CAPA That Prevents Recurrence

Part of Clinical Project Management (PM) Training: Risk-Based Quality Management.

About 1 minutes. Free, in your language, no account needed.

Correction is not corrective action

Fixing the wrong date on one case report form is a correction. Finding out why the date was wrong, and stopping it happening again, is corrective action. Most CAPA records stop at the first and are labeled as the second.

The test is simple: if the same conditions recur tomorrow, does the same error happen? If yes, nothing was corrected.

Root cause, honestly

"Human error" is where root cause analysis goes to stop. People make errors when a process invites them: a form with ambiguous fields, training that happened once eleven months ago, a system that accepts an impossible value without complaint.

Ask what made the error easy. That question has an actionable answer; "the coordinator was careless" does not.

Scope it across the trial

An error found at one site is rarely unique to that site. If a form is being misread, it is probably being misread elsewhere. A CAPA that fixes site 114 and does not ask about the other eleven sites has treated a symptom.

Effectiveness has to be checked

A CAPA closed on the day the action is completed is closed on the wrong evidence. The action was taken; that is not the same as the problem being solved.

Set a date to look again at whatever metric showed the problem. If it has not moved, the CAPA failed and the analysis was wrong. Closing it anyway produces a file full of actions that prevented nothing.