Bayesian Methods in Health Economics

Health economic evaluation in practice 193

c.0 <- c(2.38,2.38,1815.58,95.21)

c.1 <- c(7.96,7.96,1821.17,100.79)

# Computes total weekly costs

cost0 <- cost1 <- matrix(0,n.sims,J)

for (i in 1:n.sims){

for (j in 2:(J+1)){

c5[j] <- (c.0%*%m.0[i,1:(S-1),j])/sum(m.0[i,1:(S-1),j])

cost0[i,j-1] <- m.0[i,S,j]*c5[j] + c.0%*%m.0[i,1:(S-1),j]

cost1[i,j-1] <- m.1[i,S,j]*c5[j] + c.1%*%m.1[i,1:(S-1),j]

}

}

First, we define the unit cost associated with each of the first four states. Then

we define two matrices cost.0 and cost.1, each of dimension (n.sims, J)and

originally filled with 0 values. For each time point in the follow up j,wefirst

compute the cost of the state TP, according to (5.11).

Here, the R notation %*% codifies the inner product between the elements

of two v