If Only this Pseudomonas Fluoresced!
How a microbe hiding in a cleanroom was found everywhere. The conclusion of our series on contamination trends in cleanrooms.
Our previous three blogs in this series have covered environmental monitoring (EM) in Part 1, water recovery in Part 2, and product contamination in Part 3. In all these events, initial blame was placed on operators. Here in Part 4 the systemic root cause is found.
This last story ties all three of these sample types together: EM, water, and product. In one of the site’s focus factories, these samples increasingly recovered Pseudomonas fluorescens group bacterial strains over a two-year period.
These recoveries weren’t treated as a trend and a full root cause investigation was never initiated during those two years for a few noted reasons:
- The trending program focused on repeats, from the same sample type, in a 12-month period. The exact sampling location never showed up frequently in a 12-month period to be officially classified as a ‘trend’.
- Management did not view the EM hits as a risk to product. In the risk assessments for each event, analysts were encouraged to say, “there is no direct route to transfer microorganisms from the impacted surface to product”.
- The exact species varied from sample to sample, so the problem hid under the organism trending radar.
To understand this event, the team needed to take a close look into the facility’s lay out. This facility had a U–shaped hallway. Support rooms, like Clean out of Place (COP) washers, storage, and labs, were on the outside of the U. Product processing rooms, like centrifuges and filter presses, were on the inside of the U. Product flowed from one processing room (like the thaw vessels), to a storage tank (like the pooling vessels), to another processing room, back to a tank, and so on.
The tanks marked red were the two that generated Action level results during this event.
The problem was first discovered in EM samples. Almost all of them were floor samples, which really pushed that “no direct route” impact assessment. It was easy to blame the cleaning crews for not cleaning the floor properly. Management decided the risk from the floor was so low that routine sampling of the floor could be replaced with sampling wall and equipment surfaces instead. This change immediately cut down Pseudomonas recoveries. But the change only hid the problem and did not answer the question as to where the organism was coming from, enabling the organism to fester in its source until it was too late.
Recoveries in water also sprinkled in over the two years. Most of the hits were on transfer panels, so there was still a tendency to blame sampling personnel. Since Pseudomonas is known as a water organism, it was easy to claim contaminated water impacted the rooms, but it didn’t make sense why POUs were only contaminated in these rooms.
Most investigators will tour a room when it is not in operation so they can see dirty areas left behind. They may get a simple re-enactment of what occurs during sampling or product processing, but still must dig a little deeper. When investigating, one must find the time to visit the rooms while the area is active and in process. Even more than that, investigators should make sure to watch the set-up, tear down, and any manual cleaning steps that occur. In this case, the investigators’ observation, ongoing through to the end, was the crucial component to uncovering the issue.
The process was innately messy with product going everywhere, especially with centrifuge cleaning. The large centrifuge bases can’t be moved for COP. The bases were not considered direct product contact, so they also don’t get Clean in Place (CIP). The operators performed extremely thorough equipment cleaning, and not just because they had witnesses. The operators meticulously scrubbed all surfaces with cleaning and sanitizing agents.
As the very last cleaning step, the operators connect a hose to the WFI drops in the room and spray down the equipment to remove any remaining cleaning agent. They then remove the hose and hang it to drain. The hoses remain hanging until they’re needed for the next cleaning.
The hoses and wet spots on equipment (like screw holes in the centrifuge) were swabbed. They were all covered in Pseudomonas! These hoses had never been sanitized. After each cleaning, Pseudomonas was sprayed and splashed all over these processing rooms.
A quick correlation was discovered, every room that used those hoses had EM or water Pseudomonas recoveries, however rooms without hoses had never recovered Pseudomonas. Hits were more frequent in rooms where water was utilized for more intensive rinsing steps.
A few product lots came back with high Pseudomonas counts from the second red-dotted tank. The manufacturing process before the tank includes a manually intensive filter press step. Unfortunately, the investigation does not conclude here, this site went on to blame personnel, yet again, for poor filter cleaning practices.
If you have enjoyed this article check out Jon's entire series on cleanroom investigations.