← Insights

Quality systems

Why CAPAs stay open forever

July 2026 · 7 min read

Open any mature quality system and look at the CAPA log. Somewhere in it there will be an action raised 200 days ago, still marked "Open", whose owner has changed jobs twice and whose original problem nobody quite remembers. Ask about it in a management review and the room does the same small wince.

CAPA - corrective and preventive action - is supposed to be the engine of improvement in a quality system. It is the mechanism by which an organisation learns. When the log silts up, the organisation has stopped learning; it is just filing its mistakes in date order. The silting has causes, and they are the same everywhere.

Cause one: the status column is manual

In most CAPA logs, "Status" is a cell somebody types into. Which means the log shows not the state of the action but the state of the last time anyone remembered to update the cell. An action can sail six weeks past its due date while the column still says Open in a calm, reassuring grey. Nothing changes colour, nothing escalates, nothing chases.

The fix is structural, not motivational: status should be computed, never typed. Given a due date, a closed date and an effectiveness-check date, a formula can derive everything that matters - Open, Overdue, Closed, Effectiveness check due, Reopened - and an overdue action turns red the day it goes overdue, whether or not anyone is watching. You cannot forget to update a formula.

Cause two: corrective actions that are really containment

"Reworked the batch and reminded the operator" closes tickets, but it corrects nothing - it is containment (D3 thinking, in 8D terms) wearing a corrective action badge. The defect will be back, and next time it will bring an audit finding about ineffective corrective action with it.

A corrective action, properly, removes the root cause. That means the log needs a root cause field that gets a genuine answer - a 5 Whys or Fishbone reference, not "operator error" - and a corrective action that changes the system: the fixture, the work instruction, the training, the incoming inspection. The test: if the person who made the mistake left tomorrow, would the fix still protect you?

Cause three: closure without proof

The deepest failure is the quietest one. An action is implemented, the log says Closed, and nobody ever checks whether the problem actually went away. ISO 9001 (10.2.1) is explicit: you must review the effectiveness of corrective action taken. That review needs three things a signature does not provide:

• A defined check, in advance. When we implemented the fix, what evidence would tell us it worked? Usually: no recurrence over a defined window, or a measured improvement in the metric that first flagged the problem.
• A date when the check falls due - far enough out for recurrence to have had a fair chance. Thirty to ninety days is typical.
• An honest verdict. Effective: now it is genuinely closed. Not effective: the CAPA reopens - back to root cause, because the first analysis was wrong. A reopened CAPA is not a failure of the system; it is the system working.

A CAPA is not finished when the action is done. It is finished when the problem has demonstrably stayed gone.

What a working log looks like

Beyond computed status, two features separate a live log from a graveyard:

Aging you can see. Days-open should count live, per action, so the management review sees "47 days and climbing", not "Open". Average days-to-close is the single best health metric for the whole corrective action system - watch its trend, not its value.

A source breakdown. Every CAPA comes from somewhere: internal audit, customer complaint, NCR, supplier issue, management review. Count them by source and you get a Pareto of where your problems originate - which is itself the start of the next preventive action. A CAPA log that can answer "what keeps hurting us?" has graduated from record-keeping to intelligence.