Comparer des méthodes

	Analyse de survie pragmatique	Analyse de survie prospective
Domaine	Épidémiologie	Épidémiologie
Famille	Process / pipeline	Process / pipeline
Année d'origine	Conceptual framework: 1967; widespread application: 1990s–2000s	1958–1972 (foundational methods); prospective design emphasis formalized by 1980s
Auteur d'origine	Schwartz & Lellouch (explanatory vs. pragmatic distinction, 1967); extended in survival analysis literature from the 1970s onward	Kaplan & Meier (estimator, 1958); Cox (proportional hazards model, 1972); prospective design formalised in modern clinical epidemiology
Type	Observational / experimental hybrid — time-to-event analysis in real-world or pragmatic-trial settings	Longitudinal observational or experimental study design with time-to-event analysis
Source fondatrice	Ford, I., & Norrie, J. (2016). Pragmatic Trials. New England Journal of Medicine, 375(5), 454–463.	Kleinbaum, D. G., & Klein, M. (2012). Survival Analysis: A Self-Learning Text (3rd ed.). Springer. ISBN: 978-1441966452
Alias	real-world survival analysis, pragmatic time-to-event analysis, effectiveness survival analysis, PSA	prospective time-to-event analysis, prospective failure-time analysis, forward-looking survival study, prospective event-time study
Apparentées	5	5
Résumé	Pragmatic survival analysis applies time-to-event statistical methods within pragmatic or real-world settings, estimating how long patients survive, remain event-free, or retain treatment benefit under conditions of routine clinical practice. Unlike explanatory survival analyses conducted under tightly controlled trial conditions, the pragmatic variant embraces the heterogeneity, treatment switching, non-adherence, and competing events that characterise real-world patient populations, prioritising external validity over internal precision.	Prospective survival analysis is a longitudinal study design in which participants are enrolled before the event of interest occurs, followed forward in time under standardised conditions, and analysed using survival-analytic methods to estimate the time until a defined clinical endpoint — such as death, disease recurrence, or treatment failure. Because data are collected prospectively, exposure and covariate information are recorded before outcomes are known, substantially reducing recall and selection bias relative to retrospective approaches.