The importance of health policies based on data science

Despite the vast amount of data on the public health system, governments rarely uses these resources to develop and improve public policies

Fabio Kon

Brazil’s Public Health System (SUS) produces a huge amount of data from daily interactions involving various types of public health services. Specific official information systems are used to collect, store, and aggregate data on births (SINASC) and deaths (SIM), which are used by all health services in the country, including in the private sector.

Others seek to monitor what happens in between, such as consultations with health professionals at the outpatient (SIA) and inpatient (SIH) levels, notifications of health problems of interest (SINAN), and more.

Unfortunately, the vast majority of this data is rarely used by federal, state, and municipal governments to create public policies.

When they are used, they are typically based on long, tedious printed (or PDF) reports containing basic descriptive statistics, using twentieth-century tools.

Data available in real-time are generally ignored, with very rare and honorable exceptions, e.g., recent systems motivated by the COVID-19 pandemic.

The SUS is a great achievement and source of pride for the Brazilian people, but it struggles with inertia, waste, and very little encouragement of innovation.

Accessible indicators

Imagine if it were possible to create an interactive dashboard that presented data about patient flow between municipalities, identifying health regions that are not capable of adequately serving the local population.

What if this dashboard could indicate the therapeutic journeys of patients seeking healthcare from different medical specialties and professionals? What if it could be used every year (or every few months) to plan the budget for healthcare in the next period?

Another dashboard could show the geographical distribution of diseases that result in hospital admissions in a major city such as São Paulo or in the state of Rio de Janeiro.

This would make it possible to analyze the time taken for patients to travel from their homes to the hospital, and to see how this distance varies according to medical specialties, and how the result compares with other Brazilian cities and states. The planning of investments in new healthcare infrastructure could then be guided by this type of information.

What if we had machine learning algorithms capable of predicting dengue outbreaks one, two, or three months in advance, so that awareness campaigns and anti-mosquito measures could start earlier, and healthcare services could better prepare themselves to treat the public?

What if we had a better understanding of the seasonality of diseases that cause deaths at different times of the year, following the specific seasons of the various regions of the country?

What if a system could emit weekly alerts to municipal health departments on likely increases in certain diseases in the coming weeks?

Sounds like a dream, does it not?

But all the possibilities above have two things in common:

None of them are used by the public authorities in Brazil.
All of them are easy to implement, with working prototypes already developed in recent years using very little investment.

Projects and funding

I am referring to InterSCity, a project funded by the São Paulo State Research Foundation (FAPESP), the Brazilian National Council for Scientific and Technological Development (CNPq), and the Brazilian Federal Agency for Support and Evaluation of Graduate Education (CAPES) that worked on a series of health projects at the Department of Computer Science of the Institute of Mathematics and Statistics at the University of São Paulo (IME-USP).

In collaboration with health professionals and scientists, computing and data science undergraduate and master’s students created tools that proved that none of the above is a dream—it is all within reach.

You can visit the InterSCity website yourself to find out more about this health research, interact with the dashboards, and read the articles describing the studies. If you know how to program, you can download the code behind the systems and contribute to their development (they are all open-source software).

Of course, these programs serve only as demonstrations of what could be achieved and they would require improvements before they could be used on a large scale. But they are already functional systems, using real data from the SUS, and could already provide insights for public authorities.

If computer science students are capable of developing such rich and potentially wide-reaching systems, imagine what a well-organized professional team could create.

If the SUS were to intelligently invest a small portion of its budget in technological innovation based on high-level computer science and data science, it could create highly innovative software systems capable of radically transforming the quality of information available to public health managers in the country.

Would information technology (IT) solve all of the SUS’s problems?

Of course not! There are several difficulties of other natures. But better use of IT could bring major benefits, improving public policies and how the system operates at all three levels of government.

It is high time for the Ministry of Health and the state and municipal health departments in Brazil to start seriously investing in innovative projects that use computer science and data science in a modern way.

This would allow for the construction of interactive and visual online tools that can provide strategic information to help guide the planning and operation of the SUS.

It no longer makes sense to focus almost all management efforts on decades-old technologies.

So, what is the solution? There is no silver bullet. However, we should start by recognizing that investment is directly related to the reward.

Effective investment

For example, smartly investing 1% of the health budget in innovation for evidence-based management would most likely lead to savings of much more than 1%.

This would allow us to invest existing resources more effectively, offering a higher-quality service for the same cost. But what does investing smartly mean?

It is crucial for pilot projects—which can later be replicated nationwide—to be conducted by capable teams, with specialists from the various fields of health working together with computer scientists, software architects and software engineers, statisticians, economists, and designers.

It is thus essential for governments to partner with universities, research institutes, startups, and social organizations.

Any systems we develop must be open-source software, built using open industry standards and based on data standardized by the Ministry of Health.

This would enable solutions to be easily replicated across the entire country at a very low cost.

There are many opportunities afforded by science and IT in the twenty-first century, and we are paying a high cost for not taking advantage of them.

The Brazilian public deserves more than the quality of health services they are currently offered.

Fabio Kon is a computer science professor at the Institute of Mathematics and Statistics of the University of São Paulo (IME-USP) and head of the National Institute of Science and Technology of the Internet of the Future for Smart Cities (InterSCity.org).

Opinion articles do not necessarily reflect the views of Science Arena or Hospital Israelita Albert Einstein.

* This article may be republished online under the CC-BY-NC-ND Creative Commons license.
The text must not be edited and the author(s) and source (Science Arena) must be credited.