Biomedical Engineering Reference
In-Depth Information
(a)
Factors
Distribution
Fixed
Type
Scale
pH (growth)
Random
Uniform
DO%
Random
Uniform
Feed-1 rate
Random
Triangular
Temperature (growth)
Random
Uniform
N runs
10000
Random noise
S D: 31.5
(b)
Quartiles
100.0%
99.5%
97.5%
90.0%
75.0%
50.0%
25.0%
10.0%
2.5%
0.5%
0.0%
maximum
252.4
207.0
183.1
158.2
135.9
110.9
85.5
62.9
38.4
17.4
-23.6
240
210
quartile
median
quartile
180
150
120
minimum
100
Moments
Mean
SD
SE mean
Upper 95% mean
Lower 95% mean
N
110.8014
37.075438
0.3707544
111.52815
110.07464
10000
70
40
10
Tolerance intervals
-20
Parameter
Estimate
Lower TI
Upper TI
1-Alpha
Proportion
Mean
110.8014
14.17159
207.4312
0.950
0.990
Figure 5.15. MonteCarlo simulationparameters (a) and statistics (b) for the calculationof PVAC
for product quality 1. Note: Data have been normalized against the average small-scale model
performance at set point operating conditions.
5.9 REGULATORY FILING, PROCESS MONITORING,
AND POSTAPPROVAL CHANGES
The design space in the form of the acceptable ranges for the key and critical operational
parameters along with the process validation acceptance criteria should be included in
the filing. After the product is approved, process monitoring should be performed to
demonstrate that product quality and process performance attributes are within the filed
design space. As mentioned earlier, the primary benefit of the design space concept is the
regulatory flexibility. Process changes within the design space should require less
regulatory review or approval. Therefore, process improvements during the product
life cycle with regard to process consistency and throughput can be made with reduced
postapproval submissions. However, as stated in ICH Q8, “The degree of regulatory
Search WWH ::




Custom Search