Nonconformances, Data and CAPA
When something fails to meet a requirement, the organization needs to control the immediate problem and decide what further action is appropriate. Good decisions use evidence about risk, causes, and recurrence.
This lesson explains nonconformances, data analysis, and CAPA, meaning corrective and preventive action. Watch the differences between containment, correction, corrective action, and preventive action. Also notice why completing an action does not automatically demonstrate that it worked.
Control the immediate problem
A nonconformity is a failure to meet a requirement. Identify and control affected product to prevent unintended use or delivery, document the issue, and evaluate the appropriate disposition.
Determine whether investigation or notification of an external party is needed, and retain the rationale and subsequent actions. Consider already-delivered product when relevant.
A quality coordinator can organize containment and evidence. The disposition decision follows the assigned technical and quality authorities. Controlling the product creates space for a justified decision; it is not itself the final disposition.
Containment and correction
Containment limits immediate exposure. Holding a potentially affected lot is one example.
Correction removes an observed nonconformity. Reworking an item so that it meets its requirement is one possible correction, when that disposition is appropriate and controlled.
These actions address different needs. A hold can protect production or customers while the issue is evaluated. A correction can fix an affected item. Neither automatically identifies or removes the cause that allowed the problem to occur.
Corrective and preventive action
Corrective action addresses the cause of a nonconformity to prevent recurrence. Preventive action addresses a potential nonconformity and its cause before it occurs.
The distinction matters because fixing an item and changing the conditions that caused the failure are different results. It also allows the organization to act on warning signs before a requirement is actually missed.
Data analysis helps reveal both recurring failures and developing concerns. The response should be connected to the evidence, rather than selected because a familiar action is easy to complete.
Choose a proportionate response
Not every isolated issue requires an identical full CAPA project. Evaluate its significance, risk, and recurrence, and retain the reason for the response chosen.
A recurring or systemic problem can indicate a need for broader action. An isolated issue can also be important because of its consequences. Frequency alone does not tell the whole story.
The practical goal is a justified response that addresses the issue. Applying the same paperwork to every event does not replace thinking about what the evidence and risk call for.
Investigate before choosing the solution
Gather evidence before selecting a solution. Relevant information may come from product, process, supplier, service, complaint, or audit records.
Operator error is a description to investigate, not a reliable stopping point when unclear instructions, unsuitable equipment, or workload may have contributed. Look for evidence that supports the cause and relevant contributing conditions.
A cause-based action might involve equipment, design, instructions, training, or other controls. Retraining everyone is not automatically effective for every failure. Define actions, responsibilities, resources, and an implementation plan that fit the findings.
Implementation and effectiveness
Check that a proposed action does not harm device safety, performance, or regulatory conformity. Implement the action and retain evidence of what was done.
Then evaluate whether it achieved the intended result, using suitable criteria and follow-up data. Implementation evidence and effectiveness evidence answer different questions.
Closing an action on the day a procedure is revised may show that the revision happened. It may not show sustained effectiveness. Decide what evidence would demonstrate the intended improvement and review the actual result against that expectation.
Keep rework controlled
Rework also needs control. Use controlled instructions, consider adverse effects, obtain appropriate review and approval, verify against acceptance criteria, and retain the records.
An item that appears repaired is not automatically ready for release. The evidence needs to support that it meets the applicable acceptance criteria after the work.
Keep the rework decision connected to the nonconformity and the assigned authorities. That connection helps preserve both the product history and the basis for the eventual disposition.
Use data to see beyond one event
Analyze appropriate data from feedback, product conformity, process and product trends, suppliers, audits, and service activities. A trend that has not crossed a specification limit can still justify preventive action.
Imagine the same receiving defect appears in several shipments. Holding each shipment protects immediate production, but the pattern may warrant supplier investigation and corrective action.
Define how later deliveries will demonstrate effectiveness, review the results, and adjust the action if the defect continues. Handling individual events and learning from the pattern are both needed to improve the process.
What to carry into your work
Control the immediate problem, understand which type of action is needed, and investigate causes using evidence. Keep implementation and effectiveness distinct. Use trends to learn from repeated events and potential problems, and adjust actions when the results do not support the intended outcome.
This lesson draws on the AAMI and ISO practical guide to ISO 13485, sections eight point three through eight point five. It is introductory education. Applicable requirements and your company's approved procedures govern actual work.
SOURCE CREDIT AAMI/ISO 13485:2016 — Medical Devices: A Practical Guide (ISO 2017), §§8.3–8.5; purchased through ANSI. This video presents original educational explanations informed by the guide. It is not an official AAMI, ISO, ANSI or FDA course. Applicable requirements and company procedures govern actual work.