Statistical methods for elimination of guarantee-time bias in cohort studies: a simulation study

201762 citationsJournal Articlegold Open Access

Authors

In Sung Cho · Yonsei University

Ye Rin Chae · Yonsei University

Ji Hyeon Kim · Yonsei University

Hae Rin Yoo · Yonsei University

Suk‐Yong Jang · Eulji University

Gyu Ri Kim · Yonsei University

Chung Mo Nam · Yonsei University

Abstract

While both time-dependent Cox regression model and landmark analysis are useful in resolving the problem of guarantee-time bias, time-dependent Cox regression is the most appropriate method for analyzing cumulative dose effects in pharmaco-epidemiological studies.

Topics & Keywords

Inflammatory mediators and NSAID effects Estrogen and related hormone effects Pharmacogenetics and Drug Metabolism

UN Sustainable Development Goals

Good health and well-being

Publication Details

Published in: BMC Medical Research Methodology

Volume 17, Issue 1, pp. 126-126

DOI: 10.1186/s12874-017-0405-6

Field-Weighted Citation Impact: 3.06