Abstract:Background. The Expected Value of Sample Information (EVSI) measures the expected benefits that could be obtained by collecting additional data. Estimating EVSI using the traditional nested Monte Carlo method is computationally expensive but the recently developed Gaussian approximation (GA) approach can efficiently estimate EVSI across different sample sizes. However, the conventional GA may result in biased EVSI estimates if the decision models are highly nonlinear. This bias may lead to suboptimal study designs when GA is used to optimize the value of different studies. Therefore, we extend the conventional GA approach to improve its performance for nonlinear decision models. Methods. Our method provides accurate EVSI estimates by approximating the conditional benefit based on two steps. First, a Taylor series approximation is applied to estimate the conditional benefit as a function of the conditional moments of the parameters of interest using a spline, which is fitted to the samples of the parameters and the corresponding benefits. Next, the conditional moments of parameters are approximated by the conventional GA and Fisher information. The proposed approach is applied to several data collection exercises involving non-Gaussian parameters and nonlinear decision models. Its performance is compared with the nested Monte Carlo method, the conventional GA approach, and the nonparametric regression-based method for EVSI calculation. Results. The proposed approach provides accurate EVSI estimates across different sample sizes when the parameters of interest are non-Gaussian and the decision models are nonlinear. The computational cost of the proposed method is similar to other novel methods. Conclusions. The proposed approach can estimate EVSI across sample sizes accurately and efficiently, which may support researchers in determining an economically optimal study design using EVSI.

Unbiased Monte Carlo estimation for the expected value of partial perfect information

Efficient Estimation of the Value of Information in Monte Carlo Models

A Review of Methods for the Analysis of the Expected Value of Information

A Review of Methods for Analysis of the Expected Value of Information

Uncertainty and Value of Information in Risk Prediction Modeling

Estimating nested expectations without inner conditional sampling and application to value of information analysis

Calculating the Expected Value of Sample Information Using Efficient Nested Monte Carlo: A Tutorial

Calculating the Expected Value of Sample Information in Practice: Considerations from 3 Case Studies

Decision-theoretic sensitivity analysis for reservoir development under uncertainty using multilevel quasi-Monte Carlo methods

Estimating the EVSI with Gaussian Approximations and Spline-Based Series Methods

Calculating the Expected Value of Sample Information in Practice: Considerations from Three Case Studies

Optimal Unbiased Estimation for Expected Cumulative Discounted Cost.

Nested Multilevel Monte Carlo with Biased and Antithetic Sampling

Computing the variance of a conditional expectation via non-nested Monte Carlo

When are Unbiased Monte Carlo Estimators More Preferable than Biased Ones?

Automated Efficient Estimation using Monte Carlo Efficient Influence Functions

A hierarchical expected improvement method for Bayesian optimization

Microsimulation Estimates of Decision Uncertainty and Value of Information Are Biased but Consistent

The value of perfect information for the problem: a sensitivity analysis

Efficient High-Dimensional Importance Sampling

Improving the Expected Improvement Algorithm