Biomedical Engineering Reference
In-Depth Information
the ratios of the MCSE to the posterior standard deviation of these quantities
were less than 3%.
12.5
Simulations
The trial was simulated 10,000 times under many different scenarios, four of
which are summarized in Table 12.3.
Each
scenario
is
specified
by
xed
values
of
T
(s;c;q)
true
and
F
E
(s;c;q)
true
for s = 0 and 1. To obtain fixed true probabilities for all s
in [0, 1], we used several interpolation methods, allowing
T
(s;c;q)
true
and
F
E
(s;c;q)
true
to take various shapes as functions of s: The joint probabilities
E;T
(s;c;q)
true
used to generate (Y
E
;Y
T
) were computed using Equations
(12:9) and (12:10), and the resulting true utility u(c;q)
true
was obtained from
expression (12:12): Scenario 1 has a pattern similar to that of the prior means,
with (c;q) = (0.5, 0.2) optimal. In Scenario 2, smaller values of both c and q
have higher u(c;q)
true
. Scenario 3 is unsafe, with unacceptably high values
of all
T
(s;c;q)
true
compared to the upper limit
T
= 0.15. In Scenario 4,
all ecacy probabilities F
E
(s;c;q)
true
of dissolving the clot are unacceptably
small compared to the lower limit
E
= 0.50 Thus, in Scenarios 3 and 4, it is
most desirable to stop the trial early and select no (c;q) pair.
Table 12.3 shows that, under Scenarios 1 and 2, the method does a reli-
able job of selecting (c;q) pairs with higher utilities, and sub-sample sizes are
favorably balanced toward more desirable pairs. In Scenarios 3 and 4, where
no pair is acceptably safe and ecacious, the results show that the method
is likely to stop early and not select any pair. Additional simulations assess-
ing the method's sensitivity to the prior N; cohort size ; and interpolation
method, as well as using a simpler version of the model, are summarized by
Thall et al. (2011).
Search WWH ::
Custom Search