Translating diagnostic assays from the laboratory to the clinic: analytical and clinical metrics for device development and evaluation. Issue 8 (4th April 2016)

• Record Type: Journal Article
• Authors: Borysiak, Mark D.; Thompson, Matthew J.; Posner, Jonathan D.
• Is Part Of: Lab on a chip. Volume 16:Issue 8(2016)
• Page Start: 1293
• Page End: 1313
• Publication Date: 2016-04-04
• Subjects: Miniature electronic equipment -- Periodicals; Combinatorial chemistry -- Periodicals; Biotechnology -- Periodicals; 543.0813
• DOI: 10.1039/c6lc00015k
• Languages: English
• ISSNs: 1473-0197

Abstract

Abstract : This tutorial review focuses on analytical and clinical metrics related to the design and evaluation of lab-on-a-chip diagnostic devices. We aim to provide standardized definitions and practical guidance for researchers working in this area. Abstract : As lab-on-a-chip health diagnostic technologies mature, there is a push to translate them from the laboratory to the clinic. For these diagnostics to achieve maximum impact on patient care, scientists and engineers developing the tests should understand the analytical and clinical statistical metrics that determine the efficacy of the test. Appreciating and using these metrics will benefit test developers by providing consistent measures to evaluate analytical and clinical test performance, as well as guide the design of tests that will most benefit clinicians and patients. This paper is broken into four sections that discuss metrics related to general stages of development including: (1) laboratory assay development (analytical sensitivity, limit of detection, analytical selectivity, and trueness/precision), (2) pre-clinical development (diagnostic sensitivity, diagnostic specificity, clinical cutoffs, and receiver-operator curves), (3) clinical use (prevalence, predictive values, and likelihood ratios), and (4) case studies from existing clinical data for tests relevant to the lab-on-a-chip community (HIV, group A strep, and chlamydia). Each section contains definitions of recommended statistical measures, as well as examples demonstrating the importance of these metrics at various stages of the development process. Increasing the use of these metrics in lab-on-a-chip research will improve the rigor of diagnostic performance reporting and provide a better understanding of how to design tests that will ultimately meet clinical needs.