observational studies identifies predictors of sickness absence. J Clin Epidemiol. 2007;60:1105–1115.

13. Krantz G, Lundberg U. Workload, work stress, and sickness absence in Swedish male and female white-collar employees. Scand J Public Health. 2006;34:238–246.

14. Labriola M, Christensen KB, Lund T, Nielsen ML, Diderichsen F. Multilevel analysis of workplace and individual risk factors for long-term sickness absence. J Occup Environ Med. 2006;48:923–929.

15. Gimeno D, Benavides FG, Amick BC III, Benach J, Martinez JM. Psychosocial factors and work related sickness absence among permanent and non-permanent employees. J Epidemiol Community Health. 2004;58:870–876.

16. Hultin H, Hallqvist J, Alexanderson K, et al. Low level of adjustment latitude-a risk factor for sickness absence. Eur J Public Health. 2010;20:682–688.

17. Burton JP, Lee TW, Holtom BC. The influence of motivation to attend, ability to attend, and organizational commitment on different types of absence behaviors. J Manag Issues. 2002;XIV:181–197.

18. Hoogendoorn WE, Bongers PM, de Vet HCW, Ariens GAM, van Mechelen W, Bouter LM. High physical work load and low job satisfaction increase the risk of sickness absence due to low back pain: results of a prospective cohort study. Occup Environ Med. 2002;59:323–328.

19. Josephson M, Vingard E. Zest for work? Assessment of enthusiasm and satisfaction with the present work situation and health—a 1.5-year follow-up study. Work. 2007;29:225–231.

20. Marmot M, Feeney A, Shipley M, North F, Syme SL. Sickness absence as a measure of health status and functioning: from the UK Whitehall II study. J Epidemiol Community Health. 1995;49:124–130.

21. Muchinsky PM. Employee absenteeism—review of literature. J Vocat Behav. 1977;10:316–340.

22. Munch-Hansen T, Wieclaw J, Agerbo E, Westergaard-Nielsen N, Bonde JP. Global measure of satisfaction with psychosocial work conditions versus measures of specific aspects of psychosocial work conditions in explaining sickness absence. BMC Public Health. 2008;8:270.

23. Nicholson N, Brown CA, Chadwickjones JK. Absence from work and job satisfaction. J Appl Psychol. 1976;61:728–737.

24. Notenbomer A, Roelen CAM, Groothoff JW. Job satisfaction and short-term sickness absence among Dutch workers. Occup Med (Lond). 2006;56:279–281.
25. Roelen CA, Koopmans PC, Notenbomer A, Groothoff JW. Job satisfaction and sickness absence: a questionnaire survey. Occup Med (Lond). 2008;58:567–571.

26. Hultin H, Hallqvist J, Alexanderson K, et al. Work-related psychosocial events as triggers of sick leave - results from a Swedish case-crossover study. BMC Public Health. 2011;11:175.

27. Maclure M. The case-crossover design: a method for studying transient effects on the risk of acute events. Am J Epidemiol. 1991;133:144–153.

28. Maclure M, Mittleman MA. Should we use a case-crossover design? Annu Rev Public Health. 2000;21:193–221.

29. Eriksson T, Maclure M, Kragstrup J. Consultation with the general practitioner triggered by advice from social network members. Scand J Prim Health Care. 2004;22:54–59.

30. Eriksson T, Maclure M, Kragstrup J. To what extent do mass media health messages trigger patients' contacts with their GPs? Br J Gen Pract. 2005;55:212–217.

31. Maclure M. “Why me?” versus “why now?”—differences between operational hypotheses in case-control versus case-crossover studies. Pharmacoepidemiol Drug Saf. 2007;16:850–853.

32. Statistics Sweden Swedish socio-economic classification (SEI). Reports on Statistical Co-ordination. Stockholm: Statistics Sweden; 1982:4.

33. Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. Philadelphia: Lippincott Williams & Wilkins; 2008.

34. Mittleman MA, Maclure M, Robins JM. Control sampling strategies for case-crossover studies: an assessment of relative efficiency. Am J Epidemiol. 1995;142:91–98.

35. Redelmeier DA, Tibshirani RJ. Interpretation and bias in case-crossover studies. J Clin Epidemiol. 1997;50:1281–1287.

36. Dahl-Jorgensen C, Mo TO, Saksvik PO. Absence as a coping strategy for better health? (In Norwegian). Tidsskrift for Samfunnsforskning. 2002;43:3–29.

37. Hultin H, Hallqvist J, Alexanderson K, et al. Low level of adjustment latitude—a risk factor for sickness absence. Eur J Public Health. 2010;20:682–688. Published Online First: 9 February 2010. doi:2010.1093/eurpub/ckp2240.
38. McKevitt C, Morgan M, Dundas R, Holland WW. Sickness absence and “working through” illness: a comparison of two professional groups. J Public Health Med. 1997;19:295–300.

39. Karasek R, Theorell T. Healthy Work: Stress, Productivity, and the Reconstruction of Working Life. New York: Basic Books; 1990.

40. Karasek RA. Job demands, job decision latitude, and mental strain—implications for job redesign. Adm Sci Q. 1979;24:285–308.

41. Csikszentmihalyi M. Flow: The Psychology of Optimal Experience. New York: HarperCollins; 1990.

Cited By:

This article has been cited 2 time(s).

American Journal of Industrial Medicine
Is Absence Related to Work Stress? A Repeated Cross-Sectional Study on a Special Police Force
Magnavita, N; Garbarino, S
American Journal of Industrial Medicine, 56(7): 765-775.
10.1002/ajim.22155
CrossRef
Plos One
Lack of Adjustment Latitude at Work as a Trigger of Taking Sick Leave-A Swedish Case-Crossover Study
Hultin, H; Hallqvist, J; Alexanderson, K; Johansson, G; Lindholm, C; Lundberg, I; Moller, J
Plos One, 8(4): -.
ARTN e61830
CrossRef
Back to Top | Article Outline

©2012The American College of Occupational and Environmental Medicine

Login

Article Tools

Images

Share