|
KMID : 1146320190070020112
|
|
Journal of Health Technology Assessment
2019 Volume.7 No. 2 p.112 ~ p.118
|
|
|
Concordance Assessment and Satisfaction of Medical Professionals for the Artificial Intelligence Watson
|
|
Lee Kyoung-A
Kim Chan-Hee
Baek Jeong-Heum
Shim Seon-Jin
Ahn Sung-Min
Ahn Hee-Kyung
Lee Uhn
Lee Seon-Heui
|
|
|
Abstract
|
|
|
Objectives: The purpose of this prospective study was to examine the concordance rate of diagnosis and treatment between Watson for Oncology(WFO) and multidisciplinary tumor board, and toevaluate the satisfaction of medical professionals about WFO.
Methods: The subject of this studywas 126 patients with cancer and 54 medical professionals who participate in WFO multidisciplinarycare at Gachon University Gil Medical Center in Korea. Concordance rate between the WFO and themultidisciplinary tumor board was measured by the concurrence between the WFO presentation andthe final decision of the medical staffs. Satisfaction of medical professionals was measured with a questionnaire that identifies satisfaction, intention of use, and the strengths and weaknesses of WFO.
Results: In 121 cases (96.0%), the recommendation and consideration presented by WFO were consistent with the final treatment method. The overall satisfaction for WFO was 6.74±2.08 out of 10. Thestrength of WFO that medical staffs thought was found to be hospital publicity (4.11±0.78) and patient compliance increase (3.98±0.64). The weakness of WFO was that it did not consider ethnic andcultural differences (3.63±0.98) and that it did not reflect the health insurance cost in Korea (3.61±0.96).
Conclusion: WFO has a high concordance rate when deciding on treatment options, whilethere are some limitations in the reflection of national racial, regional, cultural and environmental differences and the application of patients in specific situations. It is expected to be used as a basic data fordeveloping Korean artificial intelligence Watson model by grasping the needs and improvement of localized WFO.
|
|
|
KEYWORD
|
|
|
Artificial intelligence, Multidisciplinary care, Analysis effect
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|