Skip to main content
Fig. 1 | BMC Medical Research Methodology

Fig. 1

From: Incorporating individual historical controls and aggregate treatment effect estimates into a Bayesian survival trial: a simulation study

Fig. 1

Event-free survival distribution of the historical and new control arm depending of their commensurability. On each panel, the black curve represents the hypothetical historical control survival data (Kaplan-Meier estimates) simulated from a Weibull (left panel) and a piecewise exponential (right panel) distribution. Panels a and b represent no conflict in terms of underlying survival distribution but possible non-commensurability in terms of parameters (green, red and blue curves for commensurate, negative prior-data conflict and positive prior-data conflict, respectively). Panels c and d represent non-commensurability in terms of survival distribution (green curve)

Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us