Most diagnostic tests give binary results (above or below a
single cutpoint). Some tests give continuous results (e.g. from 1 to 100). A
few tests have a categor for INTERMEDIATE and INDEFINITE results. By creating a
middle category for perhaps 10% of the patients, the accuracy of those called
POSITIVE or NEGATIVE is increased. What are some papers on this general topic
(when to design a test with an intermediate and indefinite zone).
SCITE.AI
In the realm of diagnostic testing, the traditional approach often yields binary outcomes, categorizing individuals as either positive or negative based on a single cut-off point. However, there is a growing recognition of the potential benefits of incorporating intermediate or indefinite zones in diagnostic tests to enhance accuracy. By introducing a middle category for a subset of patients, typically around 10%, the accuracy of identifying individuals as positive or negative can be significantly improved (Zou et al., 2007).
Studies such as those by emphasize the importance of receiver-operating characteristic (ROC) analysis in evaluating diagnostic tests and predictive models, highlighting the need to consider binary, ordinal, or continuous outcomes in diagnostic classification tests (Zou et al., 2007). 's work further delves into the significance of selecting appropriate cut-off values in diagnostic tests to optimize sensitivity and specificity, a crucial aspect when designing tests with intermediate zones (Lim, 2021). Additionally, research by and explores the use of linear combinations of biomarkers to enhance diagnostic accuracy, a concept that aligns with the idea of creating intermediate categories in tests to refine result interpretation (Kang et al., 2012; Hsu & Chen, 2015).
Moreover, the study by sheds light on the importance of reporting and analyzing inconclusive test results in diagnostic accuracy studies, emphasizing the need to move beyond a dichotomous presentation of results (Shinkins et al., 2013). This aligns with the notion of incorporating intermediate or indefinite zones in tests to provide a more nuanced understanding of test outcomes. Furthermore, the work by addresses the challenges of evaluating diagnostic tests with intermediate results in the absence of a gold standard, underscoring the complexities involved in assessing tests that go beyond binary outcomes (Xu et al., 2012).
In the context of medical imaging, discusses the handling of indeterminate data in assessing diagnostic performance, highlighting the common practice of using an "indeterminate" category for results that fall between clear positive and negative interpretations (Wang, 2023). This approach resonates with the concept of incorporating intermediate zones in diagnostic tests to account for cases that do not definitively fit into either category. Additionally, the study by on enhancing quantitative flow ratio-FFR revascularization decision-making strategy underscores the value of intermediate zones in refining clinical decision-making processes based on diagnostic performance (Peper et al., 2021).
Overall, the literature reviewed underscores the importance of considering intermediate or indefinite zones in diagnostic testing to improve accuracy and provide a more nuanced interpretation of test results. By incorporating these middle categories, tests can better accommodate cases that do not neatly fit into binary classifications, ultimately enhancing diagnostic precision and patient care.
References:
Hsu, M. and Chen, Y. (2015). Optimal linear combination of biomarkers for multi‐category diagnosis. Statistics in Medicine, 35(2), 202-213. https://doi.org/10.1002/sim.6622
Kang, L., Xiong, C., Crane, P., & Tian, L. (2012). Linear combinations of biomarkers to improve diagnostic accuracy with three ordinal diagnostic categories. Statistics in Medicine, 32(4), 631-643. https://doi.org/10.1002/sim.5542
Lim, C. (2021). Methods for evaluating the accuracy of diagnostic tests. Cardiovascular Prevention and Pharmacotherapy, 3(1), 15. https://doi.org/10.36011/cpp.2021.3.e2
Peper, J., Hamersvelt, R., Rensing, B., Kuijk, J., Voskuil, M., Berg, J., … & Swaans, M. (2021). Diagnostic performance and clinical implications for enhancing a hybrid quantitative flow ratio–ffr revascularization decision-making strategy. Scientific Reports, 11(1). https://doi.org/10.1038/s41598-021-85933-9
Shinkins, B., Thompson, M., Mallett, S., & Perera, R. (2013). Diagnostic accuracy studies: how to report and analyse inconclusive test results. BMJ, 346(may16 2), f2778-f2778. https://doi.org/10.1136/bmj.f2778
Wang, S. (2023). Indeterminate data and handling for assessing diagnostic performance in imaging drug developments. The New England Journal of Statistics in Data Science, 112-119. https://doi.org/10.51387/23-nejsds46
Xu, H., Black, M., & Craig, B. (2012). Evaluating accuracy of diagnostic tests with intermediate results in the absence of a gold standard. Statistics in Medicine, 32(15), 2571-2584. https://doi.org/10.1002/sim.5695
Zou, K., O’Malley, A., & Mauri, L. (2007). Receiver-operating characteristic analysis for evaluating diagnostic tests and predictive models. Circulation, 115(5), 654-657. https://doi.org/10.1161/circulationaha.105.594929
