Before reworking batches that do not meet standards or specifications, the reason for the non-conformance should be investigated. Reworked batches should be evaluated, tested, stability-tested if warranted and documented to show equivalent quality to the original process, concurrent validation often being suitable (a protocol defining the procedure, its conduct and expected results), or for a single batch a report with release once acceptable. Procedures should compare each reworked batch's impurity profile with batches from the established process, adding methods where routine ones cannot characterise the reworked batch.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.