Question

Applying concept: Consider the following hypothetical scenario...

Suppose faith-based institutions had been offering small group discussions about HIV to attempt to prevent HIV transmission in inner city areas for some time, but no one had ever performed an economic evaluation. Individuals do not have to pay money to attend these sessions. You are part of a team that is preparing to conduct an economic evaluation. Question:

Suppose all epidemiological data you could possibly need were available to develop a model, but that you wanted to develop the most parsimonious model possible and not necessarily have everyone go through the model until death. At what even might it be reasonable to end such a model? Please justify.

Answer 1

The faith based institutions have been offering small group based discussions on HIV and how to prevent its transmission to the inner city areas. However, no economic evaluation has ever been done.

Now, all epidemiological data which are needed are there. We need to develop a parsimonious model. In order to develop that we need as few a predictor variable as possible. For decision makers to have informed judgements about how to spend the finite resources on public health to prevent the spreading of the disease one must know the anticipated cost, the opportunity cost of the prevention available and it's benefits. It is suggested individual level models are better than population level models for estimating the effects of population heterogeneity. The choice of model selection also depends on the time, available resources and also relevance of previous developed models (if any). Proper epideoepidemiol model is needed to estimate the cost benefit analysis of the prevention campaigns.

These are major predictors which should be considered for developing an epidemiological model.