O. Zaballa, A. Pérez, J. A. Lozano
Understanding the disease evolution and accurately estimating associated healthcare costs are essential for effective healthcare management and treatment planning. With the availability of large amounts of information in electronic health records, there is an opportunity to gain insights into the progression of a disease and its relationship with associated costs. Existing methods have limitations when applied to real-world patient economic data due to the need for extensive domain knowledge. This work presents a probabilistic generative model that takes into account the dynamics of a disease for predicting healthcare costs. Specifically, the proposed method classifies treatments based on their temporal progression and estimates the associated costs. Experimental results in synthetic data demonstrate the learning procedure’s ability to recover the underlying generative model. We use real-world data to evaluate the effectiveness of our approach in accurately estimating healthcare costs.
Keywords: Probabilistic generative model, disease progression, healthcare cost estimation
Scheduled
Data Analysis
November 8, 2023 5:20 PM
HC3: Canónigos Room 3