The process of reporting and coding the cause of death is a vital part of health systems [1, 2]. It provides essential information for monitoring and understanding the main causes of death in a population, which is why it is primordial in the context of epidemiological surveillance. In Brazil, this process is conducted through the Mortality Information System (SIM) of the Brazilian Ministry of Health (MoH). The SIM was established in the 1970s to maintain and manage the country’s death records [3, 4]. Within Brazil’s National Health Surveillance Policy [5], the SIM platform aims to deliver mortality data across all sectors of the health system and to continuously develop health indicators for the entire country [6–8].

The Death Certificate (DC) is the foundational document for the SIM platform and is mandatory for certifying deaths in Brazil [9, 10]. The DC contains crucial information necessary for formulating and planning public health actions based on the specific needs of the population. To accomplish these objectives, physicians must be dedicated and committed to maintaining the accuracy, integrity, and reliability of the information provided in the DC. In Brazil, physicians are solely responsible for completing and ensuring the accuracy of the information contained in the death certificate.

An internationally standardized list of categorized codes for diseases and medical conditions is employed to generate the information in the medical certificate of the DC. In addition, instructions on medical certification and guidelines for classifying and determining the cause of death are utilized [11, 12]. Based on this information and the guidelines, the underlying cause of death is coded and entered into the Mortality Information System. The codes are based on international guidelines utilizing the International Classification of Diseases (ICD) [13–15]. Presently, Brazil uses the tenth version of the classification (ICD-10) to code and select the underlying cause of death, defined as the “disease or injury that initiated a succession of events and ended in death or, in the case of accidents or violence, the circumstances thereof” [16].

In Brazil, the coding of the underlying cause of death is manually performed by coders. These are professionals with diverse backgrounds working in State or Municipal health secretariats. Referred to as “coders of the underlying cause of death,” they assign an ICD-10 code to each medical description and apply international selection and/or modification rules to classify the underlying cause of death. Coders play a critical role in ensuring the generation of accurate mortality data nationwide. Such data is crucial for planning, monitoring, and evaluating policies directed at safeguarding and improving public health [17–19].

This activity requires logical reasoning, knowledge, and skills specific to the use of the three volumes of the ICD-10. Moreover, it involves studying the causes for assigning a specific code to each diagnosis described by the physician on the certificate. Selection and/or modification rules are then applied to derive the underlying cause of death [20]. To this end, the professionals are equipped with the ICD-10 volumes, specific local protocols, and the Underlying Cause Selector (SCB), a module included in the Mortality Information System [21]. As of 2024, approximately 4,000 coders were operating in Brazil, registered with the Ministry of Health [22].

However, errors and inaccuracies have been frequently observed in the completion of death certificates in Brazil. These issues compromise the classification of deaths, resulting in the loss of critical data for the health system and hindering the development of effective policy actions in this sector [20, 23, 24]. Moreover, the shortage of qualified professionals to accurately perform coding tasks has also been observed. This may result in errors in selecting the correct codes, resulting in inaccurate cause-of-death information. Additional issues, such as high coder turnover, limited training provision, insufficient supervision, fragmentation, and duplication of information also hinder the generation of mortality data [25].

Notably, the process of reporting the cause of death on the SIM platform is generally not integrated with other health information systems, which hampers data exchange among different institutions [26, 27]. This compromises the quality and accuracy of mortality statistics in Brazil, further exacerbated by data incompleteness. Consequently, it becomes more difficult to effectively analyze and monitor mortality trends [27–29].

To address these shortcomings, the Brazilian MoH adopted a strategy to enhance professional capacity by promoting training courses on the underlying cause of death coding. Originally, the courses lasted 80 h of in-person instruction. Subsequently, the courses were delivered directly at the State Health Secretariats, with the workload reduced to 64 h and then to 40 h due to challenges in excusing professionals from their work duties to participate. Since 2016, the course has been offered in partnership with IFRN/UFRN in a blended format, with a 64 h workload. Due to the COVID-19 pandemic, the course was adapted to a fully online format while maintaining the same workload. In 2023, the course was redesigned into a self-instructional format, encouraging shared responsibility among the States for continuous professional training [22].

In this context, it is important to emphasize the pivotal role of data generated during the coding process in shaping and implementing public health policies, thereby ensuring the accuracy of population mortality statistics [30]. According to Valentim et al. [31], massive health education through technological mediation constitutes an impactful approach in various scenarios, capable of influencing and contributing to the promotion of public health policies. Especially in countries like Brazil—with a continental size of 8,515,767.049 km2, a population exceeding 220 million, and over 570,000 physicians distributed nationwide—knowledge dissemination strategies are key to achieving public health policy goals more quickly and effectively [32–34].

Similarly, Moreira Teixeira et al. [18] assert that massive open education through Massive Open Online Courses (MOOCs) can be an effective and sustainable solution for the continuous training of health professionals. Studies on digital learning effectiveness, such as the meta-analysis by Means et al. [35], often indicate that online learning can be as or more effective than traditional methods, provided there is adequate attention to instructional design and human-technology interaction (HTI) [36]. Furthermore, considering the typical challenges and evaluation models of MOOCs, which frequently address issues like low completion rates [37], it is crucial to recognize their potential when specifically applied to continuous professional development (CPD). As such, MOOCs are tools that can contribute to the training required to improve the quality of cause-of-death data input into the Mortality Information System. This, in turn, enhances the understanding of epidemiological trends and the impact of health policies in Brazil.

These deficiencies in the process of cause-of-death reporting and coding in Brazil negatively impact data quality and hamper the ability to monitor and understand the leading causes of death at a national level—an obstacle to the formulation, evaluation, and monitoring of public health policies. For instance, one-third of deaths in Brazil are reported with causes that are not useful for public health analysis, commonly referred to as “garbage codes” (GC) [38].

This scenario reinforces the case for investing in adequate training for health professionals through health education—based interventions, aimed at improving data entry and the completion of death certificates issued by the Brazilian State. Recognizing the difficulties and challenges involved, the Brazilian MoH offered the “Training Course on Underlying Cause-of-Death Coding – ICD-10” to 1,533 professionals nationwide. The aim was to enhance the quality of information recorded in the Mortality Information System.

Given this phenomenon, particularly in the context of the course offered by the Brazilian MoH, this data report aimed to demographically characterize a group of course participants and their perceptions regarding the “Training Course on Underlying Cause-of-Death Coding – ICD-10.” Since this study resulted in a data report, its primary contribution lies in characterizing and sharing the data with health professionals and scholars interested in this topic of global interest, not exclusively within Brazil.

The data from the online learning module of the “Training Course on Underlying Cause-of-Death Coding – ICD-10” was analyzed using descriptive statistics, which allows for the fundamental exploration and description of the relevant properties and characteristics of the dataset. The principal resources used were: measures of absolute and relative frequencies; measures of central tendency, mean and median; measures of dispersion or spread, observing the standard deviation (std).

Drawing on the analysis model proposed by Valentim et al. [31], Equation 1 was employed to normalize the data related to enrollments and populations in each Brazilian region. Therefore, the coefficient “rate” represents the proportion of each region analyzed (normalized values per 100,000 population). Equation 1 was primarily used to construct Figure 1. The population figures for Brazil’s regions, as well as the country’s population, were elicited from the 2022 demographic census published by the Brazilian Institute of Geography and Statistics (IBGE) [43]. The following notations have been defined for the variables contained in Equation 1:rate=(xtargetxpop)nfactor(1)were:

rate: variable storing the coefficient for indicators proportional to each region or Brazil;

The online learning module of the “Training Course on Underlying Cause-of-Death Coding – ICD-10,” available on AVASUS https://avasus.ufrn.br/local/avasplugin/cursos/curso.php?id=231, recorded 1,533 enrollments, of which 1,023 identified as female (66.73%) and 315 as male (20.55%). Moreover, 195 course participants did not report gender (12.72%). Considering the overall group, 1,254 (81.8%) participants were eligible for certification in the online module of the course. As of the day the data was exported, a group of 279 participants (18.2%) had not completed 100% of the online activities.

In Figure 1, it can be seen that the course had enrolments from all five Brazilian regions. The largest number of participants was from the Northeast region, totaling 788 people (51.4%). Despite the prioritisation of placements for local authorities or administrative regions with the highest incidence of mortality coding throughout the year, the significant number of enrolments in the Northeast region can be partly explained by the historical deficit in training opportunities in previous years, as well as by the support provided by local public administration.

The online learning module of the “Training Course on Underlying Cause-of-Death Coding – ICD-10” was evaluated by a group of 927 course participants (60.47%). On a satisfaction scale ranging from 1 to 5, the course obtained an average score of 4.8 (median = 5 and standard deviation = 0.56). The same group also evaluated the course by submitting written comments in a specific field on AVASUS. Using the analysis strategy based on word clouds for visual text representation, Figure 2a was created from the comments written by course participants.

From the perspective of the course participants’ professional profile, a group of 1,135 professionals (74.04%) were found to have different occupations. Figure 2b shows the ten most frequent occupations among course participants working. At the top of the list, it is clear that the occupation “nurse” is the most frequent among course participants, with a total of 442 (38.94%) professionals in this field. Following this, also in the field of nursing, are nursing technicians and nursing aides, totaling 123 (10.84%) participants in this occupation.