Fatal work accidents in Ecuador from 2014 to 2020: How the age of the deceased worker relates to the accidents’ temporal and geographical characteristics

Abstract This article analyzes the relationship between the age of male workers deceased in work accidents and temporal (year, month, day of the week, working shifts), geographic (provinces), and economic sector characteristics in which the accidents occurred in Ecuador between 2014 and 2020. Available data on fatal accidents from the Ecuadorian Social Security Institute (IESS) were collected. The results, which report both frequencies and proportions (rates), indicate that fatal accidents have decreased in the period, although the average age of deceased workers has increased. No significant differences were found regarding the month, day of the week, and work shift, nor in the frequency of accidents or the age of the deceased workers. However, differences were found in terms of geographical areas and sectors of economic activity. This study contributes to the literature as it is the first to analyze the temporal and geographical characteristics of fatal accidents about the age of the deceased worker.


Introduction
Injuries due to work accidents represent an occupational health problem worldwide, especially those that cause worker death. [1][2][3] The International Labour Organisation estimates that 380,500 people die annually from work accidents. Thus, fatal work-related injuries remain a significant problem. However, there is a slight and progressive decrease in the Americas' low-and middle-income countries, especially in agriculture, where a rate of 9.3 fatal accidents per 100,000 workers in 2010 decreased to 8.7. Also, in services, a rate of 6.0 reduced to 5.7 fatal accidents per 100,000 workers. 4 The literature has analyzed different time factors related to mortality and fatal work accidents. For example, various studies have highlighted the importance of the worker's age as a determining factor in work-related accident mortality. [5][6][7][8][9] However, the evidence is not conclusive. Numerous authors have observed that deaths are more frequent in adult ages, 7,10,11 whereas other studies place younger populations as the group most prone to death because of the severity of the injuries caused by accident. 12,13 In addition, it is worth highlighting that men are more likely to present serious work accident injuries among the older adult population than women. 14 Research also reveals that fatal accidents are more frequent in certain seasons of the year, on specific days of the week, and at particular times of the working day. 15,16 In addition, the frequency of fatal accidents concentrates in areas with the highest population and the most dangerous economic sectors, such as construction and industries. 17 However, the distribution and rate of fatal work-related injuries differ between regions concerning the particular characteristics of each country. 18,19 Few published studies exist regarding time-related factors in fatal work accidents in Ecuador. Helping in reduce this gap, the purpose of this study was to analyze the relationship between the age of male workers deceased in work accidents and temporal (year, month, day of the week, working shifts), geographic (provinces), and economic sector characteristics in which the accidents occurred in Ecuador between 2014 and 2020. The records available in the Ecuadorian Institute of Social Security (IESS), including the last seven years (2014-2020), were collected.

Design and procedure
This epidemiological study explores and describes death cases by injuries of work accidents in the population affiliated with the IESS and covered by professional contingencies in 2014-2020 (the cutoff for 2020 was December 24). The information comes from the administrative records containing micro-data of certified work accidents from the Directorate of General Occupational Risk Insurance (DSGRT) from IESS. This Office assesses the circumstances of death according to the country's regulations.
In the period (2014-2019), 1,329 fatal accidents were reported, equivalent to 1.2% of all reported work accidents. Of these fatal accidents, we focused only on those that occurred in male employees working in the usual workplace (552 fatal accidents). Fatal accidents "in itinere," those that occurred during the worker's transfer from home to work and from work to home, have not been included (368 cases). Nor have fatal accidents been considered in places that are not the ordinary place where the worker usually carries out his activities (409 cases), for example, when they are temporarily assigned to other areas to replace another worker or for planning reasons. The reason for excluding these cases was to avoid biases in the analysis by place of occurrence, 20 since the conditions and characteristics in which this type of accident occurs vary from one to another. 21 Official reports do not detail these conditions, making it impossible to equate this type of accident to those in the usual workplace.
Of the 552 fatal accidents, 62% occurred in unskilled manual occupations (23% not defined), 94% did not specify the anatomical region of the body affected, and 68% did not include the diagnosis. These limitations did not allow us to perform a more detailed analysis.

Epidemiological variables
Collected data report on the age of the deceased worker (in years) and other temporary epidemiological data such as the year, the month, the day of the week, and the working hour in which the accident occurred. Data also inform the country's territorial data (24 provinces) and economic sectors (agriculture, construction, industry, and services). Data on the working population affiliated with Social Security (denominator) was collected from the National Employment Survey. 22

Data analysis
For the data analysis, the frequency distributions and measures of central tendency: mean, standard deviation, and 95% confidence intervals were calculated. Differences between groups through a one-way ANOVA test were estimated. In addition, overall mortality rates were estimated (for every 100,000 workers) and disaggregated by the province of the country (for every 10,000 workers) for each year of the period (2014-2020). The statistical package SPSS v.25.0. to perform the analyses was used.

Results
The total number of fatal accidents meeting the criteria in the entire period was 552. The average age of deceased workers was 38.07 years (SD ¼ 12.24; range from 18 to 79 years). Table 1 shows the frequency distribution by year and age. It can be seen that the frequency, and consequently the rate per 100.000 workers, tends to decrease while the deceased workers' age tends to increase over the years. Thus, the number of deaths per year shows an evident progressive decrease, with a reduction of À81% from the beginning of the period in 2014, with 137 fatal accidents, to the end of the period in 2020, with only 26 fatal accidents. In contrast, the mean age of deceased workers increases from 38.22 to 44.69 years. Significant differences were found in the proportion of the distribution per year (X 2 (6) ¼ 107.27, p < .001). In addition, the analysis of the mean differences (one-way ANOVA) showed significant differences in the average age of the deceased workers over the years (F (6,545) ¼ 3.04, p ¼ .007). Thus, the deceased workers in 2015 and 2018 were significantly younger than those in 2020.
As for fatal accidents per month, no significant differences were found about the age of deceased workers (F (11,540) ¼ 0.71, p ¼ .73). Nor are there significant differences in the proportion of the distribution of fatal accidents per month (X 2 (11) ¼ 17.78, p ¼ .09), indicating that the number of deaths does not differ significantly from one month to another. However, it is worth noticing that November is usually the month in which fewer workers die (See the Appendix, Table 1, for more detail).
Concerning the day of the week, the days that register more deceased workers are Tuesday and Friday, with significant differences over the rest of the days (X 2 (6) ¼ 71.35, p < .001). These two days of the week accumulate 40.8% of the total cases. However, no significant differences were found in the days of the week and the mean age of deceased workers (F (6,454) ¼ 1.31, p ¼ .25). As expected, according to the time of occurrence of the fatal accidents, most occurred between 6 a.m. and 6 p.m. since these are the hours in which more workers are exposed to risk situations. No significant differences were found in the mean age of the deceased according to the time of the accident (F (3,548) Tables 2  and 3 in Appendix for more details).
According to the analysis by province, in absolute terms, Guayas (19.4%), Pichincha (18.3%), El Oro (10.1%), Azuay (8.7%), and Manab ı (7.6%) represent 71.7% of the total cases. When comparing the number of fatal accidents in the period with the average number of workers affiliated with Social Security in each month during the period according to the province, it can be seen that the highest fatal accident rate corresponds to Napo (11.14), followed by Morona Santiago (7.69) and Sucumb ıos (7.62), while Guayas and Pichincha have the lowest rate (1.48 and 1.10, respectively). On the other hand, the average age of death shows significant differences between the 24 provinces of the country (p ¼ .010), ranging from 29.3 ± 10 years in Napo to 30.6 ± 9.4 years in Santo Domingo de Los Ts achilas, where the ages are below the average compared to 53.5 years in Galapagos and 52.3 in Carchi (See Table 2). In turn, there is a significant negative correlation between the rate of fatal accidents and the average age of deceased workers (r ¼ À.52, p < .01) (See Figure 1).
Regarding the economic sector (see Table 3), although the highest number of fatal accidents occurred in the services sector (266 cases), followed by construction (116 cases) and industry (114 cases); however, the rate of accidents in relation to the working population in each sector shows that construction is the sector with the highest fatal accident rate (0.94 per 10,000 workers). In contrast, services are the sector with the lowest mortality rate. Significant differences were found in the mean age of death according to the economic sector (F (3,548) ¼ 4.43, p ¼.004), where younger deaths are significantly higher in the industry than in agriculture and services. Figure 2 shows the trend in the decrease in fatal accidents over time in each production sector.

Discussion
This study confirms a significant reduction in annual fatal work accidents in Ecuador's male population affiliated with the IESS from 2014 to 2020. In addition, it highlights differences in the percentage distribution of deaths by factors related to time, geographic, and economic sectors. It shows variations in the average age of the deceased worker according to provinces and productive sectors. Finally, it is observed that the year 2020 is a landmark, both in terms of the historical minimum of fatal work accidents and the increase in the average age of deceased workers compared to previous years. Some research has found significant reductions in work accidents due to the Covid-19 pandemic measures. 23,24 This phenomenon relates to intermittent confinement and reductions in face-to-face working hours but not to improving preventive management in companies. 25 The frequency of deaths by month varies, although the monthly average age of the deceased workers shows similar patterns. Also, our results differ from the reality of other contexts regarding the day of the week. Some studies 26,27 showed that in developed countries, the percentage of occupational accidents decreases as the week goes on, being Mondays the day with the highest occurrence of accidents, although less severe, than other days of the week. 15,28 Our results place Tuesdays and Fridays as the most critical days. Among the possible explanations, it could be that pressure to finish the work before the weekend and accumulated physical-mental fatigue lead to the commitment of errors that resulted in fatal accidents. 28 On the other hand, there is a progressive increase in the average age of the deceased workers for the first three days of the week (Monday to Wednesday), decreasing slightly on Thursdays and increasing again on Fridays. It would be of great interest to delve into those factors associated with the working days of the week and the injured worker's age. However, this analysis is not exempt from complexity.
The high percentage of cases on Saturdays is striking. Specific economic sectors, such as construction and industry, extend the standard workweek to Figure 1. Relationship between the age of the deceased worker and the accident rate. This figure shows the scatter plot of the age of the deceased worker (Y-axis) and the accident rate of the provinces (X-axis). The line shows the negative (and significant) correlation trend between these two variables. R 2 is the coefficient of determination of the explained variance.  Saturday. In addition to being a factor of imbalance between work and personal life, 29 increasing one more day of work involves more time of exposure to risk factors of a mechanical type and an added risk in the probability of occurrence of work accidents. 30 For example, a recent study conducted in the main cities of Ecuador (Guayaquil and Quito) has revealed that weeks of more than 44 working hours constitute a very considerable risk of the probability of the occurrence of accidents among men (OR ¼ 3.05; CI 95% ¼ 1.82-5.11; p <.001). 31 According to work hours, the proportion of fatal work accidents is concentrated in hours with the most significant presence of workers, as expected. Some studies have shown that many fatal injuries usually occur in the first and last hours of the working day due to the lack of physical and mental recovery from the previous day as well as exhaustion from long working hours during the day and the week. 16,32 However, no differences were observed in terms of the mean age of dead workers by time slots.
At the moment, the possible determinants of territorial variability are unknown due to sociocultural complexity and the limitation of more disaggregated data. However, the difference in the demographic profile and diversity of the labor market for each country's region is a factor to consider. 9 This study places certain provinces with low frequency, although with more extreme values in the average ages. Notably, the Service sector has the highest frequency of fatal work accidents, which may be explained because it is the economic sector with more workers than the other economic sectors. However, it has not been traditionally considered a more dangerous sector than agriculture or construction. 33,34 Furthermore, informal employment is more latent in agriculture and construction in Ecuador. This suggests that future research should consider using adequate estimation methods by territory and economic sector to reduce the risk of bias related to differences in the number of workers by group. 35 The increase in the average age of deceased workers in 2020 could be related to administrative decisions. In 2020, the Government issued guidelines to consider deaths related to occupational exposure to SARS-COV2 as occupational accidents in health workers with COVID-19. Given that the probability of dying from COVID-19 increases with age, 36 this excess of deaths could be attributed to the rise in the recognition of COVID-19 casualties in older health personnel as occupational accidents. In addition, it is possible that, for the next few years, the average age of deceased workers will increase progressively due to the aging of the Ecuadorian population. This leads to rethinking new age management approaches in public health and safety policies.
This study is not without limitations. First, it is difficult to establish relationships between the intermediate factors in the chain of deaths and age due to the data restrictions of the source of information used, which does not provide the causes and circumstances of the work accident. This information could be of great interest for a multifactorial analysis that evidences this phenomenon with greater precision. Second, Ecuador's government's statistical data collection system is limited and incipient. Data on fatal work accidents are only available since 2014. In addition, work accidents that occur in informal work are not registered. These limitations make it difficult for researchers to describe the mortality characteristics of work accidents. Similarly, various studies show the potential under-reporting in statistical information systems in middle-income countries. 19,37 However, this limitation may be overcome by the trend analysis conducted in this document, allowing us to offer a more accurate picture of reality for a relatively long period. Future efforts to address this problem in the country will require broader data, provided the statistical information systems improve their quality. 38 Third, a more comprehensive analysis will make a significant contribution based on evidence to investigate other aspects, for example, exposure to occupational risk factors and other sociocultural determinants. Finally, the observed decrease in deaths must be taken with caution since it could mask reality, as the use of other epidemiological indicators, for example, the years of working life lost, to quantify the social burden represented by deaths attributable to injuries in work accidents. 39

Conclusion
Although the results show that fatal accidents are reduced year by year, this occupational health problem continues to be a pending task. In addition, the increasing age of the deceased workers is a circumstance that aggravates the problem mainly due to the growing aging of the active population. Our results offer a first approximation, in Ecuador, to the study of the temporal and geographical aspects of the accidents, which may be related to the age of the deceased worker. The findings can help promote the development of prevention strategies to reduce mortality in the Ecuadorian male population affiliated with the IESS, especially the elderly.