Skip to contents

Specify Borrowing Model

borrowing_details.Rd

Specify type of borrowing and, for Bayesian Dynamic Borrowing, set hyperprior for commensurability parameter tau.

Usage

borrowing_details(method, baseline_prior, ext_flag_col, tau_prior = NULL)

Arguments

method

character. The type of borrowing to perform. It must be one of: 'BDB', 'Full borrowing', or 'No borrowing'. See Details for more information.

baseline_prior

Object of class Prior specifying prior distribution for the baseline outcome. See Details for more information.

ext_flag_col

character. The name of the column in the data matrix that corresponds to the external control flag (1/0 or TRUE/FALSE). This identifies a patient as belonging to the external control cohort.

tau_prior

Object of class Prior defining the hyperprior on the "commensurability parameter". See Details for more information.

Value

Object of class Borrowing.

Details

Method

The method argument specifies the type of borrowing that will be implemented. There are currently three types of borrowing that are supported:

  • 'BDB' for Bayesian Dynamic Borrowing. In Bayesian Dynamic Borrowing, external control information is borrowed to the extent that the outcomes (i.e., log hazard rates or log odds) are similar between external and internal control populations. See Viele et. al. 2014.

  • 'Full borrowing' for pooling of historical and concurrent controls. There is no distinction between patients in the internal and external control arms. While the ext_flag_col must still be specified, it is not used.

  • 'No borrowing' for evaluating only the internal comparison, ignoring historical controls. Note that this method will filter the model matrix based on values in ext_flag_col.

Though the ultimate model specification is the same for 'Full borrowing' and 'No borrowing', both are available as options to facilitate comparison between methods.

Baseline Prior

The baseline_prior argument specifies the prior distribution for the baseline log hazard rate or log odds, depending on the outcome type. The interpretation of the baseline_prior differs slightly between methods:

  • 'BDB': the baseline_prior for Bayesian Dynamic Borrowing refers to the log hazard rate or log odds of the external control arm.

  • 'Full borrowing' or 'No borrowing': the baseline_prior for these borrowing methods refers to the log hazard rate or log odds for the internal control arm.

External Control

The ext_flag_col argument refers to the column in the data matrix that contains the flag indicating a patient is from the external control cohort. While this column is not used in 'Full borrowing', it must still be specified.

Tau Prior

The tau_prior argument specifies the hyperprior on the precision parameter commonly referred to as the commensurability parameter. See Viele et. al. 2014 for more details. This hyperprior determines (along with the comparability of the outcomes between internal and external controls) how much borrowing of the external control group will be performed. Example hyperpriors include largely uninformative inverse gamma distributions [e.g., gamma_prior(alpha = .001, beta = .001)] as well as more informative distributions [e.g., gamma_prior(alpha = 1, beta = .001)], though any distribution \(x \in (0, \infty)\) can be used. Distributions with more density at higher values of \(x\) (i.e., higher precision) will lead to more borrowing.

References

Viele, K., Berry, S., Neuenschwander, B., Amzal, B., Chen, F., Enas, N., Hobbs, B., Ibrahim, J.G., Kinnersley, N., Lindborg, S., Micallef, S., Roychoudhury, S. and Thompson, L. (2014), Use of historical control data for assessing treatment effects in clinical trials. Pharmaceut. Statist., 13: 41--54. doi:10.1002/pst.1589

Examples

sb <- borrowing_details(
  method = "BDB",
  baseline_prior = normal_prior(0, 1000),
  ext_flag_col = "ext",
  tau_prior = gamma_prior(0.001, 0.001)
)