No. It is not acceptable to take the average among the LVP units tested in each batch/lot when
following this method because the purpose of this method is to measure and limit intra-batch
variability.
“Particulate matter” refers to small, sub-visible particles. USP <788> provides two tests for
detecting such particulates–light obscuration and microscopic assay. Both are generally
accepted for use in testing LVPs and small volume parenterals (SVP) for the determination of
sub-visible particulate matter. Normally, samples are first tested by the light obscuration method;
if the sample fails the specified limits, the microscopic assay method can then be used. However,
the microscopic method can be the sole test if there is a documented technical reason or
interference from the product under test that would make the light obscuration method unsuitable
or the results invalid.
Confusion about when averaging data is and is not acceptable is probably due to the sample
preparation method for the light obscuration test (USP <788>). At least 2, 5-mL aliquots from
each sampled unit or the pooled sample (see below) are to be used in the particulate count
determination, and the results from these aliquots are to be averaged for comparison with the
specification. Note that the average is of the results from examining each aliquot and not
between units. (The results of the first aliquot examined by light obscuration are to be discarded,
and the subsequent aliquots–2 or more–are retained.) Pooling units prior to analysis is permitted
only if the volume in each unit is less than 25 mL, in which case 10 or more units may be pooled.
If the volume in the SVP or LVP is 25 mL or more per unit, single units are to be examined by
this method (USP <788>).
Results among the test units cannot be averaged because particulate matter is assumed to be
non-uniformly dispersed throughout the lot. The intent of assessing results from each individual
unit is to ensure adequate representation of the lot and to detect potential variation within a lot.
As to the number of individual units to be tested for LVP and SVP units having a volume of
25mL or more, the USP states that the number of units tested depends on “statistically sound
sampling plans”, and “sampling plans should be based on consideration of product volume,
numbers of particles historically found to be present in comparison to limits, particle size
distribution of particles present, and variability of particle counts between units”. The USP also
suggests that the total number of units tested for any given batch may be less than 10 units (for
LVP and pooled SVPs) with proper justification. This is consistent with the CGMP requirement
for statistical sampling plans (see 211.165).
