Human Identification in Mass Disasters: Analyzing Complex Tattoos in the Brumadinho Tragedy

Abstract

Background: The identification of victims in mass disasters is a challenging task, particularly when forensic teams must address fragmented human remains. Objectives: This article reports two necropsy cases from the 2019 collapse of the Brumadinho mine tailing dam, the largest humanitarian disaster in Brazilian history. Results: In both cases, identification was achieved through comparative anthropological analysis of complex tattoos. Positive identification was based on the presence of multiple coincident points between postmortem (PM) and antemortem (AM) photographic records of the tattoo designs, along with the absence of exclusionary elements. Conclusion: The authors propose that the comparative analysis of PM and AM tattoos should be more widely adopted as a low-cost, rapid identification method, particularly in complex forensic scenarios such as mass disasters and other challenging cases in the medical and forensic anthropology context.

1. Introduction

On 25 January 2019, the tailings dam of an iron ore mine owned by Vale S.A. collapsed in Brumadinho, Minas Gerais, Brazil. The city is located about 50 km from Belo Horizonte, the state capital. Approximately 12 million cubic meters of mine tailings were spilled, covering an area of 290 hectares. The spill directly affected the company’s personnel, local residents, and tourists. The disaster caused 272 fatalities, a number that includes two unborn babies. It is considered the largest occupational accident and one of the most severe environmental disasters in Brazil’s history [1,2,3].

The term ‘disaster’ is defined by the International Criminal Police Organization (INTERPOL) as an unexpected event causing the death of multiple victims and can be classified as either open or closed [4]. Disasters are classified as open when the number and identity of the victims are unknown and the lack of information requires further investigation. In contrast, closed disasters involve victims who belong to a known group with available reference information [4]. The Brumadinho disaster fits both categories, as it affected not only areas with registered workers but also residential and rural areas where there was no pedestrian traffic control [2].

To guide the process of human identification in the context of catastrophes, INTERPOL created the Disaster Victim Identification (DVI) guide. The guide provides recommendations and establishes a standard on how to conduct DVI operations worldwide. An identification is based on the confrontation (reconciliation) between postmortem (PM) data, obtained through the examination of the mortal remains, and antemortem (AM) data, obtained from healthcare professionals, public and private databases, as well as interviews with family and friends.

PM data include clothing and belongings, photographs, fingerprints, medical and dental imaging, DNA samples, and PM procedures (necropsies); on the other hand, AM data include information on clothing and belongings, medical and dental records and imaging, photographs, posts on social media, DNA samples, and particular identification marks, such as tattoos. The confrontation is considered positive when there is correspondence between PM and AM data, allowing the victim’s identity to be confirmed [4].

The most reliable identifiers to confront in the process of identification are considered primary methods: fingerprints, dental records, and DNA. Alternatively, any other identifier capable of characterizing an individual in any way is considered a secondary method, a resource mainly used to reinforce an already established identity by means of a primary method. However, depending on the context of the analysis, secondary methods may become the protagonists in the case of a non-successful identification attempt with primary methods, limited access to primary identifiers or non-existent primary identifiers. Moreover, the quality of the data is a determining factor in the choice of which identification method will be the most appropriate for each case [4].

The identification of unknown bodies is one of the most important objectives of forensic sciences [5]. This procedure is indispensable for many reasons, including its legal and administrative importance for the state, as well as the moral and humane commitment towards the deceased and their families [6]. For some people, grief begins with the suspicion of death. For others, the uncertainty of death is harder to deal with than its confirmation; in such cases, the confirmation of death is a prerequisite for grief [6,7]. Therefore, in response to the Brumadinho DVI, the Brazil’s largest task force dedicated to the search, localization, and identification of victims was created [1,2,8]. In this context, the Instituto Médico Legal Dr. André Roquette (IMLAR) of the Polícia Civil de Minas Gerais provided the infrastructure and personnel, among other resources, to carry out the collection and confrontation of PM and AM data to ascertain the identity of the victims of the disaster. With this initiative, the victims’ mortal remains could be returned to hundreds of families, allowing them to say goodbye to their beloved ones and pay their last homage [8,9].

Tattoos can be invaluable identifiers. Some of the fundamental characteristics that make tattoos a useful resource in the process of body identification include their stability and relative durability, the fact they can be easily reported during the necroscopic exam, the possibility of being found in body fragments, and the low cost and fast performance of this confrontation method [6,10,11]. These attributes are particularly significant in challenging forensic cases, including burnt human remains, extensive fragmentation, and other scenarios encountered in Disaster Victim Identification (DVI) or routine forensic investigations.

Even when dealing with a catalog tattoo reproduced by the same tattoo artist on two different individuals, there are noticeable differences in aspects such as linework, pigment intensity, and inclination. These variations arise due to the unique characteristics of each individual’s skin, the pressure applied by the tattoo artist, and other subtle factors. With the use of macro lenses for PM records and the advancements and accessibility of smartphone cameras, mostly used for AM pictures, these differences have become increasingly perceptible. Just as there is specialized expertise in authenticating a work of art or verifying the authenticity of a signature, the forensic analysis of tattoos operates on the same principle. It is precisely these distinct details that form the basis for identification through the comparison of tattoo records.

This work assesses the utilization of the comparative exam of tattoos as an identification method based on two cases of the Brumadinho DVI, carried out by medical examiners from IMLAR’s Forensic Anthropology Division (SAF). The cases presented had authorization for the use of images for exclusive scientific and teaching purposes signed by the family members responsible for removing the bodies from IMLAR, and the study was approved by the Research Ethics Committee of the Faculdade Ciências Médicas de Minas Gerais.

2. Case Reports

2.1. Case 1

The remains of one of Brumadinho’s victims, already in the gas-producing stage of the putrefaction process, were brought to the SAF/IMLAR. In the necroscopic exam (PM), a tattoo was observed, depicting the face of a male figure with hair apparently tied up and a long beard, resembling a ‘samurai’. The tattoo, mostly black-colored, was located on the topographic area corresponding to the left deltoid muscle (Figure 1).

The AM databases created by the taskforce were consulted. One of the entries contained the following AM information under the field parameterized as “tattoos”: “design/color: wizard sucking souls and Japanese inscription. Body part: left arm”.

Based on information provided by the families, the missing person matching the described characteristics was identified for comparison. The name of the missing person was searched on the social media platform Facebook, where recent antemortem (AM) pictures were found. Despite their low resolution, the images were suitable for a tattoo comparative exam (Figure 2).

After sanitizing the body and acquiring the photographic records of the unknown victim’s tattoo (PM), the images were edited using the Photoshop software (Version 25.0) to match the proportion and dimension of the missing person’s photograph (AM) without compromising the original design. Next, the confrontation between the PM (pink) and AM (yellow) images was carried out, using the software Microsoft Power Point—Version 2019 (Figure 3).

Multiple coincident points were successfully established during the comparison of PM and AM images, including the following: a drawing depicting the face of a male figure with a long beard; the topographic location corresponding to the left deltoid muscle; predominantly black color; morphology (shape, length, and width) and white color of the male figure’s eyebrows; blue color of the eyes; white color of the mustache; and the shape of the eyes, nose, mustache, and mouth (Figure 3). Additionally, the following coincident points were identified: the angles of the face’s outline (white arrows); the expression lines on the male figure’s face, including their number, morphology (shape, length, and width), disposition, and location (numbered 01 through 18); and the points of contact between expression lines 03 and 18, 04 and 09, and 10 and 11 (white dotted circles in Figure 4).

Coincident points were also observed in the morphology and disposition of the areas outlined by the expression lines on the male figure’s face (Figure 5). Additional coincident points included the number of tips, morphology (shape, length, and width), and disposition of the long beard hairs on the male figure (numbered 01 through 11); the points of contact between the long hairs (white dotted circles in Figure 6); and the combination of all the aforementioned and detailed coincident points.

Concurrently, no discordant points were identified between the PM and AM images that could be considered exclusionary. Therefore, given the authenticity of the AM data, it was determined that there were sufficient technical forensic elements to establish a positive identification.

2.2. Case 2

The remains of one of Brumadinho’s victims were brought to the SAF/IMLAR, consisting of skin fragments from the head and neck region, as well as an incomplete segment of the articulated thorax with part of the upper left limb, in the colliquative stage of the putrefaction process. During the necroscopic exam, a tattoo (PM) was observed, depicting the face of a feline, predominantly gray, next to a pink flower. Its topographic location corresponded to the left deltoid muscle. The segment of skin containing the tattoo was resected to allow for the panoramic recording of the design (Figure 7).

The AM databases created by the taskforce were consulted. One of the entries under the field labeled “tattoos” contained the following AM information: “Design/Color: left arm (shoulder)”. This was the sole detail available in the database regarding the tattoo.

Based on information provided by the families, the missing person matching the described characteristics was selected for comparison. The family provided recent photographs (AM). Despite their low resolution, the images were suitable for a tattoo comparative exam (Figure 8).

After sanitizing the body, photographic records of the tattoo (PM) were acquired. The images were then edited using Photoshop software (Version 25.0) to match the proportions and dimensions of the missing person’s photograph (AM), without compromising the original design. Subsequently, the comparison between the PM and AM images was conducted (Figure 9).

Multiple coincident points were successfully established between the PM and AM images, such as the following: the design, representing the face of a feline and a flower; the disposition between the feline face and the flower, the flower and its leaves, and the leaves themselves; the expression of the feline face representing a snarl with its mouth open and protruded tongue; the color of the feline face, predominantly gray, and the color of the flower, predominantly pink; the colors of the details on the feline face: yellow eyes and snout and pink tongue; the colors of the details on the flower: yellow details on the petals and green leaves; the location, on the topography corresponding to the left deltoid muscle; and the morphology (shape) of the right ear, eyes, snout, dental elements, and tongue of the feline, as well as the petals and leaves of the flower (Figure 9).

The following coincidences were also observed: the angles of the feline face’s outline (white arrows) and the number of canine dental elements on the feline face, both upper and lower, numbered 01 through 04 (Figure 10).

Furthermore, notable coincidences were observed in the number, morphology (shape, length, and width), disposition, and location of the expression lines on the feline face (numbered 01 through 08), as well as in the lines and angles of the tongue’s outlines (white arrows in Figure 11). Additionally, the morphology of the areas outlined by the expression lines, jaw, and tongue of the feline face was generally consistent between the PM and AM images (Figure 12, Figure 13 and Figure 14), despite differences in skin elasticity caused by thanatological processes.

Notable coincidences were also observed in the angles of the flower’s outline and its leaves (white arrows), the number of leaves below the flower (numbered 01 through 03 in Figure 15), and the morphology of the areas outlined by the flower’s petals, details, and leaves (Figure 16).

The combination of all detailed coincident points, corroborated by the absence of divergent exclusionary points between the PM and AM images, enabled the conclusion of the case. The authenticity of the AM data was also taken into account. It was determined that sufficient technical forensic elements were present to establish a positive identification.

3. Discussion

Tattoos are permanent markings created by perforating the skin and injecting ink into the dermis [10,12], a practice that dates to prehistoric times [10,13,14]. Historically, tattoos have served as identifying traits for various groups, including “barbarian tribes”, Christians, “freak show” performers, slaves, marginalized individuals, sailors, criminals, prostitutes, and military personnel [10,12,13,14]. However, the migrations, social transformations, and cultural amalgamations associated with the progression of civilization have led to a shift in the motivations behind tattooing, contributing to the deconstruction of related stereotypes [10]. Today, tattoos are often considered a form of body art, created with custom designs on various anatomical regions, carrying symbolic significance and frequently commemorating important life events for their bearers [6,10,13,15].

Tattoo pigments are deposited in the dermis layer of the skin, allowing them to remain visible even after exposure to treatments, immersion in liquids, burns, or decomposition [11,14,15,16]. In some cases, the detachment of the epidermis can even enhance the visibility of the tattoo’s colors [11]. Additionally various techniques can improve the visualization of unclear tattoos. These include radiographic examination, particularly useful for older tattoos created with metallic inks [11]; the direct application of a 3% hydrogen peroxide solution, despite its potential for tissue destruction [11]; and infrared photography, which enhances black and green pigmentation [6,11].

Tattoos have become increasingly popular in Western countries, with prevalence rates estimated at 10–16% among teenagers and 3–9% in the general population [12,13,15,17]. A recent study conducted at the Instituto Jalisciense de Ciencias Forenses (IJCF) in Mexico reported an even higher prevalence of tattoos among bodies admitted to legal medical institutes, with 45.8% of individuals aged 15 to 87 years displaying tattoos. Notably, individuals aged 20–29 years exhibited the highest prevalence, with 67.9% of males and 61.2% of females having tattoos [18]. Furthermore, the reduction in prejudice historically associated with tattoos has contributed to a rise in tattoos placed on anatomical sites that are less or not at all concealed by clothing, making them more visible. Supporting this observation, the same study revealed that 63.3% of male and 74.2% of female tattooed bodies had tattoos located on body segments considered exposed [18].

In parallel, technological advancements have increased the social visibility of individuals. The widespread availability of personal (e.g., cameras, cell phones) and collective (e.g., security cameras) electronic devices capable of producing high-quality images and videos, along with the growing use of social media platforms to share such images, has further contributed to the identification and exposure of tattoos, potential AM resource.

The increasing number of tattooed individuals, combined with the greater visibility of tattoos on the body and the enhanced exposure of individuals on social media, has expanded the possibilities for utilizing these distinctive markers. In forensic contexts, tattoo comparison has already been employed to aid in solving crimes involving living suspects and in the identification of bodies. Examples of such applications have been reported in countries including Germany, Australia, Brazil, and the United States [11,13,14,15]. Concurrently, there has been a rise in the number of unidentified bodies, attributed to factors such as the growing population of unsheltered individuals, increased migration and travel, and disasters involving multiple victims or natural catastrophes, such as the 2004 tsunami in Thailand [10]. Consequently, the potential applications of tattoos as an identifier warrant further exploration.

In the Brumadinho DVI, the immense mass of tailings transferred significant kinetic energy to the victims. This caused devastating mechanical effects, reducing most bodies to fragments. The prolonged recovery of mortal remains trapped under the iron ore tailings further complicated the process. The inevitable progression of decomposition and the environmental impact of elements in the tailings mud also damaged biological tissues. As a result, the quality of the recovered postmortem (PM) data was severely compromised.

These challenges were amplified by the large number of victims and the dual characteristics of both open and closed disasters. This combination disrupted the collection of antemortem (AM) data. Together, these factors made the identification process extraordinarily difficult. Various identifiers and methods were employed based on their effectiveness in each specific case. Their selection considered the quality of available data as well as the time and resources needed for analysis [1,2,3,8,9].

In the present case series, despite the bodies being in different stages of decomposition, the anatomic sites of the tattoos remained sufficiently preserved from the traumatic forces, allowing examination. In both cases, family reference samples and AM genetic material from the suspected victims were collected for potential comparison. Biological tissue from the bodies was also preserved for postmortem (PM) DNA extraction to enable future analyses. While DNA identification was one of the primary methods used in the Brumadinho DVI [2], this approach faces challenges, including DNA degradation due to decomposition and the harmful effects of iron ore tailings [19]. Additionally, DNA analysis requires substantial time, financial resources, and logistical efforts [6]. In cases involving multiple victims, faster and more cost-effective methods are crucial.

Given these constraints and the availability of preserved tattoos, the comparative examination of tattoos was chosen as the most suitable identification method for these cases. During necropsy, the tattoos (PM) were easily identified without requiring physical treatment for visualization. Photographic records (PM) were digitally adjusted to match the dimensions of AM images, enabling simpler and more precise comparisons.

In case 1, fingerprint identification was also performed, corroborating the results of the tattoo comparison. For case 2, while no PM data for dental or fingerprint-based methods were available, the identification by tattoos was later confirmed using dental records from another body segment brought to the IMLAR.

Regarding AM data, several challenges were noted. Tattoo descriptions were inherently subjective, influenced by the perspectives of both AM reporters and PM examiners [6,15]. Additionally, AM tattoo reports were brief, making purely descriptive comparisons inconclusive. Despite these limitations, access to up-to-date AM images was critical for the identification process [11]. Social media, while often criticized for overexposure, proved instrumental in case 1 by providing a photograph essential for comparison. Without this image, AM data collection would have been impossible.

Notably, the low quality of AM data in both cases did not hinder the effectiveness of the comparative examinations. The proactive approach of a dedicated team was key, ensuring an active search for AM materials and diligent data confrontation. The examinations compared PM and AM images to identify coincident points while avoiding exclusion of features. This method followed established practices used in comparative dental exams and the radiographic or tomographic imaging of frontal sinuses. The results were unequivocal, successfully establishing the victims’ identities.

This highlights that, while DNA is often regarded as the gold standard in Disaster Victim Identification (DVI) due to its reliability and precision, alternative methods may still hold significant value [20]. According to the most recent edition of the DVI guide, INTERPOL acknowledges that secondary identification methods, including tattoos, can, depending on the circumstances, serve as standalone identification methods [4].

Tattoos, in particular, may provide a distinct and highly reliable means of confirming identity, especially when they feature unique patterns, designs, or inscriptions that are easily verifiable. This flexibility underscores the importance of a context-driven approach in DVI, ensuring the most effective and timely identification process possible. Similarly, the existing literature supports the use of tattoo comparison as an independent method of identification [6,18]. The value of morphological methods for identifying unknown bodies is widely acknowledged [6,18].

While tattoos are often viewed as storytelling features, reflecting an individual’s life experiences and identity, they can also be regarded as objective morphological characteristics. These features persist after death and can be effectively used to confirm the identity of both living and deceased individuals [6,18].

In addition to the cases presented in this study, SAF/IMLAR documented other instances of successful identification through the meticulous AM/PM comparisons of tattoos and the analysis of medical radiological documentation [21]. This approach proved effective, not only in other cases from the Brumadinho DVI response, but also in various forensic contexts, highlighting its reliability and adaptability in complex identification scenarios.

Similar to dental analysis and identification through medical findings, these comparisons were conducted using rigorous scientific procedures, ensuring their validity, reliability, and reproducibility [21,22].

Further advancements are anticipated as automated matching and retrieval systems become more widely available and increasingly feasible for identification and forensic authentication purposes [12,23,24,25]. These systems are already utilized to assist in the identification and search for tattoos on living individuals [26,27]. Although their application in the matching and identification of postmortem tattoos demonstrates significant potential, the authors hope it will become the focus of increasing research efforts.

4. Conclusions

The authors suggest that the comparative examination of PM and AM tattoos should be more widely adopted as an identification method. While no single method is perfect for identifying a human body, each case requires the application of the most suitable approach. In many instances, tattoo confrontation can be the elected method, offering a practical and effective solution for identification. This approach relies on a procedure that identifies all available coincident points while ensuring that no features are excluded. It follows principles like those used in comparative dental record exams, as well as the radiographic and tomographic imaging of frontal sinuses.

Tattoo comparison can be sufficient for human identification, particularly when tattoos display a high degree of complexity. Conducted in a methodical, reproducible, and rigorous manner, tattoo analysis can emphasize unique and distinguishable features, demonstrating its potential as a reliable tool in forensic investigations. Adherence to established forensic standards further enhances its credibility and utility, particularly in DVI contexts and other forensic applications, making it a sufficient resource for identification purposes.