Sc.D. Knigavko V. G., Ph.D. Radzishevskaya E. B., Levchenko T. V., Solodovnikov A. S.

Kharkov national medical university,

Ukraine

The significance of bases of decision making theory for doctors training at medical universities

Having subject to own professional peculiarities every specialist in a varying degree meets the difficulties of decision making. However there are few fields for human activity where it is possible to apply decision making process as ordinary and basic professional skill. Medical activity being sequence of diagnosing and treatment is the typical representative of such field. Methods of decision making are multipurpose and universal though their successful application substantially depends on professional qualification of the specialist who must have exact understanding of peculiarities of the system he studies and must know how to lay down the task. Uninterrupted development of informational technologies and improvement of technological and industrial base put the problem of decision making theory in medical field to advanced level. This fact become apparent in all the fields of medical science relating to the activity of medical establishments, for instance, in form of creation and implementation of specialized medical programs, telemedicine technologies and projects of integrated automatization for hospitals or clinics [1]. Actually all these are subject of investigation for medical informatics (MI) that is the unbiased reason to include decision making theory to MI course for medical university students.

Kharkov National Medical University (KhNMU) with support of Department of Biomedical Engineering of Kharkiv National University of Radio Electronics (KhNURE) have developed and implemented (since 2012) the lecture «Decision making in medicine» [2]. It is argued by fact that the doctor meets the common problem in different medical tasks (patient data acquisition, diagnostics, and treatment tactics). This problem is the choice of decision method.

Although the requirements to the accuracy and reliability of the diagnosis (or diagnosis validity in other words) become more and more strict from year to year, medicine does correspond semistructured branch of knowledge that causes serious difficulties in formalization approach to decision making. On the one hand the hypothesis or final decision based on standard disease manifestation arises during medical observation, on the other hand it appears only after detailed observation. It may be a cause of adjustment of the diagnostic analysis chain and sometimes of fundamental changes depending on the obtained results. Quickness of decision-making is affected by doctor qualification and disease features in concrete patient.

Getting the information from the world around them or analyzing the occured situation doctors always refer to own knowledge and experience accumulated for years. But human memory feature is that the total amount of reproduced information is always less than amount of perceived information while technical device allows displaying and receiving information at the same size. In comparison with mechanical memorizing and representing ability human is often weaker than machine, so long as his memory during the fixation, storing and reproducing information sometimes loses it. The speed of correct solving of posed problems is very important for specialists, especially in risk conditions when timely measures and prompt actions let the doctor save patient's life. That is why in a number of cases it is necessary to use special instruments supporting the process of decision making [3]. Such instrument does not relieve human of responsibility for consequences of wrong actions. It is needed to increase the precision and speed of task solving and speed of searching for required information.

During recent years, personal computers and software are tremendously improved. Software companies create lots of applications that support and realize decision making process. This gives ability to verify solution methods by developing medical and biological processes model [4].

Students must get an explanation while studying the general theory of decision making that the term «make a decision» from the point of view of math means to make a choice from number of alternatives. Despite the variety of existed problems, there are several main stages of decision procedure: objective definition; creation of a set of alternatives; forming of criteria to compare alternatives; finding optimal solution from set of feasible solutions (optimization task). Further students become acquainted with basic terminology of the theory: «decision-making consequence», «system of preferences», «objective function», «decision», «decision-making person», and after that the classification of decision-making tasks is discussed. Students study the methods of decision-making for determined tasks, tasks in conditions of stochastic indeterminacy, and in risk conditions [5]. Taking in account the complexity of considered problem all discussed questions are supplied by examples from medical field or by situational tasks. For instance, students must understand to which category it is possible to lable the task of psychic sanity detection in an offender who is suspected of grave crime, and who tries to mitigate penalty by mental disorder simulation.  With relation to medicine there is a topical question of tasks decomposition in static and dynamic types as well.

Medical diagnosis being the kind of decision-making problems has a probabilistic nature and connected with hypothesis verification. The decision making person during the task solving tries to check his supposition or hypothesis. To do it correctly students are merged into statistical section in order to consider the notions «hypothesis», «significance level», «power of test», «test sensitivity», «test specificity», and «errors of 1-st and 2-nd type» as well.

In the end, students must pay attention to the tasks of decision-making problem which occur when goal-setting is difficult to be formulated by appropriate mathematical tool or when it is necessary to assemble group of specialists of different branches of knowledge. There are special approaches, technics, and methods for such cases. To start the process we need to define problem domain and to detect factors which influence on the task solution and also to choose methods allowing us to formulate the task so that we can get a correct way of solving. The next step is to obtain the expression connecting goal with tools of its achievement. All of these are embodied by mathematical models with different criteria like criteria of functioning, , efficiency criteria, objective function and so on.

Task can be always solved if it is possible to define connection between aim and methods. It can be done easily if we know the law of connection. Otherwise we must to find another way of mapping the problem domain to the solving technic. This law may be based on statistical researches or functional dependencies. If even this way is fallacious then the specialist should choose or develop theory containing range of statements and rules that help him to get a conception and take it as a base for construction of decision-making process. When theory does not exist we need in raised hypothesis for creation of imitating models to investigate possible ways of research.

System representation, methods of system analysis are used to help formulating the objective on a tight timetable as well as to help the specialist to analyze goals, to define applied tools and to select required information that characterizes decision-making conditions and influences on the choice of criteria and restrictions. System analysis is useful way to get an objective with methods connected to it.

Thereby, in spite of complexity the considered problem the future doctor should apply theory of decision making as theoretical basis for his profession to improve his work and to be more precise. Besides that, this theory is just an additional fragment of Artificial Intelligence and Expert Systems course, for it can be considered like introduction to the larger theme expanding medical specialists knowledge.

Literature:

1. Макаров С.С., Жидкова Т.З., Косенко Е.Ю., М.В.Зиборов, Финаев В.И. Моделирование и информационное обеспечение медицинских учреждений. – М.: МГУП, 2005. – 210 с.

2. Современные подходы к преподаванию медицинской информатики в медицинских вузах / Высоцкая Е.В., Книгавко В.Г., Полещук А.П., Радзишевская Е.Б., Солодовников А.С. // Сучасні здобутки медичної інформатики [електронний ресурс]: збірка матеріалів науково-практичної конференції з міжнародною участю (електронне видання), 13-14 червня 2013 року, м. Київ. – К.: НМАПО, 2013. – С. 113-117

3. Данильченко Т. В. Система підтримки прийняття рішень діагностики та моніторингу / Т. В. Данильченко // Моделювання та інформаційні системи в економіці. – 2015. – № 91. – С. 224-235.

4. Чудна Р. В. Математичні моделі прийняття рішень в медичній реабілітації / Р. В. Чудна, О. А. Владимиров // Збірник наукових праць співробітників НМАПО ім. П. Л. Шупика. – 2013. – Вип. 22(3). – С. 124-131.

5. Методы и модели принятия решений в условиях многокритериальности и неопределенности: монография / Э. Г. Петров, Н. А. Брынза, Л. В. Колесник, О. А. Писклакова; под ред. Э. Г. Петрова. – Херсон: Гринь Д. С., 2014. – 192 с.