Study design, Study Area and Period
An institutional-based cross-sectional study was conducted in Addis Ababa city construction sites governed under the Addis Ababa city administration construction office from March to April 2021. Of eight active construction sites in the city, four of them namely Akakie zonal stadium, Lideta secretariat, Megenagna TMC, Arada police Arada sub-city and Nefassilk secretariat were the selected for the study.
Study populations
All construction workers who were directly involved in the construction work at selected organizations were included in this study. Workers who have no direct involvement in construction work like administrative staff, supportive staff, and those workers who were absent from work for any reason during the time of data collection were excluded from this study.
Sample size, sampling procedure, and study variables
The sample size was calculated using the single population proportion formula by considering the prevalence(p) of occupational injury 38.3% from a previous study in Addis Ababa [1], 95% confidence level, and a 5% margin of error. Since the source of the population was 537which is less than 10,000, a correction formula was used. By taking the non-response rate to be 10%, the final sample size obtained was 239. Then the number of study participants to be sampled from each construction site was determined using the proportion to size allocation formula, Akakie zonal stadium (80), Lideta secretariat (46), Arada police (33), Megenagna TMC (42), Nefassilk secretariat (38) and simple random sampling technique was used to select sampled construction workers from each construction site.
The dependent variables of this study were occupational injuries whereas Socio-demographic factors (Sex, Age, Marital status, Educational status, working experience, and employment condition), Working Environmental factors (Availability of PPE, Health and safety training, health and safety supervision, and Hours worked per week or day) and behavioral factors (smoking, chewing khat, alcohol, job satisfaction and Use of PPE) were independent variables.
Operational definitions
Occupational injury: Any physical injury resulting from an accident in the course of construction work that is reported by the respondent in the past one year.
Construction worker: worker employed in the manual labor of the physical construction of the built environment and infrastructure.
Health and safety training: Short-term training given on health and safety to construction workers.
Personal protective equipment: Specialized clothing or equipment (such as goggles, gloves, earplugs, masks, helmets, face shields, boots, protective clothing) worn by employees for protection against health-related and safety hazards at the time of working hours.
Data collection instruments
Data were collected by using pretested and structured questionnaires via face-to-face interviews that were adapted from the previously studied literature with some modifications [13]. The questionnaire was translated to the local language (Amharic) in written form and then back to the English version after data collection for its analysis and processing. A questionnaire used for data collection has five parts. Part one was demographic characteristics including, gender, age, marital status, educational status, employment, working hours, and work experience. Part two includes the utilization status of personal protective equipment(PPE) which has 8 questions focused on assessing PPE use at the worksite, availability of it, type of PPE use, and training related to PPE. Part three was about occupational injury characteristics and contained 12 questions, part four was about working environment-related factors which contains 11 questions and part five was about occupational workers’ behavior and characteristics (Additional file 1).
Data collection process
Firstly, we made contact with the Addis Ababa city administration construction office to get permission with an approved ethical clearance letter obtained from the Addis Ababa University Department of emergency medicine Ethical Review Committee to conduct the study. The office accepted our request. Then, we went to each selected construction site with a permission letter and asked the managers for their cooperation to give the construction workers’ lists in their unit. Workers who had no direct involvement in construction like administrative staff and support staff, and workers absent for any reason during data collection were excluded from the study. Workers from each site who fulfill the inclusion criteria were included in the study.
Three trained BSc degree nurses were responsible for the recruitment of participants and data collection and one MSc nurse for supervision. Both data collectors and supervisors were trained for one day about the study instrument and data collection procedures. Construction workers were randomly selected from their list using the lottery method until the required number of samples was obtained. Data collectors informed workers about the study verbally and distributed the participant consent form and information sheet to those who agreed and voluntarily participated in the study. Then data collectors interviewed the participants during working hours of the day.
Participants who were not volunteer for the study were permitted not to participate. The estimated time taken to complete the questionnaire was about 40 min.
Data quality management
Semi-structured Questionnaires, which fit with the context, were prepared using an expert in English. The training was given both to data collectors and supervisors. To identify potential problem areas and unanticipated interpretations, the interview questionnaires were pretested on 12 respondents at lideta police construction site two weeks before the actual data collection and based on the pretest results, the questionnaires were adjusted contextually. The validity and reliability of the instrument were evaluated by experts in the field.
Data analysis
Responses in each question were coded for simplicity of data entry. The coded data were entered into Epi data 4.2.0 and exported to SPSS version 26.0 statistical software for data analysis. In the first step, the descriptive analysis like; percentages, frequency distribution, and measures of central tendency were computed. Both bivariate and multivariate logistic regression models were computed to see the association between independent versus dependent variables. Then factors with a p-value < 0.2 in bivariate analysis were entered into multivariate logistic regression models to control the effect of confounding factors and the p < 0.05 cutoff point was considered statistically significant for all the independent variables. Then the result was presented with text, graphs, figures, and tables.
Study guidelines- the study was performed in accordance with the relevant institutional guidelines and regulations[14]