Study design
This was a cross-sectional study conducted in Ningbo in between February 10 and 20, 2020.
Sample
A convenience sampling method was applied. First, with the help of the Health Commission of Ningbo, five designated hospitals in Ningbo that have treated confirmed or suspected COVID-19 patients were asked to participate in this survey. Second, we sent the online questionnaire to every department director of the five designated hospitals, which they then forwarded to every subordinate. Finally, HCWs forwarded the online questionnaire to their family members.
Data collection
To prevent the spread of COVID-19 through droplets or contact, we used an online-based survey via the WeChat-based survey program “Questionnaire Star” to collect data.
Participants
The inclusion criteria of this study were (1) being the next of kin of HCWs from hospitals designated for medical treatment of COVID-19 in Ningbo and (2) having access to the Internet. Exclusion criteria were as follows: (1) self-reported history of neurological disorders, mental illness, and other serious systemic disorders; and (2) self-reported substance abuse. All participants were informed of the study procedure and were invited to sign an informed consent form online upon their recruitment. Participants who passed the initial self-screening phase were asked to complete the self-administered questionnaires [25]. Of the 882 participants recruited for this study, 29 participants were excluded because they provided incomplete information, and eight were excluded because their family members took a vacation during the study period. Thus, 845 participants completed the survey. The response rate was 95.80%.
This study was approved by the Ethical Committee of Ningbo First Hospital, Ningbo, China (approval number: 2020-R042), and registered with the registry website http://www.chictr.org (registration number: ChiCTR2000030697).
Measures of dependent variables
Depressive symptoms
We employed the Chinese version of PHQ-9 to assess the depressive symptoms of HCW families. PHQ-9 is a 9-item self-report measure to assess severity of depression. Participants rated each item in accordance with the frequency of symptoms over the past 2 weeks on a 4-point scale from 0 (not at all) to 3 (nearly every day). Total scores ranged from 0 to 27, with highest scores indicating greater severity of depressive symptoms [26]. The PHQ-9 has been widely used in China and good reliability and validity of the Chinese version of PHQ-9 has been demonstrated [27]. The Cronbach’s α was 0.92 in this study. The depressive symptom was defined as a total score of ≥5 points in the PHQ-9 according to the previous study during COVID-19 epidemic [28].
Anxiety symptoms
We employed the Chinese version of GAD-7 to assess the anxiety symptoms of HCW families. GAD-7 is a self-report questionnaire that screens and measures severity of generalized anxiety disorder. Participants rated seven items according to the frequency of symptoms in the past two weeks on a 4-point scale from 0 (not at all) to 3 (nearly every day). Total scores ranged from 0 to 21, with higher scores indicating greater severity of anxiety symptoms [29]. The GAD-7 has been widely used in China and good reliability and validity of the Chinese version of GAD-7 has been confirmed [30]. The Cronbach’s α was 0.92 in this study. The presence of anxiety symptoms was defined as a total score of ≥5 points in the GAD-7 according to the previous study during COVID-19 epidemic [28].
Measures of independent variables
Demographics
The demographic characteristics included age, gender, educational level, occupation, and nature relationship with HCWs. Occupation included the following five types: (1) HCWs; (2) private sector workers; (3) government or institutional employees; (4) students; and (5) others, which consisted of freelancers, retirees, social workers, and other relevant staffs. The nature relationship with HCWs included the following four types: (1) spouses; (2) children; (3) parents; and (4) other next of kin.
Working status of family members (that is, HCWs), the participants’ knowledge of COVID-19, and COVID-19-related events in the lives of participants
The questionnaires for working status of family members (that is, HCWs), the participants’ knowledge of COVID-19, and COVID-19-related events in the lives of participants were self-developed specifically for this study, as there were no suitable scales available for measuring factors related to HCW families during the COVID-19 outbreak.
The working status of family members (that is, HCWs) included: (1) whether HCWs were in direct contact with confirmed or suspected COVID-19 patients; (2) the average working time (hours) per week for HCWs; (3) participants’ self-reported safety score for protective equipment of HCWs (safety scores ranging from 1 to 5, with higher scores indicating better protective effect); and (4) the department of HCWs, which was categorized into five types: (a) frontline departments, including the respiratory department, infectious diseases department, the intensive care unit, the fever clinic, and the isolation ward; (b) the medical technology department (imaging and laboratory departments, etc.); (c) the nursing department; (d) the logistics department; and (e) other departments.
Participants’ knowledge of COVID-19 was assessed based on responses to the following five COVID-19-related single-topic questions: (1) “Which of the following symptoms is not a common symptom of COVID-19?” with possible response options being fever, stuffy and runny nose, fatigue, and dry cough; (2) “how many days do people returning from Hubei Province need to be quarantined under observation?” with possible response options being 10 days, 12 days, 14 days, and 15 days; (3) “which of the following masks can prevent COVID-19?” with possible response options being activated carbon mask, cotton mask, sponge mask, and medical surgical mask; (4) “the known transmission routes of COVID-19 do not include which of the following?” with possible response options being contact transmission, droplet transmission, soil transmission, and aerosol transmission; and (5) “with regard to the disposal of discarded masks, what is incorrect about the following statement?” with possible response options being throw it away at any time when you run out of it; masks worn by people with fever need to be disinfected, sealed, and discarded; wash your hands immediately after handling the mask; and discarded masks should be discarded into hazardous trash cans. Of the above five questions, one point was given for each correct answer and no points were given for incorrect answers. A total score was calculated by summing points for each of the five questions, ranging from 0 to 5, with higher scores indicating a better knowledge of COVID-19.
COVID-19-related events in the lives of participants included: (1) whether there had been confirmed COVID-19 cases in families or friends; (2) whether there had been suspected COVID-19 cases in families or friends; and (3) time spent thinking about COVID-19, which was measured by the average number of hours spent thinking about the COVID-19 information every day.
Statistical analysis
The Kolmogorov-Smirnov test was used to test the normal distribution of continuous data. Continuous variables were presented as mean ± standard deviation or medians (interquartile range [IQR]) depending on the distribution of the data. Categorical variables were presented as percentages. Univariate analysis was performed using the Student’s t-test or the Mann-Whitney U-test for continuous variables depending on the distribution of the data, and the chi-square tests for categorical variables.
Multivariate logistic regression analyses were used to assess the independent association of symptoms of depression and anxiety with independent variables. Variables with a P value of < 0.20 in univariate analysis were considered potential factors for inclusion into the multivariate logistic model [31], all these variables were entered into the final model since this was an exploratory study. Model discrimination and calibration were evaluated using C-statistic and Hosmer-Lemeshow goodness-of-fit statistic, respectively.
Statistical analyses were performed using SPSS v.22.0 (IBM Corp, Armonk, NY, USA). Two-sided P < 0.05 was considered statistically significant.