Inspections, Compliance, Enforcement, and Criminal Investigations
StarKist Company 3/1/11
Department of Health and Human Services
|Public Health Service|
Food and Drug Administration
|San Francisco District|
1431 Harbor Bay Parkway
Alameda, CA 94502-7070
VIA UNITED PARCEL SERVICE
Our Reference: FEI 2915096
March 1, 2011
Mr. Ingu Park, President/CEO
323 North Shore Drive, Suite 600
Pittsburg, Pennsylvania 15212
Dear Mr. Park:
We inspected your seafood processing facility, located at 368 Main Road, Atu”u, American Samoa, on November 1 – 4, 2010. We found that you have serious violations of the seafood Hazard Analysis and Critical Control Point (HACCP) regulation, Title 21, Code of Federal Regulations, Part 123 (21 CFR 123). In accordance with 21 CFR 123.6(g), failure of a processor of fish or fishery products to have and implement a HACCP plan that complies with this section or otherwise operate in accordance with the requirements of Part 123, renders the fish or fishery products adulterated within the meaning of Section 402(a)(4) of the Federal Food, Drug, and Cosmetic Act (the Act), 21 U.S.C. § 342(a)(4). Accordingly, your canned tuna and pouched packed tuna are adulterated, in that they have been prepared, packed, or held under insanitary conditions whereby they may have been rendered injurious to health. You may find the Act, the seafood HACCP regulation and the Fish and Fisheries Products Hazards & Controls Guidance through links in FDA's home page at www.fda.gov.
During the November 2010 inspection, we also determined your compliance to the Thermally Processed Low-Acid Foods in Hermetically Sealed Containers regulations, 21 CFR 113 and identified some deviations from the regulations.
At the close of our inspection, the investigators provided Johnny R. Elgin, Quality Assurance Manager, with the form FDA 483, which presents their evaluation of your firm’s performance regarding various aspects of the HACCP and Low Acid Foods requirements. We attached a copy of the FDA 483 for your reference.
Your serious deviations are as follows:
These deviations were noted upon further review of your HACCP plan:
(1) You must conduct or have conducted for you a hazard analysis for each kind of fish and fishery product that you produce to determine whether there are food safety hazards that are reasonably likely to occur and have a HACCP plan that, at a minimum, lists the food safety hazards that are reasonably likely to occur, to comply with 21 CFR 123.6(a) and (c)(1). A food safety hazard is defined in 21 CFR 123.3(f) as any “biological, chemical, or physical property that may cause a food to be unsafe for human consumption.” However, your firm’s HACCP plan for tuna does not list the food safety hazard of Staphylococcus aureus growth and toxin formation as a result of time/temperature abuse. Your firm receives frozen, vacuum packaged, precooked tuna loins from (b)(4) which your firm thaws and holds in the unopened packaging at elevated temperatures. FDA recommends that cumulative temperature exposures, which include the thawing process and other processing steps, be limited to less than 3 hours to control the food safety hazard of Staphylococcus aureus when temperatures exceed 70°F at any point.
During the November 2010 inspection, FDA observed several racks of thawed, pre-cooked loins still in their plastic vacuum packages staged in the processing room and stacked on production tables, staged for further processing. Temperatures in American Samoa range from 76°F – 81°F throughout the year.
We have observed that your thawing process for your pre-cooked tuna loins is different from the method used for the frozen tuna. This step may need to be addressed as a separate critical control point (CCP) from your currently listed thawing critical control point.
We acknowledge receipt of a letter from Mr. John Aaron Maxfield, Quality Assurance Director, dated December 2, 2010. To address the above deviation, Mr. Maxfield states in his letter that StarKist is updating the HACCP plan to include a (b)(4) on the processing of loins. We will verify your correction at the next inspection.
(2) You must conduct a hazard analysis to determine whether there are food safety hazards that are reasonably likely to occur and have a HACCP plan that, at a minimum, lists the critical control points to comply with 21 CFR 123.6(a) and (c)(2). A critical control point is defined in 21 CFR 123.3(b) as “a point, step, or procedure in a food process at which control can be applied, and a food safety hazard can as a result be prevented, eliminated, or reduced to acceptable levels.” However, your firm’s HACCP plan for tuna fails to list critical controls points for:
o The processing steps following thawing where the tuna is exposed to unrefrigerated conditions to control scombrotoxin formation. Your plan currently lists only a critical control point for the (b)(4) process. In addition to thawing, FDA recommends monitoring the cumulative time/temperature exposures of the tuna during all unrefrigerated processing steps from your Thawing critical control point until retorting.
Our previous inspection observed that some pre-cooked fish only achieve (b)(4) internal temperatures of 130°F after the (b)(4) hour pre-cook process. Your firm should provide evidence that normal pre-cooking times/temperatures will inhibit or prevent scombrotoxin formation. In lieu of scientific evidence demonstrating that the pre-cook process will inhibit or prevent scombrotoxin formation, FDA recommends the handling steps after the unloading of the thawing tanks, including butchering, pre-cooking, post-pre-cooking processing steps, and any staging or holding periods be included in cumulative time/temperature monitoring. Your plan currently lists a critical control point for thawing, which appears to cover the time from (b)(4).
In addition, it is likely that the frozen pre-cooked tuna loins you receive and the frozen tuna you receive may require different critical limits for the cumulative time/temperature exposures.
o The (b)(4) cleaning, the (b)(4) packing process, and the (b)(4) for retorting steps to control the food safety hazard of Staphylococcus aureus growth and toxin formation in the tuna processed from frozen raw tuna. Our previous inspection observed that your fish are only partially cooled after pre-cooking and then significantly handled by employees during the cleaning and canning process which occur at ambient temperatures above 70°F. FDA’s safety guidance recommends that exposure times above 70°F be limited to 3 hours under these conditions to control Staphylococcus aureus growth and toxin formation.
(3) You must have a HACCP plan that at a minimum, lists monitoring procedures and their frequencies for each critical control point, to comply with 21 CFR 123.6(c)(4). However, your firm lists a monitoring procedure in your HACCP manual for scombrotoxin testing that is inadequate. Your HACCP manual lists that you will obtain samples for scombrotoxin testing from the (b)(4) portion of the fish. Our investigator observed that you implemented this procedure by obtaining samples from the (b)(4) portion, near the fish (b)(4); this portion of the fish is unlikely to detect scombrotoxin in a fish containing elevated scombrotoxin and therefore renders the scombrotoxin testing control approach applied ineffective and consequently inadequate for HACCP monitoring purposes.
Because scombrotoxin is generally not uniformly distributed in a fish or a lot, the validity of scombrotoxin testing is dependent upon the design of the sampling plan. The amount of sampling required to accommodate such variability of distribution is necessarily quite large. The method of collection of the fish sample is also critical. In large scombrotoxin-forming fish, the lower, anterior (forward) portion of the fish loin (not the belly flap) is likely to provide the best information about the scombrotoxin content of the fish.
(4) You must implement the monitoring procedures and frequencies that you have listed in your HACCP plan, to comply with 21 CFR 123.6(b) and (c)(4). However, your firm did not implement your frequency of (b)(4) listed at the Receiving critical control point. Your HACCP manual defines lots under monitoring procedures as (b)(4). Our investigator observed that a lot of fish examined by FDA on November 2 was combined from the vessels (b)(4). This combining of product from multiple vessels does not meet your definition of a (b)(4).
When FDA determined that (b)(4) tuna from (b)(4) were decomposed, your firm provided us with documentation listing that wells (b)(4) and (b)(4) from the (b)(4) were combined into a single lot of which the decomposed fish found by FDA weighing (b)(4) comprised only (b)(4) of that. The lot was accepted and no corrective action was taken beyond removal of the detected decomposed fish. The lot contained both product from multiple wells aboard a single vessel and product from a single container. This combination of product from multiple vessel wells and a container does not meet your definition of a (b)(4).
(5) You must verify that your HACCP plan is effectively implemented as required by 21 CFR 123.8(a). However, your firm did not adequately verify that your staff responsible for detecting decomposition in raw tuna effectively performed their assigned tasks as evidenced by your (b)(4) personnel’s failure to detect decomposition in raw tuna during the (b)(4) analysis.
During the current inspection, FDA sensory analysts detected odors of decomposition in several tuna from (b)(4) at the Receiving critical control point, and from thawed tuna lots, on the (b)(4) after examination by your (b)(4) personnel. Specifically, on November 3, 2010, FDA sensory analysts found (b)(4) decomposed (b)(4) tuna in a (b)(4) from (b)(4). Decomposed fish were also found by FDA sensory analysts on November 2, 3, and 4, 2010 on the (b)(4) after the lots were examined and passed by your (b)(4) personnel.
During the 2009 inspection of your facility, FDA’s National Sensory Expert also found decomposed fish after the fish had been examined and passed by your sensory personnel. Our expert discussed the observation with management and recommended corrections. Management indicated agreement and they would consider our expert’s recommendations.
We acknowledge that the tuna identified by FDA as decomposed were set aside and not processed for sale; however, we note that an adequate corrective action was not taken with regard to the lot of fish.
(6) You must take an appropriate corrective action when a deviation from a critical limit occurs, to comply with 21 CFR 123.7(a). However, your firm did not take an appropriate corrective action to control scombrotoxin formation when your process for tuna in cans or pouches deviated from your critical limit at the “Thawing” critical control point.
Specifically, your HACCP monitoring record (b)(4) dated 10/21/10 showed that the cumulative thaw time of (b)(4) tuna in (b)(4) was approximately 14.5 hours (b)(4) thaw water temperature measured at (b)(4) was 70°F. The cumulative thaw time of (b)(4) tuna in (b)(4) was approximately 13 hours, 50 minutes (b)(4) thaw water temperature measured at (b)(4) was also 70°F. Your firm exceeded your critical limit at the Thawing critical control point, but there were no records to show that a corrective action was taken when necessary.
Your HACCP plan lists that you will (b)(4) when your critical limits are exceeded at the Thawing critical control point. Your firm has stated that when your critical limits at the thawing critical control point are exceeded you will (b)(4) the tuna to ensure that the entire process is completed in (b)(4) hours. FDA recommends that when scombrotoxin forming species are exposed to temperatures above 70°F that the cumulative exposure time for temperatures above 40°F be limited to 12 hours. The tuna was already exposed to temperatures above 70°F during thawing and that process took more than 12 hours. (b)(4) the tuna to meet a (b)(4) deadline is not an appropriate corrective action to ensure product safety.
We note that the corrective action listed in your plan limits the corrective action cumulative time period from thawing to precooking. FDA has no scientific evidence that the precooking process inhibits or is intended to inhibit scombrotoxin formation. The precook process has historically been intended solely to facilitate the cleaning of the fish carcass, not to control scombrotoxin formation or microbial growth. FDA recommends the cumulative exposure time periods extend through all unrefrigerated processing steps to the start of the retort process unless scientific evidence can show that the pre-cook parameters inhibit scombrotoxin formation.
(7) Because you chose to include corrective actions in your HACCP plan, your described corrective actions must be appropriate, to comply with 21 CFR 123.7(b). However your corrective action plan for tuna at the (b)(4) to control scombrotoxin is not appropriate. Your HACCP plan lists (b)(4). Neither your HACCP manual nor HACCP plan indicate what happens to decomposed fish at this point. The narrative in your HACCP plan indicates that if the total of the rejected products from the sensory inspection is (b)(4) but not what happens to the individual fish detected as decomposed when the total rejected products do not exceed (b)(4).
In addition, your corrective action plan listed in your HACCP plan and HACCP manual at the (b)(4) is also not appropriate. Your HACCP plan lists (b)(4). It is inappropriate to rely on finished canned product sampling and testing to determine if there is a hazard. FDA has determined that sampling of canned or pouched tuna products does not provide reliable results representing the condition of the fish in the lot and is inappropriate as a corrective action for ensuring the safety of products that have been previously found to not meet your critical limits.
(8) You must fully document, in your records, all corrective actions taken, to comply with 21 CFR 123.7(d). However, our review of your HACCP monitoring record (b)(4) dated 10-22-10, listed a (b)(4) taken from (b)(4). Your record does not show that the (b)(4) as outlined in your HACCP manual. Your HACCP manual indicates that if any composite sample has a scombrotoxin content exceeding (b)(4), the (b)(4) samples making up the (b)(4) shall be analyzed (b)(4) for scombrotoxin content. The (b)(4) were not documented on the record in the (b)(4) column provided on the record.
We also have the following comments regarding your HACCP plan:
- Your firm’s HACCP plan for tuna lists a critical limit (b)(4) critical control point that is not adequate to control histamine formation. According to your HACCP plan, this (b)(4)critical limit applies to (b)(4) FDA’s action level for histamine is equal to or greater than 50 ppm (5mg %) each individual fish in the sample. The scombrotoxin (histamine) critical limit must be reduced accordingly if the fish collected for analysis are composited. Your firm (b)(4) samples from (b)(4) fish and tests them as a single sample. Your critical limit should reflect this and histamine levels should not exceed (b)(4). This deviation was noted upon further review of your HACCP plan.
- The critical limits for scombrotoxin testing listed in your HACCP plan and the text of your HACCP manual are inconsistent. It appears in all situations that you initially (b)(4)and then if the (b)(4) appears violative, test (b)(4) from each subsample individually. These two steps require two different critical limits, which are noted in the HACCP manual. Your plan does not reflect this and should be corrected to be consistent with your HACCP manual.
- Our investigator observed that your firm uses the (b)(4) of decomposed fish against the (b)(4) of the sample of (b)(4) fish. FDA’s guidance recommends determining the percentage of decomposition within the sample using the count of decomposed fish against the total count of 118 fish. Neither your HACCP plan nor your HACCP manual clearly describe your practice of using (b)(4). Please provide us with a scientific explanation of how using the (b)(4) of decomposed provides a comparable outcome to using the (b)(4) of decomposed fish carcasses. Your plan should be corrected to clearly describe your procedures.
Low Acid Canned Foods:
(1) In accordance with 21 CFR 113.60(a)(3), for closures other than double seams and glass containers, you must perform appropriate detailed inspections and test intervals of sufficient frequency to ensure proper closing machine performance and consistently hermetic seal production.
Specifically, your firm does not perform a meaningful destructive test on heat sealed pouches of tuna products at intervals of sufficient frequency under conditions of manufacture. For example, the frequency of burst tests that you perform on filled and sealed pouches (one 2.6 oz pouch per retort load of (b)(4) pouches) is not sufficient to detect pouches with leaking seals or weak seals that have potential to leak. You also conduct a tensile strength test on (b)(4) empty pouches on each pouch filling line approximately every (b)(4) hours during production. These test pouches do not reflect the actual conditions of manufacture because they do not contain product. Further, your firm visually inspects (b)(4) sealed pouches sampled consecutively from your pouch hot bar sealer on line (b)(4) every (b)(4) minutes during production. However, the record of this visual seam examination does not document the specific sealing head associated with each sampled pouch.
Mr. Maxfield’s letter of December 2, 2010 states that StarKist Co. now incorporates burst testing in their destructive testing protocol, in addition to tensile strength testing which the firm has historically performed. Mr. Maxfield’s letter also states that it is StarKist’s procedure to collect (b)(4) samples from a (b)(4) sealing machine, to ensure each sealing head is visually inspected, every (b)(4) minutes even though head numbers are not recorded. Mr. Maxfield claims that this practice parallels the industry standard for visual double seam checks. In addition, Mr. Maxfield’s letter states that StarKist is currently reviewing your practices company wide and will take FDA’s recommendation under consideration.
Your firm’s destructive tensile strength testing is not appropriate. The pouches that are tested are not first filled with tuna fish. Instead, the test pouches are empty 4 side seal pouches (sealed by the pouch manufacturer on two sides and 1 end) as received from the pouch manufacturer without being first subjected to the filling process. The pouch filling machine on Line(b)(4) has a tendency to contaminate the sealing area of the packer’s end during filling that can result in the contamination of (b)(4) pouches. Such contamination may not be detected by visual means and can result in leakage and contamination after filling and subsequent processing. Although your firm has now incorporated burst testing in your destructive testing protocol, this test is not performed at intervals of sufficient frequency representing all (b)(4) heat sealing heads. Your burst test is conducted on only one pouch for every retort load of (b)(4) pouches.
(2) In accordance with 21 CFR 108.35(h), upon written demand during the course of a factory inspection pursuant to section 704 of the FD&C Act by a duly authorized employee of the Food and Drug Administration, the commercial processor shall permit the inspection and copying by such employee, all records of processing, deviations in processing, container closure inspections, and other records specified in part 113, to verify the adequacy of processing, the integrity of container closures, and the coding of the products. Specifically, you did not permit inspection and copying of the following records covering the production of StarKist tuna in 3 oz and 43 oz pouches produced at your American Samoa Processing Plant between August 1, 2010 and November 4, 2010:
You are reminded that failure to comply with the requirements of Section 704 of the Act is a prohibited act under Section 301(f).
Mr. Maxfield claims in his letter that these summary reports are not required in 21 CFR 113 and is not part of your food safety program.
FDA requested the above records to determine the firm’s actual percent defects under the conditions of manufacture on Lines (b)(4). The information from these records is necessary to determine the capability of your filling and sealing equipment to consistently produce pouches of finished tuna fish with hermetic seals. During the November 2010 inspection, the investigators noted that your firm is finding more pouch defects (b)(4).
We may take further action if you do not promptly correct these violations. For instance, we may take further action to seize your product(s) and/or enjoin your firm from operating. We may also refuse admission of your tuna product under Section 801(a) of the Act (21 U.S.C. §381(a)), including placing them on "detention without physical examination."
You should respond in writing within fifteen (15) working days from your receipt of this letter. Your response should outline the specific things you are doing to correct these violations. You should include in your response documentation such as HACCP, LACF, and verification records, or other useful information that would assist us in evaluating your corrections. If you cannot complete all corrections before you respond, you should explain the reason for your delay and state when you will correct any remaining violations.
This letter may not list all the violations at your facility. You are responsible for ensuring that your processing plant operates in compliance with the Act, the seafood HACCP regulation (21 CFR Part 123), the Current Good Manufacturing Practice regulation (21 CFR Part 110), and the Good Manufacturing Practices for Low Acid Canned Food (21 CFR 113). You also have a responsibility to use procedures to prevent further violations of the Act and all applicable regulations.
Please send your reply to the U. S. Food and Drug Administration, Attention: Ms. Erlinda N. Figueroa, Compliance Officer, 1431 Harbor Bay Parkway, Alameda, CA 94502-7070. If you have questions regarding any issues in this letter, please contact Ms. Figueroa at (510) 337-6795.
Barbara J. Cassens
San Francisco District
FDA 483 dated 11-04-10
cc: John Aaron Maxfield
Director of Quality Assurance
323 North Shore Drive, Suite 600
Pittsburgh, Pennsylvania 15212
Brett B. Butler, General Manager
P. O. Box 368
Pago Pago, American Samoa 96799
Johnny R. Elgin, Quality Assurance Manager
P. O. Box 368
Pago Pago, American Samoa 96799