Physicians increasingly face the diagnostic dilemma of peripheral pulmonary lesions. Innovative bronchoscopy technologies targeting the lung periphery have developed at an accelerated pace, yet their introduction into clinical practice has been backed by variable and deficient evidence. A current and thorny problem has been inconsistent definitions of key diagnostic measures. Typical diagnostic accuracy measures, namely, sensitivity, specificity, and positive and negative predictive values, are binary outcome measures that reflect the ability of an index test to establish the presence or absence of a single target condition. In contrast, peripheral pulmonary lesions result from a range of malignant and benign processes, which has prompted the use of diagnostic yield as an outcome measure. There has been considerable variation in the definition of diagnostic yield across the advanced bronchoscopy literature. The reported diagnostic yield can increase by up to 20%, simply depending on the definition adopted by authors. A recent American Thoracic Society/American College of Chest Physicians consensus statement recommended adherence to a strict definition of diagnostic yield. The diagnostic yield numerator includes all patients with peripheral pulmonary lesions in whom a minimally invasive diagnostic procedure establishes a specific benign or malignant diagnosis that is sufficient to inform patient care, while the denominator includes all patients in whom the procedure was attempted or performed. A clear reference standard is essential and may include clinical follow-up to support final classification of nodules as benign or malignant but cannot taint interpretation of the index procedure results for the purpose of diagnostic yield calculation. The prevalence of malignancy influences diagnostic outcome measures and should be clearly reported. Finally, reporting should proceed according to established Standards for Reporting of Diagnostic Accuracy Studies (STARD) guidelines. Transparent and granular data reporting supports informative data pooling and meaningful comparisons across studies and technologies.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Diagnostic Outcome Definitions in Bronchoscopy

  • Anne V. Gonzalez

摘要

Physicians increasingly face the diagnostic dilemma of peripheral pulmonary lesions. Innovative bronchoscopy technologies targeting the lung periphery have developed at an accelerated pace, yet their introduction into clinical practice has been backed by variable and deficient evidence. A current and thorny problem has been inconsistent definitions of key diagnostic measures. Typical diagnostic accuracy measures, namely, sensitivity, specificity, and positive and negative predictive values, are binary outcome measures that reflect the ability of an index test to establish the presence or absence of a single target condition. In contrast, peripheral pulmonary lesions result from a range of malignant and benign processes, which has prompted the use of diagnostic yield as an outcome measure. There has been considerable variation in the definition of diagnostic yield across the advanced bronchoscopy literature. The reported diagnostic yield can increase by up to 20%, simply depending on the definition adopted by authors. A recent American Thoracic Society/American College of Chest Physicians consensus statement recommended adherence to a strict definition of diagnostic yield. The diagnostic yield numerator includes all patients with peripheral pulmonary lesions in whom a minimally invasive diagnostic procedure establishes a specific benign or malignant diagnosis that is sufficient to inform patient care, while the denominator includes all patients in whom the procedure was attempted or performed. A clear reference standard is essential and may include clinical follow-up to support final classification of nodules as benign or malignant but cannot taint interpretation of the index procedure results for the purpose of diagnostic yield calculation. The prevalence of malignancy influences diagnostic outcome measures and should be clearly reported. Finally, reporting should proceed according to established Standards for Reporting of Diagnostic Accuracy Studies (STARD) guidelines. Transparent and granular data reporting supports informative data pooling and meaningful comparisons across studies and technologies.