Occupational injuries and risk factors in agriculture

Since KNHANES was a survey designed to examine South Koreans' general health and nutritional status, detailed data on occupational injury were not available, although the rate of injury could be compared across different occupational categories [83]. In the EU and the US, various types of research have long been conducted to study the attributes of occupational injury in agriculture [5]. In contrast, little research has been conducted in South Korea on the attributes of farm-work related injury and risk factors. It is difficult to generalize the findings from farm-work injury research in advanced countries where the social structure differs from South Korea and animal husbandry (e.g., dairy farming) is a major farm type, to South Korea where rice farming and open field farming are the primary farm types. Therefore, it is essential to investigate the attributes of farm-work related injuries in South Korean farmers to develop criteria for managing agricultural safety and health and responses to farm-work related injury [9,10,84].

Accordingly, in this thesis a nationwide survey study was conducted to examine the scale, types, and sources of farm-work related injury. The results showed that in 2012 3.6% of farmers experienced farm-work related nonfatal injuries requiring outpatient or inpatient treatment.

This 3.6% injury rate is similar to the rate (3.2%) reported in a study that surveyed 8,064 farmers to examine the proportion of farm-work related injuries resulting in four or more off-work days in 2008 [10]. However, the 3.6% rate is much higher than the 1.3% reported in agriculture employees insured by WCI in 2012. Karttunen and Rautiainen (2013) mentioned that injury rates based on survey data tended to be higher than injury rates based on insurance records [39].

Study II may have observed a higher rate because, young farmers employed by incorporated farms tended to insure under WCI (in 2016, only 3.3% of all farmers had WCI) [37]. Additionally, since relatively few farmers may be fully aware of the concept of farm-work related injury, farmers often do not file a WCI claim when an injury occurs.

The rate of farm-work related injuries requiring hospital treatment reported by western countries was 4.4% in Sweden [36] and 9.3% in the US [31]. The figure for Finland was 5.9% [47] , which was computed based on insurance records. The injury rate computed in Study II was relatively low in comparison to the rates

reported in these studies. There are few hospitals and limited public

transportation in rural South Korea. Thus, it is difficult for elderly farmers who do not drive to go to a county or township hospital [18]; consequently, the number of injuries that met the study criteria (which included outpatient or inpatient hospital treatment) likely underestimated the true number of farm-worker injuries.

A more accurate rate of farm-work related injury in South Korea can be obtained by comparing the results of the biannual KFODIS and the injury rate computed from insurance records after Farmers' Safety Insurance (FSI) transitions from voluntary insurance for self-employed farmers and spouses to compulsory insurance that covers employees as well.

The analysis conducted to examine the association between farm-work related injury and age using KFODIS data showed that the OR of occupational injury was significantly higher for farmers age 50 or older compared to farmers under age 49 (Table 10, page 50). In a case-control study conducted with farmers in Iowa, US, the OR of occupational injury in farmers was 0.44 for ages 40-64 and 0.61 for ages 65 and older, using ages 22-39 as the reference [6]. In contrast, in Finland the rate ratio (RR 1.91, CI: 1.01-4.00) was reported to be significantly higher for ages 30-39 compared to ages 60-69, and in Sweden age-dependent differences in injury rates were not found [36]. In South Korea, a very strong association was reported between an increase in age and an increase in farm-work related injury, beginning at age 50 [10].

Similar to what Nilsson et al. (2010) reported, elderly farmers in South Korea tend to continue farming because of insufficient pension, financial difficulty, and lack of social activity experience other than farming, even if old age makes it difficult to perform farm work [36]. With aging, most individuals lose some degree of physical ability in vision, hearing, reflexes, and emergency coping skill, and are easily fatigued. The high-risk level of occupational injury among elderly farmers in South Korea can be explained by the current farming status of the country where many farmers grow dry field and greenhouse crops demanding intense physical labor and use old rotary tillers with few safety devices.

Nilsson et al. (2010) stated elderly farmers obtain safety knowledge through many years of farming experience and may learn safer methods while working [36]. In addition, Sprince et al. (2003) suggested that younger farmers are more likely to experience occupational injuries because they lack farming experience [6].

However, in the present research, an association between the length of farming experience and injury was not found.

McQuiston et al. (2012) stated that most accidents stem from a variety of reasons, including safety system errors, not just from lack of training [68]. Lehtola et al. (2008) argued that the efficiency of educational intervention in reducing occupational injury in farmers had proved to be low [85].

The risks posed by essential hazards inherent in muddy and slippery terrain and in agricultural machinery with rotating blades cannot be reduced by simply providing safety education to farmers. Particularly in South Korea, most farmers work in open dry fields or paddy fields, and are thus exposed to uncontrollable hazards such as heat waves [9]. Inconsistencies between the present study findings and Nilsson et al (2010) 's argument that farmers become better at safety management with age and that age is not associated with injury rate [36] may be a result of the unique farming environment and inadequate safety systems in South Korea compared to western countries.

Agricultural injury was associated with age, farm ownership, and nighttime working experience at statistically significant levels in the univariate analysis results. However, age, farm type, length of farming experience, number of work months in a year, and experience of working under the influence of alcohol, which were significant in the univariate model, were not found to be associated with injury in the adjusted model (Table 10, pages 51).

Chae et al (2013) reported that injury rates were significantly higher in men than in women [10]. In a review study of 10 studies examining the association between gender and farm-work related injury, the pooled OR for men compared to women was 1.68 (95% CI: 1.63−1.73) [5]. In Study II, although the proportion of the injured was approximately 1.5 times higher among men compared to women, gender differences were not confirmed in the adjusted multivariate model with farm ownership and nighttime working experience adjusted. Generally, in South Korea, female farm owners work for as many hours as male farm owners and perform physical farm work at a similar level.

As shown in Table 10 (pages 51), the OR of injury was significantly higher for farmers with nighttime working experience than those without. On the

assumption that nighttime working means overtime working, this finding is in line with a finding reported by Taattola et al (2012) that overtime working increased the likelihood for injury to occur [35]. It is believed that because of these

attributes, gender difference was insignificant in the multivariate model with farm ownership and nighttime working experience adjusted. However, gender

difference may be observed in the use of agricultural machinery, a major source of injury [86]. In future research, gender differences in the rate ratio (RR) of injury

should be explored in greater detail by carefully investigating machinery-related sources of injury and the tasks performed when injuries occurred and including these relevant variables in the logistic regression model.

Most farm owners in South Korea work full-time. Lien et al. (2006) observed a significant difference between full-time and part-time farmers in the awareness of safety management [87]. This attribute may explain the significant association between farm ownership and farm-work related injury confirmed in the thesis.

The association with nighttime working experience and farm-work related injury may be explained by lack of sleep and working in a hurry. In a study by Zhu et al (2014) [55], the risk of injury increased by 59% in farmers getting under six hours of sleep a night compared to those getting eight or more hours of sleep, and Kallioniemi et al (2011) reported that working in a hurry may be a risk factor for farm-work related injury [54]. In crop farming, there are occasions when farmers have to rush work during sowing and harvest seasons, depending on the weather and the condition of the crops. Thus, farmers often work into the night to

complete work within a restricted time frame. This attribute of farm work can explain the significant relationship between nighttime working experience and injury.

Generally, the rate of occupational injury is higher in animal husbandry such as dairy farming, pig farming, and beef cattle farming than in crop farming [47].

However, in the present research, no significant difference was found in the ORs for crop farming and animal husbandry in the unadjusted logistic regression, and farm type also did not affect injury occurrence in the adjusted model. Unlike crop farms, animal farms in South Korea are large in scale and animal farm owners and family-member farmers tend to concentrate on clerical tasks and management, leaving the physical labor like animal shepherding and barn maintenance to migrant workers. Kim et al. (2017) reported that of 407 farmers working on South Korean cattle farms, 16.5% experienced farm-work related injury in the past year [88]. Because Study II was conducted on farm owners sampled by region, the rate of occupational injury in animal husbandry was likely to be underestimated.

Working under the influence of alcohol was found not to be associated with injury. It is a very reasonable proposition that working under the influence of alcohol increases the risk of occupational injury [41,56]. To increase work efficiency and mitigate musculoskeletal burden, the culture of drinking during lunch or at break time is prevalent in crop farming in South Korea. Operating agricultural machinery in the afternoon after drinking is highly likely to increase the risk of farm-work related injury. The survey item used in Study II asked directly

about respondents' experience of doing farm work under the influence of alcohol.

Thus, it is speculated that the relationship between working under the influence of alcohol and injury was weak, because respondents may have avoided giving an honest answer to the interviewer.

In South Korea, until 2008 when KFODIS was initiated, the rate of occupational injury and disease in agriculture was computed based on the annually reported WCI records. However, WCI covers only a very small fraction of farmers who work on incorporated farms protected by WCI. Hence, the injury rates based on WCI records do not represent the true status of occupational injury in farmers,

although they may be used to compare occupational injury rates across industries (including agriculture). In this thesis, the scale of farm-work related injury and the profile of risk factors were identified by using KFODIS data obtained by a direct nationwide survey of farmers. The scale of occupational injury and disease among farmers in South Korea is expected to be assessed accurately when FSI becomes compulsory.