evppi_lrmm: Estimation of the Expected Value of Partial Perfect...

In feralaes/dampack: Package with useful decision-analytic modeling tools

Estimation of the Expected Value of Partial Perfect Information (EVPPI) using a linear regression metamodel approach

Description

evppi_lrmm is used to estimate the Expected Value of Partial Perfect Information (EVPPI) using a linear regression metamodel approach from a probabilistic sensitivity analysis (PSA) dataset.

Usage

evppi_lrmm(nmb = NULL, params = NULL, sel.params = 1, sel.gam = T,
  k = NULL, verbose = F)

Arguments

nmb	Matrix of net monetary benefits (NMB). Each column corresponds to the NMB of a different strategy.
params	Vector or matrix of parameters.
sel.params	A vector including the column index of parameters for which EVPPI should be estimated.
sel.gam	Logical variable indicating if a generalized additive model, GAM, (i.e., a spline model) should be fitted (Default = T). If FALSE, a polynomial of degree k is fitted.
k	Scalar with the order of basis functions for the spline model if (sel.gam == T) or the degree of polynomial if (sel.gam != T)
verbose	Logical variable indicating if estimation progress should be reported.

Details

The expected value of partial pefect information (EVPPI) is the expected value of perfect information from a subset of parameters of interest, \theta_I of a cost-effectiveness analysis (CEA) of D different strategies with parameters \theta = \{ \theta_I, \theta_C\}, where \theta_C is the set of complimenatry parameters of the CEA. The function evppi_lrmm computes the EVPPI of \theta_I from a matrix of net monetary benefits B of the CEA. Each column of B corresponds to the net benefit B_d of strategy d. The function evppi_lrmm computes the EVPPI using a linear regression metamodel approach following these steps:

1. Determine the optimal strategy d^* from the expected net benefits \bar{B}

   d^* = argmax_{d} \{\bar{B}\}

2. Compute the opportunity loss for each d strategy, L_d

   L_d = B_d - B_{d^*}

3. Estimate a linear metamodel for the opportunity loss of each d strategy, L_d, by regressing them on the spline basis functions of \theta_I, f(\theta_I)

   L_d = \beta_0 + f(\theta_I) + \epsilon,

   where \epsilon is the residual term that captures the complementary parameters \theta_C and the difference between the original simulation model and the metamodel.

4. Compute the EVPPI of \theta_I using the estimated losses for each d strategy, \hat{L}_d from the linear regression metamodel and applying the following equation:

   EVPPI_{\theta_I} = \frac{1}{K}\sum_{i=1}^{K}\max_d(\hat{L}_d)

   The spline model in step 3 is fitted using the 'mgcv' package.

Value

evppi A numeric vector of size one with the EVPPI of the selected parameters

References

1. Jalal H, Alarid-Escudero F. A General Gaussian Approximation Approach for Value of Information Analysis. Med Decis Making. 2018;38(2):174-188.

2. Strong M, Oakley JE, Brennan A. Estimating Multiparameter Partial Expected Value of Perfect Information from a Probabilistic Sensitivity Analysis Sample: A Nonparametric Regression Approach. Med Decis Making. 2014;34(3):311–26.

Examples

## Load mgcv package and matrixStats
library(mgcv)
library(matrixStats)
## Load PSA dataset
data(syndX)
## Net monetary benefit (NMB) matrix
nmb <- syndX[, 5:7]
## Matrix of model parameter inputs values theta
theta <- syndX[, 1:4]
## Optimal strategy (d*) based on the highest expected NMB
d.star <- which.max(colMeans(nmb))
d.star
## Define the Loss matrix
loss <- nmb - nmb[, d.star]
## Estimate EVPPI for parameter 1 (MeanVisitsA)
evppi_lrmm(nmb = nmb, params = theta, sel.params = 1, verbose = TRUE)

feralaes/dampack documentation built on Sept. 8, 2024, 11:31 a.m.