Department of Occupational Health, Graduate School of Public Health, Yonsei University, Seoul 03722, Republic of Korea.
Department of Business Administration and Data Science, CHA University, 120 Haeryong-ro, Donggyo-dong, Pocheon-si 11160, Republic of Korea.
Int J Environ Res Public Health. 2022 Dec 29;20(1):575. doi: 10.3390/ijerph20010575.
A replacement driver is a type of gig worker who provides driving services to the target point with the drunk driver's own car. This study aimed to examine the association of replacement drivers (ref: paid workers) with depressive symptoms. Information on replacement drivers was collected through online/offline surveys. Data from the 8th Korea National Health and Nutrition Examination Survey were applied to construct the control group. The Patient Health Questionnaire-9; ≥5 points was defined as depressive symptoms. Adjusted odds ratios (aOR) and 95% confidence intervals (CI) were calculated by performing multivariable logistic regression analysis. The mean age of replacement drivers was 56.11. The prevalence of depressive symptoms in replacement drivers and controls were 49.63% and 12.64%, respectively. Replacement drivers showed a higher association with depressive symptoms than paid workers (aOR 7.89, 95% CI [5.53-11.26]). This relationship was prominent in the older, low-education, and low-income groups. Linear discriminant analysis was the most effective in predicting depressive symptoms among the machine learning models. Using the replacement driver feature increased the AUC values of the models. Given the strong association between depressive symptoms and replacement drivers, in-depth studies to establish guidelines to prevent mental diseases among replacement drivers are required.
代驾司机是一种提供驾驶服务的零工人员，使用醉酒司机自己的车辆将其送到目的地。本研究旨在探讨代驾司机（参考：付费工人）与抑郁症状之间的关联。通过在线/线下调查收集代驾司机的信息。应用第八次韩国国家健康和营养检查调查的数据来构建对照组。将患者健康问卷-9 的得分≥5 定义为抑郁症状。通过多变量逻辑回归分析计算调整后的优势比（aOR）和 95%置信区间（CI）。代驾司机的平均年龄为 56.11 岁。代驾司机和对照组的抑郁症状患病率分别为 49.63%和 12.64%。与付费工人相比，代驾司机与抑郁症状的相关性更高（aOR 7.89，95%CI [5.53-11.26]）。这种关系在年龄较大、受教育程度较低和收入较低的人群中更为明显。在机器学习模型中，线性判别分析是预测抑郁症状最有效的方法。使用代驾司机特征可以提高模型的 AUC 值。鉴于抑郁症状与代驾司机之间的强关联，需要深入研究制定预防代驾司机精神疾病的指南。
