Environmental Engineering Reference
In-Depth Information
consumed and did not consume water against the number of those people who
became ill and did not become ill.
The following is a hypothetical example involving collected data on the pop-
ulation associated with the waterborne outbreak:
52 people drank contaminated water and became clinically ill. (
a
)
32 people drank contaminated water and did not become ill. (
b
)
21 people did not drink contaminated water and became ill. (
c
)
64 people did not drink contaminated water and did not become ill. (
d
)
The 2
×
2 table is constructed to display the data as given.
Did not
Drank water
drink water
Became ill
52
21
Did not become ill
32
64
a/c
b/d
=
52
/
21
32
/
64
=
4
.
95
=
5
Calculation of the odds ratio (OR):
The OR clearly establishes a strong connection between exposure (water) and
the prevalence of disease.
In an actual study, there may be a number of possible sources for the disease
agent including food, insects, and personal associations, to name a few. With
the category of food, many subsets are possible, including salads, meats, breads,
juices, milk, and so on. Each of the sources deserves consideration as a vehicle
or vector, depending on the nature of the suspected disease agent. Case-control
studies can be constructed to test any and all of the potential sources of the
disease agent. The odds ratios can then be statistically analyzed to narrow the
field of suspected sources. Usually, the statistical evaluation is performed at the
95 percent confidence level (
p <
0.05).
In the previous example of a case-control study in connection with a water-
borne outbreak, cases of the disease had been established. Now consider a
situation where the town health officer released advance information to a popu-
lation of people that a wastewater cross-connection was found to have leaked at
some point in the distribution system. These conditions may provide the oppor-
tunity for a cohort study, which is an observational study in which two or more
groups of persons who are free of disease and differ by extent of exposure to a
potential cause of a disease are compared over time with respect to the incidence
of the disease.
140
In our example, this would be a prospective investigation of a
group (cohort) of healthy people known to have been exposed to contaminated
water. The object of the study would be to follow the course of events to evaluate
the appearance of illness in the exposed population and determine if consuming
the contaminated drinking water posed a risk for illness. In the cohort study, it
