INTRODUCTION
According to the World Health Organization, social stigmatization means when individuals or groups in the immediate environment of an individual are labeled, discriminated against, or lose their status because of conditions related to a disease (1). Individuals can be subject to stigmatization for being close to an individual with a disease even when they are not sick themselves. Family members who undertake the care of the patient and healthcare employees who maintain the treatment are the groups that are subject to stigmatization (2). Stigmatization leads to severe consequences both individually and socially. It might cause stigmatized individuals to avoid treatment and thus increase mortality and morbidity rates (3,4). Additionally, it is well known that stigmatization is correlated with higher depression rates and lower self-regard in these individuals (5). Stigmatization is widespread during epidemics and that used to manifest itself for tuberculosis (TBC), human immunodeficiency virus (HIV), severe acute respiratory syndrome (SARS), and swine flu is present today for coronavirus disease-2019 (COVID-19) (5,6). Stigmatization of COVID-19 poses a severe threat to healthcare employees, patients, elders, and patients’ relatives (7). physical and verbal violence toward healthcare employees who play a key role in the fight against epidemics increased during the COVID-19 pandemic (6). This makes it even harder for healthcare employees to continue with their duties under already challenging working conditions.
In addition to the direct negative effects of the epidemic on physical health, there are many indirect negative social effects. Stigma creates divisions in the mind, leading to discrimination. In this case, a loss of status may occur according to the social, economic, and political power of the stigmatized group (8). It seems that there are two types of stigma in the stigma studies conducted on people who have had COVID-19. While the first is self-stigmatization, the second is society’s stigmatization of people with COVID-19 (9). People with COVID-19 may feel differentiated in their physical or mental abilities. This differentiation can cause social isolation and some psychiatric disorders (10). The second type of stigma, social stigma, can increase the self-stigma of people with COVID-19. As a result, existing mental problems may intensify and the functionality of individuals may deteriorate (11).
The psychometric assessment tools used in studies on stigmatization in epidemics are very rich. The literature on stigmatization of COVID-19 gets richer every day, while there are some shortcomings in terms of psychometric measurements. The researchers used the scales frequently used in past epidemics by adopting COVID-19 (7,12). However, it should be noted that this method can be problematic because every epidemic has its specific development and perception created by this development in people’s minds. For example, in any diseases such as TBC, bird flu, swine flu (H5N1), or SARS transmitted through the respiratory tract, there has never been a period in which worldwide precautions were taken and habit changes were experienced as in the period of COVID-19. This period brought many new practices, including shutting down airports to flights, closing down cafes, restaurants, and stadiums, transforming in-person education into online education, and modifying measures taken depending on age groups. Therefore, it can be assumed that developing disease-specific scales will be more effective in determining the size of stigmatization and in terms of measures to be taken. It is noteworthy that generally, a limited number of items are used in the assessment of stigmatization about COVID-19 in studies.
Moreover, as such assessments that lack validity and reliability analyses might lead to erroneous and missing results, it can be suggested that these methods are insufficient. The purpose of this study is to develop a scale that evaluates stigmatization of COVID-19 and analyzes its validity and reliability, considering the shortcomings in the literature. We believe that this scale will be important in evaluating the problems of stigmatization in the literature and developing solution recommendations.
MATERIALS AND METHODS
Development of the COVID-19 Stigma Scale
While developing the COVID-19 Stigma Scale (CSS), stigmatizations experienced by people with COVID-19 in both their close interpersonal contacts and social life were addressed. Studies on stigmatization attitudes about COVID-19 were reviewed. In addition, studies on infections with high contagiousness, such as TBC, HIV, and H5N1, with a stigmatization attitude were reviewed. While deciding on the scale items, attention was paid to making the items plain and understandable and ensuring that a single item only has a single idea or judgment. Because of the evaluations made among the researchers, areas such as labeling, stigmatization in close contact people, stigmatization in social relations, inequality, discrimination, loss of social status, and emotional responses were identified as main dimensions. It was observed that the scales developed for stigmatization were especially combined on three conceptual fields: labeling, emotional response, and loss of social status. In this framework, a 39-item form was created initially. Although these items were created considering the dimensions in design, it was predicted that every item could occur in another dimension. For example, the item “The society should respect fewer people who experienced COVID-19” includes both labeling (disrespect) and loss of social status concepts. These items corresponded to multiple situations regarding labeling (stigmatizing the individual, healthcare employees, foreigners, people in close contact, and people who experienced COVID-19). As the study’s main purpose was to determine social stigmatization against patients, in other words, stigmatization against COVID-19 patients, items outside this framework were excluded. Some similar items were combined. In the final assessment among the researchers, the number of items in the draft scale dropped to 25.
In the second stage, for content validity, opinions were sought from three academic members, two from psychiatry and one from adolescent mental health, and a nurse specializing in public health. An item was removed, and corrections were made to make the items easier to understand on the basis of the recommendations of these persons who have sufficient background and expertise in their respective fields and are outside the study team. The Davis technique was used in content validity analysis. Because of this technique, the content validity index of the scale was determined to be 91.6%.
After these arrangements, a pilot study was conducted with 10 participants. The items that were not understood well or might have caused a misunderstanding were corrected again. Thus, face validity was guaranteed for the form of 24-item CSS.
It was considered appropriate to use a Likert-type evaluation for the items on the scale. Because individuals might have demonstrated an attitude with which they might show less or partial participation rather than exhibiting a sharp attitude about the items. For every item, the participation level was configured as completely disagree (0), disagree (1), slightly agree (2), agree (3), and completely agree (4). The total scale score ranges from 0 to 76-higher scores in the scale indicate a higher stigmatizing attitude toward COVID-19 patients. The Turkish and English versions of the final version of the scale are shown as an additional file (Appendix 1A).
Inclusion and Exclusion Criteria
Among the relatives of the patients who applied to the hospital, at least primary school graduates volunteered for the study, and over 18 were included in the study. As exclusion criteria, relatives who applied for infectious diseases, chest diseases, and COVID-19 polyclinics were accepted.
Sample
Three hundred twenty-nine people voluntarily agreed to be involved in the study. Three hundred twenty-three forms filled out with complete data were evaluated. The sample’s mean age was 39.04±13.74, and mean years of education were 12.69±4.38. Most of the participants were female (n=176, 54.5%), married (n=202, 62.5%), and employed (n=188, 58.2%).
Assessment Tools
Sociodemographic Information Form: A form prepared by the researchers to evaluate the patients’ population and other characteristics (age, gender, educational level, marital status, employment status, alcohol use and smoking habits, sickness status, and living environment) was used.
COVID-19-Related Experience Form: Information form with variables including having experienced COVID-19 and having been affected by the pandemic. It was prepared by the researchers.
COVID-19 Information Questionnaire: The COVID-19 Information Questionnaire (CIQ) was prepared by the researchers in light of the information in the literature. In this questionnaire comprising 15 items, every item has a right or wrong option. Some items are reverse scored-higher scores in this questionnaire indicate a higher level of knowledge about COVID-19. This questionnaire is attached (Appendix 1B).
Ethics Approval and Consent to Participate
Ethical permission for the study was obtained from the Kocaeli University Non-invasive Clinical Research Ethical Committee (approval number: KÜ GOKAEK-2021/7.18, project number: 2021/134). After obtaining ethical permission, the study was conducted between 15.04.2021 and 10.06.2021. Permission was also obtained from the Republic of Türkiye, Ministry of Health, Directorate of Healthcare Services Management. The participants were explained the purposes of the study and were told they could withdraw from the study at any stage without any reason. Written consent was obtained from all participants who agreed to participate in the study.
Statistical Analysis
SPSS 22 (IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.) was used for statistical analysis. Descriptive statistics are given in numbers, percentages, mean, and standard deviation (SD). Because the skewness and kurtosis values of the numeric variables were between (±1), it was assumed that the data were evenly distributed (13). The correlation between the numeric variables was analyzed using Pearson’s correlation test. The independent samples t-test was used to compare the numeric variables between the two groups, and One-Way analysis of variance was used to compare the numeric variables between three and more groups. Exploratory factor analysis (EFA) was used to determine construct validity. In this analysis, the Kaiser-Meyer-Olkin (KMO) test was used to evaluate the sufficiency of the sample size, and the Barlett sphericity test was used to evaluate the conformity of the scale items to the factor analysis. Principal component analysis was used in EFA, and direct Oblimin was preferred for the factor rotation technique. While determining the factor items, the item factor load was accepted as >0.40. For their validity together, the correlation coefficient between the CIQ and CSS scores was examined. In the reliability analysis, both the total score and the internal reliability coefficient of the CSS subdimensions were determined using Cronbach’s alpha value. In addition, when item-total scores and items were excluded, Cronbach’s alpha value was calculated. For statistical analyses, the significance level was taken as p<0.05.
RESULTS
Validity of the Scale
Construct Validity
An EFA was performed using the principal component method and the direct Oblimin rotation technique for the 24-item scale. Because of the KMO and Bartlett’s tests, the KMO value was 0.929, and Bartlett’s test (χ2=4827.63; SD: 276; p<0.001) was significant. A four-factor structure that had an eigenvalue greater than one and explained 62.65% of the total variance was identified. Three items with a factor load below 0.40 were identified (items 7, 11, and 17). Factor analysis was renewed after removing these items. The scale explained 59.33% of the variance in a three-factor structure. However, with the new analysis, two more items in the scale had a factor load below 0.40 (items 12 and 23). These two items were also removed from the scale. With the renewed analysis after removing these items, a structurally valid scale comprising 19 items and three factors that explained 61.85% of the variance was obtained. The first factor explained 43.08%, the second factor explained 11.75%, and the third factor explained 7.02% of the total variance. Considering the items that constituted the factors, the first factor (1, 2, 3, 4, 5, 16, 22, 24) was called the rejection dimension, the second factor (8, 13, 14, 15, 18, 19, 20) the discrimination dimension, and the third factor (6, 9, 10, 21) the emotional reaction (Table 1).
Considering the items that failed to provide a sufficient factor load in the factor analysis made in the first stage (items 7, 11, and 17), it was observed that perception of riskiness about persons who experienced COVID-19, shopping restrictions, and feeling angry about the bad news on the media were not correlated with stigmatization. This situation suggests that there is a limit between fundamental rights, needs, and stigmatization. Additionally, people might not see other people who experienced COVID-19 as risky, or even if they do so, it seems to be not in correlation with stigma. The items that were excluded from the scale (items 12 and 23) look similar. Both items meet at a common point based on insufficiency. Removing these similar items together shows the fact that scale validity was obtained. Individuals do not establish a relationship between receiving COVID-19 and insufficiency (physical weakness) or insufficiency in work performance after receiving COVID-19 and stigmatization.
Content Validity
For this type of validity, the correlation between the total score of the scale and its subscales was calculated (Table 2). There was a significant high correlation between the CSS total score and rejection (r=0.929, p<0.001), discrimination (r=0.789, p<0.001), and emotional reaction sub-scales (r=0.733, p<0.001).
Concurrent Validity
The concurrent validity of CSS was evaluated with its correlation with CIQ. The results are presented in Table 2. There was a significant negative relationship between the CSS total score and the CIQ total score (r=-0.301, p<0.001). Similarly, there was a significant negative relationship between the CSS subscale scores and the CIQ total score (Table 2).
Discriminant Validity
In terms of discriminant characteristics, past studies have demonstrated a clear relationship between education status and stigmatization attitudes (9,10). Our study reinforces this outcome (Table 3). There is a significant negative relationship between education level and CSS total score (r=-0.254, p<0.001). The relationship between the CSS total score and pandemic experiences is shown in Table 4. The CSS total score was significantly high in those whose income level dropped post-pandemic (F=5.580, p=0.004), significantly low in those who experienced COVID-19 [t(321)=3.174, p=0.002], and significantly low in those whose family member experienced COVID-19 [t(321)=2.082, p=0.038]. These results reinforce the fact that the scale has a discriminant character.
CSS Reliability Analysis
Internal Consistency
In the internal consistency analyses, the item-total item correlation coefficients were used, and when the item was removed, Cronbach’s alpha coefficients were used, as shown in Table 1. Of the 19-item scale, the corrected item-total item correlation coefficients were between 0.385 and 0.795. The scale’s internal consistency Cronbach’s alpha coefficient was calculated as 0.91. Cronbach’s alpha coefficient was 0.90 in the rejection sub-dimension, 0.88 in the discrimination sub-dimension, and 0.68 in the emotional reaction sub-dimension.
Sociodemographic Characteristics of the Participants
The sociodemographic characteristics of the participants and their relationship with the CSS scores are shown in Table 3. While there was a significant positive relationship between CSS scores and age (r=0.129, p=0.020), a significant negative relationship was found between CSS scores and education years (r=-0.254, p<0.001). The CSS score was significantly lower in alcohol users than in the non-users [t(321)=3.933, p<0.001].
DISCUSSION
To the best of our knowledge, our study is the first community-based scale development study that intends to evaluate stigmatization against patients who experienced COVID-19. The adoption of scales that are not COVID-19 specific (5,7) or stigmatization in the items created by the researchers (15) was evaluated. Furthermore, stigma scales are presented as recommendations (16). However, these studies’ validity and reliability analyses were either not done at all or not on the desired level. The scale we developed based on this gap in the literature was shown to be a valid and reliable tool for evaluating stigmatization against patients who experienced COVID-19.
The CSS, which was completed through in-person interviews with the participants, revealed a three-factor configuration that was found to be valid and reliable for psychometric assessment. The factor constructs of the scale are generally in conformity with the conceptual foundation of stigmatized beliefs, attitudes, and behaviors. The first factor we call rejection in our scale evaluates avoiding both close and social relations, in other words, rejecting the establishing relationships with the patients. Not approving being in a close relationship with people who experienced COVID-19, not wanting to live and travel in the same environment, not wanting their kid to be in the same environment, and being concerned or uncomfortable while in the same environment are suitable for the denomination of this factor. Rejection was defined as the first stage in Rosenberg’s (17) study, which attempts to determine the social stages of epidemics. The researcher who took inspiration from the book of Albert Camus while identifying these phases suggested that society, in the first place, attempts to reject the presence of epidemics to secure themselves and protect their financial interests (14). From this perspective, rejection is the first stage in which society begins to stigmatize.
Meanwhile, this rejection continues and becomes an individual attitude in further stages. Link and Phalen (8) in 2001 reported that another important component of stigmatization is ironically “power”. It was stated that the main structure of this power comprised factors such as not approving, excluding, discriminating, categorizing, and rejecting (18,19). Considering the items under the second factor of the scale, it was found appropriate to call this sub-dimension discrimination. Items such as not wanting to meet friends who experienced COVID-19, thinking about re-evaluating the relationship after the partner experiences COVID-19, not employing people who experienced COVID-19, and showing less respect by the society are placed in a common structure about discrimination. While discrimination is an important part of all types of stigmatization, it is one of the main components of social stigmatization. Social stigmatization exists based on society’s stereotypes, prejudices, and discriminatory beliefs (18). On the other hand, a main component of discrimination, “social stigmatization” is closely related to other types of stigmatizations (primary, secondary, corporate, and structural stigmatization), and social stigmatization can be a result of and a reason for other types of stigmatizations (20). Because the COVID-19 pandemic is not limited to an individual level and impacts social habits, attitudes, and perceptions, stigmatization of COVID-19 is clearly a type of social stigmatization. Discrimination is also a component of stigmatization.
Because of the studies in the literature, it was concluded that stigmatization has phases such as “labeling”, “stereotyping”, “prejudice”, “loss of social status and discrimination” (19). These phases are not distinguished from each other with clear borders, and it is possible to see that one of them can embody one another or others at the same time. Although emotional reactions were not discussed much in the initial studies on stigmatization, Link et al. (21) emphasized that emotional reactions are an important part of the stigmatization process. When stigmatization occurs, emotions emerge in both the stigmatizer and the stigmatized. The severity of these emotions is closely related to the depth of stigmatization. The third factor in our scale clearly demonstrates the emotional reactions that emerge despite stigmatization. Items such as blaming people who experienced COVID-19, seeing them as the source of concern of society, and as the main reason for the recent unhappiness they experienced explicitly include these emotional reactions. Considering the scale factors in light of the information from the literature, it seems suitable for the conceptual content of stigmatization. On the other hand, as mentioned by the researchers, the relationship between these factors is highly dynamic, and its boundaries are not clear.
It is a known fact that humans can marginalize others with an instinct to protect themselves despite situations or events that they are not knowledgeable about (22). In other words, people create a border with another by stigmatizing them. Therefore, they feel safe in their inner world. The relationship between stigmatization and lack of information is clearly observed in both psychiatric and medical illnesses (23). For example, the less knowledgeable individuals are about schizophrenia, the more they tend to stigmatize them (22,23). It has been demonstrated that providing information about diseases reduces negative opinions and stigmatization about the disease (24,25). The moderately significant negative correlation (r=-0.301, p<0.001) between the CSS scale developed in our study and the total scores in the CIQ shows the validity of the scale. In addition, stigmatization level was high in individuals with lower levels of education (Table 3). Studies on stigmatization of the COVID-19 pandemic demonstrate that individuals with low educational levels are under social media impression and engage in stigmatization to a higher degree (10).
Behaviors such as stigmatization, discrimination, marginalization, and avoidance emerged in past epidemics such as the COVID-19 pandemic (26). Because of the potentially lethal situations, new diseases and incurable/unknown diseases are defined as factors related to an increased risk of stigmatization (27). The COVID-19 pandemic is the first pandemic that occurred in an era in which communication networks were developed to such an extent, despite having similarities to other pandemics. This pandemic we live through has created many specific conditions because of the period we live in.
Investigating the impact of these conditions, Yuan et al. (28) recently compared individuals who experienced COVID-19 with healthy controls in terms of self-stigma. This study yielded very striking outcomes. For individuals who experienced COVID-19, stigmatization areas such as social rejection, financial insecurity, internalized embarrassment, and social isolation were significantly higher than those in healthy controls. The researchers who evaluated factors related to stigmatization stated that male gender, low educational level (high school and below), unemployment, experiencing COVID-19 by a family member, and economic loss variables were closely related to stigmatization. Another study was conducted with 7411 participants from 173 countries (29). This study had a cross-sectional design, and it was found that 27.3% of the participants believed that people spoke ill or spread rumors about individuals who experienced COVID-19. However, as one of the most important study results, one out of five persons in the sample (21.9%) believed that people who experienced COVID-19 lost their dignity and status in society (29). In our study, the total score on the scale was significantly higher in people who had lower income levels in the post-pandemic period, people who did not experience COVID-19 infection, and people whose family did not experience COVID-19 than in other categories (Table 4). These results demonstrate that the scale has a predictive character. Additionally, the level of stigmatization against people who experienced COVID-19 by people who used alcohol was significantly lower than that against those who did not use alcohol (Table 3). These results can be explained by the higher education level of individuals who use alcohol (14.82±3.52 vs.11.78±4.40, p<0.001). The lack of a detailed investigation of alcohol use makes it harder to interpret the outcomes obtained in the respective field.
Conducting the study face-to-face and reaching a sample beyond the necessary sample for developing the scale are the strengths of the study. We believe that our study will provide important contributions to the literature in measuring COVID-19-related stigma, which is also considered a public health problem. These evaluations will provide a basis for further studies.
Study Limitations
However, our study has certain limitations. The COVID-19 pandemic is a dynamic process in which individuals’ perceptions and attitudes can change periodically. Conducting this study in a specific period might be a particular limitation. It should also be noted that although there were individuals from each education group in our study, the average years of education were above the country average, and the study represents the results of a single center.
CONCLUSION
With this study attended by enough persons from every educational level in society and conducted by in-person interviews, the 19-item three-factor CSS was found to be valid and reliable. For both the total and scale factors, Cronbach’s alpha internal consistency coefficients were found to be sufficient. Evaluating the stigmatization attitude against people who experienced COVID-19 and taking necessary measures in this context will play a key role in preventing possible psychiatric disorders and trauma experienced by society.
Acknowledgments: We would like to thank all our psychiatry assistants who helped collect data during the research process and our reviewers who contributed to the development of our study.
Ethics
Ethics Committee Approval: Ethical permission for the study was obtained from the Kocaeli University Non-invasive Clinical Research Ethical Committee (approval number: KÜ GOKAEK-2021/7.18, project number: 2021/134).
Informed Consent: Written consent was obtained from all participants who agreed to participate in the study.
Authorship Contributions
Concept: B.G., M.Y., Design: B.G., M.Y., Data Collection or Processing: B.G., M.Y., Analysis or Interpretation: B.G., Literature Search: B.G., M.Y., Writing: B.G., M.Y., Critical Review: B.G., M.Y.
Conflict of Interest: No conflict of interest was declared by the authors.
Financial Disclosure: The authors declared that this study received no financial support.