Rostec's subsidiary taught artificial intelligence to detect violations in providing medical aid

The national center of informatization (the NCI, is included into Rostec Group), developed for TFCMI a system based on artificial intelligence for quality control of treatment and identification of violations at delivery of health care, reported TAdviser in Rostec in September, 2019. The Moscow region became the pilot region where a system was implemented.

The developed system under the name "Expert" identifies cases of delivery of health care on the compulsory health insurances having signs of violations or errors – in drawing up documentation, expenditure of means of compulsory health insurance and treatment of patients.

The Federal Compulsory Health Insurance Fund regularly carries out inspections which conditionally are divided into technology, carried out in the automated mode, and expert. Formatno-logical control and medico-economic control belong to technology. To expert – medico-economic examination and examination of quality of the provided medical care.

The developed analysis system of cases of delivery of health care on compulsory health insurance based on AI is going to be replicated free of charge in other regions (a photo - iz.ru)

Checks of formatno-logical control provide correctness of the entered data. It is impossible to enter number instead of a name of the patient, or instead of date to enter character values. Checks of medico-economic control prevent introduction of incorrect and contradictory data on the provided medical care. For example, the male patient cannot write delivery of health care on the gynecology profile. Also at a stage of medico-economic control the correctness of input of a financial component of the register of the account is controlled. Use of set of technology checks provides correct formation of account registers for further payment of the provided medical care.

Part of checks in the compulsory health insurance system is performed in the automated mode: it is formatno-logical and medico-economic control. The artificial intelligence is connected at a stage of medico-economic examination and expertize of quality of medical care which was carried out completely earlier by experts of insurance companies and specialists of territorial departments of Federal Compulsory Health Insurance Fund. A system reveals suspicious, potentially defective cases, and further experts deal with them.

According to "hints" of a system based on artificial intelligence experts pointwise check mismatch of the medical services shown for payment, to entries in primary medical and accounting and reporting documentation. Besides, the medical record is analyzed, the clinical record then the conclusion about that is drawn is studied, treatment was how correctly and qualitatively carried out. Such approach accelerates process of checks and allows to process more data, note in Rostec.

According to Federal Compulsory Health Insurance Fund, on training of the neuronet used in this solution three months left. All account registers and acts of examinations for 2017 on one of the subjects of the Russian Federation were provided for a pilot project.

The associate director of territorial Federal Compulsory Health Insurance Fund of the Moscow region Vadim Lukashov provides statistics that in the region about 5 million cases of delivery of health care monthly are registered. Experts check about 5-8% from them. A system helps to select carefully for them those cases where the error probability is highest, Lukashov explained.

Rostec says that as a result of use of a system considerably increased speed, the accuracy, efficiency of checks and volume of the checked data. The number of the revealed errors and violations grew more than by 3 times.

Development was conducted within tender which cost is about 30 million rubles. Rostec told TAdviser that third-party software products in the created system are not used.

Group of companies also reported that the Point of view company and the Center of artificial intelligence and digital economy of the Yaroslavl state university of P.G. Demidov act as partners of the project. "Point of view" is based in Yaroslavl and is an industrial partner of YarGU. About 10 years are engaged only in algorithms based on machine learning. "Point of view" won the open competition held by NCI.

Earlier the vice-chairman of Federal Compulsory Health Insurance Fund Elena Suchkova told^[1]that the developed system based on artificial intelligence is going to be replicated free of charge in regions that everyone the compulsory health insurances territorial funds of the subjects of the Russian Federation could be connected and use the same analytical opportunities. Rostec told TAdviser that their solution is made so that regions could implement it independently.

Notes