人工智慧技術導入對醫療照護服務設計的影響

Abstract

隨著全球人口數持續成長，且人口結構高齡化，醫療照護服務的成本與效率的相關議題持續受到關注。由於病患對於健康、醫療資訊的接收管道越來越多且暢通，病患的自主意識抬頭，也促使醫療照護服務轉為以「病患為中心」的模式。本研究即根據「以病患為中心」為出發點，探討在人工智慧導入醫療照護產業下，對於服務流程的影響與轉變，輔以服務設計的架構切入AI智慧醫療發展下，對於醫療照護流程中各利害關係人的影響。 本研究以過去文獻所提出的健康照護服務設計規劃模式作為切入的基本架構，探討AI技術導入之後，會如何影響其中的投入、服務傳遞系統與產出三個階段。在透過與醫療照護服務流程中的利害關係人—服務使用者（病患）與服務提供者（醫師）針對這些議題進行結構式的深度訪談後，將受訪者針對相關議題的想法與建議進一步歸納、分析並收斂至到本研究的基本架構上。歸納分析的結果顯示，AI導入可使部分常規工作與病患分流，將有助於有限的醫護的人力做最有效的投入，投入面向則會轉而著重於醫病互動與信任關係，因為AI智慧醫療下，醫護人員與病患的互動關係是更需要被強調的服務核心，因此，AI智慧醫療的服務設計不應以完全取代醫護專業為核心目的，而是讓醫護專業輔以AI化。此外，是否納入健保與知覺到較高的效益，是民眾認同AI醫療的關鍵。健保支付制度與現行法規則是台灣目前導入AI智慧醫療的主要限制。 雖然國外已開始有研究從病患的觀點探討AI智慧醫療的效益與風險，但國內相關的研究論述目前仍較少。然而，要形塑健全的醫療照護服務系統，需要有更多的研究從病患端的角度出發，才能使科技的導入更符合病患的需求。

As the global population continues to grow and the population structure is aging, issues related to the cost and efficiency of healthcare services continue to draw lots of attention. Patients’ unimpeded receiving of medical and health information and sense of autonomy has led to the transformation of healthcare services into a patient-centered model. Based on the patient-centered service model, this study investigated the impact of application of artificial intelligence (AI) on service process in healthcare industry. With the framework of service design, changes to stakeholders (patients and clinicians) during the development of AI in healthcare were discussed. Based on the healthcare service design planning model proposed by the previous study, we focused on how the three phases of input, service delivery system and output will be affected after the introduction of AI technologies. Stakeholders’ consideration and recommendation on these issues were collected through structured in-depth interviews. The analysis showed that the introduction of AI was believed to help share the loading of clinical routine work and patient care, which would result in more effective deployments and inputs of limited medical staff. Since patient-clinician interactions were still considered as the core of healthcare service, the released labor inputs should be transferred to human-human interactions and relationships. Therefore, the role of AI in the healthcare service design should be assistive instead of completely replacing medical professionals. In addition, patients’ recognition to the use of AI were largely determined by whether higher value was perceived and whether AI-based services were included in health insurance. The health insurance system and the regulations related to medical care and patient data were considered as main restrictions on the introduction of AI in healthcare services in Taiwan. Studies on the benefits and risks of AI in healthcare from patients’ views have been published in other countries, whereas related research in Taiwan is still rare. We believe that accounting for patients’ perspectives will make the introduction of AI technologies more in line with patients’ real needs.