Assessing the Association Between Emotional Labor and Presenteeism Among Nurses in Korea: Cross-sectional Study Using the 4th Korean Working Conditions Survey.

BACKGROUND: Presenteeism has emerged as an important health-related issue and has been studied in a variety of occupation groups. This study examines the relationship between emotional labor and presenteeism in nurses in Republic of Korea.
METHODS: As a cross-sectional study, our study was conducted on 328 female nurses participating in the fourth Korean Working Conditions Survey (2015). Nurses were identified by the Korean Industry Classification Code. Multivariable logistic regression analysis was performed to explore the association between emotional labor and presenteeism.
RESULTS: Female nurses who always or sometimes hide their emotions in the workplace were found to have a high risk for presenteeism compared with female nurses who rarely hide their emotions in the workplace {odds ratio [OR] = 2.40 [95% confidence interval (CI) 1.04-5.54]; OR = 4.12 [95% CI 1.72-9.84], respectively}. Furthermore, the risk of presenteeism was higher in nurses who sometimes engaged with complaining customers compared with nurses who rarely did so, but it lacked statistical significance.
CONCLUSION: Presenteeism in nurses can cause various negative secondary effects; therefore, an alternative should be sought to mediate nurses' emotional labor to prevent presenteeism.

Introduction

The concept of presenteeism was first introduced in the mid-1990s as an important factor related to productivity within the business administration sector and is still one of the most researched concepts in the field of occupational health [1]. The term has slightly different interpretations. The most representative interpretation involves two aspects: (1) when an employee has come into work but is limited in his or her capacity to work because of health issues [2] and (2) the act of coming into work despite having health issues [3]. Although some definitional confusion will be addressed in what follows, the most recent scholarly conception of presenteeism involves showing up for work when one is ill [4]. Presenteeism is a relatively new concept. While it is personal in nature, it may also be a very serious concern for employers as employee health is directly related to a company's productivity and the employer's profit margins [5,6]. Presenteeism has also received considerable attention through research that showed presenteeism to have higher indirect economic costs because of decreased productivity compared with other related illnesses [7].

Previous research on various occupational groups has establish the prevalence and factors of presenteeism. The major occupational group for research includes healthcare workers such as nurses, doctors, and nurses' aides [[8], [9], [10]]. In particular, nurses' presenteeism seems to require special attention as it may have dire results. When a nurse comes to work with health issues, there can be a decline in concentration that may lead to a decline in the quality of nursing [11]. This has already been recognized as an issue that may place coworkers, patients, and patients' families at risk [12].

Nurses provide medical services as professionals simultaneously with emotional services, such as dealing with patients face-to-face and listening to their complaints. This characteristic of nurses' tasks is inevitably related to emotional labor. The term emotional labor was introduced for the first time in the 1980s as “the management of feeling to create a publicly observable facial and bodily display” [13] and was defined as an individual's efforts, plans, and control necessary to cater to the emotional expression that an organization demands in the interaction between people [14]. Furthermore, as competition between medical institutions intensifies, institutions seek to gain a friendly atmosphere by educating nurses on kindness and enforcing uniform standards of emotional expression [15].

Such emotional labor causes emotional dissonance, which can lead to emotional burnout, exhaustion, low level of job satisfaction, or performance and even depression [14,16]. Previous research has found that stress, burnout, low level of job satisfaction, and depression can all serve as risk factors for presenteeism [17]. This suggests that emotional labor may directly or indirectly affect presenteeism. However, there is a lack of research identifying the relationship between emotional labor and presenteeism and a lack of research on particular occupational groups, such as nurses. Therefore, this study aimed to use data from the fourth Korean Working Conditions Survey (KWCS) to identify the relationship between emotional labor and presenteeism in nurses in Republic of Korea.

Materials and methods

Study participants

This study used data from the fourth KWCS, a survey conducted by the Occupational Safety and Health Research Institute under a government mandate. The survey was conducted in 2014 and used multiarea random sampling based on all workers over the age of 15 years to be more representative of the nation as a whole. The KWCS was developed based on the European Working Conditions Survey. The KWCS is a national open-source database with safeguards to protect the participants' anonymity and privacy rights and consisted of door to door interviews targeting wage workers. Statistics Republic of Korea determined the KWCS information's reliability to increase the usage of its data. The survey's response rate was 33.0%, the cooperation rate 69.9%, and the refusal rate 14.2%. The objective is to provide a better and safer working environment by providing information on the overall working environment of the Korean people, identifying work-related factors, health effects, and accidents.

The total number of people surveyed was 50,007. All the fourth KWCS participants should respond to a questionnaire about their occupations based on the Korean Industry Classification Code; of these, the number of nurses from the total of participants was 395. All 395 nurses were adults over the age of 20 years. For the purposes of this research, all male nurses (n = 19, very few numbers), nonwage worker (n = 1), participants that provided inadequate answers (n = 47), such as “I don't know” or left answers blank) were excluded. Thus, the data of 328 female nurses were selected and used for this research.

Measurements

General and occupational characteristics

Data were collected on general characteristics, including age (20-29 years, 30-39 years, 40-49 years, or ≥ 50 years), level of education (high school level or lower, college, university, or higher), and monthly income [based on posttax amount; <1,500,000 Korean won (KRW), 1,500,000 ~ 1,990,000 KRW, 2,010,000 ~ 2,490,000 KRW, or ≥2,500,000 KRW] and occupational characteristics, including employment status (permanent or temporary), working hours per week (≤40 hours, >40 and <52 hours, or ≥52 hours), shift work (yes or no), number of employees (<5, 5–49, 50–299, or ≥300), and workplace ergonomic risk (low or high). We included workplace ergonomic risk as a confounder for analysis in our study design because musculoskeletal pain is crucial for presenteeism and also workers exposed to various ergonomic risks were vulnerable groups of presenteeism [18,19]. Ergonomic risk variable was calculated, which was the sum of the scores for the five questions: tiring or painful positions, lifting or moving people, carrying or moving heavy loads, standing, repetitive hand or arm movements, dealing directly with people who are not employees at your workplace, such as customers, passengers, pupils, and patients. There were seven original possible responses for this question: “At all times,” “Most of the time,” “¾ of the time,” “Half of the time,” “¼ of the time,” “Rarely,” and “Not at all.” One to seven point was assigned for “Not at all” to “At all times”. Then, the sum of the score were categorized into “High” group and “Low” group, based on the median value [20].

Workplace psychosocial factors

Presenteeism is affected by various psychosocial factors in the workplace. According to previous studies, as an organizational factor, job-related stress factors (such as job control, job demand, support, job satisfaction, and job insecurity), and workplace discrimination were well-known risk factors for presenteeism [17,21]. The aforementioned factors are also included in the KWCS questionnaires.

Based on the questionnaires of the fourth KWCS, the items of workplace psychosocial factor were selected: (1) job control: five questions (able to choose or change order of task, methods of work, speed or rate of work, whether you have a say in the choice of working partners, can take a break when you want); (2) job demand: two questions (working at very high speed, working with tight deadline); (3) social support: two questions (colleagues help and support you, supervisors help and support you); (4) job satisfaction: one question (Overall, what do you think about the work environment that you usually do?); (5) job insecurity: two questions (loss of job within the next 6 months, ability to find a job with a similar wage in case of unemployment); (6) workplace discrimination: 10 questions (Over the past 12 months, have you ever experienced workplace discrimination based on your age, race, nationality, sex, religion, disability, sexual orientation, education level, region, and employment status) [22,23].

Three questions about job control were yes/no questions: able to choose or change order of task, methods of work, speed or rate of work; one point was assigned for “yes”. The responses of two questions about job control (whether you have a say in the choice of your working partners, can take a break when you want) were “always”, “most of the time”, “sometimes”, “rarely”, and “never”. One point was assigned for “always”, “most of the time”, and “sometimes”. Then, the sum is calculated for each factor and categorized into “low” and “high” group, based on median value [20]. Aforementioned categorizing method was applied to other variables.

Health status

General poor worker health is well-known factor that affects presenteeism [2]. Therefore, in this study, the participants who have past 1 year medical history (cardiovascular disease, injury, depression, and insomnia or sleep disorder) related to presenteeism were classified into the group with health status “poor” [2,17].

Emotional labor

This research used two items from the fourth KWCS questionnaire to determine the level of emotional labor. One previous study that used the fourth KWCS to evaluate the health effects of emotional labor had also used the same two items by grouping them [24]. The first item was “I have to hide my emotions during work.” The original possible responses were grouped into three new possible responses: “Always” (originally “Always” and “Almost always”), “Rarely” (originally “Almost never” and “Never”), and “Sometimes.” The second item used was “I manage customers or patients who complaining angrily at work.” There were seven original possible responses for this question: “At all times,” “Most of the time,” “¾ of the time,” “Half of the time,” “¼ of the time,” “Rarely,” and “Not at all.” From these responses, “At all times,” “Most of the time,” and “¾ of the time” were grouped together as “Always”; “Half of the time”; and “¼ of the time” were grouped together as “Sometimes”, and “Rarely” and “Not at all” were grouped together as “Rarely.”

Presenteeism

The term “presenteeism” is used to describe the concept of workers coming to work even when they need to rest at home because of illness or injury [3,4]. This study classified presenteeism as a dependent variable, when a person answered “yes” to the question “Over the past 12 months, did you work at least one day when you were sick?” Many previous studies that used the fourth KWCS to evaluate presenteeism also used this question [[25], [26], [27]].

Data analysis

A Chi-square test was conducted to analyze the distribution of nurses that experienced presenteeism according to general (age, education level, and monthly income), occupational characteristics (employment status, weekly working hours, number of employees, shift work, and ergonomic risk), workplace psychosocial factors (job control, job demand, social support, job satisfaction, job insecurity, and workplace discrimination), and health status. The Chi-square test was also conducted to analyze the distribution of the independent variables “Hiding emotions at workplace” and “Engaging with complaining customers” on the distribution of presenteeism experience.

Multivariable logistic regression analysis was conducted to analyze the impact of “Hiding emotions at workplace” and “Engaging with complaining customers” on the risk of presenteeism. The adjusted odds ratio (OR) and 95% confidence interval (CI) were calculated through after correcting for general, occupational characteristics, workplace psychosocial factors, and health status. The analysis models according to the adjusted variables were as follows: (1) Model I: crude OR; (2) Model II: adjusted by age, education level, monthly income, employment status, weekly working hours, number of employees, shift work, and ergonomic risk; (3) Model III: Model II + adjusted by job control, job demand, social support, job satisfaction, job insecurity, and workplace discrimination; (4) Model IV: Model III + adjusted by health status. The statistical software SPSS 14.0 (SPSS Inc., Chicago, IL) was used for statistical analysis and p value (≤0.05).

Ethical considerations

Each survey participant in KWCS provided permission via face-to-face interview, and each participant's anonymity was guaranteed. This study was approved by the Institutional Review Board of S. University Hospital (IRB No. 2018-07-015).

Results

The distribution of presenteeism among female nurses in relation to general and occupational characteristics

As the overall prevalence, 96 nurses (29.3%), from a total of 328, responded to having experienced presenteeism. There were no significant distributional difference on other variables, but only ergonomic risk was found to be significantly different. Nurses group who exposed to high ergonomic risk showed more experience of presenteeism (33.3%) than low group (23.1%) (Table 1).

Table 1

Number of nurses with presenteeism by general and occupational characteristics

Variables	Total (n = 328)	Presenteeism		p value∗
		No (n, %)	Yes (n, %)
Age(years)
20–29	75	62 (82.7)	13 (17.3)	0.052
30–39	126	88 (69.8)	38 (30.2)
40–49	102	65 (63.7)	37 (36.3)
≧50	25	17 (68.0)	8 (32.0)
Education level
High school or lower	16	11 (68.8)	5 (31.3)	0.815†
College	119	82 (68.9)	37 (31.1)
University or higher	193	139 (72.0)	54 (28.0)
Monthly income (KRW, 10ˆ5)
<150	24	17 (70.8)	7 (29.2)	0.581
150–199	112	83 (74.1)	29 (25.9)
200–249	87	63 (72.4)	24 (27.6)
≧250	105	69 (65.7)	36 (34.3)
Employment status
Permanent	317	224 (70.7)	93 (29.3)	0.591†
Temporary	11	8 (72.7)	3 (27.3)
Weekly working hours
≤40	163	118 (72.4)	45 (27.6)	0.746
>40, <52	126	86 (68.3)	40 (31.7)
≧52	39	28 (71.8)	11 (28.2)
Number of employees
<5	81	59 (72.8)	22 (27.2)	0.748
5–49	121	88 (72.7)	33 (27.3)
50–299	86	59 (68.6)	27 (31.4)
≧300	40	26 (65.0)	14 (35.0)
Shift work
No	228	165 (72.4)	63 (27.6)	0.357
Yes	100	67 (67.0)	33 (33.0)
Ergonomic risk
Low	130	100 (76.9)	30 (23.1)	0.048
High	198	132 (66.7)	66 (33.3)

p-value < 0.05.

Based on Chi-square test.

Based on Fisher's exact test.

The distribution of presenteeism among female nurses in relation to workplace psychosocial factors and health status

Female nurses in high workplace discrimination group showed more experience of presenteeism (44.9%). Also, nurses who were in poor health status group (50.0%) showed more experience of presenteeism than good health status group (27.8%). There were no significant difference between job control, job demand, social support, job satisfaction, and job insecurity between presenteeism (Table 2).

Table 2

Number of nurses with presenteeism by workplace psychosocial factors and health status

Variables	Total (n = 328)	Presenteeism		p value∗
		No (n, %)	Yes (n, %)
Job control
Low	220	154 (70.0)	66 (30.0)	0.701
High	108	78 (72.2)	30 (27.8)
Job demand
Low	307	217 (70.7)	90 (29.3)	0.942
High	21	15 (71.4)	6 (28.6)
Social support
Low	116	85 (73.3)	31 (26.7)	0.454
High	212	147 (69.3)	65 (30.7)
Job satisfaction
Low	57	35 (61.4)	22 (38.6)	0.089
High	271	197 (72.7)	74 (27.3)
Job insecurity
Low	308	218 (70.8)	90 (29.2)	0.941
High	20	14 (70.0)	6 (30.0)
Workplace discrimination
Low	279	205 (73.5)	74 (26.5)	0.009
High	49	27 (55.1)	22 (44.9)
Health status
Good	306	221 (72.2)	85 (27.8)	0.027
Poor	22	11 (50.0)	11 (50.0)

p-value < 0.05.

Based on Chi-square test.

The distribution of presenteeism among female nurses in relation to emotional labor

Those who responded “Always” to the item “Hiding emotions at workplace” (41.7%) had more experience of presenteeism compared with those who chose the other responses [“Sometimes” (27.5%) and “Rarely” (13.4%)], and the result was statistically significant. The responses for the item “Engaging with complaining customers” did not show any significant differences [“Always” (28.2), “Sometimes” (33.9), “Rarely” (26.6%)] (Table 3).

Table 3

Distribution of presenteeism by hiding emotions and engaging with complaining customer

Groups	Total (n = 328)	Presenteeism		p value∗
		No (n, %)	Yes (n, %)
Hiding emotions at workplace
Rarely	67	56 (86.6)	9 (13.4)	<0.001
Sometimes	153	111 (72.5)	42 (27.5)
Always	108	63 (58.3)	45 (41.7)
Engaging with complaining customers
Rarely	177	130 (73.4)	47 (26.6)	0.402
Sometimes	112	74 (66.1)	38 (33.9)
Always	39	28 (71.8)	11 (28.2)

p-value < 0.05.

Based on Chi-square test.

Association between the degree of emotional labor and presenteeism

Multivariable logistic regression models were conducted to determine the relationship between the degree of emotional labor and presenteeism. Therefore, we ran four logistic regression models in sequence on the method abovementioned: (1) model I, crude OR; (2) model II, adjusted by age, education level, monthly income, employment status, weekly working hours, number of employees, shift work, and ergonomic risk; (3) model III, model II + adjusted by job control, job demand, social support, job satisfaction, job insecurity, and workplace discrimination; (4) model IV, model III + adjusted by health status.

The results of model I indicated that female nurses who responded “Always” or “Sometimes” to the item “Hiding emotions at workplace” had a higher risk of experiencing presenteeism than those who responded “Rarely” [OR = 2.37 (95% CI 1.07–5.22), OR = 4.55 (95% CI 2.03–10.17), respectively].

The association remained statistically significant after further adjusting for general and occupational characteristics [OR = 2.61 (95% CI 1.15-5.97), OR = 4.87 (95% CI 2.09-11.35), respectively] in model II, workplace psychosocial factors [OR = 2.45 (95% CI 1.06-5.66), OR = 4.25 (95% CI 1.78-10.12), respectively] in model III, and health status [OR = 2.40 (95% CI 1.04-5.54), OR = 4.12 (95% CI 1.72-9.84), respectively] in model IV.

The OR of presenteeism was higher in nurses that responded “Sometimes” to the item “Engaging with complaining customers” than those that responded “Rarely,” but did not show significant correlations in model I. These patterns were consistent across all different models (Table 4).

Table 4

The odds ratios and 95% confidence intervals of emotional labor on presenteeism

Groups	Model I		Model II		Model III		Model IV
	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Hiding emotions at workplace
Rarely	Reference		Reference		Reference		Reference
Sometimes	2.37	1.07–5.22	2.61	1.15–5.96	2.45	1.06–5.66	2.40	1.04–5.54
Always	4.55	2.03–10.17	4.87	2.09–11.35	4.25	1.78–10.12	4.12	1.72–9.84
Engaging complaining customers
Rarely	Reference		Reference		Reference		Reference
Sometimes	1.32	0.78–2.24	1.21	0.69–2.12	1.18	0.67–2.10	1.19	0.67–2.11
Always	0.91	0.41–2.02	0.72	0.31–1.67	0.68	0.28–1.61	0.68	0.28–1.62

OR, odds ratio; CI, confidence interval.

Model I: Crude odds ratio.

Model II: Adjusted by age, education level, monthly income, employment status, weekly working hours, number of employees, shift work, and ergonomic risk.

Model III: Model II + job control, job demand, social support, job satisfaction, job insecurity, and workplace discrimination.

Model IV: Model III + health status.

Discussion

This cross-sectional design research identified the relationship between presenteeism and the level of emotional labor among female nurses who participated in the fourth KWCS, such as “Hiding emotions at work” and “Engaging with complaining customers”. The more nurses were required to hide their emotions at workplace, the more susceptible they were to presenteeism. We think this result is meaningful because, as far as we know, there is a lack of research concerning the relationship between presenteeism and emotional labor for Korean female nurses.

Among the 328 female nurses in our study, the prevalence of presenteeism was 29.3%. At a focus group interview of 20 Korean nurses performed for qualitative research, all 20 nurses responded that they had experienced presenteeism [9]. In a study that included 250 Korean nurses, 45.6% responded that they had experienced presenteeism [28], and in a study that included 3,000 Swiss nurses, a high percentage (32.9%) reported they had experienced presenteeism [29]. These results indicate that the rate of presenteeism within the nursing occupation is high regardless of Eastern/Western differences.

Our study's results are in line with the findings of various previous studies. In this study, the higher the exposure to ergonomic risk and discrimination, the higher the rate of presenteeism. In studies involving only nurses exposed to ergonomic risk factors, the risk of presenteeism increased with the degree of low back pain [19]. Also, previous study about the impact of the musculoskeletal pain, closely related to ergonomic risks, on presenteeism in a fortune 100 companies showed that severity of pain showed a predominantly positive relation to presenteeism [30]. Many studies have identified the negative effects on mental health of workplace discrimination [31]. Poor mental health from discrimination, especially depression, is closely associated with presenteeism [32]. It is widely known, through previous research, that workers suffering from underlying diseases are more susceptible to presenteeism [33]. Our study also showed that poor health status group had a higher proportion of presenteeism.

Emotional labor, especially “Hiding emotions at workplace”, showed a significant relation with presenteeism. This study confirmed that the higher the level of emotional hiding, the higher the risk of presenteeism. Emotional labor such as hiding emotions can increase the risk of sleep disorders and insomnia in workers [24]. It has been found that patients with these types of disorders are prone to presenteeism [34]. Furthermore, emotional labor can function as the cause of work-related stress which leads to emotional dissonance [35]. Previous studies have found that workers with less work-related stress are less likely to experience presenteeism [36]. Emotional labor has various other negative effects, such as a low level of job satisfaction and burnout syndrome [37]. A low level of job satisfaction is known to have a significant relationship with presenteeism as well [38].

This study was one of the few studies that confirmed a correlation between emotional labor (“Hiding emotions at workplace” and “Engaging with complaining customers”) and presenteeism. Another strength of the study was the pool of participants. This study specifically used responses from nurses, increasing its significance and validity. This study also identified that risk factors, such as “Hiding emotions at workplace,” show clear quantitative dispositions and confirmed such factors to be among the first to be considered when assessing preventative measures for nurses experiencing presenteeism.

There were several limitations to this study. First, the number of participants was relatively small number using the Korean Industry Classification Code to select study participants in the fourth KWCS. Also, it was an unweighted study result, but there were many meaningful previous studies with unweighted results using KWCS [39,40]. Hence, this study has even greater implications because significant results were achieved within a small pool of participants. This approach allowed us to identify the existence of the phenomenon and construct a research model that can find a cause and effect relationship in future studies. Second, the causal relationship between presenteeism and emotional labor cannot be identified through a cross-sectional study. Yet, as there is a lack of previous research, this research can be used as a reference for future research. Third, the fourth KWCS does not include other factors, especially personal factors such as lifestyle and home circumstance, that may affect presenteeism in the questionnaire. Therefore, confounding variables could not be fully taken into account. Additionally, categorized self-report questionnaires were used, making it difficult to objectively evaluate emotional labor as an independent variable. In addition, we could not evaluate the dependent variable, presenteeism, as a tool that can be objectively evaluated like the Stanford Presenteeism Scale. KWCS questionnaire should be supplemented by reflecting these points. Nevertheless, there is an abundance of previous research based on KWCS and the European Working Conditions Survey, so we believe this research is sufficiently valuable.

This study confirmed that emotional labor, such as “Hiding emotion at workplace” is related to presenteeism in nurses. Compared with nurses who rarely hide their emotion at workplace, nurses who sometimes or always hide their emotion at workplace showed a significant rise in the risk of presenteeism. Nurses' presenteeism can bring about a decrease in concentration at work and the quality of nursing [11] and endanger the safety of coworkers [12]. The most important effect is the risk to patient safety [41]. Various studies have analyzed the factors that can affect presenteeism. Therefore, measures must be established to help decrease the risk of presenteeism through arbitration or alternatives that can manage nurses' emotional labor. Furthermore, future research to confirm definite causality is warranted.

Ethics approval and consent to participate

Approved by the Institutional Review Board of S. University Hospital.

(IRB No.SCHUH2018-07-015).

Consent for publication

Not applicable.

Availability of data and materials

The fourth KWCS data are publicly available.

Funding

This work was supported by the Soonchunhyang University Research Fund.

Author contributions

Study conception and design was carried out by SW Jung, KJ Lee, JH Lee; data acquisition was carried out by SW Jung; data analysis and interpretation was carried out by KJ Lee, SW Jung; drafting the manuscript was carried out by SW Jung; critical revision was done by JH Lee, KJ Lee. All authors read and approved the final manuscript.

Conflicts of interest

The authors have no conflicts of interests.