FAQ - Pharmaceutical/Life Sciences Particle Counting

Lasair II Particle Counter

Question
A regulatory authority has asked that we monitor at the stopper bowl area of our filling lines on a routine basis. We intend to survey the area around the stopper bowl to find the worst case location excluding the areas that are routinely accessed by operators. We don't intend to monitor zones that are routinely accessed by operators such as the loading of the stopper bowl with stoppers. Does this seem a reasonable approach? Our real poser is what actions should be taken if an action limit is exceeded at the stopper bowl area.

Answer

This is a frequent question and posed by both the FDA and the EU. The concern is that you have invested hard and long over ensuring sterility of the stopper and then leave it open to environmental conditions for relatively long periods of time.

I have always shied away from continuously monitoring at the stopper bowl, unless; as you have, a direct request to do so. There are two key issues.

1. Location of the sample point. The sample point should be placed in such a position such that the operators filling the stopper bowl (and making the routine banging on the drum to free stuck stoppers; yes we know it happens!) are not impeded by the sample probe, thus making a worst case selection difficult. Also the rationale for monitoring is to ensure the air shower under which it resides is meeting ISO5 quality, I would therefore propose putting the sample probe higher than I would when monitoring the point of fill. The rationale here is that I am measuring the quality of air, while allowing normal operations, given it is a vibratory surface I will see particles being generated and do not want to sample process generated particles, only those environmental ones generated due to operators. But it does need to be above the stopper bowl and not at some adjacent position.
2. The sampling during charging of the stopper bowl needs to be either turned off (same philosophy for powder filling) or the alarm limits increased to reflect historical data. The revision of alarm limits is the best choice and a change of alarm limits can easily be implemented using the recipe function in Pharmaceutical Net.

If the sampling then shows an alarm and you have set alarm limits inline with your normal operations, for both stopper load and feed, then you have to ask yourself, do you feel lucky? What are the risks, if it is an alert threshold exceeded, what prompted such excursion? If it was an operator intervention, then we know these are high risk activities, it may be prudent to clear out the stoppers being fed to the line, and allow an air shower to purge the bowl, it may be sufficient to just allow the condition to stabilize. If it was completely random due to a 95% limit being set then just allow a cleaning down period, changing the alarm ratio from an immediate annunciation to one where three consecutive reading must be over the limit could help in balancing out your process. The key is that alarms should reflect what you do at your site not arbitrary level issued by the EU, this is in line with their current standards and in line with future thinking.

For information on monitoring locations read our paper Choosing the Most Suitable Non-viable Sample Point Locations.