Manufacturers - 1
Without documentation of multiple mercury readings there is no assurance the temperature did not drop below the minimum temperature chosen for the evaluation. If the processor did not satisfactorily document the process conditions, determine if the processing authority recommended reprocessing or destruction of the product.
Missing entries on written retort logs are sometimes ignored. FDA has seen missing venting temperatures ignored in processing deviation evaluations, because the processing authority assumed that it was improbable that any problem occurred. Lack of recorded vent temperatures can be evaluated, by a processing authority, by performing temperature distribution tests without venting the retort at all, or venting the retort in a manner which duplicates the processing parameters documented by reliable processing records.
For example, when a processor's computerized control and recordkeeping system had a communications problem, the automated retort logs could not be generated. Instead of reverting to manual operator logging, the retort supervisor relied on the recorder charts. These processes could not be evaluated by a processing authority and all containers had to be reprocessed.
When evaluating processing deviations, processing authorities will often examine original heat penetration data and/or temperature distribution data to determine whether this data can be utilized. When such data is not applicable, simulations of processing deviations are necessary. Sometimes the simulations are more costly than the value of the affected product.
FDA dealt with a processor who had a critical factor of minimum hydration of dry beans before formulation. When the records showed less than the minimum hydration was achieved, the processor identified the incident as a processing deviation. The processing authority evaluated the deviation and determined the process should have been adequate, since hydration would have been completed during the process by exchange of moisture between the sauce and the beans. However, the processing authority had no way of knowing when the beans became adequately hydrated.
FDA has observed situations where processes were developed by one processing authority and processing deviations were evaluated by another processing authority. The evaluators did not have the original test data or protocols, and proceeded with the evaluations based on file data. FDA investigators should be suspicious of this technique for scheduled processes with critical factor deviations, or where complex formula or preparation procedures are involved.
Note: FDA does not object when a plant uses incubation results as an aid in the detection of spoilage, but will not accept incubation as the sole means to identify a problem with public health significance. Scientific techniques used to perform deviation analyses must stand alone.
Evaluation results must be reported in writing to the processor and include a brief explanation of the processing conditions evaluated, the method of evaluation employed (Ball, Stumbo, Cumulative Lethality and other methods), the specific processing parameters used for the evaluation and a summary statement of the specific records that were evaluated. The results of the evaluation must be provided. The actual Fo for the process in question and the significance of this value must be provided. The significance needs to be clarified in terms of meeting commercial sterility; less than commercially sterile, but sufficient to protect the public health; or less than minimum health. If the evaluation results indicate that a process is less than commercial sterility, but sufficient to protect the public health, then the actual minimum health value and its scientific reference need to be supplied in the evaluation report. If the product must be destroyed disposition procedures may be proposed. Document the firm's actions taken to reprocess or destroy product and report your findings in the EIR.
Written evaluation results must be in the possession of the processor prior to release of the lot into interstate commerce. If a processor releases the product prior to receiving the evaluation report a deviation from sections 21CFR 113.89 has occurred. Verbal releases are not considered appropriate, nor is applying a prior response to a seemingly identical processing deviation. Every processing deviation which is handled by evaluation must be identified in a written report.
As part of your inspection, assess the firm's handling of deviations. In cases where process deviations have occurred and their are questions regarding appropriate handling of deviations, collect the appropriate records as exhibits to the EIR. If questions cannot be resolved at the district level check with the Center to verify that deviations were properly handled. If CFSAN requests, include as part of the EIR, exhibits of firm's records and a copy of the processing authority's evaluation report for CFSAN review. Even in cases where there are no