Fuqua School of Business, Duke University, 100 Fuqua Drive, Durham, NC, 27708, USA.
Duke Margolis Center for Health Policy, Duke University, Durham, NC, 27708, USA.
J Behav Med. 2021 Feb;44(1):38-52. doi: 10.1007/s10865-020-00171-0. Epub 2020 Jul 28.
When discussing risks and benefits with cancer patients, physicians could focus on losses such as mortality rates and cancer recurrence or, alternatively, gains such as survival rates and curing cancer. Previous research has shown that the way health information is framed influences individuals' preferences and choices. We operationalized gain-loss framing as physicians' choice of words related to gains (cancer survival), or losses (cancer mortality). In an exploratory analysis, we investigated (a) whether physicians used gain or loss words as a function of their recommendation, (b) whether physicians' choice of words was associated with patients' treatment choices. We analyzed transcribed consultations with male patients who had intermediate-risk prostate cancer. Using an iterative process of gathering and evaluating words, we created gain- and loss-dictionaries. The loss-dictionary included words related to cancer death and cancer progression. The gain-dictionary included words related to survival and cure. Using Linguistic Inquiry and Word Count software, we calculated the number of words related to gains and losses in each transcript. We found that physicians who recommended immediate cancer treatment for prostate cancer (vs. active surveillance) used slightly fewer words related to losses and significantly fewer words related specifically to death from cancer. Further analysis showed that loss words were associated with the patient's choice of immediate cancer treatment. A novel method of automated text analysis showed that physicians' use of loss words was correlated with physicians' recommendations for cancer treatment versus active surveillance. Additionally, loss words in consultations were associated with patients' choice of cancer treatment.
当与癌症患者讨论风险和益处时，医生可以关注死亡率和癌症复发等损失，或者关注生存率和治愈癌症等收益。先前的研究表明，健康信息的呈现方式会影响个人的偏好和选择。我们将收益-损失框架定义为医生选择与收益（癌症生存）或损失（癌症死亡）相关的词语。在探索性分析中，我们调查了（a）医生是否根据其建议选择收益或损失词，（b）医生的用词选择是否与患者的治疗选择相关。我们分析了男性中危前列腺癌患者的转录咨询。使用收集和评估词语的迭代过程，我们创建了收益和损失词典。损失词典包括与癌症死亡和癌症进展相关的词语。收益词典包括与生存和治愈相关的词语。我们使用语言探究和词汇计数软件计算了每个转录本中与收益和损失相关的词语数量。我们发现，建议对前列腺癌进行立即癌症治疗（与主动监测相比）的医生使用的与损失相关的词语略少，与癌症死亡相关的词语明显更少。进一步的分析表明，损失词语与患者选择立即进行癌症治疗有关。一种新颖的自动文本分析方法表明，医生使用损失词语与他们建议进行癌症治疗而不是主动监测有关。此外，咨询中的损失词语与患者选择癌症治疗有关。
