AccuReg: Handling OOS Results

The Regulatory Clinic: October 1995
Handling Out-Of-Specification (OOS) Results

Problem:

I recently assumed a position as QC Laboratory Manager at a large pharmaceutical company. I was surprised to discover that no formal, documented plan for handling out-of-specification (OOS) test results exists. I want to establish an efficient and compliant plan, but am encountering significant resistance from all sides. Can you give me a concise explanation of why these OOS result plans are necessary and useful - something I can provide to upper management, our production staff, and my technicians? I feel sure some background information would help support my recommendations. Thanks.

Solution:

Without the proper tools, handling OOS results can become a show-stopping headache and a regulatory minefield. Is this a true product failure? Maybe you should resample. Or is it laboratory error? Perhaps a simple retest is in order. The result is only out by a bit, so is it possible to perform more tests and obtain a passing average? Or can the OOS result be proven to be a statistical outlier? If nothing else, what did you do the last time this happened?

You are wise to recognize that a pre-determined and documented plan is the most efficient and compliant mechanism for resolving OOS results. When utilized properly, a formal OOS result procedure also helps prevent future OOS results, and it functions as an excellent diagnostic tool for identifying and mitigating recurring problems within the laboratory.

Since the issuance of the Barr Decision in 1992, mishandling of OOS results has become an increasingly common observation on FDA 483s. With a systematic and scientific approach, a firm can now invalidate (or overcome) the OOS result within the laboratory, that is to say, an OOS result does not necessarily mean a product failure, if the error can be proven to have occurred in the laboratory. Obviously, this a far more efficient manner to dispatch an OOS result, and avoids a costly product failure investigation. However, FDA has specific expectations as to how this may be accomplished, as described in the 1993 Guideline to Inspection of Pharmaceutical Quality Control Laboratories. Here are some pointers to bear in mind as you develop an OOS result handling procedure:

Familiarize yourself with tenets of the Barr Decision, which set the stage for the basic requirements of a properly handling an OOS result. And remember, an OOS result ignored or mishandled during any validation testing can negate the whole study!
Develop a methodical and systematic investigation plan. Documentation should include a comprehensive laboratory investigation checklist. Trend analysis of routine analytical QC results should also be considered.
Plan for implementation, assessment, and follow-up of findings or recommendations from the investigation. Clearly delineate the direction of the investigation if the OOS result is confirmed. Generally this leads into a formal product failure investigation procedure.

You mentioned that your company is a large pharmaceutical firm, probably with a long history of manufacturing and a relatively high volume of production. No doubt your administration is well aware that even in the best QC labs, mistakes can happen. If you can help them realize that a pre-defined plan for handling OOS tests results will enable you to efficiently diagnose, cure, and prevent future problems, chances are they will support (and later, applaud) your effort to establish a documented plan.

Do you have a problem you'd like to see addressed in the Clinic? Send your question to accureg@regulatory.com. We'll send a pocket-size copy of the GMPs for pharmaceuticals, devices, or both, in a convenient wallet to individuals whose regulatory or compliance questions are posted in the Clinic.

7501 Northwest 4 Street, Suite 210
Plantation, Florida 33317, USA
Telephone: 954-641-6400 Fax: 954-641-6410
Email: accureg@regulatory.com

Contact webmaster