Biology Reference
In-Depth Information
pharmaceutical product complaint
depression aggravated
completed suicide
suicidal ideation
depression
suicide attempt
serotonin syndrome
0.05
0.10
0.50
1.00
5.00
10.00
50.00
Reporting Ratio
Fig. 15.3. Adverse event plot for the drug fluoxetine with the point sizes proportional to the suspect
fraction
S
ab
.
in Sec. 15.3. The other case-specific association measure considered here is the
mean number of concomitant medications
,
µ
ab
,defined as the average number of
other drugs appearing along with Drug A in all ISR's listing both Drug A and Ad-
verse Event B. Note that if all of these ISR's are pure plays, it follows that
µ
ab
=0
and that
ψ
ab
=
S
ab
=1; in fact, it is not difficult to show that
µ
ab
=0if and only
if
ψ
ab
=
S
ab
=1.
15.5. Relations between Measures
The results just presented lead us to expect, for fixed Drug A, a positive correlation
between
S
ab
and
ψ
ab
and negative correlations between
µ
ab
and both
S
ab
and
ψ
ab
as the Adverse Events B vary. Conversely, the pronounced differences in
appearance between Figs. 15.2 and 15.3 suggest the possibility of poor agreement
between the subjective association measure
S
ab
and the aggregate measures
R
ab
and
U
ab
, at least for some drugs (e.g., infliximab). Support for these suggestions
is given in Figs. 15.4 and 15.5, which illustrate the relationships seen between
two of these pairs of association measures for two different drugs.
Search WWH ::
Custom Search