A Vanilla Investigation Story

What’s worse than being blamed in the laboratory? Not having an explanation for the contamination issue that you can easily fix.

NOTE: This article is available in other languages (FRANÇAIS / DEUTSCH).

When you investigate contamination, you hope for a smoking gun. You want data screaming “Look over here, I can explain everything!”. What happens if you have two smoking guns? Sounds great, right? I experienced this scenario at a contract manufacturer for non-sterile soap products.

This site was a sensory overload. The vanilla, gingerbread, and cinnamon emanating from the formulation areas made me love the holidays. The sometimes-burnt media coming out of the laboratory autoclave elicited other feelings. Glitter made the filling lines sparkle. I was stuck validating how to clean it off.

My confidence was riding high in this role. We were a year into my tenure without any new major micro concerns. The lab was chugging along smoothly.

The feeling changed when Pseudomonas popped up on multiple product test plates.

The Smoking Guns

Pseudomonas is classically associated with water sources. And whaddya know, we worked on our water system the previous week. The failing results came from the first test session performed after the system was returned to use.

The loop construction was the root cause, right? Contamination there would impact every batch we make. We tested each point of use as soon as the system was up and running. Since those plates needed 2 more days to incubate, we immediately reviewed the work done on the system.

The contaminated test samples had something else in common. They came from three different product families – one foaming hand soap, one hand-soap gel, and two body washes. Those products are filled on completely different lines. However, they all had “Vanilla” in the name. Our vanilla raw material must be dirty! We tracked down all the raw material lots used and their respective results.

Our customers were waiting shipment for multiple quarantined batches. We threw any available resources into this investigation. We swabbed all over the raw material containers, their processing areas, and where work was done for the water system. We sent organisms to Accugenix for Identification. This was important for 3 main reasons:

• Our lab was small. We didn’t have the resources or investment capital to perform IDs onsite
• The 0-day turnaround expedited root cause decisions on these smoking guns
• Historically, many gram-negative rod species were initially classified as Pseudomonas until further research was performed. Having an identification service with a broad, up-to-date library ensured we found Pseudomonas species specific to our environment.

The Let-Down

The smoking guns were so persuasive they blinded us to evidence redeeming them.

The water system proved its’ innocence quickly. Even though product samples were tested after construction, they were formulated or filled before the work started. All bioburden tests for the system, including resamples collected for this investigation, returned negative for growth. We had no direct evidence implicating the system.

The raw materials were an interesting story. It turns out, different materials are used to create the same vanilla scent depending on the final product formulation. It’s amazing what those formulation R&D scientists can do! There were no common materials used throughout the contaminated products.

Our investigation quickly sputtered. None of our investigative IDs were closely related to Pseudomonas. In fact, we were happy with the cleanliness of the water supply and raw material areas.

The Second-Hand Smoke

Does “Smoking gun” have an opposite? Some sort of slow, hidden, mundane killer? Second-hand smoke? That was the case here.

It’s convenient for a manufacturing site to blame the micro laboratory for failing results. If microbes come from the lab, product isn’t implicated. Unfortunately, lab personnel faced these accusations in the past. Even after manufacturing root causes were found, there was still bias against the lab. Morale took a hit.

The smoking guns took attention away from the lab. Despite the business implications, we were excited to avoid the initial blame. As the investigation continued, test sessions from subsequent days returned Pseudomonas. More batches were quarantined. Almost everything in the lab was swabbed for this investigation.

What’s worse than being blamed in the laboratory? Not having an explanation for the issue that you can easily fix. I read the investigative swab plates with the lab manager. We were not in the greatest mood. We went through a huge stack of plates with no source to blame. As we went through, swabs from the water bath, the incubators, test hood, countertop surfaces, and sinks were all clean.

The very last test plate gave us our Eureka moment. We instinctively jumped to high five. A pipette! The inside of one of the pipettes was COVERED in Pseudomonas. The organism dripped down into test plates during each session. We had our definitive root cause! We worked with the pipette manufacturer to improve our routine cleaning and maintenance program, and never saw the problem again.

As my colleague Alan Hoffmeister would agree, "Don't underestimate your lab accessories".