เปรียบเทียบวิธี

	การถดถอยความเสี่ยงสัดส่วนของ Cox แบบย้อนหลัง	การวิเคราะห์การรอดชีพย้อนหลัง
สาขาวิชา	ระบาดวิทยา	ระบาดวิทยา
ตระกูล	Process / pipeline	Process / pipeline
ปีกำเนิด	1972	1970s–1980s (retrospective variant established)
ผู้ริเริ่ม	David R. Cox	Kaplan & Meier (foundational estimator, 1958); Cox (regression model, 1972); retrospective application is a design variant documented since the 1970s
ประเภท	Semi-parametric survival regression	Retrospective observational analytical study
แหล่งต้นตำรับ	Cox, D. R. (1972). Regression models and life-tables. Journal of the Royal Statistical Society, Series B, 34(2), 187–220.	Collett, D. (2015). Modelling Survival Data in Medical Research (3rd ed.). CRC Press. ISBN: 978-1439856789
ชื่อเรียกอื่น	Cox PH regression (retrospective), retrospective Cox survival model, retrospective hazard regression, Cox model on historical data	historical survival study, retrospective time-to-event analysis, retrospective follow-up survival study, archival survival analysis
ที่เกี่ยวข้อง	5	5
สรุป	Retrospective Cox proportional hazards regression applies Cox's (1972) semi-parametric survival model to time-to-event data extracted from existing records — medical charts, administrative databases, registries, or biobanks. It estimates covariate-adjusted hazard ratios (HRs) without specifying the underlying baseline hazard, making it the dominant analytic tool when the investigator works backward from already-recorded outcomes and exposures.	Retrospective survival analysis applies time-to-event statistical methods — most commonly the Kaplan-Meier estimator and Cox proportional hazards regression — to data collected from past records rather than through prospective follow-up. The researcher looks back at medical records, disease registries, or administrative databases to reconstruct each patient's journey from a defined starting point (e.g., diagnosis or surgery) to an outcome of interest (e.g., death, relapse, or hospital readmission), making it a cost-efficient approach for studying prognosis and risk factors when prospective follow-up is not feasible.