Sensitivity, Specificity, and Predictive Values in the 'Sensitivity and Specificity of Clinical Diagnostics'
Bradley A. Woodruff, Ruth Baron
Abstract
Bradley A. Woodruff, Ruth Baron
Abstract
To the Editor.— The measures of accuracy in clinical screening or diagnosis relate to the intrinsic ability of a procedure to identify correctly those persons with disease (sensitivity) and those persons without disease (specificity).1In a recent article entitled "The Sensitivity and Specificity of Clinical Diagnostics During Five Decades,"2the authors use theterms clinical accuracy for a positive diagnosis and clinical accuracy for a negative diagnosisand equate them by definition to the positive predictive value (the proportion of positive tests or diagnoses that correctly identify the presence of disease) and the negative predictive value (the proportion of negative tests that correctly identify the absence of disease), respectively. Unlikesensitivity and specificity, the predictive values are not properties inherent to the procedure itself, but vary with the prevalence of disease in the population tested. For example, the authors state that "the accuracy of the clinical diagnosis
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
To the Editor.— The measures of accuracy in clinical screening or diagnosis relate to the intrinsic ability of a procedure to identify correctly those persons with disease (sensitivity) and those persons without disease (specificity).1In a recent article entitled "The Sensitivity and Specificity of Clinical Diagnostics During Five Decades,"2the authors use theterms clinical accuracy for a positive diagnosis and clinical accuracy for a negative diagnosisand equate them by definition to the positive predictive value (the proportion of positive tests or diagnoses that correctly identify the presence of disease) and the negative predictive value (the proportion of negative tests that correctly identify the absence of disease), respectively. Unlikesensitivity and specificity, the predictive values are not properties inherent to the procedure itself, but vary with the prevalence of disease in the population tested. For example, the authors state that "the accuracy of the clinical diagnosis
Key concepts: Medicine, Sensitivity (control systems), Predictive value, Internal medicine, Engineering, Electronic engineering