Every quality manager has lived this: a corrective action closes, everyone signs off, and three months later the exact same defect walks back through the door. The CAPA reopens, the customer notices, and the confidence you spent building drains out of the room.
When a corrective action reopens, it's telling you something specific. It didn't find the root cause. It found a symptom that looked close enough.
Containment is not correction
The most common failure isn't laziness — it's stopping too early. A part fails in the field, the team sorts inventory, adds an inspection, and re-trains the operator. The defect stops shipping, so the action looks successful. But sorting, inspecting, and re-training are containment. They protect the customer today. They don't change the reason the defect was created in the first place.
Chase the cause of the cause
The discipline that separates a fix that holds from one that reopens is refusing to stop at the first plausible answer. A 5-Why done well doesn't just ask why the part failed — it asks why the condition that caused the failure was allowed to exist.
Here's a real shape it takes on the floor:
- The seal failed adhesion in the field. Why? The bond was weak.
- Why was the bond weak? Primer film thickness was below spec.
- Why was it below spec? The spray nozzle was partially clogged.
- Why was it clogged? The inline filter change interval was too long.
- Why was the interval too long? The PM standard never listed that filter.
The first four answers all produce containment actions — clean the nozzle, re-check thickness, add an inspection. Only the fifth changes a standard so the problem can't recur. Stop at why number two and you've written a CAPA that will reopen.
A corrective action that closes and stays closed almost always ends at a document, a standard, or a design — not at a person.
Verify effectiveness on purpose
The last mistake is closing a CAPA the day the change is made. A root-cause fix isn't proven until the process has run long enough to show the defect is gone. Define, in advance, what evidence closes it — a run of parts, a number of shifts, a capability study — and hold the action open until that evidence exists. It feels slower. It's the reason it never comes back.
The short version
CAPAs reopen because we mistake catching a defect for preventing it. Chase every symptom to the standard, design, or system that allowed it. Change that. Then verify the change held before you close. That's the whole discipline — and it's most of what we do on the floor with quality teams every week.
