Background
Presenteeism is most commonly defined as people who attend work in spite of their illnesses [
1]. Compared to absenteeism, the concept of presenteeism is relatively under-studied. However, presenteeism is a global phenomenon that is common among employees of all levels and has been suggested to cause a greater loss to an organisation than the costs attributed to absenteeism [
2] through productivity loss [
3], and future poor health and sick leave [
4]. Although presenteeism has most often been studied in employees, it can also be seen in non-workplace environments, for instance universities, schools and nurseries [
5].
Studies have shown that over 60% of employees have attended work while sick, rising to 90% in some studies of occupations such as physicians [
4]. Antecedents of presenteeism may include feelings of being irreplaceable, a high workload, not being able to afford to take time off, perceiving presenteeism as an organisational norm, or perceiving that you are not sick enough to justify time off work [
1,
6]. However, most research has focused on presenteeism due to chronic conditions [
3]. Presenteeism with acute infectious illnesses such as influenza can arguably present more problems to organisations due to the possibility of workplace epidemics.
Employees who continue to work despite having symptoms of an infectious illness pose a risk to others, especially to people who are vulnerable to diseases, such as patients, the elderly, and children [
7‐
9] . This may be particularly true for health or social care professionals working with vulnerable populations [
10,
11]. In the worst case scenario, presenteeism can even contribute to pandemics, as illnesses circulate within workplaces and education settings [
12]. The problems associated with presenteeism are certainly recognised within the general public: a representative survey by Canada Life Group [
13] reported that 82% of UK workers say they have become ill as a result of a colleague coming into work when they are unwell.
Given its public health importance, in this review we sought to summarise the prevalence of, self-reported reasons and statistical risk factors for presenteeism associated with an infectious illness in workplaces or educational / childcare settings. Our intention was to assist in highlighting possible public health approaches to presenteeism, avenues for future research, and to identify areas where there is potential to develop interventions to reduce presenteeism.
Methods
The reporting of this review adheres to the standards for the Preferred Reporting Items for Systematic reviews and Meta-Analyses [
14].
Search strategy
KK and RL carried out preliminary work testing a variety of different search strategies, to balance both specificity and sensitivity. These were finalised in discussions with GJR and IH. Our final search strategy used terms and associated words for ‘acute infectious illness’ and ‘presenteeism’, joined by the AND function. A copy of our search strategy in MEDLINE is included as Additional file
1. The search strategy was modified for each specific database due to differences in MeSH terms, boolean operators and wildcards. Where possible, searches were limited to articles published in the English language and excluded review articles.
Searches
The following electronic databases were searched with the predefined search strategy: Web of Science, Scopus, and OvidSp (Medline, PsycINFO, and PsycARTICLES). Web of Science and Scopus were included for their cover of the sciences and social sciences, and for the fact that the two resources together complement each other, as neither of them are all inclusive. OvidSp was chosen for its cover of health science journals, and also for its inclusion of the database PsycINFO, and PsycARTICLES.
Review process
KK and RL tested the screening process for one database prior to the full database search. This was to ensure consistency in the screening process and clarify any uncertainties about whether studies met the inclusion criteria or not. RW carried out the full search on 12th October 2018 and initial electronic searches from the different databases were combined using EndNote with duplicates identified and deleted. First, the titles and abstracts were screened for mentions of an empirical study examining presenteeism relating to an infectious illness. If it was not clear from the abstract, the study was taken to full text review. All full text versions of papers that remained potentially relevant were screened in relation to the exclusion/inclusion criteria. Those papers that met the inclusion criteria also had their reference section manually searched for any other potential studies that could be included.
Selection criteria
Studies were eligible for inclusion in this review if they meet the criteria as outlined below:
-
Population. Human population, any age.
-
Exposure. Presenteeism relating to an infectious illness. Presenteeism must be defined as going to work or school while ill.
-
Outcome. The study reported data on prevalence of people attending work or university / school / childcare with an infectious illness, OR the study reported data on risk factors associated with attending work, or university / school / childcare with an infectious illness OR the study reported data on self-reported reasons given for attending work, or university / school / childcare with an infectious illness.
-
Study design. Both qualitative and quantitative studies were eligible. Quantitative studies could be of any design. Articles that did not report on original data, e.g. review articles were excluded.
-
Other limiters: Published in the English language.
Data from the final set of studies were extracted by RW using a data extraction table which was developed for this systematic review. Data extracted included citation, country of study, study design, main characteristics of participants (sample size, mean age, % male), occupation or industry, illness, and results regarding prevalence of presenteeism of those with infectious or suspected infectious illness, and/or risk factors associated with presenteeism and/or reported reasons for presenteeism.
Quality assessment
The quality of all eligible studies was assessed using appropriate quality assessment tools for the relevant study designs. These included the CASP critical appraisal tool [
15] for qualitative and cohort studies, while the Mixed Method Appraisal Tool (MMAT) [
16] was used to assess the quality of other quantitative studies. Instead of using the original yes/ can’t tell/ no answers in checklists, slight adaptation with a final category of “low”, “unclear” or “high” risk was used for each question. Answers “yes”, “can’t tell” and “no” correspond to “low risk”, “unclear risk” and “high risk”, respectively.
Data synthesis and analysis
Because of the expected heterogeneity in study designs and outcomes, we did not plan for any meta-analyses and instead used a narrative synthesis. There is no general consensus on the best way to carry out a narrative synthesis for systematic reviews [
17]. As such we decided to use a weight of evidence approach in order to consider the quality of the studies alongside the results in order to assess the strength of evidence of their findings.
Discussion
In our review, the overall prevalence of infectious illness presenteeism ranged from 35 to 97%. Although a very broad range of estimates, even the lower end of this range is troubling and is likely to result in increased transmission of infection in a workplace or school. This range is in line with previous studies of presenteeism prevalence relating to ill health in general [
1,
41,
42]. Our review found that rates of presenteeism were generally higher in health and social care workers, which matched the results from existing literature [
4,
5]. Again, given the vulnerable populations these groups interact with, this is a source of some concern.
Reported reasons for infectious illness presenteeism could be grouped into three main themes concerning organisational characteristics, job characteristics and personal reasons. Common organisational factors included policy regarding sick leave, with a lack of flexible sick leave and strict attendance control protocol appearing to stimulate presenteeism in employees [
43,
44]. Similarly, applying a policy of sick leave allowance may encourage employees to save their allowance for family emergencies, again leading to an increase in presenteeism [
45]. Other organisational factors concerned the perception of a presenteeism culture. Pressure from organisations, supervisors and colleagues to work while ill, and the urge to maintain a positive relationship with co-workers were often given as reasons. It is possible that employees were reluctant to call in sick due to the fear of receiving negative comments from colleagues or to avoid creating tension with supervisors who may question the legitimacy of their sickness [
46]. Fear of receiving disciplinary actions was also reported, yet it is important to note that punishments were mainly anticipated, and many respondents may not have been penalized.
In terms of job characteristics, a perceived ‘lack of cover’ was widely reported, especially in respondents who were health care professionals. Employees in this sector might find it more difficult to find backups due to their specialized roles or due to general understaffing in the workplace [
32]. Because of these highly specialised roles another related reason concerned professionalism, and the fact it was their duty to provide care to patients and not to disrupt this.
Personal reasons mentioned included “fear of increasing burden on others”, and to avoid feeling guilty [
47]. Moreover, many respondents were unsure about the threshold of taking sick leaves since they were uncertain if their symptoms were severe enough for sickness absence and thought they were not infectious. It is interesting that many physicians also reported similar reasons, despite their relative expertise in this area. This could be explained by the reluctance of physicians to recognize sickness in themselves and their incongruent perceptions of illness, as they may compare their illnesses with their patients’ and conclude they are not sick enough to stay at home [
24,
38].
The risk factors tested for associations with infectious illness presenteeism could be grouped into four main themes: sociodemographic, health, influenza-related behaviour and job characteristics.
For sociodemographic factors, we found inconclusive evidence for the role of gender. This seems to reflect what is found in the general presenteeism literature in which some studies find males tend to exhibit more presenteeism [
48,
49], and others females [
50,
51]. There was some indication that age was associated with presenteeism, with those that were younger showing higher rates of presenteeism, again reflecting findings in the general presenteeism literature [
52,
53] however the results were overall inconclusive.
For health, participants’ general health or prevalence of chronic conditions such as asthma or diabetes was not associated with presenteeism, but having a healthy immune system was. This is at first sight surprising as previous evidence suggests it is those with poorer general health that are at risk of sickness presence [
54‐
56]. Employees with poor health may believe that they are compelled to work due to the time off that they already have taken [
57]. Alternatively, it may be that those who feel they have a healthy immune system believe they can fend of infectious illnesses and therefore not be a risk of transmission to colleagues.
Not surprisingly past influenza-related behaviours and intentions show some associations with infectious illness presenteeism. As is well documented in health behaviour theory [
58] peoples past behaviours and intentions predict their future behaviour, and in this case those who have had a higher number of presenteeism days in the past and intend to go to work with an infectious illness do indeed have higher rates of presenteeism.
For many of the variables which fell under job characteristics and were assessed for associations with presenteeism, the evidence base was weak. Variables such as patient type and working hours had been assessed in two studies showing no associations, but again this limited evidence base is not enough to justify any robust conclusions. By far the most important factor coming under job characteristics was occupation type. We found that working in the health care sector was a risk factor for infectious illness presenteeism, which was in line with pervious literature [
54]. Different studies showed that health care professionals, especially physicians, generally reported more sickness presence [
19,
21,
25,
27]. The daily tasks of health care workers usually involve providing care services, and the relationships between these employees and their clients or students can play a crucial role in job outcomes. It is believed that such relationships would predispose employees to sickness presence [
54].
Quality of included studies
The majority of included studies were cross-sectional, and therefore causal relationships could not be established. Although the overall quality of studies was poor with many studies at a high risk of sampling and non-response bias, we did not find an obvious trend for high quality studies to report more significant results compared to low quality studies. Contrasting results were often found and acknowledged by authors; however, attempts were seldom made to explain such inconsistencies. Some frequently reported reasons, such as lack of cover were often not tested in quantitative studies. Those reasons could be adapted and tested as variables in quantitative studies to provide a more comprehensive result. The exclusive use of self-reported data to measure presenteeism was another limitation. Although it is understandable that objective data is hard to obtain in this situation, additional questions that provide more details of sickness presence may be a better measure.
Quality of this review
This comprehensive review explored the prevalence, reasons and risk factors for infectious illness presenteeism. A major strength is that two independent reviewers went through the screening process for one of the databases, meaning subjective views and human errors were minimized. This process was then followed for the remaining databases by a third independent reviewer and any inconsistencies discussed. However, this present review suffers from various limitations. Firstly, although we searched three large databases, it is possible that we missed some articles that would have fit our criteria. Secondly a large proportion of studies were excluded for not being clear that they were measuring presenteeism because of an infectious illness. It is possible that some of these studies were incorrectly excluded. Thirdly, the quality of our review may be limited by publication bias. It is plausible that some studies with non-significant results were not published. It is also worth noting that many of our included studies had health care professionals as their target populations, as such the findings of this review may be more representative of the healthcare sector than other organisational settings. Additionally, reasons for and variables tested for associations with infectious illness presenteeism were grouped into common themes and presented in our results to ease interpretation, but in spite of our best efforts in clustering them, clear differentiation was not always possible.
Implications for research and practice
The results show that presenteeism is common in employees, leading to an increased risk of disease transmission. Longitudinal studies are now needed to establish causality among variables and provide more substantial evidence regarding risk factors for infectious illness presenteeism which can be subsequently addressed in interventions. It is also important for standardised and objective measurements of sickness presence to be formulated which will help to increase the consistency and comparability of research in this field. Risk factors that provided contrasting results, such as gender, age, and those with a small evidence base such as dependents and various job-related factors should be replicated to verify associations. Other sociodemographic variables such as ethnicity and education were not a focus in existing literature, but further study in this area would be useful. Additionally, many of the existing studies solely focused on infectious illness presenteeism in health care staff, studies should also explore this in other industries and non-workplace environments such as schools and nursery’s in order to explore specific risk factors in different settings. The lack of studies that have tested interventions to reduce infectious illness presenteeism was also striking. As evidence for the risk factors of presenteeism begins to grow, it will be important to use these to develop intervention programmes that address these and cater to the different needs of health sector vs non-health sector organisations and schools.
From the results reported here it, a fruitful avenue for such interventions may be for organisations to promote the legitimacy of taking sick leave and emphasise the negative impacts of presenteeism. At the organisational level, specific job-related risk factors, such as lack of cover, should be identified so that counter measures can be developed. Since many employees are unsure about the threshold for taking sick leave, clear guidelines should also be given regarding what to do when they are sick. To minimise external pressure, it is also crucial for bosses to cultivate an organisational culture that emphasises the importance and benefits of taking sick leave and recognises the potentially hazardous impacts of sickness presence, especially the increased risks of spreading infectious diseases to other employees. At the individual level, workload should be properly managed and monitored. Although reducing workload is not always achievable, skills, resources and techniques can be enhanced to help workers cope with job demands [
56]. Supervisors and managers should act as role models and be supportive of workers who require sick leave. Having a supportive environment could encourage illness disclosure and reduce the negative feelings associated with absenteeism, such as guilt from burdening others [
59].
Even though the above-mentioned strategies may not show immediate results, such integrated and tailored approaches could provide long term economic, social and personal benefits from reducing presenteeism. In particular focusing on the management of presenteeism now, rather than during a pandemic [
60], puts us in a stronger position for when a new pandemic does happen. Developing policies and interventions designed to reduce presenteeism in different organisations, as well as schools and nurseries, will facilitate the rapid implementation of strategies to mitigate the impacts of a new pandemic when the risks are greater to the working population and children.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.