IDENTIFY & REDUCE COGNITIVE BIASES OF DECISION AIDS

认识到病人的价值观在医疗决策中的重要性，研究人员已经开始开发决策辅助工具，告知病人与现有治疗方案相关的各种危害和益处的可能性。决策辅助工具在帮助患者做出知情决定方面具有很大的潜力。但它们也可能对患者的决定产生偏见。我们的总体目标是确定当患者使用决策辅助工具和测试方法来减少这些偏差时可能发生的认知偏差。我们已经确定了四种与辅助决策设计特别相关的认知偏差：框架效应：人们的治疗选择取决于治疗收益是否以生存或死亡率为框架。患者评价的不当影响：患者评价对治疗的可取性或不可取性影响治疗选择，独立于治疗效果的统计信息。增加第三种治疗方案后的偏好逆转：在增加第三种治疗方案后，更喜欢治疗方案a而不是B的人可能更喜欢B。结果信息的过度影响：治疗结果信息的增加可能会过度影响患者的选择。我们将采用书面和计算机调查的方式开发辅助决策材料，并对公众进行调查，其中三分之一的样本有癌症相关经历。我们的目标不是对面临困难临床情况的患者进行辅助决策测试，而是开发能够产生辅助决策的方法——制定辅助决策的一般原则，这些原则将增加患者对其治疗选择的理解，而不会引入认知偏见。

In recognition of the importance of patient values in medical decisions, researchers have begun developing decision aids that inform patients about the probability of various harms and benefits associated with available treatment alternatives. Decision aids have great potential to help patients make informed decisions. But they could also bias patients' decisions. Our overall objective is to identify cognitive biases that are likely to occur when patients use decision aids and test ways to reduce these biases. We have identified four cognitive biases that are especially relevant to the design of decision aids: Framing effects: People's treatment choices depend on whether treatment benefits are framed in terms of survival or mortality. Undue influence of patient testimonials: Patient testimonials regarding the desirability or undesirability of a treatment influence treatment choices independent of statistical information about treatment effectiveness. Preference reversal after addition of a third treatment option: A person who prefers treatment option A over B may prefer B after the addition of a third option. Undue influence of outcome information: The addition of information about treatment outcomes may unduly influence patients' choices. We will develop decision aid materials using written and computerized surveys administered to the general public, with one-third of the sample having cancer-related experience. Our goal is not to test decision aids on patients facing difficult clinical situations but, instead, to develop methods by which such decision aids can be produced - general principles for the creation of decision aids that will increase patients' understanding of their treatment choices without introducing cognitive bias.