Generally, occupational drivers have very stressful jobs because many of them undergo long working times and carry the risk of accident and injury. Furthermore, face-to-face contact with passengers is often needed in their jobs. Many previous studies have addressed the prevalence of psychiatric symptoms and disorders among occupational drivers [1-5], and in particular, subway drivers experience various types of problematic situations during their driving which can be risk factors for psychiatric symptoms or disorders [2,3]. Subway drivers are always at risk of running over a person on the tracks (person under train, PUT) [6-9]. Furthermore, they spend most of their working hours in an underground space, which can cause a sense of isolation and depression [2].

In 2007, a mental health check was conducted targeting subway drivers of a public company in Seoul by using The Korean version of the Composite International Diagnostic Interview (K-CIDI 2.1), a qualified diagnostic tool developed by the World Health Organization (WHO) [10]. The standardized prevalence ratios (SPRs) for the lifetime prevalence of panic disorder and PTSD were 13.3 (95% CI 6.6-22.4) and 2.1 (95% CI 1.1-3.4), respectively. With respect to lifetime prevalence, drivers who experienced PUT showed a significantly higher risk for panic disorder (Odds ratio = 4.2, 95% CI 1.2-16.6) and PTSD (OR = 4.4, 95% CI 1.3-16.4). For the 1-year prevalence, the drivers who experienced PUT had a significantly higher risk for PTSD (OR = 11.7, 95% CI 1.9-225.8). There was no significant value of SPR and OR in major depressive disorder [3].

After the mental health check was conducted, some interventional policies were suggested, which included compulsory physician counseling, minimizing driving hours, protecting drivers from the family of the victims, practicing two-person driving, installing a platform screen door (PSD) and so on [3]. According to the suggestions, most subway stations in Seoul have been equipped with a PSD, and the PSDs have prevented people from falling down on the railway, so PUTs are very rare nowadays [11]. Drivers do not experience PUTs anymore during their driving. However, other policies that were suggested after this survey have not yet been implemented. In particular, one-person driving still continues today although the conversion to two-person driving was one of the key issues [12].

There were several limitations of the survey in 2007. First, the drivers who changed their jobs to other departments did not participate in the survey. Furthermore, some drivers elected not to participate as well. Therefore, the response rate barely reached 86.7% [3]. The company kept a lukewarm attitude and the individuals in charge did not actively support the survey.

Although there has been an effort to improve the mental health of the drivers and their working environment, several cases ended in suicide within the past two years. Consequently, the necessity for active intervention was suggested by performing a mental health survey once again in 2013.

Therefore, this study aims to find the prevalence and occupational risk factors for major psychiatric disorders among subway drivers. Furthermore, we tried to find other factors related to psychiatric disorders different from those of the previous study in 2007.

This study shows that there is a higher prevalence of the three psychiatric disorders among subway drivers than that in the national data. In KECA-2011, the 1-year prevalence of MDD, PTSD, and panic disorder among adult males were 1.8%, 0.2%, and 0.0%, respectively [18]. Based on the KECA-2011 data, the SPRs of MDD and PTSD were 1.1 (95% CI 0.7-1.7) and 5.6 (95% CI 3.1-8.8), respectively. Therefore, the prevalence of MDD seems to be similar to that of the general population. On the other hand, the SPR of PTSD was 5.6. It means that the risk of PTSD among subway drivers was 5.6 times higher than that in the general population. Due to the healthy worker effect, psychiatric disorders are less prevalent in the working population. However, in this study, subway drivers easily have PTSD and panic disorder. The working environment might be a cause of a higher prevalence of psychiatric disorders in this group. Subway drivers have potential exposure to PUT events when driving their trains [3-9]. However, after 2007, PSDs were established in most stations in Seoul. Nowadays, there are very few PUT cases in a year. Therefore, other work-related factors, such as general work stress, hierarchical culture, organizational justice, and conflict with passengers can be causes for newly developed psychiatric illness in this group.

In the study performed in 2007 in the same company, the 1-year prevalence of MDD, PTSD, and panic disorder were 1.3%, 0.9% and 0.6%, respectively. In the present study, the 1-year prevalence of MDD, PTSD and panic disorder were 1.8%, 1.6% and 1.0%, respectively. These results show that the 1-year prevalence of the three psychiatric disorders increased relative to those found in the previous study. There can be several reasons for that. One of them might be that there was a higher response rate compared to the previous study (99.6% vs 86.7%) [3]. It means that an exclusion of a high risk group could have occurred in the previous study. Actually, 4 drivers who did not participate in the previous study (in 2007) were recognized to have a psychiatric disorder in the present study. Second, other newly developed stressful conditions, such as an increase of conflict with passengers, operating new train systems, and conflicts among workers due to a plurality of labor unions, emerged in their workplace after 2007. Third, any other support systems for high risk group were not introduced in this company in spite of researchers' recommendations in 2007 (two-driver system, supporting drivers who have experienced PUTs, etc.).

As mentioned above, the result of this study does not show a significant association between psychiatric disorders and PUT related factors, with the exception of the association between the number of PUT experiences and the lifetime prevalence of PTSD. Other work-related problems, on the other hand, are significantly associated with the prevalence of the psychiatric disorders. Two subway companies that the Seoul City government runs equipped most stations with PSDs. After that, the incidence of PUTs drastically decreased; thus, PUTs occurs less than twice a year. Therefore, other work-related problems have become more prominent in recent years, including conflicts with passengers, sudden stops due to an emergency bell, near accidents, breakdown, and PUT experience of a colleague. These stressful events derived from a qualitative study with drivers were asked. Although these problems seem to be less serious than PUT, they appear to elevate the risk of three psychiatric disorders in subway drivers.

The researchers of this study recommended the company to perform mental health exams regularly. It is expected that the high-risk group and the individuals that present symptoms will be found earlier and can be treated properly. As a result of a proposal by the researchers, in the late 2013 the company established a counseling center for employees where a doctor and a psychological counselor are available. There are three categories of interventions that can solve mental health problems among subway drivers. First is the common approach targeting all subway drivers. This includes various policies to enhance the mental health or to change perceptions. Second, active surveillance is needed for a high-risk group, for example, drivers who experienced PUT or suffered seriously from a conflict with a passenger. This aims to support vulnerable workers in order to overcome mental problems at an early stage. Third is the early intervention for drivers who show symptoms in order to prevent the exacerbation of disease and to facilitate a return-to-work process. Likewise, multi-dimensional approach seems to be needed to improve the mental health of subway drivers.

This study has several limitations. First, in this study, we could not identify drivers at a pre-disease stage. Even though CIDI was used to define psychiatric disorders, a screening tool for early detection and prevention should also be applied simultaneously or step by step. Second, we did not include other possible related disorders as outcome variables (e.g., somatoform disorders, general anxiety disorders, and substance-related disorders). Finally, a lower prevalence of psychiatric disorders resulted in larger CIs, meaning that the results of this study can vary as a result of a variation due to one or two cases. However, we used CIDI, where the target disease can be diagnosed correctly, but the survey results in a low prevalence of mental disorders.

In this study, PTSD and panic disorders were still more prevalent among subway drivers in comparison with the general population. We found that conflict with passengers, a near miss accident and a breakdown while driving can be risk factors for psychiatric disorders among subway drivers. Therefore, a prompt and sensitive approach for these high risk groups should be introduced in this company.

PTSD and panic disorder were more prevalent among subway drivers than in the general population. We found that having a conflict with passengers, a near accident, and a breakdown while driving can be risk factors for psychiatric disorders among subway drivers. Therefore, a prompt and sensitive approach should be introduced for these high risk groups within the subway company.