Skip to main content

Table 1 Posterior means and 95% credible intervals of model parameters from original calibration to targets y, and from the model that is recalibrated to \(\{y,z\}\) using a start from scratch approach as well as a sequential calibration approach

From: Sequentially calibrating a Bayesian microsimulation model to incorporate new information and assumptions

Parameter

Original posterior

New targets (start from scratch)

New targets (sequential)

Sojourn time

 \(\lambda _2\)

3.72 (2.20, 4.92)

2.38 (2.02, 3.07)

2.42 (2.02, 3.23)

 \(\lambda _1\)

2.57 (2.27, 3.06)

3.77 (3.35, 4.16)

3.80 (3.36, 4.18)

 \(\lambda _3\)

− 0.35 (− 0.96, 0.67)

0.87 (0.65, 0.99)

0.86 (0.61, 0.99)

  1. Refer to  “The microsimulation model for colorectal cancer” section for details on model component distributions and parameters