Authors: Ioannis Karvelis, Augoustinos Dimitras, Maria Adamopoulou, Konstantinos Koutsogiannis

Title: Intelligent Control Tools in Healthcare Financial Management Karvelis I., Dimitras A., Adamopoulou M., Koutsogiannis K.

Abstract

Due to the size, multiple specificities, and rapid changes in the field of service provision, healthcare economics, and healthcare service management, there has been significant growth and implementation in recent years. This is because governments in most countries are ardently seeking ways to implement cost containment policies and increase the efficiency of healthcare service delivery. It is evident that the interest in healthcare service policy and economics, apart from its undeniable scientific and research significance, also has political, social, and economic importance. This is because it is related to the major issue of the distribution of healthcare resources and, consequently, the control of the healthcare sector.

Expanding the discussion among scientists and healthcare professionals to enrich knowledge and techniques in pharmaceutical policy and economics, as a prerequisite for optimizing choices in determining healthcare priorities, distributing and efficiently using scarce resources, and, above all, ensuring the quality and effectiveness of healthcare, is now among the top priorities. Every year, the pharmaceutical expenditure of insurance funds increases by €0.5 billion, mainly due to the lack of precise measurement, monitoring, and evaluation of the services provided. Additionally, a significant factor is the lack of organization in the entire healthcare system, which hinders cost control, as there is no accurate recording of the volume of consumption of healthcare products and services. The absence of a compensation system based on economic efficiency criteria and overpricing is another significant factor in the increase in pharmaceutical expenditure.

It has been shown that computational approaches that can learn from data and make predictions based on existing data are derived from machine learning (ML). These are algorithms that operate by constructing models from experimental data, models that allow predictions based on data and also lead to drawing conclusions, thus documenting decision-making systems and highlighting correlations through trend learning displayed in the data.

In the stages of machine learning, categorization or clustering is included, techniques that organize data into groups, predefined for categorization and resulting from the "training" process for the above processes. Also, representation methods that incorporate fuzzy logic or uncertainty, as usually occurs in times of crisis, can incorporate more than one or two artificial intelligence technologies. Until the combined statistical methods lead to mathematical models as a result of comparisons or prediction correlations with significant success, as proposed in the international literature. The creation of a hybrid approach that will propose the most appropriate approach each time based on success metrics could provide alternative solutions to the significant problem of predicting pharmaceutical costs.