Manufacturers - 1
may intensify the heat stable enterotoxin.
Reworking or blending into new formulation is another method of reprocessing. Whenever performed, the scheduled process must cover the level of rework material and procedure for rework. Some original scheduled processes may allow for a certain level of rework (e.g. 2%). Some processors are unaware of the potential for rework to affect the scheduled process; and in some cases processing authorities have been unaware that processors rework products.
Determination of a new initial temperature for the reprocess is also often inadequately addressed. For example, product which must be cooled prior to reprocessing may have an initial temperature below that required in the scheduled process, or product taken from a warehouse at 500F may not receive the same process as product from the line at an initial temperature of 1200F.
Reprocessing must be properly documented. Some processors fail to fully record actions taken, especially when the original process was corrected by a reprocess. Inadequate documentation in batch records and in the deviation file, can result in uncertainty as to whether a reprocess was applied. This could jeopardize an entire production period.
ISOLATION OF AFFECTED PRODUCTION
When processing deviations are detected in the thermal processing area, affected product is usually easy to isolate. If processing deviations are detected through record review, isolation of affected production is often complicated. Isolation usually encompasses a larger hold if the affected production has been commingled with other containers during casing and storage. Failure to execute isolation is not uncommon.
Processing deviations in product control factors such as formulation, viscosity and fill weight often result in having to assess which retort loads, or time spans for continuous processes, were affected. Deviations involving continuous systems require careful product isolating techniques.
When examining records, FDA investigators should assess a processor's documentation of isolation.
Regardless of the factor that is out of control, the processor should hold product back to the previous acceptable check. For example, a vent valve malfunction, where the valve would not fully open, would involve determining how many deviations resulted from this malfunction, and when did the malfunction first occur? If a fill weight was found to be high, the problem would extend back to the last full weight check that was acceptable. The processor is responsible for bracketing the full extent of the problem.
SUBMISSIONS TO PROCESS AUTHORITIES
Submission of pertinent records and information regarding the processing deviation is critical to the evaluation by the processing authority. Appropriate product identification, code, production date, processing vessel identification, cycle or time span, quantity of product held, nature and cause of the deviation, corrective actions, copies of attending records, including critical factor records, and the status of the lot are keys to evaluation.
Some processing authorities, if not provided with all records covering critical processes in the manufacturing procedure, will base their evaluation only on those records provided by the processor. For example, if headspace is a critical factor, and the processor does not provide copies of those records, some processing authorities will assume that headspace was properly controlled. Because of this, in addition to reviewing the processing authority's evaluation report, you should also review the records the processor submitted to the processing authority with the request for evaluation. This request should identify the cause of the deviation. In some cases a processing authority may accept recorded values without analyzing the records. Consider the example of a fill weight deviation which exceeded the maximum by one ounce, but had fill weights determined at wide intervals. FDA would have expected the firm to have collected and analyzed a sufficient number of additional samples to identify the worst case. Considering statistical variance, it is doubtful that the worst case is known. If the heat penetration test data, used to clear this deviation, were based on a fill weight of one ounce over the maximum, the evaluation may be errant. Assess whether the firm has submitted the additional information needed to the processing authority.
FDA has observed numerous evaluations of deviations caused by missing ink from recorder temperature tracings, or missing entries on operator's logs. For a steam retort an incomplete temperature tracing backed-up by a retort pressure pen tracing showing no loss of pressure is often proof that the scheduled process temperature was maintained. The preferred method of handling would be operator documentation of multiple mercury readings during the occurrence. In the absence of documentation, CFSAN looks unfavorably when an evaluation is performed by choosing the lowest retort temperature which meets a safe process.