Conjoint Analysis: Overcoming Obstacles to Routine Formal Preference Assessment i

描述（由申请人提供）：一名患有局限性前列腺癌的男子面临着令人眼花缭乱的治疗选择。这些选择中的大多数提供了相似的生存机会，因此对任何特定个体的最佳选择至关重要地取决于他对各种临床和非临床治疗特征的偏好程度。帮助这些人确定他们的最佳治疗选择将减轻这种疾病对这些人的负担，这是美国国立卫生研究院使命的一部分。在研究中，衡量这些偏好的强度传统上是通过几种“效用评估”方法来实现的。不幸的是，这些现有的方法都有明显的局限性，从基本的有效性问题到对患者认知负担的实际问题，这些都对它们的临床应用构成了障碍。在设计新的消费产品时，市场营销研究者也面临着描述个人偏好的挑战，尽管是对商品和服务的偏好。在市场营销中，“联合分析”已经成为描述消费者偏好和预测消费者购买行为的一种强有力的方法。该方法简单，预测有效，可以解决现有偏好评估方法的局限性。我们相信联合分析有潜力改善临床环境中的决策。我们的主要假设是双重的。首先，我们假设，与单独的教育支持相比，基于分析的偏好评估和教育支持联合将提高前列腺癌患者的决策质量。其次，我们假设联合分析在测量前列腺癌男性偏好方面比传统的偏好评估方法，特别是时间权衡和评分量表方法具有更高的结构效度。为了验证这些假设，我们将进行两项随机对照研究。这将首先通过招募接受过前列腺癌治疗的男性并确定对他们重要的治疗属性来完成。基于这些数据，我们将利用已建立的软件平台开发基于联合分析和传统评估方法的偏好评估应用程序，用于随机试验。一项使用联合分析应用程序的随机试验将检查接受联合分析和教育干预的男性的决策质量指标，与仅接受教育干预的男性的决策质量进行比较。第二部分将使用文献中发表的指标来比较这些应用程序的标准有效性。这项工作的指导方针是最终部署即时护理决策辅助工具，以提高前列腺癌患者的决策质量。公共卫生相关性：本研究旨在提高前列腺癌个体化治疗的能力，提高前列腺癌患者的决策质量。我们将测试从消费者市场研究中提取的方法是否比现有的医疗保健研究方法更准确地识别个体患者的价值。我们将测试使用这种方法是否能提高患者做出的治疗决定的质量。

DESCRIPTION (provided by applicant): A man with localized prostate cancer faces a bewildering constellation of treatment options. Most of these options provide similar chances of survival, so the optimal choice for any particular individual is crucially dependant on the strength of his preferences for various clinical and non-clinical treatment characteristics. Helping these men identify their optimal treatment choice will reduce the burden of this illness on these men, part of the mission of the NIH. In research, measuring the strength of these preferences has traditionally been approached with several methods of "utility assessment." Unfortunately, each of these existing methods has significant limitations, ranging from basic concerns about validity to practical concerns regarding the cognitive burden placed on patients, which have presented barriers to their clinical use. Investigators in Marketing have also faced the challenge of characterizing individual preferences, albeit for goods and services, when designing new consumer products. In Marketing, "conjoint analysis" has emerged as a powerful methodology with which to characterize consumer preferences and to predict consumer purchasing behavior. Given its simplicity and predictive validity, this method may be a solution to limitations of existing preference assessment methods. We believe that conjoint analysis has potential to improve decisions in clinical settings. Our primary hypotheses are twofold. First, we hypothesize that conjoint analysis-based preference assessment and educational support will improve decision quality in men with prostate cancer as compared to educational support alone. Second, we hypothesize that conjoint analysis has superior construct validity in measuring the preferences of men with prostate cancer compared to traditional methods of preference assessment, specifically time trade-off and rating scale methods. To test these hypotheses, we will undertake two randomized controlled studies. This will be done by first recruiting men who have been treated for prostate cancer and identifying treatment attributes of importance to them. Based on these data, we will use established software platforms to develop preference assessment applications for conjoint analysis-based and traditional assessment methods, to be used in the randomized trials. One randomized trial using the conjoint analysis application will examine metrics of decision quality in men who undergo conjoint analysis, along with an educational intervention, compared with decision quality in men who undergo the educational intervention only. The second will compare criterion validity of these applications using metrics published in the literature This work is guided by the promise of eventually deploying point-of-care decision aids to improve the quality of decision making for men with prostate cancer. Public Health Relevance: This research aims to improve our ability to individualize prostate cancer care and improve decision quality for men with this disease. We will test whether a method taken from consumer marketing research can more accurately identify individual patients' values better than existing methods used in healthcare research. We will test whether use of this method improves the quality of the treatment decisions that are made by patients.