Training Has to Show Up in the Work
Training Has to Show Up in the Work
A training record can prove someone attended.
It does not prove they can do the work.
That difference matters.
Especially in Quality.
FDA's July 13 warning letter to Shimoga Chemicals is a useful reminder that quality systems are not judged by whether the right roles exist on paper.
They are judged by whether those roles are working in practice.
FDA cited data integrity concerns, quality-unit review gaps, and stability program deficiencies. FDA also requested position descriptions for key personnel, particularly within the quality unit, and evidence that trained skills are being applied in day-to-day operations, including managerial oversight mechanisms used to verify on-the-job performance.
That last part is the one leaders should underline.
Evidence that trained skills are being applied.
Not training completed.
Applied.
The Matrix Is Not the Capability
Training matrices are useful.
They tell you who was assigned what.
They show completion dates.
They make audits less chaotic.
They give someone in Quality a spreadsheet to maintain until the end of time.
But a training matrix is not capability.
It does not tell you whether the analyst knows when a result looks wrong.
It does not tell you whether the reviewer knows how to challenge incomplete data.
It does not tell you whether the quality unit has the authority, time, and expectation to stop a batch disposition decision when the evidence is not there.
It tells you the box was checked.
Boxes matter.
But boxes do not release product.
People do.
Quality Unit Authority Has to Be Real
A procedure can say the quality unit has responsibility.
That is the easy part.
The harder question is whether the quality unit is actually exercising that responsibility in the flow of work.
FDA stated that Shimoga lacked procedures for electronic data review and that the quality unit did not review electronic data before a batch was released and distributed to the U.S. market.
That is not a minor administrative gap.
That is the quality system failing at the point of decision.
The batch disposition decision is where a lot of theoretical control either becomes real or disappears.
Was the raw data reviewed?
Were unexpected injections explained?
Was an OOS result investigated?
Was the batch record telling the whole story?
Did the quality unit have enough evidence to release?
If the answer is unclear, the system is not controlled.
It is moving.
Those are not the same thing.
Stability Is a Promise You Have to Keep
Expiry dating is not a guess with confidence.
It is a commitment backed by data.
FDA cited stability program deficiencies, including lack of raw test data to support a long-term stability study, failure to place certain U.S. batches on stability, and stability samples stored in packaging different from the commercial container-closure system without evaluated equivalence.
Those details matter because stability is not just a lab program.
It is product truth over time.
Will the API maintain its quality attributes through expiry?
Can the company prove it?
Are the samples representative?
Is the container-closure system relevant?
Is the annual commitment actually happening?
A stability procedure that is not followed does not support confidence.
It supports questions.
Usually expensive ones.
Training Has to Reach Behavior
This is where leaders should be blunt with themselves.
If the same mistakes keep happening after training, the problem probably was not awareness.
It may be system design.
It may be workload.
It may be unclear authority.
It may be weak supervision.
It may be a culture where people know what the procedure says but also know what the site really rewards.
That last one is uncomfortable.
It is also common.
Training is often treated like the fix because it is easy to assign and easy to document.
But if the work environment still pushes people toward speed over evidence, the training will not hold.
You cannot train your way out of a system that rewards the wrong behavior.
You have to lead your way out.
The Operator Test
If you lead Quality, Operations, Validation, or the site, do not stop at the training matrix.
Pick a role that matters.
Then test whether the skill is being applied.
Can reviewers explain what raw data they are expected to check?
Can they identify when electronic records need deeper review?
Can analysts explain what makes an injection, result, or sequence scientifically questionable?
Can the quality unit show how it verifies batch disposition readiness?
Can stability commitments be traced from procedure to actual pulls, testing, data, and expiry support?
Can supervisors show how they confirm performance on the floor, in the lab, and in the review process?
That is the difference between training and capability.
Capability leaves evidence in the work.
The Real Work Is Managerial
This is not just a Quality problem.
It is a management problem.
People do what the system expects, reinforces, and checks.
If managers do not verify how work is performed, they should not be surprised when training records fail to translate into behavior.
If leaders do not protect time for proper review, they should not be surprised when review becomes a signature exercise.
If the quality unit does not have practical authority, they should not be surprised when batch decisions outrun evidence.
None of this requires consultant theater.
It requires basic operating discipline.
Define the role.
Train the skill.
Observe the work.
Challenge weak decisions.
Correct the system.
Verify the behavior changed.
Then keep watching.
That is how training becomes capability.
That is how quality-unit authority becomes real.
That is how a site moves from documented compliance to operating control.
The paperwork matters.
But the work has to prove it.
Website Publishing Notes
• Stage as Squarespace draft only.
• Set author to Two Pharma Guys.
• Add source URL in the post settings.
• Use the hero image and alt text provided above.
• Do not publish until Micah approves.