Sterility Is Designed Before It Is Tested
A sterility test is important.
It is also late.
Very late.
By the time the batch reaches final testing, the real sterility assurance work has already happened.
Or it has not.
That is the uncomfortable part.
You cannot test your way into a sterile process that was not designed, monitored, investigated, and controlled in the first place.
What FDA Said
On September 15, 2026, FDA posted a warning letter to Bausch & Lomb Inc. The letter was dated September 4 and followed a March 2026 inspection of the company’s Tampa drug manufacturing facility.
FDA said the facility manufactures sterile prescription and over-the-counter drug products produced by aseptic processing.
The letter describes multiple aseptic-processing concerns. FDA said the firm’s environmental monitoring and personnel monitoring program was inadequate for classified areas used to produce sterile drug products. FDA also said the firm routinely recovered microorganisms from ISO 5 aseptic areas from 2023 to 2025, including organisms associated with water sources, suggesting a moisture-related contamination issue.
FDA said production continued without adequate CAPA to address persistent microbiological contamination in ISO 5 environments.
The letter also describes restricted access barrier system design deficiencies, insufficient dynamic smoke studies, poor aseptic practices during filling operations, inadequate media fills, and inadequate investigation of adverse microbiological trends.
FDA made one point that should get every sterile manufacturer’s attention:
Sterility testing is critical, but it cannot be the sole justification for releasing sterile drug product batches.
Those are the source facts.
Here is the operator lesson:
Sterility is designed before it is tested.
Final Testing Cannot Carry the Whole System
There is a dangerous comfort in a passing sterility test.
The result is clean.
The batch is released.
The paperwork closes.
The meeting moves on.
But final testing is not a substitute for contamination control.
It is the last check in a system that should already be protecting the product.
Aseptic processing depends on layers:
Facility design.
Airflow.
Barrier protection.
Material flow.
Personnel flow.
Intervention control.
Environmental monitoring.
Personnel monitoring.
Cleaning and disinfection.
Media-fill representativeness.
Microbiology investigations.
Quality unit oversight.
If those layers are weak, a passing final test does not make the process strong.
It just means the sample did not catch the failure.
That sentence should make people uncomfortable.
Good.
ISO 5 Signals Are Not Background Noise
A recurring microbiological signal in a critical aseptic area is not a paperwork nuisance.
It is the system talking.
The question is whether leadership is listening.
FDA’s letter describes persistent recoveries in ISO 5 areas and inadequate investigation of adverse trends in water for injection, environmental monitoring, and personnel monitoring programs.
That matters because ISO 5 is where sterile product is most exposed and most vulnerable.
If signals keep showing up there, the organization cannot treat them as isolated events forever.
At some point, the question changes.
Not “Can we explain this plate?”
But:
Do we still understand this operation?
That is a different level of urgency.
Smoke Studies Are Not Theater
Dynamic smoke studies are supposed to show how airflow protects the product during real operations.
Not ideal operations.
Real ones.
Setup.
Assembly.
Interventions.
Operator movement.
Component transfer.
Line stoppages.
Awkward ergonomics.
The things that happen when the process is alive.
If the study does not reflect the actual risk, it is just cleanroom cinematography.
Nice footage.
Limited control value.
FDA’s letter focused on insufficient demonstration of unidirectional airflow protection across multiple lines, including during setup and assembly.
That is the point.
Sterility assurance depends on what happens during the messy parts of the operation, not just the cleanest version of the process.
Operator Behavior Is a Process Signal
Poor aseptic technique is not only a training issue.
Sometimes it is a design issue.
Sometimes it is a supervision issue.
Sometimes it is an ergonomics issue.
Sometimes it is a culture issue.
Sometimes it is a quality-oversight issue.
Usually, it is more than one.
If operators repeatedly reach, adjust, recover, transfer, or intervene in ways that threaten first air or product contact surfaces, the answer cannot be limited to “retrain the operator.”
Training may be needed.
But training alone does not fix a line that invites risky behavior.
It does not fix insufficient barrier protection.
It does not fix poor equipment placement.
It does not fix weak supervisory oversight.
It does not fix a quality unit that sees the signal too late.
The behavior is data.
Use it.
Media Fills Have to Represent the Real Operation
Media fills are supposed to challenge the aseptic process.
That means they have to represent the contamination risks of the process.
If commercial batches are larger, interventions are more complex, or the line has specific setup risks, the simulation needs to reflect that reality.
Otherwise the organization gets a comforting answer to the wrong question.
FDA’s letter criticized media fills that were substantially smaller than commercial batch sizes and did not sufficiently represent contamination risks associated with the aseptic lines.
That is not a small technicality.
A weak simulation can create false confidence.
False confidence is expensive in sterile manufacturing.
Quality Has to Be Present Before Release
The quality unit cannot only be strong at release.
It has to be strong during the operation.
During design review.
During monitoring strategy.
During intervention assessment.
During trend review.
During deviation investigation.
During CAPA design.
During media-fill design.
During management escalation.
By the time the batch record lands on the desk, many of the important sterility decisions have already been made.
Quality has to be close enough to the process to influence those decisions while they still matter.
That is not micromanagement.
That is sterile manufacturing.
The Real Control Strategy Is Built Upstream
The practical questions are not complicated.
They are hard because they force decisions.
Can the line design protect exposed sterile product during actual interventions?
Do dynamic smoke studies show protection under real operating conditions?
Are EM and PM locations capable of detecting meaningful risk?
Are microbiological trends investigated across time, area, organism, and process condition?
Are recovered organisms identified deeply enough to support root cause?
Are media fills representative of commercial batch size and intervention risk?
Does the quality unit have enough authority and presence to stop weak operations before release?
Are CAPAs changing the process, or just closing the event?
That is where sterility assurance lives.
Not at the end.
Upstream.
The Test Is Not the Strategy
Sterile manufacturing does not reward wishful thinking.
It rewards design.
Monitoring.
Discipline.
Escalation.
Investigation.
Oversight.
CAPA that actually changes the process.
The sterility test matters.
But it is not the strategy.
The strategy is the system that protects the product before the test ever happens.
If that system is weak, the final test is carrying weight it was never built to carry.
That is not inspection readiness.
That is hoping the last gate saves you.
In sterile operations, hope is not a control.
Website Publishing Notes
• Stage as Squarespace draft, not published.
• Author must be set to Two Pharma Guys.
• Add SEO title, meta description, URL slug, excerpt, source URL, and image alt text.
• Use the finished hero image at `outputs/sterility-is-designed-before-tested-hero.png`.
• Confirm the draft appears in Blog Content before considering staging complete.