Around the Globe: Latin America
Improving Use of Epidemiological Data to Serve Patients Better: A Colombian Initiative
Julián Camilo Avendaño Gallo
Amalia Catalina Lehmann Oliveros
Juan David Cano Sierra
Andrés Ramírez
MSD, Colombia
E

pidemiological data are important assets, which support government controls and measures to ensure appropriate public health interventions. The Integrated System of Information of Social Protection (SISPRO, its Spanish acronym) provides an online free-access system of databases with information on healthcare quality, assurance, financing, and other aspects of healthcare services in Colombia. This information management system, achieved through digital technologies, has provided multiple benefits for patients and the healthcare system in Colombia.

As the 1970s came to a close, the Colombian Ministry of Health identified the need for standardized, timely information regarding health services provided to the Colombian population. Coming at a time of digital revolution and transformation, this need led to the development of an information system that could provide the broad social benefits of integrated health and health information to all Colombians. Creation of SISPRO began in 1979, when the Ministry of Health defined standards for responsibilities, classification of information, periodicity of reports, clarity, and destinations for epidemiological information collected by healthcare providers in Colombia.

However, it was not until 2001 that the ministry defined the system’s design, regulations, implementation, and information management procedures.

In 2007, the acronym SISPRO officially appeared for the first time in a ministerial act, and the Ministry of Health, solely responsible for administering the system information, formally implemented and deployed SISPRO online in 2011. Currently, the ministry is working to improve the quality of reported data and data-processing tools and to encourage more widespread use of the platform as a public asset.

Use and Limitations of the SISPRO Database

Data in SISPRO are stored in dynamic cubes within a unique online warehouse, which allows users to access information in the form of registries, reports, comparative analyses, graphs, maps, and other resources (Figure 1). These cubes are multidimensional representations of related data sources. Data in SISPRO are obtained from a variety of public sources, including the Ministry of Health and Social Protection, the National Administrative Department of Statistics (DANE), the National Institute of Health (INS), the Administrative Office of General Health Social Security System Resources (ADRES), and the National Health Superintendence.

Graph showing data flow and analysis in SISPRO
Figure 1. Demonstration of data flow and analysis in SISPRO. Developed by MSD Colombia based on SISPRO Modelo Conceptual del SGD.
The SISPRO database allows users, even those with no expertise in data analysis, to perform their own basic data queries. Raw data are analyzed by directly accessing information cubes, which allow the use of queries to export requested data.

Data analyses on the information processed and stored in the epidemiologic data module of SISPRO support monitoring the Colombian healthcare system’s efficacy in providing services to patients and were designed (among other reasons) to improve public health policies of concern to healthcare providers, national and regional regulators, researchers, and other stakeholder communities. An additional benefit for the healthcare system in Colombia has been a better understanding of specific diseases and their prevalence in the country as reflected by the number of research articles developed using SISPRO as their main data source (some examples here, here, and here).

While SISPRO has been designed to make data easily accessible, the sheer amount of data stored in recent years means that a simple search can take 10 minutes to an hour. This challenge reflects a need to generate novel data-mining tools to more efficiently handle and organize these data. These tools will allow for better understanding and use of data published by the Colombian Ministry of Health to support the medical, research, and marketing strategies of various companies in the country. These data analyses will be critical to ongoing efforts to raise awareness of the value of public data in Colombia and Latin America.

A bill has been presented to the Colombian Congress that includes the creation of a new system (Unique and Integrated Public System of Health Information, or SPUIS, its Spanish acronym), which expects to integrate the most recent big data, digital image processing, and other artificial intelligence tools to improve the use of economic, scientific, clinical, technical, and administrative data to support better decision-making and transparency of public policies and resource management in the country. This bill recognizes the limitations of SISPRO and the opportunities for improvement in terms of data validation and analysis and expects that the government will take a more active role in ensuring these conditions for the data available to the government and the public.

Role of Industry in Supporting Access to Health Data Power

The pharmaceutical industry is an important ally in identifying ways to overcome data access challenges and in developing better regional information hubs. For Colombia, integration of technology and the acknowledgment of data as a primary asset offer significant opportunities. The nation can address specific needs of its population and advance toward a public policy model that integrates all relevant variables and stakeholders and prioritizes the well-being of its people.

After considering the complexities of data analysis and collection identified in other countries implementing similar strategies, the local subsidiaries of a multinational pharmaceutical company in Colombia and Mexico collaborated to develop a data-mining tool to address some of these concerns.

This data-mining tool organizes and visually presents data on a disease of interest and then analyzes its patterns in the country. It allows the ingestion, integration, and consumption of data for all cross-functional areas that could potentially find value in these analyses for strategy development. Although the government was not involved in the process, the tool was developed by the company based on the public data available in SISPRO and is currently available for internal use only in Colombia.

Comprehensive dashboards of patient discharges, diagnoses, healthcare services provided, prescriptions, and procedures (at the national, state/department, city, and healthcare facility levels) are only some of the essential features of this tool. This analysis describes a patient’s journey by identifying medications, laboratory tests, diagnostic imaging, specialist referrals, numbers of visits to emergency rooms, lengths of hospital stays, etc. It also provides the number of patients being treated by a specific institution.

One example application is within medical affairs organizations. After identifying priorities for the organization, the search begins with the tool (Figure 2). The initial query outcomes can be extracted in MS Excel files for subsequent transformation and publication through any visual analytics tools (e.g., Power BI) to improve analysis of and decisions based upon these data.

Graph of of data flow and outcomes of the data-mining tool
Figure 2. Example of data flow and outcomes of the data-mining tool. Developed by MSD Colombia.
Through this comprehensive approach, local teams can build a solid medical and research strategy that evaluates how the healthcare system is behaving. It also helps to prioritize institutions that require educational resources and those in which basic research, clinical trials, or real-world evidence studies will benefit the most significant number of patients.

Although the accomplishments of the Colombian government in developing a centralized epidemiological data repository are significant, there is still an unmet need at the national level to ensure adequate infrastructure and maintenance of tools for data analysis and visualization. Meeting this need has become an imperative due to the massive amount of data and data sources available today. Patient-centered analysis instruments, such as the one described in this article, will allow pharmaceutical companies and healthcare providers to collaborate more efficiently with the government in identifying and implementing suitable control and preventive measures to ensure the population’s health and safety. Using this tool to understand the current status of the healthcare system and the patient’s journey through it represents just one step toward supporting a more equitable and adequate healthcare system in Colombia through solid, objective data analyses.

Acknowledgments:
The authors thank Marta Ines Agudelo Murcia for her support on the logistics and leadership crucial to this article, and Charles McNair for his editorial support.