Experienced Probabilities Increase Understanding of Diagnostic Test Results in Younger and Older Adults
Experienced Probabilities Increase Understanding of Diagnostic Test Results in Younger and Older Adults
复制标题
DOI：
10.1177/0272989x17691954
复制
发表时间：
2017-08-01
影响因子：
3.6
通讯作者：
Spaniol, Julia
中科院分区：
文献类型：
--
作者：
Armstrong, Bonnie;Spaniol, Julia
背景随着年龄的增长，医疗检查的频率增加。解释医学测试的结果涉及后验概率的估计，例如阳性预测值（PPV）和阴性预测值（NPV）。外行和专家通常都不善于估计后验概率时，相关的统计数据进行沟通。目前的研究探讨了经验格式是否会改善后验概率判断年轻人和老年人，相对于描述格式。法80名年轻人（年龄17-34岁）和80名老年人（年龄65-87岁）完成了一项实验任务，其中通过描述或经验提供了关于2种虚构疾病的医学筛查试验的信息。描述性格式的参与者阅读包含统计信息的段落，而经验格式的参与者观看代表性病例的幻灯片，说明疾病的相对频率以及阳性和阴性测试结果的相对频率。结果相对于描述格式，年轻人和老年人在经验格式中做出了更准确的后验概率估计。在描述性格式中，PPV被高估，NPV被低估。无论格式类型，参与者报告说，他们更愿意依靠医生代表他们作出医疗决定，而不是自己。讨论这些发现表明贝叶斯推断中的预防经验差距，并且它们提出了增强年轻和老年患者医疗风险沟通的可能途径。
Background. With advancing age, the frequency of medical screening increases. Interpreting the results of medical tests involves estimation of posterior probabilities such as positive predictive values (PPVs) and negative predictive values (NPVs). Both laypeople and experts are typically poor at estimating posterior probabilities when the relevant statistics are communicated descriptively. The current study examined whether an experience format would improve posterior probability judgments in younger and older adults, relative to a description format. Method. Eighty younger (ages 17-34 y) and 80 older adults (ages 65-87 y) completed an experimental task in which information about medical screening tests for 2 fictitious diseases was presented either through description or experience. Participants in the descriptive format read a passage containing statistical information, whereas participants in the experience format viewed a slideshow of representative cases that illustrated the relative frequency of the disease as well as the relative frequency of positive and negative test results. Results. Both younger and older adults made more accurate posterior probability estimates in the experience format, relative to the description format. In the descriptive format, PPVs were overestimated and NPVs were underestimated. Regardless of format type, participants reported that they would prefer to rely on a physician to make medical decisions on their behalf compared with themselves. Discussion. These findings are indicative of a description-experience gap in Bayesian inference, and they suggest possible avenues for enhancing medical risk communication for both younger and older patients.