手法を比較

	多施設用量反応解析	用量反応分析
分野	疫学	疫学
系統	Process / pipeline	Process / pipeline
提唱年	1992 (foundational trend methods); refined 2000s–2010s	Conceptual roots 16th century; modern epidemiological application mid-20th century
提唱者	Greenland & Longnecker; extended by Orsini et al.	Paracelsus (conceptual foundation); formalized by John Snow and later Bradford Hill
種類	Quantitative epidemiological analysis	Quantitative analytical method
原典	Greenland, S., & Longnecker, M. P. (1992). Methods for trend estimation from summarized dose-response data, with applications to meta-analysis. American Journal of Epidemiology, 135(11), 1301-1309.	Rothman, K. J., Greenland, S., & Lash, T. L. (2008). Modern Epidemiology (3rd ed.). Lippincott Williams & Wilkins. ISBN: 978-0781755641
別名	pooled dose-response analysis, multicenter exposure-response analysis, multi-site dose-response modeling, collaborative dose-response study	exposure-response analysis, concentration-response modeling, dose-response modeling, DRA
関連	2	4
概要	Multicenter dose-response analysis estimates the quantitative shape of the relationship between a graded exposure and a health outcome by pooling data or effect estimates across two or more study centers. Using flexible regression tools such as restricted cubic splines or fractional polynomials within a two-stage meta-analytic framework, it characterizes whether the relationship is linear, supra-linear, threshold-based, or J-shaped — providing far greater statistical power and generalizability than any single center could achieve alone.	Dose-response analysis quantifies the relationship between the magnitude of an exposure (the dose) and the probability or rate of an outcome (the response). It is a core analytical strategy in epidemiology and toxicology, providing evidence that increasing exposure systematically increases — or decreases — the risk of disease. A demonstrated dose-response gradient is one of Bradford Hill's classic criteria supporting causal inference.