Tutorial on survival analysis and its application to the assessment of competing risks in clinical data analysis.
Andrew C Hooker (1), Joachim Grevel (2)
(1) Uppsala University, Uppsala, Sweden, (2) BAST Inc Ltd, Kington HR5 3DJ UK
- Theory of survival analysis.
- Application to competing risks, multistate models and patient benefit.
Overview/Description of presentation:
Part 1: Fundamentals. We present standard survival analysis theory with a focus on pharmacometrics analysis, including the production of visual predictive checks and interpretation of results. We also present how one can design/optimize and simulate clinical trials with (repeated) time-to-event as the endpoint of interest. Competing events will be introduced and why they complicate the analysis and need to be considered if present.
Taking dropout from a clinical study into account as a competing risk and not as “intended treatment” or right-censoring.
Multistate models as alternative to exposure-response models for the identification of patient benefit in different treatment arms of clinical trials.
Step-by-step guide through the methodology:
- a) Organising a clinical study into a multiple state system
- b) Counting transition frequencies between states
- c) Displaying transitions with non-parametric cumulative intensity transition functions and cumulative incidence functions
- d) Setting up parametric models for sojourn times and transition intensities and estimating model parameters .
- e) Selecting covariates and estimating their coefficients.
- f) Evaluating goodness-of-fit graphically.
- g) Testing hypotheses related to treatment effect on patient benefit.
Conclusions/Take home message:
Part 1: Survival analysis can be very useful but will typically not account for competing events, which may bias the analysis.
Part 2: The methodology to assess clinical trials as multistate systems is not new . Specifically in oncology, patient benefit should be quantitatively measured besides treatment efficacy. The proposed methodology and the published results will help patients, payers and caretakers make more informed decisions about potentially life-prolonging treatment options.
 Asanjarani A, Liquet B, Nazarathy Y. Estimation of semi-Markov multi-state models: a comparison of the sojourn times and transition intensities approaches. Int. J. Biostat. https://doi.org/10.1515/ijb-2020-0083
 Weiss GH, Zelen M. A semi-Markov model for clinical trials. Journal of Applied Probability 1965; 2, No. 2: 269-285