Λέσχη Μαθηματικών

Analysis of time-to-event and longitudinal data: A multi-state modeling approach

Τσισμεντζόγλου Χρυσούλα

Abstract: In medical research, Survival Analysis and Longitudinal Data Analysis are distinct statistical fields used to study time-evolving data. However, the complexity of processes, such as disease progression, necessitates a joint analysis of time-to event and longitudinal data.

By employing multi-state modeling, transitions between different disease states can be effectively modeled over time. Incorporating longitudinal data—such as repeated measurements of patient characteristics or biomarkers—further enhances this modeling, providing deeper insights into these transitions. This joint approach allows for a comprehensive assessment of how baseline and time-varying factors influence the risks of moving between states, thereby improving the understanding of disease progression patterns and enhancing predictions for clinical outcomes.

Βιβλιογραφία

1. P. K. Andersen, N. Keiding, Multi-state models for event history analysis. Statistical Methods in Medical Research, 2002.
2. D. Collet, Survival Data in Medical Research. Third edition, CRC, 2014.
3. P. J. Diggle, Analysis of Longitudinal Data. Springer, 2005.
4. P. Hougaard, Multistate Models: A Review. Life Data Analysis, 5, Springer, 1999, pp. 239-264.
5. J. Klein et al., Handbook of Survival Analysis. CRC, 2014.
6. J. Le-Rademacher et al., The Utility of Multistate Models:A Flexible Framework for Time-to-Event Data. Epidemiologic Methods, 2022.
7. R. Weiss, Modeling Longitudinal Data. Springer, 2005.