手法を比較

	多施設共同診断精度の検証研究	多施設共同コホート研究
分野	疫学	疫学
系統	Process / pipeline	Process / pipeline
提唱年	2003 (STARD statement first published; updated 2015)	Mid-to-late 20th century (widespread adoption 1970s–1990s)
提唱者	STARD Group (Bossuyt, Reitsma et al.)	Developed incrementally through large collaborative epidemiological projects (e.g., Framingham Heart Study consortium expansions, 1948 onward; EPIC study, 1992)
種類	Observational diagnostic study design	Observational longitudinal study
原典	Bossuyt, P. M., Reitsma, J. B., Bruns, D. E., Gatsonis, C. A., Glasziou, P. P., Irwig, L., Lijmer, J. G., Moher, D., Rennie, D., de Vet, H. C. W., Kressel, H. Y., Rifai, N., Golub, R. M., Altman, D. G., Hooft, L., Korevaar, D. A., & Cohen, J. F. (2015). STARD 2015: An Updated List of Essential Items for Reporting Diagnostic Accuracy Studies. BMJ, 351, h5527.	Rothman, K. J., Greenland, S., & Lash, T. L. (2008). Modern Epidemiology (3rd ed.). Lippincott Williams & Wilkins. ISBN: 978-0781755641
別名	multisite diagnostic accuracy study, multicenter DTA study, multicenter index test evaluation, STARD multicenter study	multisite cohort study, multi-centre cohort, collaborative cohort study, pooled cohort study
関連	5	6
概要	A multicenter diagnostic accuracy study evaluates how well an index test (e.g., a biomarker, imaging modality, or clinical prediction rule) identifies a target condition when conducted across two or more independent clinical sites. By recruiting patients from diverse settings, it produces estimates of sensitivity, specificity, and likelihood ratios that are more externally valid than those obtained from a single center, and it enables explicit assessment of how test performance varies across sites, patient populations, and operator skill levels.	A multicenter cohort study follows defined groups of participants at two or more geographically or institutionally distinct sites over time to estimate incidence, identify risk factors, and quantify associations between exposures and outcomes. By pooling data from multiple centers, it achieves statistical power and population diversity that single-site designs cannot match, making it the workhorse of large-scale epidemiological and clinical research.