Abstract:
OBJECTIVE To systematically summarize the key points for design of diagnostic test and conclude standardized reporting specifications for journal articles.
METHODS Drawing on the Standards for Reporting of Diagnostic Accuracy Study (STARD) and the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool, the key methodological components of diagnostic accuracy studies were systematically delineated within the context of health care-associated infection (HAI) in clinical settings. Focusing on two dimensions presented by the design of diagnostic test research and the result of diagnostic test, a structured pathway for manuscript preparation was summarized.
RESULTS A comprehensive, standardized writing framework for diagnostic test was established. Addressing the prevalent challenges such as selection bias in participants, inappropriate or delayed application of golden standards, inadequate reporting of blinding, and non-standardized model fitting of multiple indicators for combined diagnosis, the targeted prevention strategies that were integrated into the design phase were proposed. Furthermore, the standardized templates for presenting baseline characteristics, diagnostic performance (including single and parallel/serial multi-indicator models) and stratified analysis of results were formulated.
CONCLUSION Strict adherence to the design principles and reporting standards of diagnostic test can effectively mitigates various research biases and enhance the scientific rigor, transparency, and reproducibility of clinical diagnostic studies, which lays a robust evidence-based foundation for prevention and control of HAIs, formulation of clinical guideline and medical decision-making.