UPDATE : Wednesday, August 5, 2020
HOME Policy
Radiology technologies using AI can get reimbursement: report 
  • By Song Soo-youn
  • Published 2019.01.10 13:10
  • Updated 2019.01.10 18:19
  • comments 0

A new guideline will allow hospitals to receive extra reimbursement for artificial intelligence (AI) in radiology if they helped improve treatment outcomes, the health insurance authorities said.

The Health Insurance Review and Assessment Service (HIRA) unveiled the final report of “Study on evaluation guidelines for AI-based medical technology in radiology,” submitted by the Korean Society of Radiology on Tuesday.

A physician explains about a medical image using IBM Watson, an AI-based supercomputer, at Gachon University Gil Medical Center. (Credit: Gachon University Gil Medical Center)

According to the report, most AI software programs that analyze radiology tests are “existing technologies.” However, particular AI technologies that offer new kind of information are subject to the HIRA’s review for new medical technology. AI that predicts a patient prognosis and a new imaging biomarker based on AI will be reimbursable, the report said.

The report noted that AI programs that merely improve physicians’ work efficiency would not be reimbursable and that the better work efficiency would be the reward itself.

In particular, the HIRA should verify the accuracy of the AI before providing the reimbursement and distinguish between what the AI does and does not do, the report said.

The report suggested HIRA providing extra reimbursement or indirect reward, setting up new reimbursement category, or recognizing reimbursement corresponding to the physician’s work.

The report also raised the need to indirectly reward the use of AI systems such as giving a certification to medical institutions, designating emergency medical institutions, and offering subsidies to evaluate medical service quality.

“Physicians recently unveiled a technology where an AI system immediately analyzes the emergency brain CT scans and place images expected to have a serious abnormality at the top of the waiting list for physicians to read them first,” the report said. “In such case, the patient ironically waits for a longer time, and it is irrational to demand additional cost for the patient who waited longer. Then, it would be difficult for HIRA to provide reimbursement for individual AI-based tests.”

Instead, HIRA would do better to reward the hospital’s overall improvement in emergency care, the report added.


<© Korea Biomedical Review, All rights reserved.>

Other articles by Song Soo-youn
iconMost viewed
Comments 0
Please leave the first comment.
Back to Top