Healthcare and Medicine Reference
In-Depth Information
medical understanding and decision making are relative risks, attributable
risks, and etiological fractions (attributable risk percent).
In analytical studies, ratios , as already mentioned above, are fractions
relating two entities (two different health events or their related character-
istics) in which the entity in the numerator is not reproduced (with some-
thing else added) in the denominator. Odds (cross product ratios, relative
odds) are ratios of the probability of occurrence of an event to that of
nonoccurrence, or the ratio of the probability that something is one way
to the probability than it is another way: In our case of a case control
study, ad/cb.
Relative risk (relative benefit increase or reduction) is a fraction relat-
ing incidence of events in one group of subjects, such as those exposed to
some factor, to another group or groups nonexposed to it. Relative risks are
quantifications of the strength of the causal relationship . In the case of a
study of noxious factors, the higher the relative risk, the stronger the causal
association. If the relative risk of a presumably noxious factor is lower than
1.0, the factor (agent) has a protective effect, not a noxious one. Cohort
studies (observations of new events over a certain period of time) of groups
(cohorts) to be compared serve this purpose.
In our example:
80 /100,000
20 /100,000
=
4.0
Hence, relative risk may be called also a ratio of rates or rate ratio.
Odds ratios most often exposures/non-exposure frequencies in cases
to exposures/non-exposures frequencies in non-cases are another kind of
estimation of the strength of a causal association derived from case-control
studies.
In our example:
a/d
b/c
80
× =
50
4,000
1,000
=
× =
4.0
20
50
N.B. If case-control studies are based on incident cases and if the
proportion of subjects exposed is similar as in the cohort study, both cohort
and case-control studies yield comparable estimations of the strength of
causal association.
Search Pocayo ::




Custom Search