Business Intelligence in supporting strategic health management: an experience report

Business Intelligence no apoio à gestão estratégica em saúde: um relato de experiência

Business Intelligence en apoyo a la gestión estratégica en salud: relato de experiencia

Esta obra está bajo una Licencia Creative Commons Atribución 4.0 Internacional.

In recent decades, due to technological advances, data accumulation in all fields of knowledge increased. Such information, converted into practical knowledge, promote major disruptions in public or private companies and, thus, are considered a valuable resource¹.

However, considering the costs of collection and storage, the data obtained must have relevant information to be extracted. Thus, after obtaining the data, it is necessary to classify them so hypotheses can be propose and tested¹. Moreover, a given information may have an expiration date and loses its value if certain actions are not implemented at that specific time, which may result in losses of many kinds².

Since 1991, the Brazilian Ministry of Health has an Informatics Department of the Brazilian Unified Health System (DATASUS) and has developed more than 200 systems³. Currently, the amount of data stored is greater than the ability to turn them into information useful for decision making. Moreover, the data are obtained in the most diverse sources and formats, making it difficult their dynamic management and analysis⁴.

Nevertheless, government health agencies in Brazil suffer from difficulties in the systematic analysis of data due to several obstacles, such as a shortage of human resources and budgets. Thus, the agencies often fail to deliver timely responses to public health demands⁵.

To transform data into information, the U.S. company Microsoft launched in 2015 a business intelligence service called Power BI (Business Intelligence), which can be used in several areas of knowledge to extract, integrate, and centralize data from many sources and formats.^6,7

This study aimed to report the use experience of Power BI by utilizing data from a State Department of Health of Western Amazon, Brazil.

This is an experience report of the use of the Microsoft Power BI program, based on an experimental approach of the database of tuberculosis cases in the state of Acre, from 2010 to 2020. The databases were provided by the State Health Department of Acre (SESACRE) in spreadsheets. For the incidence calculation, we used population estimates provided by DATASUS³.

To enable the control panel of Power BI, also known as dashboard, all data were universalized to the same standard to enable communication between them. For variables to respond together, connections were made sequentially between them. This stage in the process is complex because any mistaken connection can generate incongruous data. At the end of this procedure, the program provides a link to access the data.

This type of study that involves data from the public domain without the identification of individuals does not require approval from the Research Ethics Committee.

Power BI offers a simple and intuitive interface. To illustrate the interface, we created a link (http://tiny.cc/tbacre), in which is possible to try the tool by using a computer or mobile phone.

On the home screen, there is an access list to several pre-programmed databases. The initial step is choosing an available database. As an example, we chose the first option – “Sociodemographic Profile” (Figure 1).

Figure 1
Home screen with access to different databases.

On the “Sociodemographic Profile” display screen there is a menu on the left with pre-inserted filters: year, mesoregion, microregion, and cities (Figure 2). The second step is selecting the desired filters to view the data. By the graphs in the example, we observed different sociodemographic profiles, according to age group, gender, year, and location. In the lower left corner of this screen, there is an option to return to the home screen that allows access to the other previously programmed bases, with new options of filters and graphics.

Figure 2
Display screen of the “Sociodemographic Profile” database.

The programming of Power BI may require intermediate knowledge in computer science and/or specialized technician. However, by overcoming the difficulty of programming, we observed a customizable program that greatly assists database understanding and automatically updates as new data is entered.

The concept of Business Intelligence (BI) emerged in 1958 and it was applied to any type of organization, such as industrial, scientific, or governmental⁷. In the 1990s, BI was defined as a generic term that addresses concepts and methods to improve evidence-based decision-making⁸.

Power BI synchronizes services such as spreadsheets, websites, social networks, and other sources. According to Microsoft, the service is an “unified and scalable platform for enterprise and self-service BI, which is easy to use and helps gaining deeper insights into the data.”⁶

In short, using BI makes analysis faster and more practical, since it is unnecessary to search for data spread across several programs and platforms. Thus, it is easier to understand the data and to draw more assertive strategies.^2,4

Among other advantages, the program features case mapping, real-time information sharing, and remote working options⁹. Power BI can be useful in scientific communication by facilitating the visualization of data in scientific journals. The “COVID-19 Panel”, widely viewed since 2020, is based on this tool and allowed managers agility in combating the disease and optimizing resources.^9,10

A disadvantage of Power BI is the difficulty of programming the databases. However, training, even virtual training, can solve this issue. Once this programming is performed, an unqualified person is able to operate the tool, giving autonomy and agility to the work teams.

We recommend using Power BI especially in agencies that require agile evidence-based decision making.