Data Mining in HIV-AIDS Surveillance System : Application to Portuguese Data

Area of Science:

• Public Health Surveillance
• Infectious Disease Epidemiology
• Health Informatics

Background:

• Human Immunodeficiency Virus (HIV) remains a significant global health challenge, necessitating robust surveillance systems.
• Effective monitoring of HIV-Acquired Immunodeficiency Syndrome (AIDS) cases is crucial for healthcare planning and policy development.
• Administrative delays in reporting diagnosed cases complicate efficient and effective HIV surveillance.

Purpose of the Study:

• To identify key factors contributing to administrative delays in reporting HIV-AIDS cases within the Portuguese surveillance system.
• To evaluate the performance of machine learning models in analyzing reporting delays.

Main Methods:

• Employed machine learning algorithms including Multilayer Perceptron (MLP), Naive Bayesian (NB), Support Vector Machines (SVM), and K-Nearest Neighbors (KNN).
• Analyzed data from the Portuguese HIV-AIDS surveillance system to assess reporting delays.

Main Results:

• Multilayer Perceptron (MLP) achieved the highest classification accuracy, precision, and recall.
• Results indicated generally homogeneous administrative and clinical practices across reporting entities.
• The study identified specific factors contributing to administrative delays in the HIV-AIDS case reporting process.

Conclusions:

• Machine learning models, particularly MLP, are effective tools for analyzing HIV-AIDS surveillance data and identifying reporting delays.
• The findings suggest a need for standardized, nationwide guidelines to reduce reporting delays.
• Transversal implementation of these guidelines across all stakeholders is recommended to improve the efficiency of the surveillance system.