Standards

Hello, this is my first question to the community. When addressing nonconformities and corrective actions, we are asked to "review the effectiveness of any corrective action taken". How are people doing this? This seems to be a sticky point for us. Would additional training, document review/update and the identification of risks and opportunities address this point? I'd welcome your input on this.

 

There are three factors to consider for each effectiveness review

1. Method: What technique will be used to gather data to determine CAPA effectiveness?  A method exists by which you can measure just about anything. Examples: Audit, Spot Check, Sampling, Trend Analysis, Monitoring, Product Review. Deviation/nonconformity database lookbacks are not a recommended ER data gathering method.  Not all events are immediately detected or captured as a deviation, and this type of lookback is a lagging indicator of lack of success. Audits and spot checks can likely yield more immediate results.
   
    


2. Criteria:  What quantitative or qualitative criteria will determine your CAPA success?  A success rate of 100% is not always a realistic result of CAPA implementation.  A CAPA which results in a significant stepwise change (e.g. 30% RFT → 75% RFT) could be considered very successful.
   
    


3. Timeline: It is suggested to select the shortest timeline necessary to gather the required amount of data needed to make an effectiveness determination.  If your timeline is long, consider dividing the period into segments, with each having an effectiveness review on a portion or subset of data.

It is part of the 8D and this step needs to demonstrate or measure that the root cause has been effectively addressed.  It depends on the business and the defect, but it could be as simple as testing to try to turn on and off the root cause  It also could be verified through a period of extra inspection after the corrective action has been implemented using an I-chart to record 100% inspection or gaging for at least 30 days or 3 different manufacturing dates.  Another option is to do a process audit to ensure people is actually following the new steps or processes implemented. 


On the other hand, you could evaluate the solution based on whether the solution is technical or behavioral.  Behavioral solutions are those that depend on the operator to make a decision. Then, evaluate if the solution detects the defect or prevents it from happening, and whether this happens at the source of the root cause or downstream. You could then use the PFMEA to score the risk.