Page 55 - Assurance of Sterility for Sensitive Combination Products and Materials

Aseptic processing 43

of aseptic processing, we are preventing the action of contamination. The
complexities of proving a negative, or the lack of existence of microbes, are
more challenging than proving a positive.
Detection of contamination or sterility is difficult. Nondestructive detection
of microbes is difficult and, in many cases, not feasible. Sterility testing (USP
<71>) is destructive and therefore requires a limited sampling number. The
current sampling and testing practice do not result in a statistically rele-
vant assurance of sterility. Environmental monitoring of clean room air and
surfaces is imperfect, with significant variability in results and capability to
obtain meaningful results. Environmental contaminant recovery and testing
might not be complete and sample sizes and locations are often limited due
to probe placement and instrumentation. In addition, environmental mon-
itoring material placement can involve significant and, in some cases, risky
human intervention into critical aseptic processing areas.
There is a lack of observable correlations. Current approaches to aseptic pro-
cess control do not provide a means to measure sterility assurance in aseptic
processing. The correlation between the observed and the desired (or un-
desired) outcome is relatively poor. Air profile (smoke) studies, differential
pressures, air velocity, first air principles, ascending and descending clean
room classifications, and aseptic technique are designed to prevent micro-
organisms from contacting and contaminating sterile surfaces or contents.
These methods are not surrogates for an indication of contamination. In
other words, we do not know for certain the level of disruption in these
controls is needed to contaminate the contents of a container.
Aseptic processing involves significant variables. Aseptic processing is com-
plex. It involves the interplay between environment, microorganisms,
contamination vectors, complicated manufacturing methods, and human
behavior. Human behavior is variable and unpredictable. Training, moni-
toring, performance feedback, coaching, supervision, observation, human
factor consideration, ergonomic process design, barriers, and automation
can only help reduce variability and the effect of variability but never elim-
inate the inherent variability of human performance. Therefore, any process
that depends on human intervention and activity will be inherently variable.
The impact of the failure is severe. Failure of the aseptic process can result
in contamination of the sterile product and could impact patients signifi-
cantly. Not all microbial contamination observed in the environment, on
personnel, or on product contact surfaces will result in microbiological con-
tamination of the product and not all microbiological contamination of the
product will result in patient health issues. However, because the product