In this community-based study, the first of its kind in Bangladesh, the majority of small metal workshop workers had experienced at least one injury in a one-year timeframe. We found that laceration, musculoskeletal disorders (MSD) and eye problems were among the most cited injuries. Our study has also revealed the significant gaps in OHS-related knowledge, awareness, and safety practices among metal workers in Bangladesh.

High annual rate of injuries and their causes and consequences

This study found a very high annual rate (~ 83%) of occupational injuries among metal workers. This finding is similar to other studies conducted in Bangladesh (prevalence of musculoskeletal symptoms among metal workers during 1 year preceding the survey is 85%), India (prevalence of occupational morbidity is 60% among iron/steel workers and annual injury rate is 100% among welders), and Uganda (prevalence of occupational injuries among welders is ~ 88% within 1 year prior to the study) [12, 15, 18, 19]. As compared to construction workers, the rate of injury is higher among workers in our study [20]. Lacerations, MSDs, and eye problems were among the most cited injuries in this study.

The incidence of injuries was not associated with sociodemographic factors including age, education, occupation, marital status, income, or knowledge score in our study. In the case of other occupational injuries (among construction workers), age and education were found to be significantly associated with the history of injury in Bangladesh [20]. As most of the workers in our study were the owner of the factories, the age of the factory was possibly equivalent to working experience or service year. In our study, workers with 10 years or above of working experience were more likely to encounter injuries during 1 year preceding the study. This is in line with the findings from the construction workers of Ethiopia- the likelihood of injury occurrences among those who worked for more than 5 years was higher than among those who worked for a few years [21]. We found that the risk of injury was higher in factories with more than three workers, suggesting that it might be difficult to maintain occupational safety in large factories as there is a lack of proper safety practices and monitoring by the local authority. A similar finding is reported from iron, steel and metal manufacturing industries in Ethiopia where occupational injuries were more common in medium or large size industries than the small size industries [22].

As a consequence of widespread injuries irrespective of sociodemographic factors, a large proportion (~ 70%) of metal shop workers lost several days (median 20 days) of work and suffered financially. Financial suffering can affect workers in the lower-income category the most. The nonexistence of occupational health insurance in Bangladesh could make the low-paid victim families vulnerable to bankruptcy.

Basic knowledge of OHS is not reflected in practice

We found that almost all workers had correct basic knowledge about OHS and a large proportion of workers reported that their workplaces were safe and healthy. However, the average practice score was not satisfactory among respondent workers. In other words, awareness did not lead to actions. This is evident by the fact that most workshops did not have training programs, safety guidelines, first aid boxes, and regular risk assessment programs. As expected, a higher level of education was associated with a higher practice score since education can of course raise awareness. Interestingly, higher family income was associated with a lower practice score might be because higher incomes are received by highly experienced workers who are confident in avoiding injuries. An in-depth qualitative study is warranted to deeply understand the gap between the knowledge and practice among metal workers in Bangladesh.

We documented that the owners of the shops had higher practice scores as compared to the other employees. This could be comparable to some extent to a study conducted among particleboard workers in Ethiopia where permanent workers had higher knowledge scores as compared to temporary workers. Consequently, permanent workers had better workplace practices [16].

Unsatisfactory PPE practices

Even though a higher proportion (~ 76%) of workers reported using eye goggles, more than half of them suffered from eye problems. The reason for eye goggles not being able to protect the eyes could be the low quality of eye goggles or using the alternatives of eye goggles such as sunglasses without differentiating among them. For instance, a high proportion of welders in Nepal were found to use sunglasses (not recommended PPE) as these are cheap and comfortable [23]. Most workshops (80.49%) in our study did not have machines to move objects and therefore manual handling could be linked with a higher incidence of MSD in our study.

Specific information on the type of PPE (such as sandals or closed-toe shoes or work boots, clear or polarized glasses or goggles or face shields, and cloth masks or industrial masks) would have been more informative. This was beyond our survey scope because we wanted to keep the questionnaire as simple as possible for the workers with a low level of education. However, from the researchers’ experience at the field level, masks were typically loose-fitting cloth masks, eye wears were mostly clear goggles or face shields. Workers typically wore sandals instead of any shoes let alone safety boots [24]. Given the socioeconomic context of the country, wearing industry-grade masks and any sort of hearing protection are uncommon in the metal workshops at the community level in Bangladesh.

Non-existence of safety training programs and monitoring by the local authority

We found that almost all respondents (nearly 70% of metal shop workers) were unaware of OHS-related governmental rules and regulations. Most shops did not have written OHS policies or procedures. Importantly, programs for safety training, risk assessment, and awareness were nonexistent in most shops that participated in this study. Proper institutional policy and practice, such as the use of PPE and having institutional training, are associated with a lower risk of injuries, according to studies in India and Iran [15, 25]. The scenario is different in developed countries like Australia where the majority (87%) of employers in the manufacturing sector reported that they provided some health and safety training for each worker in the past 12 months [26].

Lack of implementation of existing national policy

The Government of Bangladesh adopted the “National Occupational Health & Safety (OHS) Policy 2013”, the first of its kind in the country [27]. The policy covers all industries in the formal-informal sector of Bangladesh including factories, establishments, trade and commerce, the agriculture sector, and all other workplaces to develop the condition of OHS everywhere. Overall, the policy aims to reduce the deaths, injuries, and diseases due to workplace hazards so that the workers’ productivity increases. The obligations of the policy include to pre-inform all workplace employees about potential accidents, health and safety risks, providing basic training for workers, using appropriate technology, and infrastructural development to ensure safety. However, our study findings indicate that the implementation of the National OHS Policy 2013 is not widespread. Owners’ associations are responsible to ensure medical care and compensation for accidents and rehabilitation of injured workers, according to the policy. Employers are also obliged to formulate their own specific OHS policy for their workplace and follow that. These are not practised in the informal, unregistered, and community-based participant workshops of our study as a consequence of negligible monitoring of social and workplace compliance. Insurance policies and social safety nets are not present to compensate for occupational health hazards although the National Plan of Action on OHS 2021–30 of Bangladesh includes the introduction of an employment injury insurance scheme by 2026.

The OHS practices in developed and western countries, for example in Australia, show a positive scenario on the other hand. There are some factors that contribute to OHS practices and performances in developed countries including legislative and regulatory framework, innovative OHS initiatives, increased awareness and Government support. All businesses in Australia must comply Work Health and Safety Act 2011 and Work Health and Safety Regulations 2011, whereas every business must have a policy for managing OHS. SME owners and managers recognise the penalties they face for failure to maintain healthy and safe working environments in Australia. They are also aware of the potential costs of death and injury to both their employees and their organisations. In large businesses, formal OHS management teams are engaged to help provide the necessary compliance with regulations to ensure the health and safety of the workforce [28].

Policy implications of our findings

Bangladesh has formulated its first OHS profile in 2015 presenting the OHS-related policy landscape and regulatory frames; OHS-related institutions and collaborative mechanisms; state of OHS inspection; and occupational diseases, hazards, and risks for workers. The OHS profile 2015 reemphasizes the use of PPE, the necessity of training, and the formulation of a health and safety committee. Although the informal sector accommodates a large number (85%) of employees in Bangladesh [5], there are no major attempts to improve workplace safety in the informal sector. Our findings can contribute as a supplementary document to the National OHS profile by providing important facts on the informal metal sector. Proper workplace policy is needed to ensure good workplace practice including workers’ safety training, risk assessments, documentation, and availability of first aid, PPE and necessary equipment. Safety net coverage can be introduced to financially harmed and injured workers. Some of these protocols can only be seen in Readymade Garment (RMG) sectors as per international stakeholders’ demand. A complete framework needs to be developed focusing on small industries and workshops to minimize occupational health hazards and ensure social safety nets for the ignored metal workers. Age and size of factories were identified as significant factors associated with occupational injury. Therefore, necessary provisions in the OHS policy are needed to maintain safety at large and old factories. A cost-benefit analysis in the shipbuilding industry of Bangladesh showed that investments in the workplace and environmental safety (new clinic, PPE, training) decrease injuries and increase efficiency [29]. Such analysis should be done in the metal industry too in the near future. The National Plan of Action on OHS 2021–30 suggested OHS coverage to SMEs and informal economy workplaces, in which we are still lagging. Furthermore, in the context of Bangladesh, some simple and inexpensive ways should be promoted to improve workers’ health and safety as well as the health and wellness of their families. These include proper housekeeping, sanitation, and hand washing.

Limitations of the study

This study has a few limitations. First, cross-sectional study design and recall bias are the main limitations. Second, our study findings may not be generalizable to all community settings in Bangladesh. However, the socio-demographic factors of this study area are similar to the other 33 district towns in Bangladesh [13]. Third, as most of the respondents are the owner and paid permanent employees, it might be difficult to generalize the results for temporary or vulnerable employees. This is important because they are generally less skilled or less trained having low knowledge scores [16], which means that they are more likely to get injured in the workplace. Fourth, the knowledge-related questions only covered very fundamental issues. Knowing the attitude of the workers besides knowledge and practice is also important to formulate policies but the attitude of the workers is not covered in this study. Fifth, we fully acknowledge that, by nature, there is no classic referent group for comparison, as might be possible with resources for a larger study design.