The outbreak of coronavirus disease 2019 (COVID-19) has brought significant impact to the mental health of the public. According to the China Mental Health Survey, a national epidemiological survey with 32,552 respondents, the prevalence of depressive disorder in general population of China is 6.8%. As for anxiety disorder, the prevalence is 7.6%. ^{Reference Huang, Wang and Wang1} However, according to a survey after the first month of COVID outbreak, moderate-to-severe stress, anxiety, and depression were noted in 8.1%, 28.8%, and 16.5%, respectively, in general public. ^{Reference Wang, Pan and Wan2}

Medical professionals are usually exposed to high occupational stress, which makes them more vulnerable to mental disorders such as anxiety, depression, and occupational burnout than general populations. ^{Reference Tyssen and Vaglum3} Previous studies have shown that medical workers experience a relatively high level of psychological disturbances, and the levels of stress, anxiety, and depression among this population seem to be much more severe than expected. In a systematic review, the estimated prevalence of depression or depressive symptoms among resident physicians was 28.8% (ranging from 20.9% to 43.2%). ^{Reference Mata, Ramos and Bansal4} Another study in England found that anxiety and depressive symptoms were prevalent in 29% and 27% senior health professionals, respectively. ^{Reference Caplan5} Researchers in China showed that the prevalence of anxiety symptoms among physicians was 25.5%, and the prevalence of depressive symptoms reached 28.1%. ^{Reference Gong, Han and Chen6}

In addition to the stress from their daily life and work, medical workers usually encounter more life-threatening stress than general populations. One previous study found that the prevalence of psychiatric morbidity was approximately 75% among health workers during the outbreak of severe acute respiratory syndrome (SARS). ^{Reference Chong, Wang and Hsieh7} In 2019, an outbreak of a novel coronavirus pneumonia (COVID-19) occurred in China. More than 1700 health workers in China have been infected by COVID-19. Facing this large-scale infectious public health event, medical staff are under both physical and psychological pressure. Hospitals are usually areas with a high incidence of infection. Medical staff not only have to face their own negative emotions, such as fear and worry about the disease, but also need to face all kinds of clinical work during the epidemic. Undoubtedly, these pressures increase the mental and psychological burden of medical staff.

During the peak of the epidemic, all Chinese people experienced a variety of worries, fears, and anxieties in the face of the growing epidemic and the overwhelming confusion of the daily news. For the medical staff, the confusing news undoubtedly increased their fear and worry, such that they worried every day about issues, such as whether the patients they come into contact with are infected, being infected themselves, and transmitting the virus to their family.

Somatic symptoms are a common physical response to stressful information, and previous research claimed that somatic complaints were significantly associated with stressful life incidents and information. ^{Reference Hart, Hodgkinson and Belcher8,Reference Ford9} Emotional distress and arousal toward stressful information or incidents frequently give rise to bodily sensations. Thus, the response is associated with increased activity of the sympathetic nervous system, resulting in many bodily changes, including sweating, pounding heart, tachycardia, shaking, muscle tension, shortness of breath, dizziness, etc. ^{Reference Ursin10} It is proposed that heightened somatic and physiological conditions induced by stressful events may increase the total amount of time that stress has a “wear and tear” effect on the human body, ^{Reference Melamed, Ugarten and Shirom11} while the extent of somatic discomfort triggered by social media and the impact of social media on the mental health and well-being of medical professionals is largely unreported.

Research on the influences of social media on the psychological well-being of the population has increased recently. Social media has become fundamental in the way many people and organizations communicate and share opinions, ideas, and information. Previous researchers claimed that exposure to information loaded with feelings of confusion, uncertainty, anxiety, fearful, desperation, and sadness and negative information usually leads to less trust in health recommendations and increased psychological disturbances. ^{Reference Nagler12,Reference Zhu, Wu and Miao13} The impact of the Internet, particularly social media, on psychological disturbances is a topic of growing interest that has drawn national attention to this topic. To some extent, social media’s influence on psychological disturbance should be considered a public health problem, and how public health approaches might be used to address this influence is a relevant issue. In this article, we discuss the role of social media on psychological disturbance among medical workers from a public health perspective. At the same time, medical personnel often have a higher prevalence of mental health disorders. Consequently, we examined Chinese physicians’ anxiety and depressive symptoms as well as evaluated their mental health disorder history and social media use during the outbreak of COVID-19.

Results

Basic Characteristics

Data collected from 460 (131 male and 329 female) medical professionals were analyzed. For marriage status, most of the participants were married (264/460; 57.4%), and 38.5% of them were single (177/460). Most of them claimed that they did not have any religious belief (388/460; 84.3%). A total of 94.1% of medical professionals were from the Han ethnic group (433/460), and 5.9% of the participants were from ethnic minority groups (27/460). Other detailed characteristics of the included patients are summarized in Table 1.

Table 1. Demographics information of the participants

For mental health disorder history, most of the participants reported no history of mental health disorder (324/460; 70.4%); 18.3% of them suffered from 1 mental health disorder (84/460); 6.7% reported comorbidity with 1 mental health disorder (31/460); and 4.6% reported comorbidity with 2 mental health disorders (21/460).

To specifically examine our 3 hypotheses, hierarchical regression analyses were used to examine participants’ reports of (1) the history of mental health disorders, (2) confusion triggered by social media, and (3) somatic discomfort triggered by social media. In these regression analyses, the covariates age, education, religion, marriage situation, and income level were entered into Step 1; mental health history and comorbidity were entered into Step 2; confusion triggered by social media was entered into Step 3; and somatic discomfort triggered by social media was entered into Step 4. For these analyses, all variables were centered.

Hypothesis 1 was supported, indicating that participants’ mental health disorders were positively associated with depression and anxiety. In other words, participants who reported having mental health disorders were more likely to report depression and anxiety.

Hypothesis 2 was that participants’ increased feelings of confusion triggered by social media would be positively associated with anxiety or depression.

Hypothesis 3 was that participants’ increased somatic discomfort triggered by social media would be positively associated with anxiety or depression.

When considering covariates and mental health comorbidity, both of the variables were positively related to anxiety or depression. In other words, participants who reported more comorbidity with mental health disorders and more negative feelings triggered by social media also reported severe anxiety or depression (see Tables 2 and 3). The first step included age, education level, marriage, sex, major, marriage status, and income level. This model was significant, accounting for 1.7% of the variance in anxiety. Mental health comorbidity was added in the second step and was a significant independent predictor of anxiety, and this model accounted for 20.7% of the variance in anxiety. Confusion triggered by social media was added in the third step and was a significant independent predictor of anxiety, and this model accounted for 35.8% of the variance in anxiety. Somatic discomfort triggered by social media was added in the last step and was a significant independent predictor of anxiety, and this model accounted for 49.8% of the variance in anxiety.

Table 2. Predictors of depression

Table 3. Predictors of anxiety

Regarding depression, the first step included age, education level, marriage, sex, major, marriage status, and income level. This model was not significant, accounting for 1.3% of the variance in depression. Mental health comorbidity was added in the second step and was a significant independent predictor of anxiety, and this model accounted for 21.7% of the variance in depression. Confusion triggered by social media was added in the third step and was a significant independent predictor of depression, and this model accounted for 31% of the variance in depression. Somatic discomfort triggered by social media was added in the last step and was a significant independent predictor of depression, and this model accounted for 40.4% of the variance in depression.