Study participants and questionnaires
The study population comprised Chinese living in mainland China. The snowball sampling method was used to invite potential study participants. Through the WeChat messaging App, the investigators invited an initial group of 10 individuals to participate, who were chosen to ensure broad representation of age, gender, education level, occupation and city. This first set of invitees then forwarded the invitations to 10 of their contacts whom they considered suitable, and this second set forwarded the invitation in the same way. Participants filled in anonymous basic information online via the Questionnaire Star (https://www.wjx.cn), and as long as they did not report a history of serious mental illness, they were asked to provide informed consent and were able to continue to the three questionnaires (see below). The study was approved by the Ethics Committee of West China Hospital, Sichuan University. Invitees were allowed to complete the survey from 4 p.m. on February 1, 2020 until midnight on February 4, 2020.
Instruments
First, participants filled in a custom-designed questionnaire that collected sociodemographic information about sex, age, educational level, occupation, family residence location, and family income. The questionnaire also asked about infection with SARS-CoV-2 (in the respondent or relatives), time spent searching for information about the virus everyday, history of contact with the epidemic area (Wuhan City), and presence of cases in the respondent’s community (Table 1).
Table 1. Demographic and clinical characteristics of the study cohort (n = 1588).
Characteristic
|
Mean±SD
|
Subgroup
|
n (%)
|
Age (year)
|
33.68±11.96
|
18-29
|
652 (41.1)
|
|
|
30-39
|
466 (29.3)
|
|
|
40-49
|
290 (18.3)
|
|
|
50-59
|
130 (8.20)
|
|
|
≥60
|
50 (3.10)
|
Family income coefficient
|
0.84±0.55
|
|
|
Sex
|
|
Male
|
526 (33.1)
|
|
|
Female
|
1062 (66.9)
|
Education level
|
|
Senior high school or lower
|
136 (8.30)
|
|
|
Technical
|
351 (22.1)
|
|
|
Bachelor
|
900 (56.7)
|
|
|
Postgraduate
|
205 (12.9)
|
Residence
in Hubei province
|
|
Yes
No
|
140 (8.80)
1448 (91.2)
|
Suspected COVID-19
|
|
Yes
No
|
256 (16.1)
1332 (83.9)
|
History of contact with epidemic area (Wuhan City of Hubei Province)
|
|
Yes
No
|
323 (20.3)
1265 (79.7)
|
Living in communities with COVID-19 cases
|
|
Yes
No
|
331 (20.8)
1257 (79.2)
|
Time spent searching for information about COVID-2019 (h/day)
|
|
1-2
3-4
5-6
7-8
>8
|
766 (48.2)
307 (19.3)
171 (10.8)
232 (14.6)
112 (7.1)
|
Family income coefficient = family income / number of people in the family
Then participants filled out the Mandarin versions of the six-item Kessler psychological distress scale (K6), the Simplified Coping Style Questionnaire (SCSQ), and the Social Support Rating Scale (SSRS).
The Mandarin version of the K6, which has been validated in the World Mental Health Survey [20], comprises six questions that ask respondents to rate how frequently they have felt ‘nervous’, ‘hopeless’, ‘restless or fidgety’, ‘so depressed that nothing could cheer you up’, ‘everything was an effort’, or ‘worthless’ during the past 30 days [21]. Items are rated on a five-point scale, with 0 indicating an absence of the symptom and 4 indicating that the symptom was always present during the past 30 days. The final K6 score can range from 0 to 24, with higher scores (≥13) indicating higher levels of psychological distress [22].
The SCSQ [23], based on the 'Ways of Coping' questionnaire [24], is a 20-item self-report that includes dimensions of active coping (12 items) and passive coping (8 items). Responses are given on a four-point Likert scale (0=never; 3=very often). The instrument has been used frequently in China, with high reliability and validity [23].
The SSRS is a 10-item self-report that assesses the level of an individual’s social support over the past year [25]. This measure consists of three subscales: subjective support (4 items), objective support (3 items), and utilization of support (3 items). Subjective support refers to perceived social support, meaning that people feel supported, cared for and helped by family members, friends and colleagues [e.g., Question: How many close friends do you have? Responses: (1) None, (2) 1-2, (3) 3-5, or (4) 6 or more]. Objective support refers to visible, practical and direct support (e.g., financial or other tangible resources that you received when you needed help). The utilization of support reflects the degree of social support used [Question: How do you get help when in need? Responses: (1) I am self-reliant, (2) I seldom ask for help from others, (3) I sometimes ask for help from others, or (4) I often ask for help from relatives and friends]. The total SSRS score ranges from 12 to 66 points, with higher scores indicating higher level of social support. The SSRS has shown good reliability and validity, with Cronbach's α ranging between 0.89 and 0.94 [25].
Quality control
Only one set of surveys was accepted from the same Internet Protocol address, and surveys were not accepted if the time to complete all questionnaires was less than 120 seconds. Surveys did not request any identifying information.
Statistical analysis
All statistical analyses were performed using SPSS 21 (IBM, Armonk, NY, USA). Exploratory data analysis was conducted using frequencies for categorical variables and mean values for continuous variables. Where appropriate, odds ratios (ORs) were reported.
Differences in demographic characteristics, coping style and social support between respondents who suspected or did not suspect that they themselves had COVID-19 were assessed for significance using the independent two-samples t test or the chi-squared test as appropriate.
To identify predictors of high psychological distress, we classified respondents into those with high psychological distress (K6 score ≥13) and those with low psychological distress (K6 score ≤12) [21].
To identify factors influencing high psychological distress among respondents who did not suspect that they had COVID-19, we performed simple binary logistic regression and backward stepwise multiple logistic regression. The dependent variable was the dichotomous classification of low or high psychological distress. The model was constructed with the following covariates: age, sex, educational level, family income coefficient (total family income/number of family members), residence location (Hubei province or other), history of contact with the epidemic area (Wuhan City) or not, time spent searching for information about COVID-19 per day, and questionnaire scores for positive coping style, negative coping style, subjective support, objective support and utilization of support. The least significant variables were removed one at a time until only significant variables (P < 0.05) remained.
Logistic regression was not performed on data from respondents who suspected that they had COVID-19, since only one of them showed low psychological distress.