The expert predicted that accreditation will become more difficult when using AI for… » Medvestnik


The Ministry of Health is assessing the capabilities of artificial intelligence (AI) methods in creating (generating) new test tasks and situational cases for the accreditation of medical and pharmaceutical workers in accordance with the specification of the assessment tool (specialty, discipline, nosology, labor functions), the deputy told RIA Novosti minister of health . This can make the procedure more difficult, requiring doctors to spend more time educating themselves.

Creating a high-quality assessment tool for accreditation is a labor-intensive process that includes not only development, but also mandatory quality assessment, and the involvement of a large number of professionals from among the teaching staff of medical universities, scientific organizations and representatives of the medical community. According to Semenova, such a toolkit has been created, but banks of assessment tools need constant updating, especially in the context of the fact that clinical recommendations and equipment are regularly updated, and the regulatory framework is changing.

“Order of the Ministry of Health of Russia No. 709n dated October 28, 2022 “On approval of the Regulations on the accreditation of specialists” contains such a concept as a “Unified database of assessment tools,” the contents of which are now being filled in manually by the Methodological Center for Accreditation of Specialists,” commented the editor-in-chief of Semyonova’s statements on the Pravo-med portal .ru . — I believe that this is a rather labor-intensive professional activity that requires highly qualified specialists in many specialties. I don’t know how paid it is.”

In addition, the Unified Database of Assessment Tools requires regular updating so that those accredited do not relax and follow the path of memorizing the correct answers, the expert added. In his opinion, if AI begins to generate non-repetitive test tasks and situational cases, they will become more difficult, and accordingly, the labor costs of doctors and nurses to prepare for accreditation will increase. In these conditions, the need to allocate working time for self-education becomes urgent, Panov stated.

At the same time, in his opinion, the market will be able to offer certain services based on the same AI, which will help those accredited in correctly performing tests and cases. “If now, for a fee, it is already possible to create a portfolio with minimal labor costs, then why won’t something similar appear as a kind of counterbalance to the efforts of the Ministry of Health to update the Unified Database,” the lawyer suggested.

The prerequisites for this already exist. The domestic medical neural network iMed scored 82% in a complex accreditation test of doctors, passing the test with 5 points, the developer said. As part of the experiment, specialists from the National Association of Medical and Pharmaceutical Organizations invited the neural network to pass a test developed by the methodological center for accreditation of specialists of the First Moscow State Medical University named after. THEM. Sechenov for assessment of competencies and accreditation of the healthcare system. It consists of 80 questions, each of which has four possible answers, of which only one is correct. As a result of the experiment, the iMed neural network correctly answered 66 questions.



Source link