Skip to main content
NCBI home page
Archives of Public Health logoLink to Archives of Public Health
. 2025 Nov 26;83:319. doi: 10.1186/s13690-025-01776-0

Effects of mining disasters on hospital morbidity in the populations of Mariana and Brumadinho, Minas Gerais, Brazil

Emerson Pessoa Vidal 1, Rita de Cássia Costa da Silva 2,✉, Paola Zucchi 2
PMCID: PMC12752124  PMID: 41299784

Abstract

Background

Disasters are events that spread throughout the affected regions’ surrounding areas. These events had immediate and long-term effects on Mariana and Brumadinho. This study aimed to assess the impact of mining disasters on the Unified Health System (SUS), based on the profile of hospitalization services used, the number of hospital admission authorizations, total costs, and the number of deaths in the municipalities of Mariana and Brumadinho before and after the disasters, using the chapters of the International Classification of Diseases (ICD-10).

Methods

Statistical analyses were performed using Student's t-test, which verified the significance of the means before and after the disaster regarding SUS hospitalization services.

Results

The results indicate an increase in hospitalizations in the ICD chapters related to mental and behavioral disorders, perinatal conditions, and abnormal symptoms and test findings. In Brumadinho, the data point to an increase in the number of hospitalizations and deaths due to parasitic diseases, the latter also occurring in Mariana.

Conclusions

In any case, the disaster generated needs that would not have been noticed or observed before, which were mainly reflected in infectious and parasitic diseases, suggesting that changes in the environment can cause serious changes in the vector cycle. These data are evident in both Mariana and Brumadinho in terms of the number of deaths, but Brumadinho had a much higher than average increase due to the COVID-19 pandemic. Although daily life in both cities has been altered, the human consequences in Brumadinho were more severe than those in Mariana. Such environmental changes increase healthcare and SUS costs.

Keywords: Healthcare Systems, Disaster Management, Mortality, Health Expenditures, Hospitalization

Text box 1. Contributions to the literature
• This is a unique study examining the relationship between hospitalization, expenditure, mortality, and diseases in the Unified Health System.
• The study aims to determine which diseases predominate in hospitalizations and how much is spent on mining disasters. Understanding the link between diseases and hospitalizations reveals which diseases are most prevalent in mining disasters.
• Recognizing public spending makes it possible to understand why specific health sectors urgently need funding in times of crisis.
Open in a new tab

Background

Between 2015 and 2019, two major mining disasters occurred in Brazil between 2015 and 2019. The first occurred in Mariana, Minas Gerais, with the collapse of the Fundão dam, which resulted in the death of 19 people and the release of more than 40 million m3 of tailings. These residues were transported through the Rio Doce Basin, affecting a 668-km stretch of water, the largest stretch of water recorded to date [1, 2]. The main areas affected were the district of Bento Rodrigues, Paracatu de Baixo, and other communities along the Doce River. The population comprises small, rural, traditional communities based on family farming, artisanal fishing, and small businesses.

The second disaster occurred when a dam at the Córrego do Feijão mine in Brumadinho, Minas Gerais, burst and released 12 million m3 of mining waste, directly affecting the company’s administrative area and surrounding communities, resulting in 272 deaths [3, 4]. The main location affected was the community of Córrego do Feijão and areas near the Paraopeba River. Population profile: Mixed community with urban and rural residents; strong economic dependence on mining (direct and indirect jobs at Vale).

This study focused on the cities where the disaster originated: Mariana and Brumadinho, to verify their unique and common aspects. Radiation was not the study topic; instead, the focus was on the center of origin.

The effects of these disasters on public health were immediate, mainly due to deaths, injuries, and environmental, social, and human impacts. However, in addition to short-term damage, monitoring possible changes in the morbidity and mortality profile of the affected populations is important. Such changes may increase the demand for health services, resulting in the need for more medical consultations, tests, and hospitalizations [5, 6].

Mariana and Brumadinho had several things in common: they were small communities whose cultural identity and way of life were severely impacted; social ties were broken, and forced displacement occurred; the communities were dependent on mining as the economic backbone; and massive environmental damage was observed to rivers, soil, and wildlife, which directly affected families’ livelihoods.

Therefore, health services must ensure that individual and collective treatment and care actions are consistent with prevention and health promotion practices. Health units, especially in the affected regions, must strive to understand the urgent needs of residents, especially the most vulnerable [7]. Knowing these communities’ real needs allows us to develop more effective and sensitive responses to their demands.

In a brief summary, the Brazilian healthcare system is mixed, combining a universal public sector, financed mainly by taxes and social contributions, with a supplementary private sector, aimed at those who can pay directly or purchase health insurance plans. It is financed by federal, state, and municipal resources and follows universality, comprehensiveness, equity, decentralization, and social participation principles. Decentralized management with shared responsibilities (Ministry of Health and state and municipal secretariats) The services offered include primary care (Family Health Strategy, vaccination, women’s health, and prevention), specialized and hospital care, urgent and emergency care (SAMU, UPA), epidemiological, health, and environmental surveillance, and strategic programs (immunization, HIV/AIDS, transplants, and free provision of medicines). It accounts for 70%–75% of the country’s healthcare coverage and is the only option for most of the population [8, 9].

The private sector, known as supplementary health care, serves approximately 25% of the population, mainly the middle and upper classes, through health insurance plans and direct payments (“out of pocket”). Models of operation include individual, corporate group, and membership plans. Relationship with the SUS: Although complementary, many private hospitals provide services to the SUS through contracts and agreements (particularly in highly complex cases) [9].

Hospital morbidity is of unique importance in this scenario because it consumes a large amount of financial resources in the healthcare sector and reflects the demand for local hospital care from a regional center population with significant projection [10]. Hospital morbidity is a valuable tool for understanding the epidemiological profile of certain population segments. It assists in disease assessment, health condition surveillance and control, health service access and use, and health planning [11, 12].

Thus, it is crucial to understand how mining disasters impacted the Unified Health System (SUS) to guide possible actions in similar situations. This article aims to assess the impact of mining disasters on the unified health system using the chapters of the International Classification of Diseases (ICD-10) to assess the profile of hospitalization services used—number of hospital admission authorizations (HAA), total costs, and number of deaths in the municipalities of Mariana and Brumadinho before and after the disasters. The study is relevant because it provides the necessary knowledge for managers to take action and implement operations to mitigate and lessen the consequences of possible new occurrences of this magnitude.

Method

This is a quantitative, longitudinal study that investigates the use of hospitalization services in the municipalities of Mariana and Brumadinho, Brazil.

We obtained records referring to hospital admission authorization (HAA), hospital deaths, and total amounts paid by SUS for hospitalizations to access the data.

Data were extracted from the Hospital Information System (SIH/SUS) and the Mortality Information System (SIM/SUS), which are publicly available on the DATASUS portal.

Data were accessed in September 2021 to ensure reproducibility. Variables related to morbidity, hospital mortality, and expenses were selected and restricted to the period from November 2013 to November 2017 for Mariana and January 2017 to January 2021 for Brumadinho. The information was treated in aggregate form, without manipulation that would enable the identification of individuals. Analyses were conducted with scientific rigor and transparency in accordance with the ethical principles of public health research. The total value variable used the real currency for the transaction, which on the current date of 06/25/2025 was quoted at R$5.56 for each US dollar (USD).

Using the month in which the two disasters occurred as a reference, the behavior of these variables was checked two years before and two years after the disasters. The data were obtained from the Hospital Information System (HIS), which is available on the SUS Information Technology Department website (http://datasus.saude.gov.br). Data were collected monthly and recorded in a Microsoft Office Excel spreadsheet from 2013 to 2017 for Mariana and from 2017 to 2021 for Brumadinho. The data were organized into two periods, before and after, using the month in which the disaster occurred as a reference, so that in Mariana, it was considered before and after November 2015, and in Brumadinho, it was considered before and after January 2019. The November 2015 and January 2019 months were identified as zero months and were therefore excluded. Data were collected in November 2021 from the Ministry of Health’s DATASUS Tabnet system. The ICD-10 chapters were used to name SUS hospital morbidities.

The obtained data were statistically analyzed using Student’s t-test for independent samples to compare the means before and after the disaster. To obtain the t-value and p-value of the compared means, the calculations were performed using IBM SPSS Statistic software version 22. Student’s t-test was used to compare the means of HAAs, total value, and number of deaths before and after the disaster to analyze whether there was a significant difference existed between the means. Thus, if the period studied shows consistent averages, i.e., with low variability and differing significantly from each other, a p-value of less than or equal to 0.05 is expected. This result (p ≤ 0.05) indicates statistical significance, i.e., the differences observed are less than 5% likely to have occurred by chance. A two-tailed Student’s t-test was used to statistically verify the differences between the means, adopting a 95% confidence interval. The average was calculated based on the 24 months before and after the disaster.

In addition, time series calculations were performed using the joinpoint regression model, in which a set of segments with different slopes, joined by points of change (joinpoints), form a time series. Each segment shows the monthly percentage change (MPC) or absolute monthly change (MC), and for the series as a whole (in the case of more than one segment), the average annual percentage change (AMPC) or average annual change (AMC).

In accordance with the recommendations of National Health Council (NHC) Resolution 466 of December 12, 2012, the ethical principles of research involving human beings were respected. Approval from the research ethics committee was waived because the study was conducted using publicly available secondary data, without the possibility of identifying individual pieces of information. The study also included a Declaration of Responsibility signed by those responsible for the research.

This study used only public-domain secondary data obtained from the DATASUS Department of Informatics of the Unified Health System. As this information is aggregated, non-nominative, and does not allow for individual identification, informed consent is required from the participants. According to Resolution No. 510/2016 of the National Health Council, research using publicly available and unidentified data does not require prior submission to a Research Ethics Committee.

Results

When analyzing the number of HAAs in Mariana, four ICD 10 chapters showed an increase in monthly averages in the subsequent period compared with the previous period. The mental and behavioral disorders chapter had an average of five admissions per month and p = 0.001; diseases of the ear and mastoid apophysis one admission per month and p = 0.025; some conditions originating in the perinatal period seven admissions per month and p = 0.035; and symptoms, signs and abnormal findings of clinical and laboratory examinations six admissions per month and p = 0.03. This behavior suggests a possible effect of the disaster on the local morbidity profile. All HAAs had a significant p-value, confirming the possible effect of the disaster (Table 1).

Table 1.

Hospital Admission Authorization (HAA) before and after the mining disaster according to ICD-10 chapters in Mariana and Brumadinho, Brazil

Mariana Brumadinho Mariana Brumadinho
Before After Before After p-value (2-tailed significance)
Monthly average in the period
N		 Monthly average in the period N Monthly average in the period N Monthly average in the period N
I. Infectious diseases 17 10 8 14 0.001 0.001*
424 246 184 331
II. Neoplasms 15 18 13 12 0.066 0.846
360 438 303 297
III. Blood diseases 3 3 1 1 0.268 0.594
63 77 28 24
IV. Endocrine diseases 14 9 3 3 0.001 0.594
345 210 77 69
V. Mental disorders 3 5 2 3 0.001* 0.049*
60 112 51 78
VI. Diseases of the nervous system 7 6 1 2 0.675 0.054
158 147 24 46
VII. Diseases of the eye and adnexa 2 1 6 2 0.061 0.034
48 26 145 45
VIII. Ear diseases 0 1 0 0 0.025* 0.422
5 15 3 6
IX. Diseases of the circulatory system 43 47 18 18 0.336 0.829
1040 1.124 435 443
X. Diseases of the respiratory system 30 35 20 13 0.177 0.002
726 848 471 315
XI. Diseases of the digestive system 26 28 16 13 0.315 0.068
613 666 387 323
XII. Skin diseases 4 5 7 9 0.246 0.264
99 121 174 206
XIII. Diseases of the musculoskeletal and connective tissue system 7 7 3 3 0.528 0.336
179 164 62 73
XIV. Diseases of the genitourinary system 30 27 18 18 0.252 0.829
728 657 437 429
XV. Pregnancy. childbirth and puerperium 74 65 27 28 0.088 0.362
1776 1.554 644 683
XVI. Some conditions in the perinatal period 5 7 3 5 0.035* 0.022*
116 161 77 112
XVII. Congenital malformations 2 2 0 1 0.920 0.006*
58 57 11 26
XVIII. Symptoms. signs and abnormal test findings 4 6 2 3 0.031* 0.030*
105 150 46 81
XIX. Injuries. poisoning 37 41 14 20 0.153 0.001*
877 976 342 481
XX. External causes of morbidity and mortality
94 0 0 0
XXI. Factors that influence health status 12 7 3 5 0.030 0.086
293 173 88 116
Open in a new tab

Source: Prepared by the authors

1The values were approximated in order to reduce or remove the decimal places, so the p-value was calculated using the exact values

*Values considered statistically significant

In Brumadinho, the ICD-10 chapters that indicate a possible effect of the disaster include:infectious and parasitic diseases, with an average of 14 monthly hospitalizations and p = 0.001; mental and behavioral disorders, 3 monthly hospitalizations and p = 0.049; some conditions originating in the perinatal period, 5 monthly hospitalizations and p = 0.022; congenital malformations, deformities, and chromosomal anomalies, 1 monthly hospitalization and p = 0.006; symptoms, signs, and abnormal findings of clinical and laboratory examinations, 3 monthly hospitalizations and p = 0.030; injuries, poisonings, and some other consequences of external causes, with 20 hospitalizations and p = 0.001. All these HAAs are statistically significant, confirming the possible effect of the disaster (Table 1).

When analyzing the total value of hospitalizations, it can be seen that in Mariana, there was a statistically significant difference in the chapters diseases of the circulatory system, with an average value of R$60,963.67 per month and p = 0.014, and diseases of the respiratory system, with an average value of R$39,495.21 per month and p = 0.036, in the period after the disaster. These findings indicate a possible increase in hospital costs associated with these illnesses after the dam collapse (Table 2).

Table 2.

Total value before and after the disaster according to ICD-10 chapters in Mariana and Brumadinho, Brazil

Mariana Brumadinho Mariana Brumadinho
Before After Before After p-value (2-tailed significance)
Monthly average in the period
Total for the period
R$		 Monthly average in the period
Total for the period
R$		 Monthly average in the period
Total for the period
R$		 Monthly average in the period
Total for the period
R$
I. Infectious diseases 26.695,10 27.700,85 19.412,43 59.356,65 0.864 0.004*
660.682,31 664.820,29 465.898,36 1.424.559,68
II. Neoplasms 34.529,57 46.049,17 30.896,81 28.991,12 0.084 0.723
828.709,58 1.105.180,06 741.523,42 695.786,77
III. Blood diseases 1.807,60 1.430,35 759,29 897,88 0.506 0.724
43.382,33 34.328,50 18.222,84 21.549,05
IV. Endocrine diseases 9.687,30 6.821,97 3.317,14 1.577,71 0.272 0.256
232.495,29 163.727,36 79.611,45 37.865,06
V. Mental disorders 2.949,32 1.669,46 255,45 569,61 0.002* 0.108
70.783,63 40.067,09 6.130,79 13.670,64
VI. Diseases of the nervous system 5.087,96 4.241,79 1.259,48 1.660,30 0.462 0.547
122.111,11 101.802,88 30.227,59 39.847,20
VII. Diseases of the eye and adnexa 2.689,86 1.506,97 4.446,52 3.782,52 0.037 0.659
64.556,54 36.167,31 106.716,44 90.780,46
VIII. Ear diseases 221,13 379,59 72,71 259,08 0.306 0.224
5.307,00 9.110,20 1.745,03 6.217,99
IX. Diseases of the circulatory system 60.963,67 81.591,61 54.876,18 70.076,88 0.014* 0.088
1.463.128,11 1.958.198,59 1.317.028,27 1.681.845,13
X. Diseases of the respiratory system 29.454,51 39.495,21 15.242,46 18.678,06 0.036* 0.330
706.908,31 947.885,06 365.819,02 448.273,39
XI. Diseases of the digestive system 18.073,43 17.359,78 12.365,27 11.929,68 0.764 0.818
433.762,26 416.634,74 296.766,58 286.312,22
XII. Skin diseases 2.952,01 3.558,80 3.681,97 2.935,26 0.571 0.501
70.848,22 85.411,25 88.367,37 70.446,17
XIII. Diseases of the musculoskeletal and connective tissue system 15.977,97 7.705,93 5.336,36 5.939,83 0.006 0.736
383.471,16 184.942,25 128.072,57 142.555,92
XIV. Diseases of the genitourinary system 20.480,66 22.661,45 14.338,75 13.981,40 0.640 0.932
491.535,90 543.874,79 344.129,97 335.553,71
XV. Pregnancy, childbirth and puerperium 37.723,57 33.484,86 16.875,50 17.394,13 0.119 0.653
905.365,60 803.636,69 405.012,11 417.459,02
XVI. Some conditions in the perinatal period 16.713,32 16.343,18 11.533,97 12.488,23 0.941 0.750
401.119,67 392.236,25 276.815,34 299.717,53
XVII. Congenital malformations 6.787,91 8.154,56 432,00 2.809,78 0.625 0.256
162.909,83 195.709,55 10.367,97 67.434,76
XVIII. Symptoms, signs and abnormal test findings 3.426,22 3.461,01 2.809,78 5.242,49 0.969 0.413
82.229,22 83.064,23 89.585,90 125.819,64
XIX. Injuries, poisoning 43.602,49 42.216,73 21.012,95 25.515,79 0.806 0.192
1.046.459,77 1.013.201,56 504.310,85 612.378.93
XX. External causes of morbidity and mortality
8.792,76 0 0 0
XXI. Factors that influence health status 1.932,79 5.014,94 3.016,52 6.676,13 0.060 0.086
46.386,91 120.358,50 76.670,60 160.227,22
Open in a new tab

Source: Prepared by the authors

1The values were approximated in order to reduce or remove the decimal places, so the p-value was calculated using the exact values

* Values considered statistically significant

Only the infectious and parasitic diseases chapter showed a statistically significant difference in the total value of hospitalizations after the disaster in Brumadinho, with an average of R$59,356.65 in monthly expenses (p = 0.004) (Table 2).

The number of deaths was statistically significant only for the group of infectious and parasitic diseases in Mariana, with an average of 2 deaths per month and a p-value of 0.006 in the period after the disaster.

Similarly, the only category with a statistically significant difference in the number of deaths in Brumadinho was infectious and parasitic diseases, with an average of 2 deaths per month (p = 0.039) (Table 3).

Table 3.

Deaths before and after the disaster according to ICD-10 chapters in Mariana and Brumadinho, Brazil

graphic file with name 13690_2025_1776_Tab3_HTML.jpg

Open in a new tab

Source: Prepared by the authors

Figure 1 corroborates the tables above, with mental disorders, ear diseases, perinatal conditions, abnormal signs and findings, and diseases that are statistically significant after the disaster, showing that the disaster resulted in post-disaster changes that demonstrate an increase in the number of hospitalizations in the city of Mariana. Infectious diseases, endocrine diseases, and factors related to health status indicate an increase in hospitalizations before the disaster.

Fig. 1.

Fig. 1

Open in a new tab

Hospitalizations according to morbidity before and after the disaster in the city of Mariana/MG

Infections diseases, mental disorders, perinatal conditions, congenital malformations, abnormal signs and findings, injuries, and poisonings increase the number of post-disaster hospitalizations (Fig. 2). Morbidities diseases of the eyes and appendage diseases of the respiratory system show an increase in hospitalizations in the pre-disaster period.

Fig. 2.

Fig. 2

Open in a new tab

Hospitalizations according to morbidity before and after the disaster in Brumadinho/Mg

Table 4 shows that endocrine diseases have an AMPC/AMC of −0.74, which means that there was an annual percentage change during the study period; mental and behavioral disorders had an AMPC/AMC of 2.41, i.e., an increasing trend; ear diseases had an AMPC/AMC of 0.03, increasing; some perinatal conditions had an AMPC/AMC of 1.03, and abnormal symptoms, signs, and findings had an AMPC/AMC of 0.14, which is also increasing.

Table 4.

Trend in the AIH rate (per 100,000) from November 2013 to November 2017, Mariana

Total
Period AMPC/AMC 95% CI
Endocrine diseases, Nov-13 to Nov-17 −0.74 −5.07 to 3.78
Mental disorders Nov-13 to Nov-17 2.41 1.42 to 3.41
Ear diseases Nov-13 to Nov-17 0.03 0.01 to 0.05
Some conditions in the perinatal period Nov-13 to Nov-17 1.03 0.02 to 2.06
Symptoms, signs, and abnormal findings Nov-13 to Nov-17 0.14 0.04 to 0.24
Open in a new tab

Source: Prepared by the authors

Table 5 shows that infectious diseases have an AMPC/AMC of 2.84, which is increasing; some conditions in the perinatal period show growth with an AMPC/AMC of 0.07; congenital malformations with an AMPC/AMC of 0.04 show growth; symptoms, signs, and abnormal test findings with an AMPC/AMC of 0.09; and injuries and poisoning with an AMPC/AMC of 0.87.

Table 5.

Trend in AIH rates (per 100,000 inhabitants) from November 2013 to November 2017, Brumadinho

Total
Period AMPC/AMC 95% CI
Infectious diseases Jan-17 to Jan-21 2.84 1.00 to 4.71
Some conditions in the perinatal period Jan-17 to Jan-21 0.07 −0.04 to 0.18
Congenital malformations Jan-17 to Jan-21 0.04 0.01 to 0.08
Symptoms, signs, and abnormal test results Jan-17 to Jan-21 0.09 −0.02 to 0.21
Injuries, poisoning Jan-17 to Jan-21 0.87 0.23 to 1.52
Open in a new tab

Source: Prepared by the authors

Figure 3 shows that endocrine, nutritional, and metabolic diseases present three trends: from Nov-13 to Jun-15 with MPC/MC of −0.65 (CI: −3.53 to 2.31); from June 2015 to November 2015 −15.56 (CI: −43.74 to 26.74); and from November 2015 to November 2017, with an MPC/MC of 2.58 (CI: −0.19 to 5.42).

Fig. 3.

Fig. 3

Open in a new tab

Rate of AIHs due to endocrine, nutritional, and metabolic diseases from November 2013 to November 2017 in the city of Mariana

Mental and behavioral disorders presented from November 2013 to November 2017 with an MPC/MC of 2.41 (CI: 1.42–3.41) (Fig. 4).

Fig. 4.

Fig. 4

Open in a new tab

Rate of AIHs for mental and behavioral disorders from November 2013 to November 2017, in the city of Mariana

Diseases of the ear and mastoid process have an MPC/MC of 0.03 (0.01–0.05), as shown in Fig. 5.

Fig. 5.

Fig. 5

Open in a new tab

Rate of AIHs due to ear and mastoid diseases from November 2013 to November 2017, Mariana

The Fig. 6 shows some conditions originating in the perinatal period with an MPC/MC of 1.03 (CI, 0.02–2.06).

Fig. 6.

Fig. 6

Open in a new tab

Rate of AIHs in some conditions originating in the perinatal period from November 2013 to November 2017, Mariana

Symptoms, signs, and abnormal findings from clinical and laboratory tests with an MPC/MC of 0.14 (range, 0.04–0.24) (Fig. 7).

Fig. 7.

Fig. 7

Open in a new tab

Rate of AIHs in symptoms, signs, and abnormal findings from clinical and laboratory tests from November 2013 to November 2017, Mariana

Figure 8 shows that injuries, poisonings, and some other consequences of external causes show an MP/MC of 0.29 (CI −0.21 to 0.60) from November 2013 to November 2017.

Fig. 8.

Fig. 8

Open in a new tab

Rate of AIHs in injuries, poisonings, and some other consequences of external causes from November 2013 to November 2017, Mariana

Infectious and parasitic diseases had an MPC/MC of 1.29 (CI 0.2–2.8) from January 2017 to February 2020 and an MPC/MC of 8.22 (CI 1.47–15.42) from February 2020 to January 2021 (Fig. 9).

Fig. 9.

Fig. 9

Open in a new tab

Rate of AIHs in some infectious and parasitic diseases from January 2017 to January 2021, Brumadinho

Figure 10 shows that congenital malformations, deformities, and chromosomal anomalies from January 2017 to January 2021 had an MPC/MC of 0.04 (CI 0.01 to 0.08).

Fig. 10.

Fig. 10

Open in a new tab

Rate of AIHs in congenital malformations, deformities, and chromosomal abnormalities from January 2017 to January 2021, Brumadinho

As shown in Fig. 11, symptoms, signs, and abnormal clinical examination findings from January 2017 to January 2021 with an OR/RR of 0.09 (CI −0.02 to 0.21).

Fig. 11.

Fig. 11

Open in a new tab

Rate of AIHs XVIII-Symptoms, signs, and abnormal findings from clinical and laboratory tests from January 2017 to January 2021, Brumadinho

According to Fig. 12, injuries and poisonings have an MPC/MC of 0.87 from January 2017 to January 2021 (CI 0.23 to 1.52).

Fig. 12.

Fig. 12

Open in a new tab

Rate of AIHs in XiX – Injuries, poisonings from January 2017 to January 2021, Brumadinho

Discussion

The mining disasters in Mariana and Brumadinho have had a significant impact on the quality of life and health conditions of the population, representing a drastic change in the health risk scenario. Severe environmental deterioration has altered the reality of these communities, with the emergence and worsening of various morbidities associated with exposure to toxic tailings [7].

Analysis of the HAAs revealed that some illnesses were statistically significant in the period before the disaster, indicating that the health services did not adequately identify or treat pre-existing problems. In Mariana, infectious diseases, endocrine diseases, and factors influencing health status stood out. Diseases of the eye and appendage and the respiratory system were more prevalent in Brumadinho. For example, in the case of endocrine diseases [13], lifestyle factors and genetic predisposition explain part of the profile [13]. Diseases of the respiratory system, eyes, and appendages [14] may be related to chronic exposure to dust and contaminated water, both of which are common in mining areas.

Hospital admissions increased in both municipalities after the disasters. In Mariana, statistically significant increases were observed in the ICD chapters relating to mental and behavioral disorders, ear diseases, perinatal conditions, and abnormal clinical and laboratory findings. Increases were observed in infectious and parasitic diseases, mental and behavioral disorders, perinatal conditions, congenital malformations, abnormal clinical and laboratory findings, injuries, poisonings, and other external causes in Brumadinho. A significant increase in mental and behavioral disorders and abnormal clinical and laboratory findings was observed in both municipalities.

With regard to the total cost of hospitalizations in Mariana, a statistically significant increase in spending on circulatory and respiratory diseases was observed in the post-disaster period. In Brumadinho, the highlight was infectious and parasitic diseases, with a significant increase in costs (monthly average of R$59,356.65 and p = 0.0004). This agrees with Espalhardo et al. (2015), who highlighted that health expenditures tend to increase after disasters [14]. In the post-disaster period in Brumadinho, we also had another major event that may have influenced the data: the COVID-19 pandemic, which, due to its emergence and unique and divergent characteristics from the influenzas that had appeared until then, ended up being classified under two ICDs, 1, some infectious and parasitic diseases, and 10, respiratory diseases, such as bronchitis and pneumonia.

There was a statistically significant increase in deaths only for infectious and parasitic diseases, both in Mariana and Brumadinho. This finding reinforces Freitas’s (2019) warning that parasitic diseases and other illnesses, such as respiratory infections, diarrheal diseases, hypertension, and diabetes, increase after major disasters [5]. These deaths in Brumadinho are likely related to the spread of COVID-19 throughout Brumadinho and around the world.

 Disaster situations can aggravate chronic health conditions due to the momentary collapse of the healthcare system, leading to the neglect of care for these diseases due to multiple factors, such as the inability to meet demand [7].

The Student's t-test was used to compare the means of HAAs between the pre- and post-disaster periods. The analysis indicated statistically significant differences in several categories, indicating an increase in the hospital demand after the event. In Mariana, although the number of HAAs for infectious and parasitic diseases (ICD I) was higher in the period prior to the disaster, deaths from these causes increased in the post-disaster period, indicating inadequate treatment in the face of an outbreak that had already begun before the disaster.

In Brumadinho, data point to an increase in HAAs, expenses, and mortality related to infectious and parasitic diseases, revealing the effects of the disaster on this group of diseases. As described by Freitas et al. (2019), exposure to contaminated water and the collapse of health services favor the increase of these diseases, often manifested by diarrhea, fever, and skin lesions [5, 15]. Nevertheless, the interruption of epidemiological, sanitary, environmental, and occupational health services resulting from the disaster [1] has increased the demand for regular health services.

In addition, the Brumadinho disaster coincided with the emergence of COVID-19, a disease initially associated with ICD-10 Chapter I (infectious diseases) before the creation of specific codes. Thus, a part of the impact observed in this chapter is also related to the pandemic, which aggravated local health conditions.

The dam collapse caused an ecological imbalance that may have altered the vector cycle, favoring the spread of dengue, Zika, chikungunya, and schistosomiasis [16, 17]. Studies have shown a correlation between this environmental impact and the increase in parasitic diseases in Brumadinho (Freitas et al., 2024; Rocha et al., 2021). The proliferation of vectors intensifies the occurrence of infections, with dengue being the most prevalent [18, 19]. Dengue fever is a notable outcome of dam collapse [19, 20]. Rodrigues et al. (2020) confirmed an increase in the number of arboviruses, particularly dengue and chikungunya, in the municipality after the disaster.

The interpretation of the results requires caution due to the transition period in the classification of COVID-19 in health information systems. In the early months of 2020, cases were often recorded under the chapter on infectious and parasitic diseases or under generic codes before the creation of the specific codes U07.1 (COVID-19, virus identified) and U07.2 (COVID-19, virus not identified). This process may have artificially inflated the number of hospitalizations and deaths attributed to this group of diseases. On January 31, 2020, the World Health Organization (WHO) published emergency codes, which became mandatory in several countries in April of the same year. Therefore, the increases observed during this period should be interpreted with caution because they reflect not only the occurrence of COVID-19 but also changes in official coding, making it difficult to distinguish between the actual epidemiological evolution and the effects of the standardization process.

The results indicate a highly complex post-disaster scenario. The reduction in the hospital admissions may reflect the loss of installed capacity of the local SUS, the interruption of health services, or even the migration of patients to other municipalities [21]. The installed capacity loss must be analyzed. Unlike the number of beds, which quantifies the number of beds available for hospitalization, it is a broader concept that refers to a system, unit, or network’s maximum potential for health service production, considering physical structure + human resources + equipment + available supplies. Thus, a sudden disruption of the service network, destruction and physical interdiction, loss of human resources, immediate overload of demand, and environmental and logistical impact can occur. In Mariana, a temporary interdiction of basic units and overload in the municipal hospital reduced the number of available beds. However, the installed capacity decreased even further as human resources and equipment in areas affected by mud and metallic dust decreased. In Brumadinho, the disaster caused direct physical damage to the city’s hospital and displaced health professionals, reducing installed capacity, even when physical beds still existed. In addition, there was a sudden increase in demand (rescue, victims, mental health), which widened the gap between beds and actual response capacity. The increase in health spending is consistent with the network’s emergency reorganization, including expenses related to medications, logistics, contracting of external services, and reconstruction of damaged structures [21–23]. Increased mortality may be associated with worsening health conditions, delayed access to services, psychological distress, and chronic disease aggravation without continuous monitoring [24]. Population dispersion and the temporary collapse of the healthcare network may have increased the lethality of previously managed conditions [25, 26].

Exposure to disasters can aggravate chronic health conditions caused by factors such as extreme temperatures, food and water shortages, and physical and emotional trauma. Individuals with mental illness or disabilities, people of low socioeconomic status, and those without regular access to health care are at greater risk of mental illness or disability. Many of these individuals may be displaced to evacuation centers after disasters, where relief teams manage their medical conditions [27, 28]. The exacerbation of chronic conditions contributes substantially to the public health burden after disasters [27], suggesting a close relationship between ICD-10 chapters such as mental and behavioral disorders and circulatory system diseases in terms of the worsening of preexisting conditions.

Mental and behavioral disorders showed a significant increase in the number of HAAs in Mariana and Brumadinho. This result is consistent with studies that point to a higher prevalence of depression, post-traumatic stress disorder, anxiety, suicide risk, and sleep disorders after disasters [29, 30]. In Brumadinho, Loyola Filho et al. (2022) observed a 123% increase in spending on psychotropic drugs, indicating a high incidence of psychiatric problems [31].

The prevalence of psychiatric symptoms in the municipality is above the national average [32, 33], which agrees with the findings of systematic reviews that confirm higher rates of psychological disorders in disaster-exposed populations [34]. These findings reinforce the importance of paying special attention to the mental health of the affected communities.

Hearing problems may be related to the presence and ingestion of arsenic, lead, and mercury by the population [35, 36], which can lead to hearing loss due to exposure to these metals. Mercury may also be associated with other health problems, such as weakness, fatigue, weight loss, tachycardia, dizziness, visual field constriction, and coma. Chronic exposure to mercury can result in sensitivity disorders in the extremities, anxiety, depression, and insomnia, as well as an increased risk of cardiovascular disease [37], which can lead to death [38]. In the case of ear and mastoid diseases, a significance of 0.025 was observed in the HAAs in Mariana. Other possible correlations with ear diseases and the disaster may occur through contact with contaminated water, which may have increased the number of infections, including otitis externa and otitis media, as the presence of microorganisms can aggravate otological diseases. Stress and trauma can cause hearing impairment, which increases perception because of associated neural sensitization.

Although there was no increase in hospitalizations for circulatory system diseases, there was an increase in costs in Mariana, which may indicate greater clinical severity or the need for more complex treatments in the post-disaster period [39].

Respiratory diseases, such as factors associated with asthma, chronic obstructive pulmonary disease (COPD), and respiratory symptoms, showed an increase in costs, although the number of hospitalizations did not significantly increase, consistent with exposure to toxic dust described in the literature [15, 30, 40].

In both municipalities, conditions originating in the perinatal period were significantly associated with hospitalizations, indicating a possible effect of disasters on maternal and child health. The stress of pregnancy in emergency contexts, associated with health service interruption, can compromise fetal development and lead to prematurity and low birth weight [41–43]. According to Mrejen et al. (2020), during the perinatal period, stressful conditions during pregnancy can lead to premature birth and low birth weight [43].

Similarly, a statistically significant incidence of congenital malformations was observed in Brumadinho, which may be associated with the adverse conditions experienced by pregnant women exposed to intense stress. Such situations have a significant impact on maternal physiology and fetal development [43]. This is an atypical situation, marked by a complex interaction of factors that influence perinatality. The aggravating factors include exposure to heavy metals and toxic substances, lack of access to prenatal care, epigenetic mechanisms, and inadequate food and nutrition.

This risk becomes even more concerning in cases of metal contamination. The scientific literature corroborates this association because transplacental exposure to mercury, for example, is considered one of the most dangerous by the World Health Organization (WHO) and can cause serious damage to the fetal nervous system [44]. Postnatal symptoms include intellectual disability, loss of vision and hearing, strabismus, seizures, and limb deformities [45, 46].

Another ICD that is relevant in the context of the disaster is symptoms, signs, and abnormal test results, which are expressed in Mariana through a higher number of cases of physical violence, according to the Rio Doce project report, which is one of the descriptors of this ICD. Other authors, such as Faria (2019), also attest to a higher number of cases of alcohol use, domestic violence, and suicide in the city [18, 47]. Moreover, Brumadinho had a significant number of HAAs for this ICD, which may be related to clinical laboratory tests or even physical violence-related injuries. According to Malta (2020), violence and accidents are responsible for more than 4.8 million deaths worldwide. In Brazil, 150,000 premature deaths occur annually [48].

The statistically significant increase in the number of HAAs in the chapter on injuries and poisoning may be related to various types of poisoning, whether accidental, occupational, or self-inflicted [18, 47]. According to Porr (2024), socio-environmental disasters, such as the one that occurred in Brumadinho, are associated with an increased risk and attempts at suicide, having a serious impact on the affected population’s mental health [49]. The possible exposure to mining waste and heavy metals is an additional risk factor that should be considered in the analysis of disaster-related health hazards [35, 38, 50, 51].

The use of calculations, such as the t-student and Jointpoint regression, allowed us to understand that the trends and significance remain regardless of the calculations used. This confirmed what we had in mind, that is, the t-student calculation was sensitive in perceiving that the differences in certain diseases were not mere chance. This is clear from the data and confirmed by the scientific literature. Thus, the data were understood to have value, if not for confirmation, then for corroboration of what had been elicited by the data and published articles that address the same topic.

Based on the developed and analyzed information, a series of actions must be promoted with a view to the well-being of communities and the efficiency of health services. Coordination with government bodies (municipal, state, and federal) must be established so that recovery and care can be better implemented. Initiatives that can provide tools to assess and strengthen health resilience, such as the implementation of health surveillance services, because morbidities, such as infectious and parasitic diseases, have a strong impact in disaster situations; the development of contingency plans that can address business continuity with risk assessment, alternative routes, contact with suppliers, and reserves of critical supplies, thereby maintaining health services [52–54].

Conclusion

Data analysis revealed a statistically significant increase in HAAs in the ICD chapters related to mental and behavioral disorders, perinatal conditions, abnormal symptoms, and test findings in both municipalities. In Mariana, there was also an increase in ear diseases, whereas in Brumadinho, there was an increase in infectious and parasitic diseases, congenital malformations, injuries, and poisonings. The total number of hospitalizations was significant in the chapters on mental disorders and circulatory system diseases in Mariana and in the chapters on respiratory system diseases in Brumadinho. Both municipalities recorded a statistically significant increase in the number of deaths from infectious and parasitic diseases. These results show that disasters, such as those in Mariana and Brumadinho, have a strong impact on mental health and contribute to the proliferation of infectious diseases, possibly associated with environmental imbalance and an increase in vectors.

In conclusion, the parasitic diseases resulting from the Mariana disaster led to a health paradox: fewer hospitalizations, higher public spending, and more deaths. This demonstrates that the analysis of isolated indicators can be used to conceal serious realities. Studies of this nature are essential to support more prepared, intersectoral, and resilient institutional responses to new disasters. However, in Brumadinho, the paradox does not arise because hospitalizations, spending, and deaths are consistent, leading to the conclusion that the disaster may have led to greater action on the city and a greater extent of damage, even though the COVID-19 pandemic occurred during the period in question. This disease appears to have a major impact on such disasters.

The use of a second calculation to confirm the initial findings revealed that our findings are valid and verifiable. Moreover, some diseases indeed stand out in such disasters.

The study’s limitations include the exclusion of patients receiving supplementary health care and those with diseases that only benefit from outpatient care. The use of secondary data is subject to underreporting, errors in completion, and changes in information systems over time. Additionally, the records reflect the use of health services, which may not fully express the occurrence of diseases in the population.

Acknowledgements

Nothing to declare.

Abbreviations

COPD

Chronic Obstructive Pulmonary Disease

HAA

Hospital Admission Authorizations

HIS

Hospital Information System

ICD

International Classification of Diseases

MH

Ministry of Health’s

NHC

National Health Council

SUS

Unified Health System

CNS

National Health Council

PTSD

Post-Traumatic Stress Disorder

WHO

World Health Organization

Authors’ contributions

EPV: Conceived and designed the study, collected the data and wrote the manuscript RCCS: Conceived. reviewed and approve it for publication PZ: Conceived, reviewed and approve it for publication.

Funding

The study received no funding, being financed entirely by the researchers.

Data availability

Availability of datasets used and/or analyzed during the current study is available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

According to the recommendations of the Resolution of the National Health Council (CNS) No. 466, December 12, 2012, the ethical principles of research involving human beings were respected, and the approval of the research ethics committee was waived, since the study was conducted from secondary data, of public access, without the possibility of individual identification of the information. This ethics committee is called the Ethics and Research Committee of the Federal University of São Paulo, and chaired by Doctor Álvaro Pacheco Silva e Filho and composed of the research team Paola Zucchi and Rita de Cássia da Silva and signed and filed in the Department of Translational Medicine, with contact e-mail: medicinatranslacional@unifesp.br, which states waiver of analysis and submission to the committee and only signature of the declaration of responsibility by the members/researchers. All methods were carried out in accordance with the relevant guidelines and regulations and following the guidelines informed and required by the committee.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Availability of datasets used and/or analyzed during the current study is available from the corresponding author upon reasonable request.

Articles from Archives of Public Health are provided here courtesy of BMC

RESOURCES