荷兰普通人群对人工智能皮肤癌检测应用的支付意愿调查
本研究嵌入随机对照试验,于2022至2023年邀请荷兰11621名普通人群参与者,通过问卷评估其对三种场景(单独AI、AI联合远程皮肤科医生、面诊皮肤科医生)的支付意愿(WTP)。1953人完成问卷,结果显示AI联合远程皮肤科医生的WTP中位数显著高于单独AI,但低于面诊皮肤科医生;年龄增大是唯一与所有场景WTP正相关的特征。定性分析揭示相似WTP可源于不同理由,WTP并不总反映感知重要性。
要点
- 普通人群对AI联合远程皮肤科医生的支付意愿中位数为20欧元,高于单独AI的5欧元。
- 面诊皮肤科医生的支付意愿中位数为50欧元,显著高于两种AI场景。
- 年龄增大是唯一与所有场景支付意愿正相关的参与者特征。
核心结论
- 相比单独AI应用,个体对涉及皮肤科医生的服务赋予更高货币价值。
- 支付意愿不一定反映感知重要性,相似估值可源于不同理由。
研究设计与样本
- 基于随机对照试验的问卷调查,邀请荷兰11621名参与者,最终1953人完成问卷(完整应答率16.8%)。
- 参与者平均年龄58.4岁(SD 13.1),57.6%为女性,95.4%自报浅色皮肤类型,49.7%高教育水平,79.4%有既往应用使用经验。
主要发现
- AI联合远程皮肤科医生的支付意愿中位数为20欧元(IQR 0-45),显著高于单独AI的5欧元(IQR 0-20),p < 0.001。
- 面诊皮肤科医生的支付意愿中位数为50欧元(IQR 22-75),显著高于AI联合远程皮肤科医生,p < 0.001。
- 年龄增大是唯一与所有场景支付意愿正相关的参与者特征。
- 定性分析显示低支付意愿可能反映缺乏信任、感知无附加价值或认为服务应免费。
临床怎么用
- 在推广AI皮肤癌检测应用时,可考虑结合远程皮肤科医生参与以提高用户支付意愿。
- 了解用户对AI应用的信任和感知价值有助于制定合理定价策略。
局限性
- 完整应答率较低(16.8%),可能影响样本代表性。
- 支付意愿基于假设场景,可能与实际支付行为存在差异。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
BACKGROUND: Rising skin cancer incidence increases pressure on the healthcare system. Artificial intelligence (AI)-based mobile health applications (mHealth apps) are being offered to laypersons from the general population to be used on their own initiative. It is unclear if, and how much, they are willing to pay (WTP) for these apps and how they decide on their valuation. OBJECTIVES: The primary objective was to investigate individuals' relative WTP for AI-based mHealth apps for skin cancer risk detection. Secondary objectives were to explore the reasons underlying WTP using a mixed-methods approach. METHODS: This survey-based study embedded in a randomized controlled trial, invited 11,621 participants in the Netherlands between 29-08-2022 and 23-02-2023. Participants were offered a questionnaire about their experiences with an mHealth app, their WTP in three hypothetical scenarios (AI alone, AI with teledermatologist involvement, and physical dermatologist consultation), and could provide open-text explanations. Associations between participant characteristics and WTP were assessed using regression analyses, and qualitative analysis was applied to open-text responses to capture underlying reasoning. RESULTS: A total of 1,953 participants completed the questionnaire (16.8% complete response rate; mean age 58.4 years [SD 13.1]; 57.6% female; 95.4% self-reported light skin type; 49.7% high educational level; 79.4% prior app use). Median WTP was significantly higher for AI combined with a teledermatologist compared to AI alone (€20 [IQR 0-45] vs €5 [IQR 0-20], p < 0.001), but lower than for a physical dermatologist consultation (€50 [IQR 22-75], p < 0.001). Increasing age was the only participant characteristic positively associated with WTP across all scenarios. Qualitative analyses revealed that similar valuations could range from pragmatic to fundamental and reflect different rationales, namely low WTP could indicate lack of trust or a perceived lack of added value, but also the belief that services should be freely accessible. CONCLUSIONS: Individuals place the highest monetary value on dermatologist involvement compared to AI-based apps alone. However, this study suggests that WTP does not always reflect perceived importance, as similar valuations could arise from different underlying rationales. Future studies could explore which features of these apps users value most and how to improve engagement.
引用
- Smak Gregoor AM, Wehrens R, van der Vall N, et al. Willingness to pay among individuals from the Dutch general population for artificial intelligence based apps for skin cancer detection: a survey-based mixed-methods study. Br J Dermatol. 2026. PMID: 42563543. doi:10.1093/bjd/ljag327
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。