Nonfatal Occupational Injuries among Workers in Microscale and Small-Scale Woodworking Enterprise in Addis Ababa, Ethiopia

Background Microscale and small-scale industries have been widely expanded in low-income countries, including Ethiopia, as a job opportunity for young workers, which makes workers vulnerable to injury. Woodworking is one of the high-risk jobs in this sector due to the use of hazardous tools and machineries. Therefore, the aim of this study was to estimate the prevalence of injury and associated contributing factors of this sector. Methods A cross-sectional study design was conducted among 634 workers selected from 194 microscale and small-scale woodworking enterprises in Addis Ababa. Data were collected using a structured interview questionnaire and observation checklist from February to March 2016. Occupational injuries were documented according to the ILO operational definition. Descriptive statistics and multivariable analyses were used to characterize the data and to identify the factors associated with injury at a p value <0.05, respectively. Result A total of 625 (98.6%) workers were interviewed. The prevalence of occupational injury was 92 (14.7%) in the past 12 months. Workers with khat chewing behavior (AOR: 2.25, 95% CI (1.04, 4.85)), job dissatisfaction (AOR: 2.89, 95% CI (1.75, 4.76)), work-related stress (AOR: 4.79, 95% CI (1.69–13.58)), job categories (AOR: 3.52, 95% CI (1.08, 11.41)) and workplace characteristics such as unguarded machines (AOR: 3.32 (1.21, 9.11)), and inadequate work space (AOR: 3.85 (1.14, 13.04)) were significantly associated with occupational injury. Conclusion The prevalence of nonfatal occupational injuries among workers in this study was substantially high. Workers' behavior, psychosocial issues, and work-related characteristics played a causal role in the occurrence of occupational injury. Therefore, workers' safety protection and behavioral intervention should be initiated.


Introduction
Microscale and small-scale enterprises (MSEs) are expanding strategically in low-income countries as a means of job opportunity for young workers [1]. ese enterprises are expected to increase from 14.6% in 2015 to 22.8% by the end of 2020 in Ethiopia [2]. e Ethiopian government has given special attention to the expansion of MSEs due to the additional benefits for industrialization [2][3][4][5]. According to the Ethiopian MSE Development Strategy, individuals not completing formal education, vocational training graduates, and less-skilled citizens are advised to form cooperatives of MSE with the goal of creating about 4 million new jobs in the Ethiopian Growth and Transformation period [2,4]. e MSEs are classified into microscale and small-scale enterprises based on their number of employees, being less than 6 and 6-30, respectively [6]. Micro and small woodworking enterprises are categorized under the manufacturing sector. ough the MSE are considered as an informal economy, these business units employ more than 5.5 million people with the majority being youth [5][6][7][8].
e machineries used and the working conditions in the woodworking activities are very dangerous and may lead to serious occupational injuries [9]. e presence of rotating devices, unguarded cutting, or shearing blades is the sources of injury in wood workshops [10]. Since the sector is considered as an informal economy, the rights of workers, labor inspection services, access to occupational health services, and injury recording or notification were significantly compromised [5]. e prevention of occupational injury needs appropriate monitoring, regulation, capacity, and commitment of workplace administrators [11,12]. Several studies from low-income settings reported higher prevalence of occupational injuries in the MSE manufacturing sector. Occupational injury prevalence studies in Ethiopia [13][14][15], Ghana [12,16], and Zimbabwe [10] revealed that workers are highly vulnerable to injury in this sector. Studies also showed that work-related conditions [13,14,16] such as contact with machinery [16], work-related stress [10], job dissatisfaction [14], and inadequate work space [17] were significantly associated with occupational injury. In addition, workers' behavioral characteristics such as chewing khat [14], smoking [14,16] and lack of personal protective equipment use [14,15] were also showed significant association with injuries.
Most of the above studies were conducted in the general MSE manufacturing sectors but little is known specifically in the woodworking sector. Hence, evidence for planning intervention activities in the micro and small woodworking has been historically limited. erefore, the aim of this study was to estimate the prevalence of occupational injuries and the associated factors among workers in the microscale and small-scale woodworking enterprises.

Materials and Methods
Institution-based cross-sectional study design was used which lasted from February 17 to March 14, 2016, among workers of microscale and small-scale woodworking enterprises in Addis Ababa. It is the capital city of Ethiopia, which has 10 subcities (boroughs). Addis Ababa was selected due to the presence of a large number of woodworking enterprises and the engagement of work force compared to other towns in Ethiopia [8,18]. A total of 3,339 and 613 microscale and small-scale woodworking enterprises were found in Ethiopia and Addis Ababa, respectively [19].
ere is no previous study in Ethiopia on the prevalence of nonfatal occupational injuries among workers in microscale and small-scale woodworking enterprises. Hence, prevalence of injury of 50% was used to estimate the sample. A sample size of 634 workers was calculated using a singlepopulation proportion formula with the assumption of 50% prevalence, 5% margin of error, 95% confidence level, 1.5 design effect, and 10% added for nonresponse. e sampling procedure was done through two steps. First five subcities out of the 10 subcities were selected using a lottery method. en, the sample size was allocated to microscale and small-scale woodworking enterprises using proportional allocations to the size to determine the numbers of study subjects required from each selected subcity. A sampling frame (list) of enterprises was taken from the Addis Ababa City Administration Bureau of Trade's central database [19]. Subsequently, the enterprises were selected randomly with Excel software until the required sample size was obtained. A total of 634 workers who were directly engaged in the woodworking production process (423 from micro and 211 from small enterprises) were selected to participate in the study. Workers who were absent from work for any reason during the time of data collection were excluded from the study.
A pretested and structured questionnaire was used to collect information from workers, and an observation checklist was used to collect the work environment data. e interview questionnaire and observation checklist were developed based on published literature and previous studies of similar setting [13,[20][21][22][23][24][25]. e questionnaire included assessment tools of the Marlin Company and the American Institute of Stress Scale [22], the Generic Job Satisfaction Scale [23], and ILO operational definition for occupational injury [21]. Both the Stress Scale and Job Satisfaction Scale were validated in previous studies and translated into Amharic, the local language [20]. In the present study, nonfatal occupational injuries were an outcome variable. e injury types were classified based on the International Statistical Classification of Diseases and Related Health Problems [26]. Workers were interviewed about their sociodemographic, behavioral, and work-related characteristics: sex, age, work experience, salary, education (illiterate, read and write, primary, secondary, college, and above), sleep hours per day, khat chewing (yes/no), cigarette smoking (yes/no), alcohol drinking (yes/no), psychosocial factors (work-related stress and job satisfaction), type of employment (permanent/temporary), job category (painter, machine operator and carpenter), work hour per week, night work (yes/no), use of personal protective equipment (yes/ no), and training on health and safety (yes/no). In addition, the observation checklist was used to examine the adequacy of working space (yes/no), the presence of an obstacle on the floor (yes/no), the adequacy of lighting (yes/no), the presence of machine guarding (yes/no), the presence of reachable emergency power cutoff switch (yes/no), the presence of functional fire extinguisher (yes/no), and availability of personal protective equipment (PPE) (yes/no).
One public health practitioner served as a supervisor, and three occupational health and safety professionals participated as data collectors. Two days of training were given for the data collectors concerning the study objective, tools, approach, and ethical issues. A pretest was done among 5% of the sample size in other subcities to check the consistency. Interviews were conducted in a private space or room to maintain the privacy of the respondent and to avoid disturbance. Observation data were collected with a checklist on the same day after interviewing the workers. Data collection activities were carried out with close follow-up by the principal investigators and the supervisor. e data were entered into an electronic template of Epinfo version 7.1 and analyzed using SPSS version 20. e data from the interview and observation were analyzed independently.
ere were many independent variables from the interview; hence, variables with a p value below 0.3 with the outcome variable in the bivariate analysis were entered into the multivariable logistic regression model [20,27,28]. e required assumptions of the logistic regression were checked with Hosmer and Lemeshow fit test statistics. Odds ratio (OR) with 95% CI was used to declare the presence and the strength of association in the multivariable analysis.
e study protocol was reviewed and approved by the Institutional Review Board of Addis Ababa University Research and Ethical Committee of the School of Public Health. Permission was granted by each enterprise owner before approaching workers. Informed written consent was obtained from each participant for the study and dissemination of the results. Participation in this study was voluntary. Confidentiality was ensured for the collected information through the use of coding.

Operational Definitions and Measurement
Self-reported occupational injuries include injuries that occurred due to incidents at the workplace during the past 12 months; and those injured workers should have been away from work at least one day because of the injury, in addition to the day of the incident [21,24]. Alcohol drinking is consumption of any alcohol more than twice a week [13]. Khat (pronounced "cot") is a stimulant drug derived from a shrub (Catha edulis) that is native to East Africa and southern Arabia. Leaves of the khat shrub are typically chewed and held in the cheek, like chewing tobacco, to release their stimulant chemicals [29]. Khat chewing refers to chewing khat leaves before or while working more than twice a week.
Smoking is referred to smoking at least one stick of tobacco cigarette each day. Workplace stress status is a result with a total score of 26 or more for work stress presence and less than 26 for no work stress [22]. Job satisfaction status is a result with total score 32-50 as satisfied with the job and less than 32 as no job satisfaction [23]. Adequate work space is the working area with enough clear space (walkways and around cupboards, storage, or doors) to allow physical actions needed to perform the task [30]. Obstacle-free floor include floors maintained to be free of slip and trip hazards (trailing cables, uneven edges or broken surfaces) and walkways clear and free from tools and materials [30]. Adequate light includes lighting is sufficient to allow workers to see machinery movements, controls, displays, and to move about easily and free of flicker and glare [30,31]. Guard is a part of machinery specifically designed to provide protection by means of a physical barrier [32]. Personal protective equipment (PPE) includes goggles, helmet, and face shield and gloves, boots, and specialized clothing that is designed to protect parts of the body including the eyes, face, hands, figure, and feet [9].

Sociodemographic Characteristics of Respondents.
A total of 625 respondents were interviewed from 162 microscale and 32 small-scale woodworking enterprises (416 and 209 workers from microscale and small-scale workers, respectively) with a response rate of 98.6%. e majority of the participants were male (94.6%), age less than 30 (75.5%), and with a work experience of less than 5 years (64.2%) ( Table 1).

Behavioral and Psychosocial Characteristics of
Respondents. Of the total respondents, 8.6% drank alcohol, 8.8% chewed khat, and 4.3% smoked cigarettes. e overall use of PPE was low (20.8%). Out of the total PPE users, the majority (82.4%) used only goggles. e reasons for not using were identified as lack of availability (75.8%), discomfort to use (21.2%), and the use of PPE limiting work performance (3.0%). Regarding psychosocial characteristics, 25.9% and 3% of the participants reported being dissatisfied with their job and had work-related stress, respectively.

Prevalence of Nonfatal Occupational Injuries.
e prevalence of nonfatal occupational injury among the 625 participants was 14.7% during the past 12 months. is prevalence among microscale and small-scale enterprises was similar (14.9% and 14.4%, respectively). About half (52%) of the injuries resulted in the workers being absent from work for more than 3 days. Cut or laceration were the most prevalent type of injury (63%), and fingers were the most affected body parts (68.5%).
Most injures happened due to contact to exposed machinery (68.5%) while cutting, flattening, and smoothing timbers. e highest number of injuries was reported on Mondays (33.7%) and in the morning (51.1%) ( Table 3).

Factors Associated with Nonfatal Occupational Injury.
Each variable was analyzed using bivariate logistic regression, and variables with a p value of less than 0.3 were fitted to the multivariable logistic regression. Variables including work experience, monthly salary, educational status, sleeping hours per day, employment pattern, working hour per week, night work, health and safety training, and type of enterprise were not fitted (p < 0.3). In the multivariable analysis, individual characteristics such as khat chewing, job satisfaction, and work-related stress and job category were significantly associated with the occurrence of nonfatal occupational injury at p < 0.05 (Table 4).
Similarly, variables of the work environment characteristics such as unguarded machines and inadequate work space were significantly associated with the occurrence of occupational injury (Table 5).

Discussion
e prevalence of nonfatal occupational injuries in this study was 14.7% (147 per 1000 workers). Previous studies in Ethiopia had reported higher injury prevalence ranging from 452 to 808 per 1000 exposed workers per year in the microscale and small-scale enterprises [13][14][15]. It could be difficult to compare our study prevalence estimate with the above studies due to the use of different case definitions to   count an injury. e use of a definition "at least one day absence from work after an injury" in this study might lower the prevalence in the current study. An injury definition in the workplace can significantly affect the prevalence estimate [33]. To avoid overreporting of minor injuries like scratches and abrasions, most countries used a reportable injury definition as "which can leave the victim absent from work for three and more days" [34]. However, such definition might be difficult to use in an informal sector where the majority of the workers were temporary employees and had no insurance access. is study indicated that the upper limbs, including hands and fingers were the most commonly affected parts of the body. Several studies also reported similar findings among the manufacturing workers [13,14,35,36]. Most of the tasks in woodworking were performed by the use of machinery and hand tools. Hence, hands are frequently in contact with sharp tools which increase the risk of injury. e majority of the injuries reported in the current study were one-time occurrences. It is similar to other occupational injury studies [20,36,37]. is could be due to workers who might learn a preventive behavior after experiencing injury [38] or after major injury, workers may be afraid to go back to work again. at is why, the ILO recommended injury recording, notification, and injury surveillance system at the workplace [39].
Contact with machinery was found to be the main cause of injury in this study. A significant percentage of the machines had unguarded parts, and workers used defective hand tools in the current study. Similar results have been reported in other studies at small-and medium-scale industries [36,37]. Injuries due to table saw use were found to be the most common as compared to other types of machineries among injured participants. is could be because of the frequent use of these machines in the woodworking enterprises. Hence, the probability of getting injured might be increased while operating these machines without appropriate safety measures [40,41]. e majority of MSE had inadequate working space and no guards on hazardous parts of machinery which showed significant association with Journal of Environmental and Public Health 5 injury inccident in the study. is finding is consistent with a study in Malaysia [17]. Inadequate work space could limit worker movement and increase involuntary contact of workers with unguarded machines or hazardous objects. e injuries could occur while the workers became close to the rotating motion of tools [42]. e current study revealed that individual and behavioral characteristics are associated with occupational injury. e probability of injury occurrence among khat chewers, machine operators, and carpenters was significantly high. Other studies in Ethiopia have showed an increased risk of injury among khat chewers [14,43]. Khat is classified as addictive substance which stimulates and increases excitement that later can affect the central nervous system [29,44]. Studies of cognitive tests demonstrated that khat chewers perform worse as compared to nonusers [45] and the levels of anxiety and depression were higher among chewers than nonchewers [46]. ese health effects might impair worker concentration and increase the likelihood of mistakes and occurrence of accidents at work.
Psychosocial characteristics such as work-related stress and job dissatisfaction of workers were significantly associated with injury in the present study. Workers who were not satisfied with their job had an injury risk nearly three times higher than workers who reported to be satisfied with their job. is finding was similar to a study conducted on small-scale and medium-scale industries [14,37,47]. Unsatisfied individuals might increase the likelihood of making mistakes while working in their job which leads to an injury [48]. Similarly, workers who experienced stress in the work place had more than four times risk to injury. A similar finding was reported among textile factory workers in Ethiopia [20]. e nature of woodworking, poor working condition, the lack of appropriate safety equipment, and lack of safety training for protection might be stressful for workers [22,49]. Prior research studies indicated that such stressful condition could alter either the mental or physical nature of workers and impair workers focus which can lead to commit error [49,50]. In addition, the safety behavior of workers could be affected by emotional stress [51].
e use of standard case definition for an injury, high participation rate, the use of standard measurement tools, and involvement of many enterprises were the strengths of this study. However, the findings were based up on selfreporting and injuries were not clinically confirmed. Since MSEs were informal sectors where temporary employment is a common practice, a healthy worker may be affected and severely injured workers may not return to their job.
Generally, the prevalence of nonfatal occupational injuries among workers in MSE woodworking enterprises is substantially high. Contact with unguarded machineries and poor working condition were the major causes of injury  Journal of Environmental and Public Health incident. Behavioral characteristics of khat chewing, psychosocial characteristics, and job category were significantly associated with occupational injury. erefore, workplace safety improvement and behavioral change communication activities should be included in the microscale and smallscale woodworking enterprise development strategy.

Data Availability
Data and other required materials can be submitted if requested.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
HM was involved in the write-up of the research proposal, supervised data collection and data entry, as well as data analyses, and wrote the manuscript. YT and MG were involved in the write-up of the research proposal and the data analyses and writing the manuscript.