Knowledge of the causes of mortality is an integral part of the foundation of public health policies, strategies and programs [1]. Death certification plays a central role in the collection of demographic and epidemiological data worldwide. Not only does it identify the causes of mortality, it is also a fundamental tool for guiding public health policies.

For several years now, high-income countries, notably in Europe and North America, have been deploying electronic death certification systems to improve the efficiency, accuracy and speed of the information collected [2,3]. For example, France introduced such a system in 2007 with the CertDc project to gradually replace traditional paper forms [3]. In Portugal, a similar system has been in place since 2013, having made a 100% successful transition from paper to electronic certification after 2 years.

The advantages of such a system are multiple. Firstly, digitizing the process eliminates manual transcription errors, but also speeds up data processing, as shown by a French study on the contribution of electronic certification [4].

In Africa, where the civil registry system is still largely dominated by traditional methods, the need for modernization is pressing. The United Nations Economic Commission for Africa (UNECA) report on death registration highlights that of the 46 World Health Organization (WHO) member states in the African region, only 1 can provide high-quality cause-of-death data (Mauritius), and 3 others can provide low- or medium-quality data (ie, South Africa, Seychelles and Zimbabwe) [5]. Nevertheless, we noted a few rare initiatives, such as Burkina Faso’s implementation of an electronic death certificate at Bobo University Hospital [6].

On the other hand, Mali like many other African countries, faces similar challenges, notably insufficient coverage of the civil registry system, delays in the transmission of paper death certificates and frequent errors in data collection.

These aspects reduce the quality of national and international mortality statistics and limits their value for health planning and health policy development [7]. Mali is no exception, despite its high maternal, infant, child, infant-child and neonatal mortality rates [8]

In Mali, death certificates vary across health facilities and even within the same hospital. Some services lack a standard form, relying on handwritten notes. This inconsistency limits mortality data availability and accuracy, with records often incomplete, containing only basic details like name, sex, date of death, and a vague diagnosis (eg, cardio-respiratory arrest). Archiving is neither systematic nor organized, and copies are not always retained. Deaths are often tracked by marking “Deceased” or “DCD” in patient registers.

To overcome these problems, we need adapted and harmonized methods and tools that comply with certain norms and standards, and the advantage of using digital tools in our context is no longer in doubt [8-10].

The hospital “Le Luxembourg” in Bamako initiated a system to digitize and integrate the death certificate into its Health Information System (HIS) to improve mortality data management. Despite its technical capacity and nationwide patient intake, it faced challenges similar to other health facilities in collecting and using mortality data. Recording 556 deaths annually, the hospital wants to develop an electronic death certificate for better archiving and data use, as detailed in this implementation report.

The implementation aimed to improve the completeness and accuracy of mortality data, accelerate the timeliness of death reporting, and facilitate the downstream use of data to inform public health planning and research.

This project was part of a thesis internship, led by a student with support from the Center for Expertise and Research in Telemedicine and E-Health (CERTES), which maintains the HIS. Implemented at the hospital “Le Luxembourg,” it involved system analysis, requirement collection, and specifications drafting. Development took three months, carried out by two CERTES developers and the student.

For the successful implementation of the project, a team was established, composed of the head of the public health and medical informatics department of “Le Luxembourg” Hospital, the trainee, the developers of the Cinz@n (HIS), the users represented by doctors (ie, residents, general practitioners, and specialists), nurses from various departments, and the hospital registrar.

All users were involved from the design phase through interviews aimed at understanding their needs. Physicians participated in the validation of the medical component, while administrative staff contributed to the design of the printable certificate. Their involvement facilitated acceptance of the new process.

The project required no additional budget as the hospital already had the necessary infrastructure. The module is maintained internally by the hospital’s medical informatics department, with updates managed by a trained technical focal point. With maintenance limited to software support, the solution is sustainable without external funding. Its design also allows for easy deployment in other hospitals using the same HIS.

This project, integrated into routine hospital activities, aimed only to digitalize the death certificate and enable the secure storage and structured use of mortality data. No sensitive or identifiable patient information was used, and no direct interaction with patients or their relatives occurred. In accordance with local guidelines, no ethical approval was required. Data confidentiality and security were strictly maintained.

This manuscript complies with the implementation reporting guidelines [11].

After the design and development phases, we implemented a new model of death certificate at the hospital “Le Luxembourg,” which included an administrative (Multimedia Appendix 1) and medical (Multimedia Appendix 2) component. This multifunctional death management module (Multimedia Appendix 3) is now integrated into the HIS. Before implementation, from a methodological standpoint, we adopted an iterative approach, starting with an initial prototype tested in one hospital department, which was then improved through several short cycles, incorporating user feedback at each stage.

User feedback at the end of each training session and two months postimplementation indicated a high level of satisfaction. This was especially attributed to the ease of access to patient information and the automatic generation of certificates.

The project timeline is presented in Figure 2.

To access the death management module, users require a username and password. Key features include creating death certificates, listing deceased patients, viewing certificates awaiting prevalidation and validation, accessing archived certificates, statistics, and settings.

Users can search for a patient using an ID number, name, or phone number before issuing a death certificate.

The “Establish a Death Certificate” menu integrates both administrative and medical components. Administrative data (Multimedia Appendix 4) are automatically retrieved from the patient’s record. The medical section (Multimedia Appendix 5) also pulls relevant information, including reason for admission and comorbidities, with ICD-10 used for cause-of-death coding.

The validation process follows a structured approach. Automatic alerts remind doctors to correct pending certificates, and pop-ups confirm actions before validation. This validation process ensures data quality, with data being easily obtained and managed from the dedicated menu. Users can filter data and export reports in the Microsoft Excel format for further analysis.

A dedicated interface allows the civil registrar to access the administrative section of the certificate and process official death declarations. By searching the deceased’s ID number, the registrar records details of the declarant (ie, relative) before confirming the declaration. Once completed, the entry becomes inactive and validated.

Within three months of implementation, a total of 62 death certificates were recorded. Although the volume appears modest, the key value of the system lies in the structured storage and availability of comprehensive mortality data.

The success of the project was driven by strong leadership, particularly the involvement of key institutional figures such as the heads of the medical informatics department and the medical board. Integrating the module directly into the existing HIS, rather than creating a parallel system, greatly facilitated adoption. Initial resistance from users was overcome through training and support. The inclusion of ICD-10 coding improved data quality and encouraged broader use of the system. Academic clinicians engaged early, motivated by access to structured data for research. The availability of IT infrastructure also played a critical role. However, limited staff numbers in departments hindered real-time data entry, underscoring the need to consider workforce constraints in future implementations.

Our results emphasize the need to standardize mortality data collection. We developed Mali’s first electronic death certificate, aligning with WHO recommendations.

This improvement in data quantity and quality was found by Lefeuvre D et al [12] in a study on the quality of electronic and paper death certificates in France in 2010. Our results highlight challenges such as inconsistent certificates, missing forms, incomplete data, and confidentiality concerns.

The implementation of ICD-10 for coding causes of death improves data reliability and allows international comparability [13].

Reliable mortality data support effective health policies. Accurate coding aids in reducing maternal and child mortality, while the new database improves data archiving and accessibility.

Electronic data storage is undeniably better than traditional storage, optimizing record search and use [14,15]. Our structured approach incorporating an alert system and a multilevel correction and validation process, serves not only to secure data confidentiality but also to elevate the overall data quality, enabling greater precision and accuracy, all of which are part of data quality criteria [16].

Our work introduced many doctors to ICD-10, which is not currently part of medical education or patient records in Mali. For civil registrars, digital certificates and automatic searches improved death declarations. The statistics module in Cinz@n enables multicriteria searches and better data analysis.

From an implementation perspective, this work confirmed that successful digital health integration relies on strong institutional leadership, alignment with existing systems, and user engagement. Embedding the module within the current HIS facilitated adoption and minimized resistance, especially when paired with targeted training and ongoing support. The integration of ICD-10 coding improved data quality and encouraged broader use of the medical record. The involvement of academic clinicians, driven by their need for structured data, further enhanced project ownership. However, limited staffing constrained timely data entry, highlighting the need to consider workforce capacity in similar initiatives. Overall, the project illustrates that a user-centered, standards-based, and system-integrated approach can enhance the adoption and sustainability of digital health tools.

One limitation of the project is the absence of a formal long-term evaluation of its impact on the quality of mortality data

The lessons we have learned are

Following our work, we recommend: