PROCESS LIFECYCLE VALIDATION - Risk Management Applications in Pharmaceutical and Biopharmaceutical Manufacturing - page 188

as 1%-2% would be a critical step unless there is a downstream blending step
that assures content uniformity, especially if the particle sizes of the ingredients
are not well controlled. Mixing a sparingly soluble ingredient into a solution at
nearly saturated concentrations is typically a critical step and parameters such
as mixing speed, rate of addition, and solution temperature may be CPPs. In an
RA for tablet manufacturing [21], several processing steps were viewed as high
risk (scores of 4-5 on a 5-point scale): compression (tablet), granulation (wet
and dry), mixing-blending, and pelleting. Measuring or weighing and primary
packaging were viewed as lower risk (scores of 3 or less on a 5-point scale).
8.7.2.4 Normal Operating Range and Proven Acceptable Range For an exist-
ing product, comparing the normal operating range (NOR) to the proven accept-
able range (PAR) should be considered when performing an RA of potential
CPPs. The assessment should also include consideration of the interaction of
potential parameters. The NOR is typically the range specified in the master
batch record, whereas PAR is demonstrated during development or characteriza-
tion/validation studies. An acceptable product is produced within the PAR range
[5]. It is paramount to know the variability of both the control parameters and
measurements at the target (set-point) and the limits. The variability of control
parameters and measurements will determine the level of risk. The probability of
exceeding the NOR/PAR limit and the consequence (i.e., severity) will determine
the level of criticality. A comparison of the NOR and PAR will typically reveal
one of the following general situations:
(I) When PAR is unknown : In this case, the PAR has not been identified or his-
torical information does not provide substantiation of acceptable ranges broader
than the NOR. It may be possible to establish the PAR from historical experience
with the process (e.g., from investigations). Another possibility is to assume that
the NOR and PAR are the same.
Figure 8.3 illustrates the case when NOR equals PAR and for two different
distribution patterns around the set points, A and B. This example shows how
control variability around the set-point is important. The likelihood or risk of
exceeding the operating limit for scenario A is much less than scenario B based
on the variability of the control from the set-point.
A higher risk of reaching the limit is apparent when the parameter's variability
is greater and/or the set-point approaches the limit. For an existing process, if
PAR is unknown and variability of the parameter is
a) Lower (i.e., scenario A in Figure 8.3) compared to the limit then it is safe
to assume that PAR is equal to NOR. There is little value in determining
the PAR value as the risk or probability is extremely low of ever reaching
the approved limit.
b) Higher and/or close to the limit (i.e., scenario B in Figure 8.3), the likeli-
hood of reaching the limit becomes high and the parameter is considered a
CPP. In this case, it would be value added in determining the PAR, which
may be beyond the NOR.