智慧醫材臨床應用之法律責任

The Legal Liability of the Clinical Application of Medical Devices Based on Artificial Intelligence

陳鋕雄(Chih-Hsiung Chen)

《領導護理》 21卷3期 (2020/09) Pp. 1-15

https://doi.org/10.29494/LN.202009_21(3).0001

摘要

基於人工智慧科技的智慧醫材在近年短期內成為熱門應用領域，逐步進入醫療臨床場域。由於智慧醫材特性與一般醫材之差異，使其使用上的法律責任風險有所差別。本文依兩種分類方式區分智慧醫材。首先是依是否需經政府查驗登記，以及是否需要經過臨床試驗，而區分智慧醫材的法律責任。愈經過臨床試驗及查驗登記的醫材，上市後的產品責任風險較低。但純以軟體為主的醫材，有可能被認定為服務而非產品。其次，本文依使用上是否需要醫護人員介入而區分智慧醫材，說明輔助醫護人員臨床決策的醫材，仍由醫護人員負最終法律責任。而智慧醫材若有瑕疵，由於不容易與其他產品進行比較，不容易成立設計瑕疵，較容易成立的是標示瑕疵。對於新興科技的不完美產生的風險，醫療機構與醫護人員應注意相關的買賣及服務契約條款，以分散或排除相關的法律風險。

關鍵字

人工智慧 ； 智慧醫材 ； 產品責任 ； 輔助臨床診斷醫療器材 ； 標示瑕疵

並列摘要

Medical devices based on the technology of artificial intelligence have been a hot topic in recent years and have gradually adapted in the medical clinical practice. Because of the different characteristics of these smart devices, the application risks of legal liability are different for these devices. This article divides smart devices by two classification methods. The first one classifies smart devices depending on the requirements of governmental premarket approvals and clinical trials. Devices go through the paths of clinical trials and premarket approval will have less risks of product liability. However, the software-alone devices would be deemed as service rather than products. Second, this article classifies smart devices depending on the intervention of medical personnel in the use of the devices. Medical personnel is responsible for malpractice caused by smart devices that assist medical therapists for decision-making. Smart devices are not easy to be proofed for design defect because of the difficulty to compare with other products. Label defect is easier to be proofed. For the risks caused by the imperfection of innovative technology, medical institutions and healthcare practitioners should pay attention to the clauses of sale or service contracts in order to distribute or exclude the relevant legal risks.

並列關鍵字

artificial intelligence ； smart devices ； product liability ； clinical decision-making assisting device ； label defect

參考文獻

Alemzadeh, H., Raman, J., Leveson, N., Kalbarczyk, Z. & Iyer, R. K. (2016). Adverse events in robotic surgery: A retrospective study of 14 years of FDA data. Plos One, 11(4), 1-20 . doi: 10.1371/journal.pone.0151470

Arkes, H. R., Shaffer, V. A., & Medow, M. A. (2008). The influence of a physician’s use of a diagnostic decision aid on the malpractice verdicts of mock jurors. Medical Decision Making, 28(2), 201–208. doi: 10.1177/0272989X07313280

Tupasel, A., & Nucci, E. D. (2020). Concordance as evidence in the Watson for oncology decision‑support system. AI & Society: Knowledge, Culture and Communication. https://doi.org/10.1007/s00146-020-00945-9

王若樸（2019 年 4 月 17 日）．醫療影像 AI 實例：臺北榮總：AI 只花30 秒就能自動從數百張 MRI 影像找出腫瘤！準確率達 95%．取自 https://www.ithome.com. tw/news/129883

中國醫藥大學院附設醫院（2020 年 8月 16 日）．乳房超音波腫瘤 AI判讀門診系統演示．取自https://www.cmuh.cmu. edu.tw/CMUHVideo/Detail?no=316