# Only probabilities available to calculate CE

In conducting a CE analysis to chose whether to implement a vaccine program or not, we have sots but there is no absolute efficacy. The info we have are probabilities,such as 90% that kids would get the vaccine should the program be implemented and the cost for . The rest of the information are also probabilities that a child may or may not get the disease in each year of a 3 year time period. Do I need efficacy and what type of measure should it be in to run the CE analysis with a 3% discount rate?

1 comment
• Dear Gilda,

The probability that a child gets the disease can be used as the effectiveness measure, but in this case the analysis might be reduced to a cost-minimization analysis.  Here's what I mean:

1.  If you have the cost of treating children who become sick, then you can calculate the total cost per child in the vaccinated sample and in the unvaccinated sample.  The cheaper strategy is the cost-minimizing strategy, and the analysis becomes a cost-minimization analysis (as opposed to a cost-effectiveness analysis).  Effectiveness is implied in the probability of illness.

2.  If you do NOT have the cost of treating children who become sick, then you can use "Cases of illness averted per 1,000 children vaccinated"  (or something similar) as the incremental effectiveness measure and the incremental cost of vaccinating 1,000 children to calculate the ICER = cost per case of illness averted.

I guess you could discount the cases of illness averted each year by 3%, although I cannot recall that I have seen this done.  Maybe someone else has a comment that could address this.

Melissa