Probability of Outbreak Given Surveillance Data

Although newer Bayesian detection algorithms, such as PANDA and BARD (described in Chapters 18 and 19),

In our model, we assume that testing of individuals with diarrhea will confirm the existence of an outbreak, if it exists, in three days. We note that direct testing of water in Glasgow did not produce a definitive answer for weeks. This process can be accelerated but at present it is unclear whether in practice it can occur in a more rapid time frame.

figure 29.4 Influence diagram model for boil-water advisory decision. CB indicates costs and benefits of the possible outcomes of the model.

compute the posterior probability of an outbreak from surveillance data directly, older methods do not.3 In the next chapter, we discuss methods to estimate the probability of an outbreak from surveillance data when newer detection algorithms are not available (by far the more common situation). For our discussion here, we simply use the numerical result that we obtain in that chapter: p = 0.0410.

Was this article helpful?

0 0

Post a comment