Unit information prior for adaptive information borrowing from multiple historical datasets

Abstract

In clinical trials, there often exist multiple historical studies for the same or related treatment investigated in the current trial. Incorporating historical data in the analysis of the current study is of great importance, as it can help to gain more information, improve efficiency, and provide a more comprehensive evaluation of treatment. Enlightened by the unit information prior (UIP) concept in the reference Bayesian test, we propose a new informative prior called UIP from an information perspective that can adaptively borrow information from multiple historical datasets. We consider both binary and continuous data and also extend the new UIP to linear regression settings. Extensive simulation studies demonstrate that our method is comparable to other commonly used informative priors, while the interpretation of UIP is intuitive and its implementation is relatively easy. One distinctive feature of UIP is that its construction only requires summary statistics commonly reported in the literature rather than the patient-level data. By applying our UIP to phase III clinical trials for investigating the efficacy of memantine in Alzheimer's disease, we illustrate its ability to adaptively borrow information from multiple historical datasets. The Python codes for simulation studies and the real data application are available at https://github.com/JINhuaqing/UIP.

Unit information prior for incorporating real-world evidence into randomized controlled trials

“…The parameter M is affected by the level of consistency between the observational studies and the current RCT, 22 for which the more inconsistency the smaller value of M. Scenarios 1 to 3 show that the patient population does not impose a strong influence on the values of M.…”

“…Let scriptD k denote the underlying individual patient data corresponding to scriptS k, and L false( k false) false( θ k falsefalse| scriptD k false) refers to the likelihood function for θ k. According to Jin and Yin, 22 we first introduce the unit information I U false( θ k false) for θ k,which represents the observed Fisher information at a unit sample level. The UIP to θ in the current RCT is then defined as follows:where I U false( θ ^ k false) refers to the unit information evaluated at θ ^ k, and false( γ 1 , … , γ K false) are pre-specified hyper-parameters with γ k = prefixtruemin false( 1 , n k / n false) so that a study with a larger sample size would be assigned a higher prior weight.…”

“…The only concern now is to specify I U false( θ ^ k false) for the kth observational study. Jin and Yin 22 suggested performing a case-by-case derivation to obtain the explicit form of I U false( θ ^ k false) based on the type of outcome and the corresponding model. By contrast, we provide a more general alternative to determine the unit information.…”

A group‐sequential randomized trial design utilizing supplemental trial data

“…An alternative approach to reducing required sample size is to incorporate supplemental or historical information. There is existing literature discussing borrowing of supplemental data 11‐16 . Static borrowing methods are the simplest approach, but they are suboptimal as they do not allow for any re‐evaluation of the initial assumption about exchangeability 13 .…”

“…There is existing literature discussing borrowing of supplemental data. [11][12][13][14][15][16] Static borrowing methods are the simplest approach, but they are suboptimal as they do not allow for any re-evaluation of the initial assumption about exchangeability. 13 Dynamic approaches are data-driven and do not require that the extent of borrowing be prespecified.…”