There are two important tasks in clinical epidemiology and evidence-based medicine, which are to generate evidence based on clinical needs; Combining evidence to carry out evidence-based clinical practice. In these two parts of the work, the key is to grasp the core problems of the research quickly, clearly and accurately. To better meet this need, evidence-based medicine proposes a PICO framework for problem decomposition. The PICO framework is very practical and efficient for the combing of interventional studies, especially randomized controlled research questions. Because the core of interventional research itself lies in who to intervene (patients), how to intervene (intervention), what measures to use as comparison (comparison), and where is the difference in the outcome brought by intervention and control (outcome). However, for diagnostic research, the application of this framework has great limitations. Unlike answering treatment, prognosis, etiology, etc. focusing on causal associations between interventions, exposure factors and outcomes, diagnostic studies are concerned with whether a certain indicator can distinguish patients or subjects who are in different states. As a result, causality is weakened, and more cross-sectional data is used to construct statistical associations that do not rely on causality.