Artigos

INTRODUCTION
The growing understanding of the Social Determinants of Health (SDH) has highlighted the influence of socioeconomic factors, such as income, wealth, working conditions, education, among others, in determining a variety of health outcomes, including those related to worker health1–3. The SDH are related to the conditions in which people are born, live, and die, and allow us to explain how health is sensitive to the environment in which one lives and serves as a parameter for the pursuit of social justice4,5.
The theme of the social determination of workplace accidents (WA) has been addressed in various ways in the literature. Several studies have evaluated individual-level risk factors (sociodemographic, occupational, working conditions, health conditions, among others) that are associated with a higher occurrence of WA6–8. In a recent national survey on the health status of the Brazilian population, men, young people, black individuals, those with lower education levels, and rural workers showed higher prevalence of WA8. In another study, although stable, the mortality rate from WA in Brazil remained high compared to high-income countries, with black, mixed-race, indigenous individuals, people with low education levels, and those from the North and Northeast regions showing the highest rates9. In the informal sector, higher chances of WA were identified among men, young people, workers in the meat or poultry sector, and those performing heavy physical labor10.
The cited results demonstrate that there is a relationship between health-disease-WA processes and suggest that some conditions are associated with a higher occurrence of these injuries. However, most research aiming to identify factors related to WA underestimates the structural social determination of these occurrences. This finding was already present in a publication dated 1999, when Lima and his collaborators reported a greater contribution of contextual socioeconomic characteristics in the occurrence of WA compared to individual characteristics of workers11. In view of such factors, for the authors, the prevention of WA necessarily should involve a general improvement in living and working conditions, greater appreciation of workers, and greater investments in training and remuneration11.
In work developed in the Asian continent, it was shown that structural socioeconomic inequalities can influence Occupational Safety and Health and (OSH) outcomes at different levels, revealing that the social development of a territory can improve OSH outcomes. However, accelerated economic growth, without adequate investments in OSH or updated regulations, can also result in increased morbidity and mortality from WA12, a finding that emphasizes the importance of considering the role of structural socioeconomic inequalities of territories in the formulation of public policies in this area. Another study developed in Poland also emphasized the influence of socioeconomic conditions on the number of WA in the construction industry13, reaffirming the association of structural SDH with working conditions.
In the national scenario, although the scientific literature extensively explores the connection of structural SDH in public health, the investigation directed at the association of these factors with worker health outcomes, particularly in relation to WA metrics, remains an underexplored field. Therefore, little is known about the relationship between indicators of socioeconomic inequalities and income distribution and morbidity and mortality resulting from WA in Brazil.
This study aimed to analyze the relationship between indicators of socioeconomic inequalities and mortality from WA among Social Security beneficiaries on a national level.





METHODS
This is an ecological study conducted with secondary data obtained from public access databases made available by the Ministry of Social Security of Brazil.
The unit of analysis was the Brazilian federative units (FU), and the outcomes evaluated were the mortality rates from WA among Social Security beneficiaries in Brazil, by FU. The main independent variables were three indicators of socioeconomic inequalities: Theil-L Index, Gini Index, and the Municipal Human Development Index (MHDI).
The number of deaths from WA and the average number of work-related ties in the reference year were extracted from the Historical WA database of the Statistical Yearbook of WA (AEAT InfoLog)14. The years 2010 to 2022 were considered.
Mortality rates were calculated by FU and expressed per 100,000 work-related ties, as recommended14. This indicator, calculated according to the following equation, measures the relationship between the total number of deaths from WA recorded in the year and the population at risk of accidents14:
Mortality Rate = (Number of deaths from WA among Social Security beneficiaries) / (average annual number of work-related ties) x 100,000.
The use of mortality rates from WA was chosen over incidence rates because underreporting of these events, both fatal and non-fatal, is a significant problem in Brazil15. However, fatal WA are more investigated due to their severity and often involve authorities such as emergency services, coroners, police, and OSH authorities, making underreporting more difficult16.
The Theil-L index, Gini index, and the Human Development Index (HDI), measures used to assess the level of inequality in each state, were obtained from the Atlas of Human Development of the United Nations Development Programme (UNDP)17. This platform provides statistical information on human development, including population, education, housing, health, work, income, and vulnerability across various levels of Brazilian territory. The extracted data refers to the 2010 Demographic Census.
The Theil-L and Gini indices are indicators that measure income distribution inequality in a territory, both expressed from 0 to 1, with higher values indicating worse income distribution (greater inequality). The MHDI evaluates human well-being, being a synthetic measure of the health, education, and longevity conditions of a population; it ranges from 0 to 1, with values closer to 1 indicating better human development18. More information about the indices can be obtained on the UNDP portal17.
The variation in mortality rates from WA among FUs was investigated using descriptive epidemiology: statistical range (A), standard deviation (?), and interquartile range (IQR). The strength of the association between the average mortality due to WA in each FU and Theil-L, Gini, and MHDI indices was measured by Spearman's correlation test, expressed through the correlation coefficient (Spearman's rho). This coefficient measures the strength and direction of the association between two variables and ranges from -1 to 1, with rho close to 1 representing a strong positive association, rho close to -1 indicating a strong negative association, and rho close to 0 suggesting an absence of a linear association. For interpreting the magnitude of the correlations, the following classification was adopted: rho ? 0.3 (weak magnitude correlation), rho > 0.3 to ? 0.5 (moderate magnitude), and rho > 0.5 (strong magnitude)19.
All tests were performed using Jamovi software, version 2.2.5, with a significance level of 5% (p-value < 0.05).
The study used secondary, aggregated data from public bases, with no possibility of identifying subjects, and was therefore not submitted to the CEP/CONEP system, in accordance with CNS Resolution No. 466/2012.

RESULTS
Between 2010 and 2022, 33,111 deaths from WA were reported among beneficiaries of the General Social Security System (RGPS) in Brazil, resulting in an average mortality rate of 6.08 deaths per 100,000 employment ties among Social Security beneficiaries.
The mortality rate varied substantially across the FUs, with the Federal District (3.64) and Paraíba (4.08) showing the lowest average rates of work-related deaths during the period, while Mato Grosso (15.08) and Rondônia (12.94) recorded the highest rates (Figure 1).
INSERT FIGURE 1.
This variation in mortality rates among the territories analyzed was corroborated by descriptive analysis, which revealed an amplitude of 11.44 (ranging from 3.64 to 15.08), an interquartile range of 3.27, and a standard deviation of 2.61 (Figure 2). The variation in WA mortality rates across Brazil's FUs, as well as the independent variables, is represented in Figure 2.
INSERT FIGURE 2.
There was a moderate negative correlation between the WA mortality rate and the Gini Index (rho = -0.458, p = 0.016) and between the same indicator and the Theil-L Index (rho = -0.489, p = 0.010) (Table 1). No statistically significant correlation was found between the mortality rate and the MHDI (p = 0.685).
INSERT TABLE 1.
Both the Gini Index and the Theil-L Index showed significant negative correlations, indicating a downward trend in the associated scatter plots. This visual dispersion between mortality rates by FU, according to the three socioeconomic inequality indicators investigated, can be observed in Figure 3.
INSERT FIGURE 3.

DISCUSSION
This study aimed to determine whether socioeconomic factors of the territory are associated with WA mortality among RGPS beneficiaries in Brazil. The findings revealed the presence of heterogeneity and variation in WA mortality rates among the FUs in the country, consistent with the variation in socioeconomic conditions also observed in the inequality indicators. In particular, a negative correlation was observed between the WA mortality rate and the Gini and Theil-L Indexes, suggesting that in regions with lower economic inequality, the WA mortality rate tends to be higher. This finding indicates that income concentration and socioeconomic disparities in the territories are associated with the WA mortality rate in the country.
The underreporting of WA and fatal WA emerges as a significant reality in the Brazilian context20. Therefore, a plausible hypothesis for the seemingly contradictory association between economic inequality indices and mortality rates from WA may be related to the underreporting of these events in areas characterized by greater socioeconomic disparity. In regions with higher inequality indices, the lack of resources and precarious access to health services may be directly impacting the records of WA, leading to a possible underestimation of these events. Similarly, the scarcity of oversight and control infrastructure in these areas may be contributing to the underreporting of cases, especially when considering the higher risk of fatal occurrences in environments with greater labor informality and, consequently, conditions conducive to inadequate investigation of these events, or even no investigation and/or notification in cases of workers not covered by WA insurance.
Cultural, political, and educational issues deserve emphasis and may be influencing the adoption of preventive practices for WA and perceptions of OSH in different FUs in Brazil20,21. This is because, in many corners of the country, the underreporting of WA reflects the complex power relations that permeate labor relations, highlighting the absence of citizenship and basic rights for a significant portion of the working population. In contexts where oversight is scarce and workers' voices are often silenced, the lack of records is often not merely an administrative issue but an indication of a system that marginalizes those who already live in precarious conditions20-22. Furthermore, public policies aimed at OSH and worker health vary between regions and FUs, affecting the outcomes of preventive measures and oversight in different areas of the country23,24.
The normalization of this underreporting, therefore, may represent a reality in which social protection and labor rights are not guaranteed, perpetuating a cycle of vulnerability and exclusion. Thus, the structural inequalities that characterize these regions may reflect not only a lack of effective public policies but also a culture of disrespect that has taken root, making citizenship a distant ideal for many. It is hypothesized that in territories with greater socioeconomic inequality, there may be less awareness among professionals and companies about the importance of notifying WA to Social Security – which would, to some extent, explain the lower reporting of WA and fatal WA in these regions.
One possible explanation for the observed negative correlation is related to the higher concentration of micro-regions with industrial jobs, where significant industrial agglomerations are primarily located in the Southeast and South regions, and more weakly in the Northeast, North, and Midwest regions25,26. These locations tend to attract investments, resulting in an expansion of job opportunities, thereby increasing the likelihood of WA occurrence. These regions with greater industrial significance tend to house other companies of different natures, some of which may have more hazardous work environments or expose workers to more significant occupational risks.
Mascarenhas et al. also support the above-mentioned viewpoint by highlighting that the highest concentration of occupational accidents in Brazil coincides with the economic development level of the territories. They observed that economically developed areas tend to register a higher number of emergency care cases due to work-related injuries, attributing this to the concentration of industries and maintenance and repair services. These same authors also point out that the high proportion of rural workers, especially in FUs dominated by agribusiness, is another factor that increases the frequency of WA in regions with economies focused on large-scale agricultural production27.
Historically, records of WA tended to reflect more severe cases in industrial sectors, while in less industrialized regions, the lack of reporting was common28. With the growth of the service economy and the precarization of work29,30, these reporting practices may have been further affected. At this point, it is known that about two-thirds of the economically active population is outside the formal labor market, which excludes them from social protection through WA insurance (currently referred to as GILRAT - Contribution of the Incidence Rate of Labor Incapacity Arising from Workplace Risks).
Therefore, the greater formalization of the labor market in some territories may also be contributing to increasing the statistical visibility of fatal WA in some regions. In turn, in regions with a higher prevalence of informal work, there may be greater underreporting, leading to an underestimation of these events and, consequently, a lower magnitude of these occurrences. The increase in informal work in Brazil, rising from 39.8% in 2012 to 43.4% in 2019, may have influenced the results of this research31. This trend often implies a lack of labor rights, precarious living conditions, denial of citizenship principles, and perpetual reproduction of poverty and social inequalities7,32. Informal workers, even when RGPS beneficiaries such as freelancers or self-employed, do not fit the Social Security definition of WA according to Social Security legislation (Article 19 of Law 8.213/1991). There are also cases of "false cooperatives" or "pejotização," which are not linked to accident insurance, falling into a zone of social invisibility. Although some cases may be illegal and subject to correction, most workers in these conditions fall outside the definition of accidents adopted by Social Security.
Likewise, the rising forms of labor precarization – such as cooperativization, pejotização, platformization, among others29,30 – may also have affected the accident indicators presented here, potentially contributing to a scenario of “underreporting” (actually, underestimation) of these events. At this point, some questions need to be raised regarding the assessment of these events: Are the definitions of WA used by Social Security adequate for the current reality of the Brazilian labor market? How have the records evolved in previous years? What recent legislative changes may have encouraged or hindered this practice? All these issues need to be questioned and considered by both academia and public authorities, given the current scenario and the transformations that the job market has undergone/undergoes.
Regarding inspection and control by the competent authorities, it is important to note that labor inspection, carried out by the Labor Inspection, an agency linked to the Ministry of Labor, is unequal across the country, with greater presence in some regions23,32. The higher presence of the Labor Inspection in FUs areas of the Southeast and South regions could, among others, be a possible explanation for the lower underreporting of work accidents in these locations. The relationship between the number of companies and the number of auditors, as highlighted in the specialized literature32-34, suggests that instead of effective inspection, we face a situation of inefficiency and ineffectiveness, where inspection becomes insufficient to guarantee worker safety and adequate documentation of accidents.
In a national survey investigating the mode of operation of labor inspection, it was emphasized that, besides the distance of locations from the inspection center in FUs, the efficiency of the inspection conducted by labor inspectors is also influenced by the density of commercial or industrial establishments and the grouping of geographic micro-regions across the country32. In another study, it was found that labor inspection actions are generally not comprehensive, do not focus on sectors with higher occupational morbidity and mortality rates, and that there is an insufficient number of inspectors primarily dedicated to OSH33. This data is corroborated by Fragoso Junior and Garcia, who point out that there is no adequate prioritization by labor inspection of sectors that fall among those with the highest WA mortality indicators in Brazil34. The performance of inspectors occurs without adequate material and budgetary structure and, at times, without the necessary technical training32,33.
It is also important to highlight that a national study showed a high level of poor quality in the records of the causal relationship of fatal WA in Brazil between 1998 and 2013, reaching more than half (67.7%) of the notifications of deaths from external causes classified as accidents. All regions, except the Southeast, were associated with missing or inconsistent records in a study aimed at identifying factors related to the quality of recording fatal WA in the Mortality System in Brazil35. A possible solution to strengthen inspection and protect workers would be more effective collaboration between Labor Inspection and the Occupational Health Surveillance teams (Visat). By joining efforts, it would be possible to expand inspection coverage and interventions, overcoming the current limitations caused by the shortage of labor inspectors.
Additionally, recent legislations – such as the Accident Prevention Factor (FAP) – have been identified as mechanisms that may encourage the underreporting of these events by companies, especially those whose cases have a greater impact on social security funds36. The reduction in the GILRAT insurance premium for companies, made possible by the FAP, when they report a lower number of accident notifications, is cited as an incentive for underreporting. This occurs because companies seek to reduce the costs associated with these events, avoiding increased taxation36.
Internationally, a study conducted in China also examined the relationship between contextual socioeconomic inequalities and occupational injuries and found that factors such as industrial production value, infrastructure investment, gross domestic product, and social aspects such as medical beds, fiscal expenses, and employment percentage significantly impacted the occurrence of occupational injuries. That study also identified a curvilinear relationship between socioeconomic development and the number of WA deaths, indicating an S-shaped pattern12. In summary, in the Asian scenario, as socioeconomic development increased, there was a tendency for a reduction in the number of WA deaths, reflecting possible improvements resulting from investments in infrastructure, more robust regulations, and possibly greater awareness in the OSH context. However, upon reaching a critical point of development, this relationship stabilized or even reversed in some localities, resulting in an increase in these occurrences12.
The findings of this research corroborate possible differences in work opportunities and conditions, causing workers to be exposed to different WA death risks in the FUs of the country. Our findings are consistent with the study by Menegon et al., which, although differing in method, design, and population adopted here, also highlighted that there are population groups (for example, men, blacks, mixed-race individuals, indigenous people, people with low education) and regions of the country where there is a higher concentration of mortality due to WA9.
It is worth noting that this study did not find a statistically significant correlation between the WA mortality rate and the MHDI. This implies that, at least in this dataset and analysis, WA mortality rates seem to be more related to socioeconomic disparities than to overall human development.
This study adds to the literature information on the relationship between socioeconomic inequality indicators of territories and WA mortality in Brazil and can contribute to understanding the association between socioeconomic factors, structural inequalities, social determinants in worker health, and WA mortality in Brazil. However, one of its main limitations, due to its ecological design, is the possibility of confusion between correlation and causality; although a negative association between mortality from WA and socioeconomic inequality indicators has been observed, this does not necessarily imply that one condition causes the other. Moreover, the data were represented only through aggregated measures, which summarize the characteristics of individuals within a group of interest. This provides only an overview, as an association observed at the aggregate level between variables does not necessarily reflect the relationship that exists at the individual level36. The sample size may also have influenced the results, such that the broad geographic coverage of the FUs may not reflect, for example, differences between municipalities in the country. The cases considered were only those of workers covered by the GILRAT of Social Security. Therefore, studies with a more expanded level of disaggregation (municipalities, macro and micro-regions, health districts, among others) are encouraged to compare with the results presented here, especially regarding the inclusion of a significant portion of informal workers and other categories not covered by Social Security insurance.
It is reiterated that the limitations related to the registration, access, and quality of data, as discussed earlier, may also have resulted in an underestimation of the reality of WA in some regions, especially in areas with higher informal labor markets. Using 2010 data as a reference for the Theil-L, Gini, and MHDI indices may also have provided a limited perspective of the current reality of Brazilian territories. These indicators are not calculated specifically for the formal workforce, so the inequality relationships that emerge from these indicators may not fully and/or directly explain the risks of WA to which RGPS workers are exposed. However, they represent the most updated official survey published by UNDP based on the Census to date and are widely used for the formulation and guidance of national public policies and other studies in the fields of public health, social sciences, and economics, due to their ability to synthesize complex information about inequality and human development.
CONCLUSION
The socioeconomic inequality indicators, specifically the Gini and Theil-L indices, showed significant negative associations with the mortality rate from WA among RGPS beneficiaries in Brazil. This association suggests that in regions with greater inequalities, the mortality rate from WA appears to be lower.
The results identified here are counterintuitive, and the possible reasons behind this seemingly inverse relationship may lie in the underreporting/underestimation of WA, a phenomenon frequently highlighted in Brazilian literature. Among other factors discussed throughout the manuscript, failures in identifying WA cases in regions with greater socioeconomic inequality, a higher degree of informality in the labor market, the expansion of precarious work in the country, lack of oversight infrastructure, and recent changes in legislation may also have contributed to this distortion.
This underscores the importance of improving the accuracy and comprehensiveness of reporting these events for a more reliable understanding of the relationships between socioeconomic inequalities and mortality from labor-related causes at the national level.










REFERENCES
1. World Health Organization. A conceptual framework for action on the social determinants of health. Available from: https://iris.who.int/handle/10665/44489. Accessed January 18, 2025.
2. Galvão ALM, Oliveira E, Germani ACCG, et al. Determinantes estruturais da saúde, raça, gênero e classe social: uma revisão de escopo. Saúde e Soc 2021; 30: e200743.
3. Borde E, Hernández-Álvarez M, Porto MFS, et al. Uma análise crítica da abordagem dos Determinantes Sociais da Saúde a partir da medicina social e saúde coletiva latino-americana. Saúde em Debate 2015; 39: 841–854.
4. Melo FCC, Costa RFR, Corso JMD. A conceptual model for studies on social determinants of health in Brazilian municipalities. Saude e Soc 2020; 29: 1–14.
5. Dantas SVAD, Miranda MG, Dusek PM, et al. Uma avaliação do Programa Bolsa Família. Interações Campo Gd 2018; 19: 713–726.
6. Malta DC, Stopa SR, Silva MMA, et al. Acidentes de trabalho autorreferidos pela população adulta brasileira, segundo dados da Pesquisa Nacional de Saúde, 2013. Cien Saude Colet 2017; 22: 169–178.
7. Takala J, Hämäläinen P, Saarela KL, et al. Global Estimates of the Burden of Injury and Illness at Work in 2012. J Occup Environ Hyg 2014; 11: 326.
8. Malta DC, Bernal RTI, Vasconcelos NM de, et al. Acidentes no deslocamento e no trabalho entre brasileiros ocupados, Pesquisa Nacional de Saúde 2013 e 2019. Rev Bras Epidemiol; 26. Epub ahead of print 2023. DOI: 10.1590/1980-549720230006.supl.1.1.
9. Menegon FA, Kupek E. Mortalidade por acidentes de trabalho no Brasil: análise de tendência temporal, 2006-2015. Rev Bras Saúde Ocup; 46. Epub ahead of print 2021. DOI: 10.1590/2317-6369000025219.
10. Rios MA, Nery AA, Andrade Rios PA, et al. Fatores associados a acidentes de trabalho envolvendo trabalhadores informais do comércio. Cad Saude Publica 2015; 31: 1199–1212.
11. Lima RC, Victora CG, Dall’Agnol M, et al. Associação entre as características individuais e sócio-econômicas e os acidentes do trabalho em Pelotas, Rio Grande do Sul, Brasil. Cad Saude Publica 1999; 15: 569–580.
12. Wei J, Lu L, Zhao D, et al. Estimating the influence of the socio-economic inequalities on counties’ occupational injuries in Central China. Saf Sci 2016; 82: 289–300.
13. Ho?a B, Nowobilski T. Analysis of the Influence of Socio-Economic Factors on Occupational Safety in the Construction Industry. Sustain 2019, Vol 11, Page 4469 2019; 11: 4469.
14. Brazil. Historical database of the Annual Statistical Report on Work Accidents (AEAT InfoLogo). Ministry of Social Security. Available from: http://www3.dataprev.gov.br/aeat/inicio.htm. Accessed January 18, 2025.
15. Hennington ÉA. Cordeiro R. Trabalho, violência e morte: miséria da existência humana. Curitiba: Appris; 2018. Cien Saude Colet 2019; 24: 3595–3595.
16. Galdino Batista A, Sousa Santana V, Ferrite S, et al. Registro de dados sobre acidentes de trabalho fatais em sistemas de informação no Brasil. Cien Saude Colet 2019; 24: 693–704.
17. Programa das Nações Unidas para o Desenvolvimento (PNUD). Atlas do Desenvolvimento Humano. Atlas Brasil. 2023. Available from: http://www.atlasbrasil.org.br/. Accessed January 18, 2025.
18. Carreira GB, Dutra VGP, Da Silva JHCM, et al. Social inequality, human development and fertility pattern in Brazil, 2000-2010. Rev Bras Saúde Matern Infant 2019; 19: 217–232.
19. Cohen J. Statistical power analysis. Curr Dir Psychol Sci. 1992;1(3):98-101.
20. Rodrigues AB, Santana VS. Acidentes de trabalho fatais em Palmas, Tocantins, Brasil: oportunidades perdidas de informação. Rev Bras Saúde Ocup 2019; 44: e8.
21. Melo MAS, Ferreira M, Coleta D, et al. Percepção dos profissionais de saúde sobre os fatores associados à subnotificação no Sistema Nacional de Agravos de Notificação (Sinan). Rev Adm em Saúde; 18. Epub ahead of print 5 June 2018.
22. Silva ALF, Diniz MB. Padrões de concentração regional da indústria de transformação brasileira. Nov Econ 2020; 30: 407–454.
23. Mesquita J, Machado H, Santana VS, et al. Situação da Rede Nacional de Atenção Integral em Saúde do Trabalhador (Renast) no Brasil, 2008-2009. Rev Bras Saúde Ocup 2013; 38: 243–256.
24. Galdino A, Santana VS, Ferrite S. Os Centros de Referência em Saúde do Trabalhador e a notificação de acidentes de trabalho no Brasil. Cad Saude Publica 2012; 28: 145–159.
25. Neto AM, Raphael De Oliveira Silva F, Severian D. Região e indústria no Brasil: ainda a continuidade da “desconcentração concentrada”? Econ e Soc 2020; 29: 581–607.
26. Silva ALF, Diniz MB. Padrões de concentração regional da indústria de transformação brasileira. Nov Econ 2020; 30: 407–454.
27. Mascarenhas MDM, Freitas MG, Monteiro RA, et al. Emergency room visits for work-related injuries: characteristics and associated factors - Capitals and the Federal District, Brazil, 2011. Ciênc Saúde Colet (Impr) 2015; 20: 667–678.
28. Santana V, Nobre L, Waldvogel BC. Acidentes de trabalho no Brasil entre 1994 e 2004: uma revisão. Ciência & Saúde Coletiva. 2005;10:841-55.
29. Melges F, Calarge TCC, Benini ÉG, Pacheco APDC. A nova precarização do trabalho: um mapa conceitual. Organizações & Sociedade. 2022;29:638-66.
30. Miranda CML, Araújo JB, Mioto BT. Precarização do trabalho e nova informalidade no urbano: permanências e transformações. Cadernos Metrópole. 2023;26:9-17.
31. Fundo Brasil. Mapeamento do trabalho informal no Brasil. Fundo Brasil, editor. Fundo Brasil; 2022. Available from: https://www.fundobrasil.org.br/wp-content/uploads/2023/01/mapeamento-trabalho-informal-corte1-1.pdf.
32. Vasconcelos FD. Atuação do Ministério do Trabalho na fiscalização das condições de segurança e saúde dos trabalhadores, Brasil, 1996-2012. Rev Bras Saúde Ocup 2014; 39: 86–100. Accessed January 18, 2025.
33. Mattos E, Politi R, Fuschini G. Fiscalização no mercado de trabalho e informalidade nos municípios brasileiros. Nov Econ 2018; 28: 103–126.
34. Fragoso Junior A, Garcia EG. Transporte rodoviário de carga: acidentes de trabalho fatais e fiscalização trabalhista. Rev Bras Saúde Ocup 2019; 44: e3.
35. Galdino A, Santana VS, Ferrite S. Fatores associados à qualidade de registros de acidentes de trabalho no Sistema de Informações sobre Mortalidade no Brasil. Cad Saude Publica 2019; 36: e00218318.
36. Shimizu HE, Bezerra JC, Arantes LJ, Merchán-Hamann E, Ramalho W. Analysis of work-related accidents and ill-health in Brazil since the introduction of the accident prevention factor. BMC Public Health. 2021;21(1).
37. Morgenstern H. Ecologic studies in epidemiology: concepts, principles, and methods. Annu Rev Public Health. 1995;16:61-81.