Associations between the working experiences at frontline of COVID-19 pandemic and mental health of Korean public health doctors

From February 2020 to the present (as of April 2021), PHDs continue to be dispatched to the working locations of frontline such as rural public health centers, airport quarantine stations, and correctional facilities. At frontline of COVID-19 pandemic, PHDs conduct acquisition of respiratory specimen for diagnostic tests, epidemiological investigations to identify the paths of movements for confirmed-positive patients, and participation in patient care in the inpatient quarantine units, among others. The inclusion criteria were 1) Male doctors who had been serving active duty of PHD for Republic of Korea and had been members of Korean Association of Public Health Doctors as of August 2020 [population size = 1917], 2) Prior or current experiences of working at COVID-19 front-line between February and August of 2020 with duties of respiratory swab collection, epidemiologic investigation of the path of the confirmed patient for COVID-19, making triage (among the active monitoring, self-isolation, COVID-testing) of the people with possible recent contacts with confirmed individual, managing and treating COVID-positive patients, and 3) Those who willing to voluntarily participate in this web-based survey on August 2020. Cases who cannot satisfy all of these criteria were not allowed to participate the current study. The minimum number of necessary samples to satisfy the desired statistical constraints (confidence level = 95%; margin of error = 5%; population proportion = 50%; population size = 1917) calculated using the web-based ‘Sample Size Calculator’ program (https://​www.​calculator.​net/​math-calculator.​html) was 321. Also taking into account the response rate of members of Korean Association of Public Health Doctors (≈ 18–19%), the final sample size was determined as 350 (out of the population size = 1,917).

During this pandemic, it is suggested that data collection should be done off-site and on an online platform to prevent the further spread of COVID-19. Promotional documents for the current study was posted on the website of the Korean Association of Public Health Doctors (http://​kaphd.​org) and also delivered to the members of Korean Association of Public Health Doctors (N = 1917) by way of the mobile message on August 7, 2020. After reading the promotional document delivered, PHDs could click the web-based link for the online survey and could participate in the current study by responding to the questionnaires followed. Informed consent was indicated by clicking on the “start” button on the front page of the online survey. The survey was anonymous, and confidentiality of information was assured. Study participation was run by first-come, first-served manner for a total of 350 participants. In other words, after the 350 research participants have entered their answers into the web-based questionnaire, acquisition of further responses (or further study participation) by way of the web-based questionnaire was closed. Accordingly, this online survey was conducted from August 7 to August 18 of 2020 (as the cumulative number of study participants reached the target sample size of 350; Fig. 1). The current study was approved by the clinical research ethics committee of Seoul National University College of Medicine and Hospital (IRB No. 2007–152-1143). All methods were performed in accordance with relevant guidelines and regulations.

In the current study, items of the online survey (supplementary information: S1) were designed to explore possible associations between PHD mental health (depressive mood, anxiety, perceived stress, and presenteeism) and elements of their dispatch experience at the COVID-19 front line (working environment, workload, and perceived threat of COVID-19 infection). Most of all, assessments of mental health for PHDs was conducted using the four self-reports. First, depressive symptoms were measured using the Korean version of the Patient Health Questionnaire-9 (PHQ-9; S1. Part 5) [31‐33]. Second, anxiety symptoms were measured by way of the Korean version of the Generalized Anxiety Disorder-7 scale (GAD-7; S1. Part 6) [34]. Third, perceived stress was explored using the Korean version of the Perceived Stress Scale (PSS; S1. Part 7) [35, 36]. Fourth, self-efficacy in one’s current PHD medical practice (work-related self-efficacy) was measured using the Stanford Presenteeism Scale-6 (SPS-6) [37] adapted for the COVID-19 pandemic (S1. Part 8).

Moreover, working experiences at frontline of COVID-19 and demographic factors of PHDs were also examined. First, basic demographic data including age, marital status, number of children, education level, and years of clinical experience were collected (S1. Part 1). Second, questions were asked about the working environment at the COVID-19 frontline, including timepoints/duration, urgency of dispatch, amount of training prior to dispatch, workload at the front line, and ease of acquiring protective equipment (S1. Parts 2 and 4). Third, PHDs were asked about the location and type of facility where they worked (S1. Part 2). Fourth, satisfaction with the experience of working at the frontline, such as subjective workload, perceived capability of participating in work-related decision making, satisfaction with the salary, and intention to volunteer for further dispatch, was examined using a 10-point Likert scale (S1. Part 4). Fourth, the perceived threat of COVID-19 infection was evaluated using 10 items, modified from Um et al. [38], a study that aimed to evaluate medical doctors’ perceptions of the risks associated with the 2017 Middle East Respiratory Syndrome (MERS) pandemic (S1. Part 3).

To examine possible associations between mental health (depressive mood, anxiety, perceived stress, and work-related self-efficacy) and working experiences at frontline of COVID-19 pandemic (working environment and psychosocial aspects) for PHD using the logistic regressions, related variables were recoded into binary or categorical formats (Table 1). First, mental health of PHD (S1. Parts 5–8) were classified into 1) non-depressive (PHQ-9 total score < 10) vs. depressive (PHQ-9 total score ≥ 10) [31‐33], 2) not anxious (GAD-7 total score < 10) vs. anxious (GAD-7 total score ≥ 10) [34], 3) not distressed (PSS total score < 18) vs. distressed (PSS total score ≥ 18) [35, 36], and 4) lower (SPS-6 total score < 19) vs. higher (SPS-6 total score ≥ 19) work-related self-efficacy [37]. Second, demographic (S1. Part 1) and working environmental (S1. Part 2 & items 4–1-1 to 4–1-4 of Part 4) variables were transformed into the categorical format (as written in Table 1). Third, Likert scale responses for the perceived threats of COVID-19 infection (S1. Part 3) were binary-transformed (never/not so much/unsure grouped as ‘NO’; possibly/certainly regarded as ‘YES’). Fourth, numerical rating responses for satisfaction of working experience at frontline (S1. Part 4) were recoded as YES (> 5) or NO (≤ 5).

Using the univariate binary logistic regression analyses, demographics or variables of working experiences at frontline that significantly explained mental health [non-depressive vs. depressive, non-anxious vs. anxious, non-distressed vs. distressed, and lower vs. higher work-related self-efficacy] were retrieved (all Ps < 0.05; Table 2). Finally, multivariate logistic regression analyses using an R package of glmulti [with options of: level = 1 (no interaction considered); method = “h” (exhaustive search); crit = “aic” (Akaike information criterion as criteria of model selection); fitfunction = “glm” (use glm function for regression analyses)] found best model of explaining mental health among all possible combinations of working experiences and demographics that were significant (P < 0.05) in the univariate binary logistic regression (Table 3; refer to S2. R code for multivariate logistic regression in the supplementary information). Data were analyzed using R version 4.0.2 (R Core Development Team, 2020).

Among the 1917 PHDs asked to participate, 350 (18.3%) completed the survey. Table 1 summarizes the demographics and dispatch experiences of all study participants. Participants were males aged 24–34 years. Most were medical school graduates (94%) who had not yet completed their internship training (62.9%), and many were in their first year as PHDs (41.7%). With notification of impending dispatch to the COVID-19 frontline, usually within 7 days of departure (77.1%), they were provided with an average of 3.96 h of education. They were then dispatched mostly to the COVID-19 screening center (85.1%) within (86.6%) or outside (64.9%) their pre-station PHD location, including the Daegu metropolitan area (35.1%; a major outbreak region from February to March of 2020).

At the dispatch location, the PHDs worked an average of 36.6 h per week and performed an average of 19.7 nasopharyngeal swabs per day. Participant ratings were as follows (average points out of 10 possible): subjective workload, 5.7; perceived ability to participate in decision making, 4.2; satisfaction with monetary compensation, 3.5. Furthermore, the score for the intention to volunteer further for the COVID-19 front line was 3.9 points out of 10 (on average). Most were afraid that they would get infected (82.6%) and transfer COVID-19 to someone else (86%). They also felt more distress during COVID-19 frontline duty than during their other duties as a PHD (75.4%). Nonetheless, the PHDs accepted the risk of caring for COVID-19 patients (84.6%).

With regard to self-reported mental health status as of August 2020, of all participating PHDs (N = 350), 53 (15.1%) reported a moderate or severe depressive mood (PHQ-9 total score ≥ 10) and were classified in the depressed subgroup [33]. With respect to anxiety, moderate or severe anxiety (GAD-7 scores ≥10) was reported by 40 PHDs (11.4%; anxious subgroup) [34, 39]. Also, 160 PHDs (45.7%) reported higher perceived stress (PSS total score ≥ 18; Figure S1(A)) and were classified as distressed [36]. Finally, 121 PHDs (34.6%) exhibited lowered work-related self-efficacy (SPS-6 total score ≥ 19, Figure S1(B)) or presenteeism [37]. Binary logistic regression analyses were performed using the glmulti function in R software to ascertain the effects of the dispatch experience at the COVID-19 frontline on the likelihood that a PHD would suffer depressive mood, anxiety, perceived stress, or lowered self-efficacy at work [40]. Results of the univariate binary logistic regression analyses are demonstrated in Table 2. Also, explanatory variables comprising the best multivariate logistic regression model (with minimum value of Akaike information criterion) in predicting the mental health outcome of PHDs (Table 3) are presented below, with adjusted odds ratios (AORs) and 95% confidence intervals (CIs) [40].

First, higher odds of depressive mood (PHQ-9 total score ≥ 10) were found among those with longer working hours during dispatch [AOR = 2.51 (35–45 h/week, 95% CI = 1.10–6.03, P = 0.033) and 3.24 (> 45 h/week, 95% CI = 1.34–8.20, P = 0.01)], working at the COVID-19 screening center (AOR = 6.07, 95% CI = 1.61–40.50, P = 0.022), and who perceived that people would avoid them because of their COVID-19 duty (AOR = 2.10, 95% CI = 1.04–4.20, P = 0.037). By contrast, an adequate supply of protective equipment (AOR = 0.32, 95% CI = 0.15–0.65, P = 0.002) and a proactive response to perceived threats, such as “I can accept the risk of caring for COVID-19 patients” (AOR = 0.35, 95% CI = 0.16–0.77, P = 0.008), were associated with lower odds of depressive mood.

Second, PHDs with moderate or severe anxiety (GAD-7 total score ≥ 10) reported severe levels of perceived threats, as expressed by “I have little chance of survival if I were to get COVID-19” (AOR = 8.41, 95% CI = 1.53–48.60, P = 0.013) and “My family and friends are worried they might get infected with COVID-19 through me” (AOR = 6.50, 95% CI = 2.05–30.26, P = 0.005). An adequate supply of protective equipment for COVID-19 duty was related to lower odds of anxiety (AOR = 0.38, 95% CI = 0.17–0.84, P = 0.015).

Third, higher perceived stress (PSS total score ≥ 18), including higher perceived threats such as those expressed by “I feel more stress during COVID-19 duty than during other tasks” (AOR = 2.04, 95% CI = 1.03–4.11, P = 0.042) and “If it were possible, I would resign from COVID-19 duty” (AOR = 2.48, 95% CI = 1.46–4.24, P = 0.001), was associated with assignment to a COVID-19 screening center (AOR = 2.90, 95% CI = 1.39–6.48, P = 0.006). By contrast, PHDs who were provided with adequate protective equipment (AOR = 0.36, 95% CI = 0.18–0.70, P = 0.003) and who were satisfied with the monetary compensation (AOR = 0.55, 95% CI = 0.31–0.98, P = 0.045) were at lower risk of perceived stress. Finally, PHDs with lowered self-efficacy at work or those exhibiting presenteeism (SPS-6 total score ≥ 19) felt more stress during COVID-19 duty compared to other assignments (AOR = 4.58, 95% CI = 2.32–9.93, P < 0.001); a willingness to further volunteer for COVID-19 dispatch was associated with lower odds of presenteeism (AOR = 0.47, 95% CI = 0.26–0.82, P = 0.009).

In the current cross-sectional online survey that enrolled 350 PHDs (= 18.3% of population number), perceived distress, lowered self-efficacy at work, anxiety, and depressive mood were reported by 45.7, 34.6, 11.4, and 15.1% of the public health doctors, respectively. This result is in concordance with other studies that demonstrated higher prevalence of mental symptoms in healthcare professionals at frontline of pandemic situation. After working experiences at frontline of COVID-10 pandemic, about 24.7–50.4%, 19.8–44.6%, and 21.9–71.5% of healthcare professionals in China or Italy reported depressive mood, anxiety, and perceived distress, which were higher prevalence compared to those at baseline (=prior to the working experiences at frontline of COVID-19 pandemic) [41‐44]. In addition, lowered self-efficacy at work (or a lower sense of personal accomplishment) was also reported from 21.4–22.7% of healthcare professionals of Libya during COVID-19 pandemic [45, 46]. For the cases of MERS outbreak, 26.6% of doctors who dealt with the MERS outbreak in South Korea [38] and 27.5% of doctors responding to the SARS outbreak in Taiwan exhibited depression [47]. During the SARS outbreak, 89% of healthcare workers who were in high-risk situations in Hong Kong reported psychological symptoms [48]. Specific viral characteristics of COVID-19, e.g., its high infectivity and often-fatal outcome in older populations, might have played a role in the experiences and feelings of healthcare workers (including PHDs). Although similarities between these cases and PHDs are evident, some characteristics unique to PHDs and their working conditions should be considered. First, PHDs are younger-aged (20s–30s) male doctors who are enrolled in mandatory military service. The average age of the study population was 28.9 years (SD = 2.11 years). Second, the intrinsic nature of the PHD system, i.e., a non-voluntary working force who must follow government orders and are stationed far from home, could have affected the psychology of these frontline physicians in a different way from other healthcare workers. These circumstances are unique and should be documented.

Multivariate logistic regression demonstrated that insufficient supply of personal protective equipment during the working experiences at frontline of COVID-19 pandemic might be a predictor of depressive mood, anxiety, and distress for PHDs. In addition, draft to the COVID-19 frontline of screening center and longer weekly workhours were also included as predictors in the best multivariate binary logistic regression model of depressive mood for PHDs. The current study result is in concordance with recent studies of healthcare workers in German or Italy during COVID-19 that showed associations between working at COVID-19 frontline versus higher level of perceived stress or post-traumatic symptoms [44, 49]. Further, other studies that examined mental health of medical professionals at frontline of COVID-19 pandemic also demonstrated increased odds of mental health worsening (depressive mood, anxiety, or burnout) in presence of working environments such as insufficient supply of personal protective equipment [50‐54], endorsed barriers at working [55], increased physical and psychological workload [51, 56, 57], reduced amount of day-off and successive period of active duty after 24-h shifts [50]. Collectively, sufficient and timely supply of personal protective equipment and efficient distribution of rest period between the active duty are required for protection of mental health for PHDs at frontline of pandemic.

In addition, psychosocial aspects of working experience at frontline of COVID-19 also significantly predicted mental health of PHDs. First, perceived threat of infection and fear of death comprised the best multivariate binary logistic regression model of predicting anxiety of PHDs. These results are in accordance with cases of medical healthcare workers that demonstrated positive associations between the severity of perceived threat for COVID-19 or being at risk of contact with COVID-19 patients versus reports of depressive symptoms, anxiety, and insomnia [44, 58, 59]. Fear of contracting COVID-19 at work is associated with higher emotional exhaustion and depersonalization [45] and could be one of the most prominent distress for medical professionals with frequent contacts with COVID-positive patients [54, 57, 60]. Second, perceived stigma from family and friends (worries for possible transmission of infection through PHDs at frontline) and rejection from neighborhood [ex. local residents forcefully entered a PHD’s house and sprayed disinfectants as the PHD had just finished duty at frontline of pandemic and returned to his original workplace as PHD (https://​www.​seoul.​co.​kr/​ news/​newsView.​php?​id=​20200319500158&​wlog_​tag3=​naver)] predicted anxiety and depressive mood of PHDs, respectively. Another study of healthcare workers in Libya also showed significant associations between the stigmatization and depressive symptoms and anxiety [61]. Therefore, greater psychosocial support from family, friends, supervisors and better cooperation between colleagues at workplace are important [15, 57]. Also, proper educational training on COVID-19 to the healthcare professionals in addition to the clearer dissemination of disease-related information to the general population could help to reduce anxiety in PHDs on the frontline of COVID-19 pandemic [15, 18].

Protection of mental health for healthcare professionals at frontline of pandemic situation is critical, not only for successful termination of COVID-19 pandemic in the long term [62] but also for prevention of suffer of PHDs and military medical staffs from post-traumatic stress symptoms after deployment [63, 64]. Also, the current study revealed possible protective factors of mental health for PHDs with working experiences at frontline of COVID-19 pandemic. First, satisfaction for monetary compensation was associated with lower level of distress. This is in line with another recent study that demonstrated positive association between the monetary compensation and willingness of nurses to care for patients with COVID-19 [65]. Therefore, application of the remuneration system for PHDs that better reflects the risks of working environments [65] would be more beneficial in enhancing the willingness of PHDs to volunteer duty at frontline of pandemic. Second, proactive coping strategy of altruistic acceptance for managing infected patients at frontline and willingness to volunteer further duty at frontline were predictive of less depressive mood and higher work-related self-efficacy, respectively. As a matter of fact, prolonged exposure to the working environment of pandemic frontline could result in biological injury (of COVID-19 infection) and psychological injury (of mental health worsening) of PHDs [66]. A recent study of medical professionals in Wuhan of China emphasizes the importance of timely mental health supports for frontline workers during pandemic [43]. Psychological supports [67] that provide strategies of positive coping (altruistic acceptance of work-related risks, motivation to learning different skills, and humor) [15, 67], psychological debriefing [15, 68], and regular monitoring of mental health [67] aiming to delivering proper treatment by psychiatrists would be effective in protecting mental health of PHDs at frontline. In addition, short-term and long-term plans to support mental health of healthcare workers during and after the COVID-19 pandemic is warranted [54].

The current study has some limitations to be addressed. First, some possible associated factors of the working experiences at pandemic frontline, such as experience of quarantine during this pandemic and perceived social supports from family/supervisors and colleagues/neighbors [15] were not covered in this study. Second, duration of data collection (August 2021) was relatively short. Third, approximately half of the participants were still actively working in the COVID-19 field at the time of the survey, with other half already had moved from the frontline of COVID-19 at time-point of study participation.