betaFeedback

WorksMathematics

Statistical methods for extrapolation from one population to another: a review

Abstract

Background: Extrapolation refers to statistical approaches that support inference on treatment effects in a target population using data from completed clinical trials conducted in a source population. This is particularly important when only limited data are available in the target population. Extrapolation may be based on multiple sources of information, including pharmacokinetic, pharmacodynamic, mechanistic, or clinical outcome data. This review focuses on methods that enable borrowing of information from one population to support inference in another. Methods: We conducted a review of statistical methodologies for extrapolation, focusing on approaches that incorporate outcome data from source populations to support inference in target populations. Methods were identified across a range of settings, including extrapolation across age groups and across different clinical indications. Literature searches were conducted using the Science Citation Index Expanded database within Web of Science and included all English-language records available up to 30 June 2025. Results: The review identified 218 extrapolation approaches from 159 papers, which can be broadly classified into three main categories: Bayesian, frequentist, and causal inference methods. These methods vary in their assumptions, implementation, and applicability across different extrapolation settings. Conclusions: This review provides a comprehensive and up-to-date overview of statistical methods for extrapolation, offering a foundation for further methodological development and applied research. It also discusses key considerations for method selection and current limitations, and highlights priorities for future research, including the development of new extrapolation methods and clinical trial designs that facilitate extrapolation.

Paper details

vlnc:W_fk3rgynf
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
arXiv nonexclusive license
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…