IMPROVING THE ALGORITHM FOR MANAGING THE PROVISION OF MEDICAL SERVICES BASED ON COMMITTEE METHODS

Abstract

In matters of management in social and economic systems, medical issues are becoming increasingly important, and the processes of diagnosis of diseases are becoming the most important task of various health care institutions. Incorrect diagnosis of diseases leads not only to the fact that the patient's treatment becomes ineffective, and sometimes even harmful, but also leads to a significant increase in the costs associated with expensive procedures for further analysis and treatment of the patient. It is for this reason that it is necessary to improve the algorithm of the diagnostic procedure, including two qualitatively new blocks based on the developed methods of latent analysis procedures based on Committee and discriminant analysis. Methods of Committee and discriminant analysis as an auxiliary device are also effective and effective in solving the problem of diagnosing any diseases, including neurological and vascular diseases, as particularly dangerous and significantly impairing the standard of living of the population and leading to increased mortality, as well as worsens the statistical results of health care institutions. The improved algorithm of the diagnostic procedure proposed in the article is supplemented by the authors with two stages, which are added if the picture of symptoms is blurred and the preliminary diagnosis is not presented, or when several diagnoses fit the existing symptoms. This usually occurs due to conflicting, redundant information that corresponds to a variety of diseases. In this case, the method of differential diagnosis is used, which consists in re-interviewing, examining the patient, studying his medical history, prescribing additional clinical, functional and laboratory tests, which in some cases are excessive and even harmful to the patient, as well as increase the total cost of treating the disease, and usually increase the costs of the health care institution as a whole.. To resolve this situation, you can use discriminant analysis methods, which will allow you to find out the necessary diagnosis by assigning factors to a particular disease. However, often there is a situation related to the inconsistency of data that does not allow us to solve the problem of discriminant analysis using standard methods. In this case, you can use current mathematical methods, such as decision trees, the support vector method, and neural networks, but they are poorly interpreted and difficult to understand. therefore, we propose a method of analysis, which is improved by the authors in order to reduce the size of the problem, which helps reduce the number of factors necessary for the diagnosis of diseases. This method allows you to eliminate the inconsistency of data and unambiguously make a diagnosis, some features of which can later be identified by other standard clinical or instrumental methods. Thus, we can conclude that the use of methods of Committee and discriminant analysis in the diagnosis of diseases is not only an auxiliary, but also a mandatory attribute, in order to improve the quality of medical services. Moreover, the main methodological position of the application of these methods is to use the search for latent risk factors for diseases, which allows increasing the proportion of timely diagnosed patients. Also, do not forget about the provision of correct information collection and processing.