An out-of-specification (OOS) result in the testing of a biopharmaceutical product can necessitate investigations, rework, closeouts, costly delays, and can carry regulatory implications. These costs and delays can be frustrating and may also, at times, be unnecessary. For example, testing may indicate an OOS result that turns out to be false because the testing methodology is not robust enough, specifications are tighter than necessary, or testing is being conducted for an attribute that is not a regulatory requirement. Similarly, legacy drugs may suffer "methodology creep", that is, accumulating over their lifecycles more and more testing that is not relevant to compliance.

As these examples indicate, unnecessary costs often originate in analytical testing. Costs, however, can be reduced by rigorously questioning the purpose of each analytical test and pushing the understanding of compliance deeper into the organization. Given the large number and complexity of tests that large molecule drugs must undergo, those cost savings can be significant.

Analytical methods should address the specific requirements of compliance. At every stage of development, ask what must be demonstrated, what is compliant, and whether the test data are presenting crucial information about the material or product. Except for the database of information maintained about what goes into the final product, any other information should be regarded as nice, but not necessary, to know. In evaluating the relevance of testing methods for legacy products, look carefully at exactly what was approved and why for the product. If "method creep" has occurred, roll it back. In some cases, this finding could lead to working with FDA to eliminate unnecessary tests or information. In addition, be smart about which tests, product attributes, and process parameters are submitted to FDA. One can avoid getting stuck with the cost and time of having to submit data or maintain attributes that are not relevant to compliance or product quality.

There are essentially three sets of analytical testing that need to be done. Those that occur during development through Phase II clinical trials, those during Phase III clinical trials through approval and the entire lifecycle of the marketed product, and as part of in-process testing. Each set entails distinctive challenges and requires the ability to assess risk in order to determine whether a particular test is genuinely necessary.