Desarrollo y aplicación de un sistema de inferencia difusa de mandani como sistema de ayuda a la decisión en farmacoeconomía

Resum

Pharmacoeconomics tries to find the most effective treatments (cost / effective) between the various possibilities available. This principle of efficiency is the same as that collected various documents of the Committee of Medical Ethics, so it should not pose an ethical dilemma between the implementation of pharmacoeconomic analysis - quality - and healthcare practice The term "Fuzzy Logic" has been translated in various forms (lógica borrosa, difusa, blanda). It refers to the part of the logic and mathematics that works with variables with a high degree of uncertainty. Uses expressions that are not fully false are not entirely true, but have a degree of certainty between these two extremes. It has a powerful theoretical development and numerous applications in fields as varied as technology, computer science, social science, economic analysis, and of course the biological and health sciences. In decision-making in health not only have to consider the huge scientific production that occurs in health sciences but also economic aspects. However, these aspects are not always adequately addressed in the scientific literature and may, not be extrapolated to our environment. Therefore, decisions in the field of health is usually performed with a high degree of uncertainty. Fuzzy logic and fuzzy inference systems are of great importance in this field. The objective of this study is to develop a support system pharmacoeconomic decision applicable to a health care provider based on fuzzy logic - specifically according to the method of Mamdami - and check its performance in several representative cases of the actual conditions in which decisions must be made. It has designed an IF-THEN matrix with four fuzzy variables to determine the most appropriate choice: i) Probability of successful treatment, technology or process to be analyzed; ii) Cost per case of success; iii) Cost per failure iv) price reduction (by negotiating with suppliers, handling of medication, patient group or other methods) Accordance the value of these fuzzy variables were defined three possible linguistic values (High (H), medium (M) or low (L)) to express the appropriateness of the choice of appropriate option. The combinations of possible values of these variables generates 81 possible decision rules. These decision rules are grouped into a new variable that we call diffuse ranking with seven possible values: very favorable, unfavorable, slightly unfavorable, neutral, favorable, slightly favorable, very favorable. Various simulations of the matrix was carried out to verify the internal consistency of the results. After verifying internal consistency, was applied to several real situations, prototype of case decisions on healthcare provider, for verifying external coherence of the system. The designed system has internal and external consistency, and can be useful for making pharmacoeconomic decisions for a healthcare providers.