Study area

The study area is found in the Sierra Madre Occidental in the Cañones Duranguenses subprovince formed by large-area plateaus, mainly associated with canyons and high mountain ranges with canyons (^{Comisión Nacional para el Conocimiento y Uso de la Biodiversidad [CONABIO] & Secretaría de Recursos Naturales y Medio Ambiente de Durango [SRNyMA], 2017}) in the municipality of Pueblo Nuevo, southwest of the state of Durango, Mexico. The area includes sawmills established in the forest region of El Salto that carry out production process of squared timber, both for private or ejido properties.

Methods

According to SEMARNAT (^{Secretaría de Medio Ambiente y Recursos Naturales, 2014}), in the "Centro-Occidente" forest supply basin of Durango, 20 main jobs are identified in a typical sawmill in the region, consisting of an average workforce of 26 people. Based on this information, the jobs were categorized according to their function in the following segments: a) machinery and equipment operators (sawyer and bandsaw, wood cutter, and wood-mizer operator); b) assistant operator (sawyer assistant, assistant operator for bandsaw and wood-mizer, cleaning assistant and c) general assistants (to help lift logs, turn wood, gather sawdust, to help piling and lifting wood).

A structured survey was used to create a database on the incidence of occupational accidents, which was applied to workers by direct interviews in sawmills. The survey was based on the document “Recording and notification of occupational accidents and diseases: A code of practice” of the International Labor Organization (^{ILO, 1996}). This survey included four categories: a) personal information of the worker (age, schooling, occupational safety training, years of work experience and years of service); b) place of accident at the company; c) body part affected by injury or accident: upper extremity segments (shoulder, neck, head, forearm, arm, wrist and hand), back segment, abdomen, hip and waist, and lower extremity segments (legs, knee, ankle, foot and toes); d) identification of the agent (material and causal) and mechanism that caused the injury or accident.

Sampling

A random sampling was applied to all sawmills in the region of El Salto, Durango, where the target population is represented by the workers. On this matter, 48 sawmills, 13 pallet and wooden packing box factories, 19 wooden packing box factories and three sawn timber processing industries (SEMARNAT, 2014) are recorded in El Salto sub-basin; however, the number of workers in these companies is unknown. When the size of the target population is unknown, the sample size is estimated with the following formula (^{Badii, Castillo, & Guillen, 2008}):

where,

n = population sample size

z = desired degree of confidence (95 %, equivalent to 1.96)

p = probability of success or expected proportion (0.05)

q = probability of failure 1- p (0.95)

e = acceptable limit of sampling error (3.0 %).

The result of the formula was a sample of 203 workers; however, it was expanded to 300 surveys, of which 100 were targeted to each of the three job categories (operators, assistant operator and general assistant), distributed in 26 sawmills and pallet factories that had an injury in the last five years. Surveys were applied from September to November 2020.

Statistical analysis

With the survey data, summary tables were prepared using descriptive statistics (mean, mode, standard deviation, frequency and proportion to summarize the variables age and years of work experience), as well as cross or contingency tables with non-parametric inferential statistics (Chi-square [Ӽ²] tests of association and independence to test the degree of relationship between two categorical variables) (^{Janicak, 2007}). For this, both the asymptotic method and Fisher's exact test were used as long as more than 20 % of the expected frequencies had values lower than 5 (^{Sharpe, 2015}). Significance level was set at 5 %. The job category (operators, assistant operators and assistant helpers) was related to safety aspects, risks and accidents at work (place of accident, body part affected by injuries or accident, material agent and causal agent). Data were analyzed using the SPSS statistical package version 19 (^{IBM Corp., 2010}).

Worker profile in sawmills

The average sawmill worker in the region of El Salto, Durango is 35 years old, has three years of experience in the job, and four and a half years of service in the company (Table 1). As the job position is more hierarchical, the worker's age and years of experience are higher.

Work experience can offer a biased view of skills, since time spend in the same job position, no matter the qualification required for the job, generates a presumption of training that is verifiable in the short term if it is exercised within the company, but difficult to assess outside the company (^{Aguilar del Castillo, 2016}). Also, subcontracting or temporary hiring of workers by sawmill management generates greater labor turnover, making it difficult to a large extent to qualify them (^{Bardomás & Blanco, 2018}). The above explains that found in this study where not only greater work experience is considered to perform a labor function, but it is also associated with greater age of the worker in positions of greater responsibility. Low work experience recorded is perhaps due to higher labor turnover in sawmills in the region of El Salto, Durango.

^{Pimenta Parente, Scherer, Zimmermann, and Fonseca (2009}) mention that the higher the number of years studied, the better the performance of people in different neuropsychological tasks that require thinking, perception and comprehension. ^{Schettino et al. (2020}) point out that the lower the schooling, the more difficult it is to perceive labor problems in the workplace and to adapt to technological innovations, therefore schooling is an important socioeconomic variable for cognitive development of workers. In this study, 2 % of workers represent the no schooling segment (Table 2), suggesting that most workers understand, comprehend and interpret instructions in work activity and are aware of dangers and safety associated with their work (^{Mitchual, Donkoh, & Bih, 2015a}).

Safety training and occupational hazards

Although only 32 % of the workers received training on safety at work, 71 % reported that they know how to act in an accident at work; 56 % consider their work area to be safe. The relationship between job category and perception of safety (P < 0.05) stands out, where the higher the hierarchy, the higher the perception of safety; in addition, it is interesting to note that the higher the job position, the more interesting the tendency of the worker to be trained and to act in an accident at work (Table 3). This can be explained because workers with more years of service have a higher hierarchy and have probably witnessed more work-related accidents, so their behavior should be better when facing this type of incidents.

^{Kwame, Kusi, and Lawer (2014}) indicate that workers in the sawmill industry usually acquire their skills and experience over the years, but do not have the safety perspective to carry out their work. In this sense, ^{Pucci, Nión, and Ciapessoni (2013}) mention that the notion of risk is a social construct that brings into play a multiplicity of interests and representations; damage assessment depends on what is represented as acceptable danger and risk thresholds. On the other hand, risk management refers to the management of uncertainty through a process of organizational learning, given that workers who do not have established models of behavior to follow, must construct mechanisms and attitudes on the fly to face these situations. This reasoning explains the fact that workers who have no training on safety at work say they know how to act or react in an accident at work.

Accidents causing injuries to workers

In the last five years, 95 % of the sawmill workers in the region of El Salto, Durango have experienced one to five occupational accidents that resulted in injuries of varying severity; of these, 33 % required medical leave with work incapacity from two days to four months. The relationship between job position and worker incapacity is significant (P < 0.05); the higher the hierarchy, the higher the number of medical leaves due to incapacity, which implies that injuries to machine operators are more serious than those suffered by helpers (Table 4).

^{Alcántara de Cerqueira and de Freitas (2013}) believe that the ability to work is the result of a dynamic process between the individual's resources in relation to his or her work, undergoing changes due to various factors, including age. In this sense, chronological aging is considered one of the determining factors of functional aging; therefore, the older the age group, the greater the possibility of losing the ability to work and suffering an accident.

Regarding the place where accidents occurred, most of the correspond to production areas (96 %), where the interaction people-materials-machinery is present and the greatest possibility of finding cutting, rolling, transmission mechanisms, high noise and dust levels due to the proximity of saws (^{Silva-Lugo et al., 2020}).

^{Bello and Mijinyawa (2010}) mention that, in sawmills in southwestern Nigeria, 90 % of accidents with injuries occurred in yards and production areas. On the other hand, in sawmills in Kenya, ^{Ogoti-Mong’are, Mburu, and Kiiyukia (2017}) reported that 86 % of accidents and injuries also occurred in the same areas. In both cases, injuries were because of logs rolling into the sawmill, during sawmilling, equipment maintenance, and lumber stacking. This is consistent with the results of this study regarding the site with the highest occurrence of injuries reported by forestry workers.

Types and effects of worker injuries

In sawmills, workers suffer injuries mainly caused by hits, contusions or crushing (56.7 %) from handling logs and sawn timber; open wounds (15.7 %) caused by contact with sharp or rotating edges; and sprains, strains and muscle tears (14.7 %) from false movements. Serious traumatic accidents such as fractures, amputations and eye loss represent 3.6 %.

The job category is related to the type of injury suffered (P < 0.05); the frequency of hits and crushing is higher among helpers, because they handle materials more often than equipment operators; while the incidence of open wounds is higher in operators because of their proximity to saws and, consequently, they have a higher risk of amputation. Fingers (35 %) and hands (20.3 %) are the most affected body parts (Table 5).

^{Bamidele et al. (2011}) found that, in sawmills in southwestern Nigeria, approximately one-third of injuries occur on the hands of workers who ever operate a cutting machine. ^{Kwame et al. (2014}) reported that the most common accidents in sawmills in Tamale Metropolis, Ghana, included cutting injuries (36.67 %), fractures (21.67 %), and sprains (8.33 %). In the present study, hand injuries accounted for 20 %; however, considering only finger injuries, the incidence is 35 % with a higher prevalence in equipment operators, which is significantly consistent with that found in Nigerian and Ghanaian sawmills.

The agents leading to accidents were material handling (53.7 %), followed by machinery and equipment operation (32 %), tool handling (10.3 %) and work environment (4 %). The mechanisms that caused accidents and injuries correspond mainly to getting stuck by moving objects (30 %), getting hit by moving objects (23.3 %), falling objects (14.3 %) and false movements (13.7 %) (Table 5). There is a relationship between jobs and the causal agent and mechanism of the injury (P < 0.05) indicating that injuries depend on the degree of contact that the worker has with a causal agent in his/her place of work, and both the mechanism that led to the accident and the type of injury will be manifested to that degree (Table 6).

^{Jones and Kumar (2004}) analyzed a database on injuries to workers in the sawmill industry in Alberta, Canada, for the period 1997-2002, and found that the greatest number of injuries resulted from getting hit by stationary objects and falling objects (30.4 %), bodily overexertion when lifting or pushing objects (27.9 %) and getting stuck by fixed and moving objects (16.4 %). The differences in the mechanism that led to injuries in this study with the reference studies may be due to the higher technological level of Canadian sawmills, where possibly aspects of protection of moving parts and equipment maintenance lead to a drastic reduction in accidents due to getting stuck by moving objects. Finally, ^{Alcántara de Cerqueira and de Freitas (2013}) mention that most sawmill workers focus their fears on two high-risk factors as generators of injury mechanisms: band saw breakage and getting hit by logs or sawn timber.