OBJECTIVES

: Access to medical care in remote regions of Russia is limited. The sparsely populated villages are served by mobile teams of doctors once a month, or not served at all. Telemedicine technology allows for the availability of medical care remotely. The aim of the work is to evaluate the effectiveness and safety of using AI in the telemedicine system MeDiCase for automated collection of anamnesis and complaints, the formation of preliminary diagnostic hypotheses and the transmission of information to the doctor for decision-making.

METHODS

: The survey of residents of remote settlements of the Republic of Karelia, Kamchatka Krai, Oryol Oblast of Russia was conducted. Average age 59,9 ± 15,7 years. A tree-like binary questionnaire provides a collection of complaints and anamnesis, after a preliminary analysis of the AI data, the doctor receives a list of possible diagnoses with their justification. The doctor decides whether to route the patient or correct the previously prescribed treatment.

RESULTS

: Surveyed 1247 patients. The diagnosis of arterial hypertension was revealed in 69.2% of the population, previously it was made in 57.26%. Broncho-obstructive syndrome was detected in 7.61%, the diagnosis was made earlier in 3.61%. COPD was detected in 14.9% versus 5.93%. Diabetes - 17%, previously - 11.1%. The decisions of the doctor and AI on routing coincided in 92.1% of cases, and in 1.6% AI estimated less urgent routing. System safety is 98.4%

CONCLUSIONS

: Artificial intelligence in the telemedicine system allows to increase the availability of medical care in remote locations, provides high-quality collection of anamnesis and complaints, and make an informed decision on patient routing.