COVID-19-related stigma and its' inuencing factors: a rapid nationwide study in China

Background (cid:0) COVID-19 poses a signi ﬁ cant challenge to global public health. During the pandemic, COVID-19 patients and people in outbreak areas have suffered from stigma associated with the disease. This study aimed to evaluate the prevalence of COVID-19-related stigma toward COVID-19 patients and people from the city of Wuhan in China and assess the association of COVID-19-related stigma, health literacy, and sociodemographic characteristics. Methods (cid:0) A cross-sectional survey covering 5,039 respondents was conducted in 31 provinces in China using a convenience sampling method. Binary logistic regressions were used to identify the factors associated with COVID-19-related stigma. Results (cid:0) Among the participants, 122 (2.4%) reported themselves and 254 (5.0%) reported the communities they lived in held a stigmatizing attitude toward COVID-19 patients, respectively. likely to stigmatize Wuhan.

example, people with higher education levels and HIV-related knowledge were less likely to stigmatize HIV patients [7,15]. This may be due to the fact that people with more HIV-related knowledge had a better understanding that they were not likely to get infected with HIV through social interactions (such as handshake, hug, and cheek kiss). However, emerging infectious diseases that are evolving in nature and have uncertain transmission patterns often cause panic among individuals and communities, as was seen with SARS, H1N1, and COVID-19. The transmission of certain infectious diseases through social interaction can ignite stigma toward disease-related groups [14] following the introduction of socialdistancing policies to prevent such diseases. Previous studies have noted that social distancing measures may affect the attitudes of individuals and communities toward people with stigmatizing conditions, and may lead to stigma [14,16]. However, few studies have shed light on the relationship between knowledge and stigma in emerging infectious diseases that can be transmitted through social interactions.
Health literacy is usually de ned as an individual's ability to obtain and process health information and take appropriate action [17]. Knowledge is an important dimension of health literacy [18]. Previous studies investigating the relationship between health literacy and stigma have mostly focused on mental illnesses and chronic diseases, and have shown that patients with low health literacy were more likely to feel stigmatized [19][20][21]. Few studies have investigated the relationship between health literacy and stigma toward infectious diseases that require social distancing.
Studies on stigma related to infectious diseases have revealed that it is not only individual patients who face stigma from infectious diseases, but can also be entire racial or ethnic groups who have or are perceived as having a higher likelihood of being infected [22]. Wuhan, the capital of Hubei Province, was the most severely affected area during the COVID-19 epidemic in China. In order to control the spread of COVID-19, the Chinese government took unprecedented measures, including locking down Wuhan, and requiring all Wuhan residents who migrated to other provinces before Wuhan was locked down to receive nucleic acid tests. Among the con rmed COVID-19 cases in many provinces, a considerable portion were imported cases from Wuhan [23]. Despite the government and media calling for tolerance, the development of a stigmatization towards residents of Wuhan was inevitable. For example, in some communities, residents of Wuhan were not allowed to enter and suffered unfair treatment. Therefore, this study aims to investigate both stigma faced by COVID-19 patients and stigma faced by residents of Wuhan.
The aims of this study are 1) to evaluate the prevalence of stigma during the COVID-19 outbreak in China and 2) to assess the association of stigma, health literacy, and sociodemographic characteristics during the COVID-19 epidemic.

Study design and participants
This was a national cross-sectional survey conducted in 31 provinces, municipalities, and autonomous regions in China (except Hong Kong, Macao, and Taiwan). The questionnaire used in this survey was developed for this study (Additional le 1). Tools to measure stigma in this study were referred to a previous study published in the BMC Public Health [24]. Two online focus groups were conducted to discuss the questionnaire design, with six people with public health and medical backgrounds in each group. Two independent experts with background in public health reviewed and further developed the questionnaire. We conducted 30 online face-to-face interviews with respondents of different ages and education levels to pre-test the questionnaire. The questionnaire included sociodemographic characteristics, COVID-19-related stigma, and health literacy during COVID-19 epidemic. We set up logic questions to check the validity of the data.

Sample selection
The respondents included in this study were aged over 16 years old and could read Mandarin. We conducted convenience sampling in 31 provinces, and 100-200 families were selected from each province. The family member from each family whose birthday was closest to the survey date was asked to ll in the questionnaire to ensure randomness in sampling. We encouraged younger family members to assist elderly family members in completing the questionnaire, if necessary. We conducted over-sampling for ethnic minority groups. We also over-sampled for Wuhan, as it was the epidemic outbreak center. We intentionally balanced respondents from urban and rural areas while conducting this survey. The nal effective sample size from all 31 provinces was 5,039.

Data collection
Data were collected using the web-based questionnaire that was re ned by the focus groups and pretesting process. The survey was administered from March 1 to March 16, 2020. Before the investigation, investigators received online trainings, and thusly they were responsible for quality control. Respondents could ll in the questionnaire by scanning QR codes or clicking the questionnaire link on smartphones, tablets and other mobile devices. Before lling in the questionnaire, respondents were informed that this was an anonymous study and they could participate voluntarily. This investigation did not provide compensation to the respondents. The Ethics Committee of the School of Public Health at Zhejiang University reviewed and approved this study.

Data analysis
All data were analyzed using IBM SPSS Statistics Version 23.0 for Windows. Descriptive analyses included means for continuous variables and percentages for categorical data. Chi-square tests were conducted to compare COVID-19-related stigma between groups. Binary logistic regression analysis was used to examine the association of the independent variables with COVID-19-related stigma. All comparisons were two tailed. The signi cance threshold was p-value < 0.05.

Health literacy
Questions on health literacy about COVID-19 were adapted from previous studies [25,26] and measured using two questions: (1) Do you agree that "it is di cult for me to nd correct and comprehensive information about COVID-19", (2) Do you agree that "it is di cult for me to understand information I got about COVID-19". Each question was answered using a 5-point Likert scale ranging from 1 to 5 (1=Strongly disagree; 2=Disagree; 3=Fair; 4=Agree; 5=Strongly agree).

Stigma
Questions on COVID-19-related stigma were adapted from previous studies [24,27]. Four questions, including stigma toward COVID-19 patients and residents of Wuhan at the individual and community levels were used, respectively. The study participants who reported holding stigmatizing views were classi ed as "stigmatized", while those who reported holding no stigmatizing views were classi ed as "not stigmatized". People who lived in Wuhan were automatically exempted from stigma questions related to residents of Wuhan.

Social demographic characteristics
The sociodemographic characteristics comprised gender, age, education, ethnicity, urbanicity, and monthly household income. According to the data of con rmed COVID-19 cases in 31 provinces o cially announced by the Chinese government as of March 1, the 31 provinces were divided into four areas: low case areas, low-medium case areas, medium case areas, and high case areas.

Results
A total of 5,039 participants (Table 1) with an average age of 33.0 (SD=12.5) were included for analysis. Most of them were female, were of Han ethnicity, received senior high school education, had a monthly household income above 705 United States dollars (USD), and lived in a medium case area.
At the individual level (Table 2), the majority (70.2%) of participants reported they felt compassion for and desired to help COVID-19 patients, 1,045 (20.7%) reported they felt compassion for COVID-19 patients but tended to avoid them, 29(0.6%) expressed their unwillingness to help COVID-19 patients, and 93(1.8%) expressed fear of COVID-19 patients. Less than one percent of participants expressed their unwillingness to help residents of Wuhan and 74(1.6%) expressed fear of residents of Wuhan. At the community level, 254(5.0%) participants reported their communities rejected COVID-19 patients, and 475(10.3%) participants reported residents of Wuhan were rejected by their communities. Approximately one-third of participants reported that they had di culties nding comprehensive and correct information about COVID-19, and 759(15.0%) of the participants reported that it was di cult to understand the information they received about COVID-19. Figure 1 shows the number of cumulative con rmed COVID-19 cases from the 31 provinces on the investigation data (March 1, 2020). Figure 2 illustrates the proportion of individual stigma toward COVID-19 patients in each province. People living in Hubei, Anhui, Guizhou, Tianjin and Yunnan provinces had a relatively high stigma percentage of over 4% of the population. Figure 3 shows that more than 4% of the respondents living in Guizhou, Yunnan, and Qinghai provinces expressed a stigma toward residents of Wuhan. The proportion of reported stigma toward residents of Wuhan in Henan, Shanxi, Ningxia, Chongqing and Zhejiang provinces was between 3% and 4%.
As shown in Table 3, the prevalence of stigma toward COVID-19 patients among people over 50 was signi cantly higher than that of people under 20 (5.1% vs. 1.2%, p 0.001). Compared with people who had a junior high school or lower degree, people with a college or higher degree reported lower levels of stigma toward COVID-19 patients (2.0% vs. 4.0%, p=0.01). Minorities showed a higher level of (3.6% vs. 2.2%, p=0.024) stigma toward COVID-19 patients than did Han respondents. Participants who felt it was easy to nd and understand information about COVID-19 expressed lower stigma toward COVID-19 patients than did those who felt it was di cult (1.4% vs. 3.7%, p 0.001; 1.5% vs. 4.5%, p 0.001). Individual stigma towards residents of Wuhan was more prevalent among male than female respondents (3.4% vs. 1.8%, p 0.001) and was relatively high among those who felt it was hard to understand COVID-19-related information (4.4% vs. 1.8%, p 0.001).

Discussion
To our knowledge, this is the rst nationwide study investigating COVID-19-related stigma in China. Our study revealed the prevalence of stigma toward COVID-19 patients and residents of Wuhan at both the individual and community levels during the epidemic. Consequently, our results veri ed the correlation between better health literacy and lower stigma during emerging infectious diseases outbreaks and showed the difference in stigma in regions with different COVID-19 epidemic severities on a large scale across the country. Additionally, we found that socio-demographic factors, such as residency, gender, age, and ethnicity, affected COVID-19-related stigma.
Overall, the prevalence of stigma was low in China during the COVID-19 pandemic and most participants had a positive attitude towards COVID-19 patients and residents of Wuhan people. Noticeably, our study showed that participants reported stigma from communities was signi cantly higher than individual stigma, which might be affected by the social desirability effect, meaning that participants' responses concerning themselves may be biased in order to meet social expectations and moral standards. Social desirability has been identi ed in previous studies on measuring individual stigma towards people with mental illness [28,29]. Stigma from communities is not unique to China, but has also been reported in the United States, Australia, Nepal and other countries [30], which deserves more attention in future studies.
Our study added to the literature by exposing the negative association between health literacy and stigma during an emerging infectious disease. Our ndings veri ed that, during the COVID-19 pandemic, there was a signi cant association between health literacy and COVID-19-related stigma. Higher COVID-19related health literacy, speci cally, a better ability to nd and understand information, might help reduce stigma toward COVID-19 patients and residents of Wuhan. Our ndings are consistent with previous studies, which identi ed a correlation between health literacy and stigma on mental disease [31,32]. Consequently, in the eld of infectious diseases, higher literacy concerning one disease may possibly help eliminate stigma. Additionally, it has been suggested that health literacy interventions, such as educational lectures to improve public knowledge and literacy, could help reduce stigma in the eld of mental health [33]. Thus, further studies are needed to verify effective measures to reduce stigma during an emerging infectious disease.
In addition to health literacy, our research found that people in different regions held differing degrees of stigma toward COVID-19 patients. In general, provinces which were close to Wuhan, such as Anhui and Chongqing, and provinces with more ethnic minorities, such as Yunnan and Guizhou, had higher levels of stigma toward COVID-19 patients. Similarly, the proportion of respondents who held stigma towards residents of Wuhan was relatively high in provinces close to Wuhan, such as Henan, Chongqing and Shanxi, and provinces with more ethnic minorities such as Qinghai, Yunnan, Guizhou and Ningxia. The danger appraisal hypothesis states that an individuals' perception of danger would make them choose a safer social distance [34]. Another study on SARS-related stigma conducted in Hong Kong showed that living in a geographical location which was close to an area with a large number of cases could increase stigmatizing attitudes [35]. Speci cally, residents living on the block with the most SARS patients reported holding the highest level of stigmatizing attitudes [36]. Similarly, in our study, people living in areas severely affected by the COVID-19 pandemic were at higher risk of social interaction with potential COVID-19 patients. Thus, they might expect to maintain a longer social distance and have less social interaction with potential COVID-19 patients, and therefore may hold higher levels of stigma.
Our study showed the in uence of sociodemographic characteristics on COVID-19-related stigma, which might help identify subgroups who were more likely to stigmatize others during an infectious disease epidemic. In our study, females were more tolerant towards residents of Wuhan, while people over 40 years old and ethnic minorities were more likely to stigmatize COVID-19 patients, which is consistent with previous studies [15,30,37]. A previous study revealed that groups with higher education and income levels had lower levels of stigma toward patients with related diseases [20]. However, this difference was not found in our study. One possible reason for this may be that, during COVID-19 pandemic, China conducted a large-scale publicity campaign through traditional and social media, such as China Central Television (CCTV), WeChat o cial accounts and short video platforms [38], which may have helped reduce barriers related to education and economic status in accessing adequate information concerning COVID-19.
There are some limitations in this study. First, this is a cross-sectional study, so it can not verify the causal relationship between the stigma-related variables. Second, the research data relies on the selfreporting of survey participants. Participants' responses regarding their stigma attitudes may be biased due to social desirability [28]. Third, we chose a snowball sampling method rather than a representative sampling method, due to the social-distancing policies in place during our investigation. However, we over-sampled ethnic minorities and ensured both the balance of urban-rural samples and the randomness of each sample in each household during the survey to reduce related bias.

Conclusion
In summary, our ndings suggest that COVID-19 patients and residents of Wuhan suffered stigma at both the individual and community levels, although the proportion of respondents holding stigma was not high. Provinces closer to Wuhan had relatively higher levels of stigma toward COVID-19 patients and residents of Wuhan. There was a correlation between better health literacy and lower levels of stigma during the COVID-19 outbreak. Tailored interventions are encouraged to improve health literacy and consequently to reduce stigma toward both COVID-19 patients and residents of Wuhan at both the individual and community levels, respectively.

Declarations Ethics approval and consent to participate
The protocol for this study was approved by the Ethics Committee of the School of Public Health, Zhejiang University. All participants were informed of the background, aims, anonymous nature and length of the survey. Participants were well informed that completing the questionnaire signi ed their informed consent.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analysed for the current study can be made available from the corresponding author on reasonable request.

Competing interests
The authors declare no con ict of interest. The funding body has no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Authors' contributions XZ, XW, and HZ made substantial contributions to the study design and supervised the data collection. TJ, LL, YZ, and YP contributed to the data collection and interpretation. TJ wrote the substantial parts of manuscript. All authors critically revised, reviewed, and approved the nal version the manuscript.     Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.