Comparison of depressive symptoms among healthcare workers in high-risk versus low-risk areas during the first month of the COVID-19 pandemic in China

Abstract

Introduction: The psychological health of healthcare workers (HCWs) has become a significant concern, particularly during the initial stage of a pandemic. This study compared the depressive symptoms among HCWs in high-risk areas (HRAs) and low-risk areas (LRAs) with matching demographics. Methods: A cross-sectional study was employed to compare the depressive symptoms (Patient Health Questionnaire score ≥ 10), workplace environment characteristics, the Health Belief Model (HBM) and socio-demographics of the HCWs working in HRAs and LRAs in several accessible regions (mainly Hubei Province and Guangdong–Hong Kong–Macao Greater–Bay–Area) in China. Eight hundred eighty-five HCWs were recruited for unmatched analysis between March 6 and April 2, 2020. After matching with occupation and years of service using a 1:2 ratio, 146 HCWs in HRAs and 290 HCWs in LRAs were selected for matched analysis. Subgroup analyzes were performed using two individual logistic regressions to delineate the associated factors in LRAs and HRAs, respectively. Results: HCWs in LRAs (Prevalence = 23.7%) had 1.96 times higher odds of depressive symptoms than those in HRAs (Prevalence = 15.1%) after adjusting for occupation and years of service (p < 0.001). Significant differences in workplace environment characteristics (p < 0.001) and the 5-dimension of the HBM of HCWs (p < 0.001 to p = 0.025) were found between HRAs and LRAs. Logistic regression showed that workers with years of service between 10 and 20 years (OR:6.27), ever had contact with COVID-19 patients (OR:14.33) and had higher scores of “perceived barrier” of HBM (OR:4.48) predicted depressive symptoms in HRAs while working in pneumology departments and infectious disease units (OR:0.06), and high “self-efficacy” in the HBM (OR:0.13) was a protective factor against depressive symptoms. Contrarily, in LRAs, those HCWs who worked in ICUs (OR:2.59), had higher scores of “perceived susceptibility toward the COVID-19 outbreak” (OR:1.41), “perceived severity of the pandemic” (OR:1.25), and “perceived barriers of wearing masks” (OR:1.43) in the HBM predicted depressive symptoms. High “cues to action” (OR:0.79), and better “knowledge” (OR:0.79) in the HBM were protective factors against depressive symptoms. Conclusion: The risk of depressive symptoms of HCWS was double in LRAs than in HRAs in the first month of the COVID-19 pandemic. Furthermore, salient predictors for depressive symptoms among HCWs in HRAs and LRAs were very different.