Assessment of the cause for a non-compliance is a structured series of branching questions. The answers to these eliminate possibilities and eventually highlight a root cause. This process is described, including roles and responsibilities, timing, and efforts at prevention.

In the case of an apparently anomalous data point, assessment of whether it really is an outlier using statistics are done. If so, assessment of an observable cause is done to find the blunder. If one is not found, the process for handling outliers is followed.

Areas Covered 

Why Should You Attend 

There are many causes for a bad result. The main symptom is a result outside is +/- 3σ. This is a GLP non-compliance event. A root-cause / failure analysis must be done. What are the steps? There are both immediate in-laboratory and outside-of-laboratory steps.

When there is a non-compliant result, immediate steps involve both the laboratory personnel and management. Internally, a review of the specific conditions for the bad result must be assessed. This includes checking control charts to see if trends in performance occurred. All logbooks and recordkeeping related to that analysis must be examined carefully. An initial report to the lab manager is made.

Meanwhile, all regular operation of that method must stop and customers must be informed. Root-cause teams are set up to assess the preliminary assessment as well as control charts and other operational data to see if a cause is clearly shown. Timing of the non-compliance is compared to these records.

Closure is done through issuing a root-cause report with corrective actions. Revalidation is done. Customers are informed.

There are also several types of situations for a laboratory operating under Good Laboratory Practices (GLP) where assessing whether results are acceptable and must be used or unacceptable and can be ignored. Results with blunders, the possibility of outliers, and other questionable results can be tricky and a source of serious non-compliance events. Blunders are the easiest type of bad result.

Documenting blunders must be done thoroughly, though. Outliers are more problematic, but can be clearly dealt with. Outlier testing is crucial.