From February 11 to 24, 2020, a cross-sectional survey was conducted in 2,484 participants using convenience sampling in mainland China. A link of the self-administrated questionnaire developed by the Environmental Health Institute at China Medical University was distributed via WeChat, one of the most popular social media platforms in China. The inclusion criteria of participants were as follows: at least 18 years of age; able to comprehend and complete the online questionnaire in Chinese independently; willing to participate in the study and; and able to provide signed informed consent. The exclusion criteria of the participants were as follows: having a history of the severe psychological disease (e.g., schizophrenia, bipolar disorder, mental disorders caused by epilepsy); having current psychological therapy; and having been diagnosed with diseases which limited the participation in this survey.

The validated questionnaire contained self-developed questions about panic, psycho-behavioral responses to the COVID-19 pandemic, and the risk perception of the COVID-19 pandemic. The questionnaire took about 20 min to complete. All of the procedures of this survey followed the Helsinki Declaration as revised in 1989 and the protocol authorized by the Ethics Committee of China Medical University (ID: 2020048).

The gender (male, female), age, marital status (married, other), occupation (government worker/healthcare worker/employee of an enterprise/student/other [e.g., commercial personnel, soldier, teacher, etc.]), education (college and below, bachelor's degree, master's degree, or above), and monthly income (≤ 5,000 yuan [ ≤ US $725.19], 5,001–10,000 yuan [US $725.34–1,450.39], and >10,000 yuan [>US $1,450.39]) were collected as demographic characteristics.

A self-developed question was applied to measure the public panic among the public. Public panic was collected using a yes or no question: “I was panicked during the COVID-19 pandemic” with choices including “yes” or “no”.

Behaviors in the past 2 weeks, such as “Taking temperature repeatedly”, “being nervous in a crowd”, “being suspicious of infection in the family”, “being worried about the future”, “being afraid of the recurrence of the COVID-19 pandemic”, “being constantly reminded of the COVID-19 pandemic by the surroundings”, “being suspicious that infected people are around”, and “avoiding gatherings during the holidays” comprised the collected psycho-behavioral responses. The psycho-behavioral responses were assessed by self-developed “yes” or “no” questions.

The perception of risk during the COVID-19 pandemic was assessed with a multiple-choice question. The choices included worries about high infectivity of COVID-19; lack of effective therapies; the wide impact of the COVID-19 pandemic; lack of medical supplies and healthcare workers; sequelae after recovery from COVID-19; extensive announcements, rumors, and nervousness; extensive media coverage; and news of the pandemic on the internet.

This study carried out the statistical analyses with SPSS version 23.0 statistical software for Windows (IBM Corporation). The relationship between public panic and other variables was explored with chi-square tests. The risk factors and preventive factors of public panic were explored with multivariable logistic regression analysis. Associations were considered significant when there was a two-tailed p < 0.05.

The prevalence of public panic during the earliest stage of the COVID-19 pandemic was 581/2,484 (23.39%). The results from the chi-square tests on public panic are shown in Table 1. The participants who were not married (1,240/2,484, 49.92%) had a higher prevalence of panic (p = 0.022). Differences in participants panic by education level (p = 0.034) and occupations (p < 0.001) were also significant.

The psycho-behavioral responses and corresponding prevalence of public panic are also presented in Table 1; Figures 1, 2. Taking temperature repeatedly (990, 39.86%), being nervous in a crowd (1,600, 64.41%), being suspicious of infection in the family (869, 34.98%; p < 0.001), being worried about the future (1,196, 48.15%), being afraid of the recurrence of the COVID-19 pandemic (611, 24.60%), being constantly reminded of the COVID-19 pandemic by the surroundings (589, 23.71%), and being suspicious that infected people are around (712, 28.66%; p < 0.001), and avoiding gathering during the holidays (2,328, 93.72%; p = 0.001) were all significant factors of public panic.

Participates' risk perception of the COVID-19 pandemic and the corresponding prevalence of panic are shown in Table 1; Figures 3, 4. Worries of high infectivity of COVID-19 (2,036/2,484, 81.96%); lack of effective therapies (1,780, 71.66%); wide impact of the COVID-19 pandemic (1,710/2,484, 68.84%); sequelae after recovery from COVID-19 (719/2,484, 28.95%); extensive announcements, rumors, and nervousness (350, 14.09%; p < 0.001); extensive media coverage about the pandemic (276/2,484, 11.11%); and news of the pandemic on the internet (436/2,484, 17.55%; p < 0.001) were significantly associated with public panic.

The results from the multivariable logistic regression are presented in Table 2 and Figure 5. Psycho-behavioral responses, such as taking temperature repeatedly (odds ratio [OR] 1.388, 95% CI 1.130–1.706), being nervous in a crowd (OR 1.494, 95% CI 1.146–1.948), being suspicious of infection in the family (OR 1.416, 95% CI 1.128–1.775), and being worried about the future (OR 1.979, 95% CI 1.563–2.504) increased the chances of panic. Among those with worries about high infectivity of the COVID-19 (OR 1.429, 95% CI 1.055–1.934), lack of effective therapies (OR 1.402, 95% CI 1.099–1.789), and wide impact of the COVID-19 pandemic (OR 1.281, 95% CI 1.016–1.616) also increased the chances of panic. Inversely, avoiding gatherings during holidays (OR 0.313, 95% CI 0.209–0.469) decreased the chances of panic.

Our study found that the prevalence of public panic was 23.39% during the earliest stage of the COVID-19 pandemic in China. The prevalence of panic was higher than the prevalence before the pandemic (5.6–13.2%) (44–47). The pandemic has seriously threatened people's daily life and greatly disrupted the normal social order, resulting in tremendous changes in population health, and inducing the prevalence of public panic. It has been found in recent studies that the pandemic can trigger the negative emotions, such as sadness, worry and fear, and increase public's vulnerability to mental disorders, such as panic, social anxiety, depression, and post-traumatic stress disorder (48–52). During this vital period, the examination of public panic and the associated factors are indispensable, but also scarce.

Results from our study indicate that panic is significantly associated with psycho-behavioral responses. The psycho-behavioral responses, such as taking temperature repeatedly, being nervous in a crowd, being worried about the future, being afraid of the recurrence of the COVID-19 pandemic, being suspicious of infection in the family, and being worried about the future, were the key risk factors of public panic. Whereas, avoiding gatherings during holidays could decrease the possibility of public panic. The public has paid more attention to personal protection because of the severe impacts of COVID-19 in China. It is common for the public to adopt social distancing measures, mask wearing, frequent hand washing, and canceling social activities during the pandemics to reduce disease spread (53, 54). While the public health measures were necessary to prevent the spread of COVID-19, the impact on public behaviors and risk perception damaged mental wellbeing (55, 56). Previous research also reported that social distancing behaviors, a common response to avoid infection, were associated with negative psychological responses (57, 58). The rapid changes during the COVID-19 pandemic raised problems in social order and development of finance, resulting in risks threatening people's health, stable income, and living quality, triggering further public panic and other mental disorders.

The results of our survey were in agreement with the findings in recent studies that during the public health emergencies, the disruption of daily routine, the uncertainty of epidemic, the threats to wellbeing, and the wide range of disease-related information were the main factors associated with psychological distress including fear, anxiety, depression, and panic (59–61). Strengthening the understanding of psychological responses and their impact on mental health, reducing public panic, and effectively managing irrational responses have become prominent issues that need to be resolved globally in order to reduce the detrimental impacts of the pandemic on people's psychological health. Therefore, psychological interventions and guidance on behaviors during public health emergencies should be developed and popularized among the public to help them regulate negative psychological changes and to cope with the pandemic with appropriately rational behaviors.

In addition, the public panic was associated with risk perception of the COVID-19 pandemic. In this study, worries about the high infectivity of the COVID-19, worried about lack of effective therapies, and wide impact of the COVID-19 pandemic have increased the public's risk perception of the COVID-19 threat. Cognitive responses are associated with behavior coping and perception of the pandemic (62, 63). Perceptions of the pandemic, such as the mode of transmission, symptoms and therapies of the disease, proper self-protective behaviors, the importance of early detection, and treatment of mental disorders, should be enhanced among the public through official channels (64–66). Besides, the worries about potential risks of tough treatment experience, huge medical cost, long time of recovery, delay of study plan, or unemployment raised from infection of COVID-19 might be part of the influencing factors of panic.

In addition to the variables as above, the outcomes of the chi-square test also indicated that the prevalence of panic was also higher among the participants who were concerned about sequelae after recovery from COVID-19; extensive announcements, rumors, and nervousness; extensive media coverage; and news of the pandemic on the internet. The methods of acquisition of health information were also significantly associated with poor mental wellbeing and public's knowledge and attitude about the COVID-19 pandemic (11, 67, 68). While the various sources of pandemic-related information provide convenience for the public to access information in a timely manner, the accuracy and authenticity of the health information were hard to guarantee, resulting in a wide diversity of psychological-behavioral responses, or even triggering public panic (69, 70).

The general public is vulnerable to excessively high-risk perception and uncoordinated behavioral responses during a public health emergency (71, 72). Negative psychological impacts raised by inappropriate risk perception of the COVID-19 pandemic can seriously affect the public's normal life and mental health (9, 10, 32). Although people should protect their health, they should also adjust their risk perceptions and enhance their coping mechanisms during epidemics (73, 74). A recent study found that <50% of the public had sufficient and appropriate risk perception of the COVID-19 pandemic to ensure their wellbeing (75). Educational programs could resolve the misconceptions about the COVID-19 pandemic (76). It is vital for the public to perceive the pandemic within a controllable range with reliable sources of health information and to cope with pandemic-related stressors in healthy ways with available instructions from psychological professionals during public health emergencies.

This study has several limitations. First, since the prevalence of panic is analyzed with a self-developed question, the generalizability of the outcomes is limited to other populations. Second, this survey was conducted during the earliest stage of the COVID-19 pandemic. Therefore, the collected data can only represent this special period, and the generalizability of the results to the other stages of the COVID-19 pandemic might be limited. Third, the results are limited by selection bias because the survey was conducted on the WeChat platform and the participants were only smartphone users. Finally, the results may be limited due to unmeasured confounders, such as gender-based violence and family violence.