Bayesian Methods in Health Economics

158 Bayesian Methods in Health Economics

sigma.e[t] <- exp(logsigma.e[t])

# Prior distributions

mu.c[t] ~ dnorm(0, 1.0E-6) # mean costs (log scale)

logsigma.c[t] ~ dunif(-5,10) # log st. dev. for costs

mu.e[t] ~ dnorm(0, 1.0E-6) # mean QALY (logit scale)

logsigma.e[t] ~ dunif(-5,10) # log st. dev. for QALYs

beta[t] ~ dunif(-5, 5) # regression between (e,c)

}

}

Notice that we need to define the nodes lambda[t] to model the precision of

the conditional distributions for c

∗

it

. Obviously, its value is set to the reciprocal

of the conditional variance.

Wesavethecodeinthefileacupt.txt and run the model in JAGS using

the following code.

library(R2jags)

dataJags <- list("n","c0","c1","e0","e1")

filein <- "acuptRCT.txt"

params <- c("mu.c","mu.e","beta","sigma.e","sigma ...