Biology Reference
In-Depth Information
FIGURE 9.1
Distribution of Ascaris worm numbers per individual in a study in the
Pulicat Lake region of Southern India.
30
Many individuals are uninfected while others are
infected by large numbers of worms.
only result in very small
changes in
prevalence. As such, prevalence is a poor measure of the impact of
a control program (
Figure 9.2
).
The ecological or epidemiological process generating the negative
binomial model of aggregated distributions of worm numbers per host,
where most harbor few worms and a few harbor many, are many and
varied.
26
One set of generative processes, and perhaps the most likely, is
where the infection of each host is governed by random infection events
with Poisson mean number of worms in an individual M, but where the
mean differs in value for each host in a manner described by a flexible
continuous distribution such as the gamma. This within and between
host variability would generate a negative binomial at the population
level.
27
Other processes, such as clumped acquisition of worms, will also
affect the distribution of worms in an individual and at the population
level.
28
and perhaps undetectable
e
e
