Menu
Scientific studies that translate data into disease prevention, treatment personalization, healthcare resource optimization, risk identification, among many other issues that yield tangible answers through data analysis via scientific research.
Artificial intelligence can make significant strides in health prediction within the societal context, and our pursuit of more knowledge and innovation is constant.
The Brazilian Association of Health Insurance Providers (Associação Brasileira de Planos de Saúde – Abramge) reports that, while direct costs associated with patient care increased 19.3% between the years of 2019 and 2022, insurance claims increased a total of 81.8% during the same period, accounting for a total of R$10.9 billion.
The significant increase in total claims in recent years has shifted the priority of healthcare providers toward detecting and preventing fraud to stabilize expenses and ensure the sustainability of its operations. Fraud prevention is also essential to limit price increases to final consumers.
Recent innovations in artificial intelligence have allowed substantial progress to be made in detecting fraudulent activity in healthcare. We are currently conducting research to develop fraud detection algorithms that use structured data to identify anomalies and alert about possible fraudulent claims. These algorithms can detect anomalies considering one or a combination of characteristics of the claim submitted, including historical information on both patients and healthcare providers.
The algorithm performance in detecting when individual cases are not consistent to the usual pattern observed for health insurance claims in the data is proportional to the quantity and quality of the information provided for the training of the model.
Whenever fraud cases have been previously identified and validated, this additional information is used to improve the algorithm’s performance and identify insurance claims with a similar pattern to confirmed fraud cases.
Instituto IA Saúde conducts research in healthcare applying artificial intelligence to large volumes of data. Our main goal is supporting decision making and the implementation of solutions to improve the efficiency and access to health in Brazil.
The broad health data available allows us to address complex research questions and guide decisions in a variety of contexts. Data analysis and machine learning algorithms are powerful tools to maximize resource allocation efficiency and aid management in health facilities. For example, machine learning can be used to plan the availability and allocation of medical professionals and hospital beds, to manage patient wait times, and even to monitor high-cost and complex cases.
Artificial intelligence can also be applied to practical clinical problems, by identifying patient risk and readmission probability, or even supporting doctors with diagnosis.
