Accuracy in the Diagnosis of Melanoma & the Impact of Double Reading

病理学家的评估是决定癌症是否存在的金标准。然而，社区病理学家在高达17%的病例中错误地诊断了黑色素瘤，而在2%-13%的病例中完全错过了它。我们将评估这些诊断错误的程度、来源及其对美国人口的影响。还将评估可能减少错误的技术，如双重阅读和继续教育。为了实现我们的目标，病理学家专家小组将使用德尔菲方法在5个测试集中建立病例诊断。超过100名美国病理学家将在两个阶段独立审查两个测试集，间隔6个月。我们的具体目标是： 1.量化病理学家在解释色素性病变活检标本时的错误程度并评估其可能来源。 2.评估由两名或多名病理学家对所有或部分病例进行独立的双重阅读是否可以提高判读准确度。 3.开发和评估个性化的基于网络的教育干预，以减少错误。 4.量化假阴性和假阳性解释的影响，以及减少短期患者护理错误的策略对美国人群的影响和相关资源利用。 总之，我们将评估病理学家对黑素细胞病变解释的准确性，强调发育不良痣和早期黑色素瘤的分类，以前的研究揭示了诊断错误的程度。这项大型多中心社区病理学家研究解决了一个日益重要的临床问题。我们的研究团队包括黑色素瘤病理学、测试准确性的统计测量以及临床医学诊断准确性研究方面的国际专家。拟议的工作是创新的，因为我们将超越简单地量化存在的错误。我们将研究与不准确诊断相关的患者和病理学家特征，以确定诊断错误的可能原因（例如，检测不充分与分类不充分）。然后，我们将评估通过使用双重阅读和教育计划来改善医疗保健服务的方法，该教育计划将在随机干预研究中进行测试。最后，我们将使用基于模型的分析，从我们的数据中预测美国人口水平上不准确诊断的临床意义。

A pathologist¿s assessment is the gold standard for deciding whether cancer is present or absent. However, community pathologists incorrectly diagnose melanoma in up to 17% of cases, while missing it altogether in 2%-13%. We will evaluate the extent of these diagnostic errors, their sources, and their impact on the U.S. population. Techniques that may reduce errors, such as double reading and continuing education, will also be evaluated. To achieve our goals, an expert panel of pathologists will establish a diagnosis for cases in 5 test sets using a Delphi approach. More than 100 U.S. pathologists will independently review two test sets during 2 phases, separated by 6 months. Our specific aims are: 1. To quantify the extent and evaluate possible sources of pathologists¿ errors in interpreting biopsy specimens of pigmented lesions. 2. To assess whether the addition of independent double reading by two or more pathologists on all or a subset of cases can improve interpretive accuracy. 3. To develop and evaluate an individualized Web-based educational intervention to reduce errors. 4. To quantify the implications of false negative and false positive interpretations and the impact of strategies to reduce errors on short-term patient care and associated resource utilization within the U.S. population. In summary, we will evaluate the accuracy of pathologists¿ interpretation of melanocytic lesions, emphasizing the classification of dysplastic nevi and early stage melanoma, where previous studies reveal a concerning degree of diagnostic error. This large multi-center study of community pathologists addresses a topic of growing clinical importance. Our research team includes international experts in melanoma pathology, statistical measurement of test accuracy, and studies of diagnostic accuracy in clinical medicine. The proposed work is innovative in that we will go beyond simply quantifying the existence of errors. We will study patient and pathologist characteristics associated with inaccurate diagnoses in order to identify possible reasons for diagnostic errors (e.g., inadequate detection vs. inadequate classification). We will then evaluate methods for improving health care delivery by using double reading and also an educational program that will be tested in a randomized intervention study. Finally, we will use a model-based analysis to make projections from our data to the clinical implications of inaccurate diagnoses at the U.S. population level.