Optimization of modern medicine raises questions of improving the nomenclature of laboratory research and economic evaluation of the laboratory. The desire to reduce the cost of laboratory research turns into an increasing alienation of the laboratory from the clinical doctor and the patient. The decision to recognize the pathology and prescribe or change treatment measures in 26—30% of cases depends on the diagnostic information obtained in the laboratory. It should be noted that the main reason for the sharp increase in financial costs for the treatment of patients in recent years is the inability to appoint an adequate laboratory examination. There is a need to correct appointments because of the unfounded appointments. This correction can be considered as a «preanalytic validation». The search for criteria for the development of the «preanalytic validation» protocol has led to the identification of many unjustified laboratory assignments. As an example, the situation with the determination of glycated hemoglobin (HbA1c) in the laboratories of state medical institutions. According to our observations, unsubstantiated prescriptions account for more than a third of all HbA1c studies (36%). Of these: «screening», in almost healthy people — 25%, performed at single border values of glucose up to 6.1 mmol/l — 21%, which gave a normal result haphazard («random») — 12%, unscheduled, performed «for hospitalization» — 2%, performed «on the list» with a «broad» examination — 1%, not justified — 1%, etc. All these data allow us to say that currently the use of the HbA1c test is not quite adequate and requires adjustment by specialists who control the treatment and diagnostic process. You need to enter the typical template comments to focus the attention of the attending physician about the possibilities of diagnosis of a particular disease, as well as about the discrepancy between the appointments made for laboratory tests and the diagnosis made. «Machine learning», as an element of artificial intelligence in the work of medical information systems (for example, EMIAS), can program the optimal basic examination of patients in accordance with the diagnosis code of the «International classification of diseases of the 10-th revision». And the «preanalytic validation» built into the information system will be an additional filter for a large flow of assignments. Doctors of clinical specialties will be informed about new research methods, and the management of medical organizations will be provided with the necessary statistical information to solve the volume of laboratory examination of patients.