Abstract: COVID-19 epidemic is threatening, accurate and rapid diagnosis and screening methods are particularly important. Numerous domestic scientists have been actively conducting relevant diagnostic trials and researches. This study reviewed the registration information of more than 10 diagnostic trials and discussed the common problems. This article focuses on how to use the PICOS principle to construct research questions for the COVID-19 diagnostic trial. The design points of the diagnostic test such as the setting of the gold standard, the representativeness of the subjects, the calculation of the sample size, the determination of synchronous, independent and blind method are explained in detail. The purpose of this paper is to provide suggestions for the design of COVID-19 diagnostic trials for researchers, help them to reduce and avoid bias in the top-level design stage, to complete high-quality clinical research, and provide evidence-based medicine for clinical diagnosis and treatment.