• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Food

  • Print
  • Share
  • E-mail

FDA's Evaluation of the Seafood HACCP Program for 1998/1999

December 8, 2000

Dear Colleague:

Attached is an evaluation of FDA's Hazard Analysis Critical Control Point (HACCP) Program for seafood for calendar years 1998 and 1999. HACCP is a science-based system of preventive controls for food safety that commercial food processors develop and operate to identify potential problems and keep them from occurring. The evaluation covers implementation of the HACCP program by about 3600 U.S. seafood processors, most of which are small businesses, and which process over 350 species of fish. The FDA HACCP program was designed to increase the margin of safety that U.S. consumers already enjoyed and to reduce those illnesses that do occur to the lowest possible levels.

The report demonstrates substantial progress by the seafood processing industry in implementing the full range of preventive controls that became mandatory in December 1997, as measured against a baseline study performed in 1992. Continued steady progress characterizes the change between 1998 and 1999. A significant majority of processors are doing well on most of the individual elements of the program, and over half have succeeded in all elements – a difficult standard to achieve. Thus far, implementation of these state- of-the-art preventive controls by the seafood processing industry contributes a significant increase in the margin of safety for consumers of these products.

Nonetheless, this report also shows that gaps exist, and that certain segments of the industry are clearly lagging behind. Accordingly, FDA will intensify its inspection efforts to focus particularly on those seafood processors that: a) need to control for pathogens; b) need to control for histamines (that can cause allergic-type reactions); and c) still have not completed their HACCP plan.

Additional actions already taken or in-process include:

  • Improved guidance and training to the industry and regulators on control of pathogens and histamine;
  • Development of an inspector certification program that emphasizes knowledge of controls for pathogens and histamine;
  • Development of guidance for fishing vessel operators to address proper handling of fish that can form histamine;
  • Development of guidance for aquaculture operators to prevent pathogen contamination of aquaculture sites;
  • Increased emphasis on compliance by foreign processors and increased surveillance of imports.

Because some of these changes were already phased in during calendar year 2000, we would expect to see additional progress and may also make future refinements once data from the inspections in 2000 are available.

In conclusion, this report shows that the seafood HACCP program has already increased the margin of safety for American consumers. This state-of-the-art program is increasingly being adopted and properly implemented by a broad and diverse array of seafood processors. But this report also identifies segments of the industry that need particular focus and intensified government oversight. FDA has already started to do this, and will accelerate its efforts in calendar year 2000.

 

Sincerely yours,

 

Philip C. Spiller
Director, Office of Seafood
Center for Food Safety and Applied Nutrition

 


 

FDA's Evaluation of the Seafood HACCP Program for 1998/1999

Fish Logo - Get Hooked on Seafood Safety - Office of Seafood - Food and Drug Administration

EXECUTIVE SUMMARY

The purposes of this evaluation are to review the extent to which processors of commercial seafood made progress toward the effective implementation of FDA's mandatory Hazard Analysis Critical Control Point (HACCP) regulations (21 CFR 123) during the first two years of the program (1998-1999) and to identify program areas where FDA should focus regulatory resources. The data reviewed in this evaluation are primarily the results of unannounced regulatory inspections of processors subject to the regulations.

The regulations reflect new processing standards for human food safety based on the application of preventive controls at "critical control points" in the processing system. The underlying premise of HACCP is that where preventive controls have been scientifically established and, therefore, are known to work, they will prevent food safety hazards from occurring in the finished product so long as they are properly applied and monitored. A mandatory HACCP system for seafood should increase the margin of safety over that which previously existed.

An FDA survey of the highest risk segments of the seafood industry in 1992, six years before the effective date of the seafood HACCP program, showed that processors often did not employ adequate preventive controls. A knowledge of food safety hazards and controls was not generally a prerequisite to the commercial processing of seafood. The importance of such knowledge is underscored by the fact that there are over 150 commonly consumed species that are subject to a range of potential hazards, including pathogens, histamine, marine parasites, and chemical and physical contaminants. The hazards may derive from the environment (e.g., aquatic habitat), or from conditions or processes in the processing facility.

Before the seafood HACCP program, it was unusual for FDA to initiate regulatory action against a seafood processor solely for deficiencies in preventive controls, or conditions in the plant. More typically, plant conditions would cause FDA to sample products to determine whether the products were contaminated so as to be adulterated under the law. Compliance rates based on this strategy were typically over 90%. Under the new program, deficiencies in preventive controls, i.e., HACCP deficiencies, could cause processors to be out of compliance even in the absence of contamination in product samples. Thus, compliance figures relating to HACCP requirements measure progress toward a new and much more rigorous regulatory standard, from which processors were generally starting from scratch. Initial post-HACCP compliance figures were expected to be much lower than traditional compliance statistics, then to rise steadily.

FDA's implementation strategy for the early years of the program was to issue written advisories to processors to provide feedback on their progress based on the first round of inspections (i.e., the first year), followed by written warnings based on the second round of inspections (i.e., the second year) in the absence of adequate progress. Seizure or injunction would occur based on the third round of inspections (third year) if firms did not respond adequately to the written warnings. Regulatory action would be taken in any year if an imminent health hazard was found. This "staged" strategy was based on the view that mastering science-based hazard analysis and preventive controls is an educational process and is unlikely to be fully mastered immediately.

This evaluation reviews industry progress from two perspectives: program elements and food safety hazards. The first, program elements, reviews percentages of processors that succeeded in implementing the most significant elements of a HACCP system for all hazards in 1998 and 1999, on an element-by-element basis. The second reviews percentages of processors in 1998 and 1999 that succeeded in implementing the same elements of a HACCP system but related to the control of specific food safety hazards. Both views enable a review of progress from one year to the next and, where data on preventive controls exist from the 1992 survey mentioned previously, from the pre-HACCP era.

These views also facilitate the identification of problem areas. The program element review provides information on program aspects of the HACCP system that processors appear to have had the most difficulty with, so FDA can focus its efforts on understanding the reasons for those difficulties and on obtaining improvements. The hazard-by-hazard review helps identify seafood industry segments that are lagging behind in implementation, so agency efforts can also be focused toward those segments. Various industry segments can be identified by the types of hazards they should be controlling.

The grading system upon which the percentages are based is rigorous. For example, in order for a processor to receive credit for drafting an adequate HACCP plan, the processor must have drafted all five of the most significant components of the plan adequately (i.e., identification or description of hazards, critical control points, critical limits, monitoring procedures and recordkeeping procedures for all of the hazards that apply to that processor). Likewise, for a processor to be credited with overall adequate implementation of its HACCP plan, the processor must have succeeded in all three of the most significant aspects of implementation (i.e., monitoring, recordkeeping, and corrective actions for all of the hazards that apply to that processor).

Before HACCP (i.e., the 1992 survey mentioned previously), an average of 12% of the processors surveyed (i.e., processors of certain types of relatively higher risk seafood products) employed the monitoring and recordkeeping components of preventive controls in a way that is analogous to that those which are now required under the HACCP program. In the first year of the program, for these processors the percentages for the same types of preventive controls typically jumped five-fold, to an average of 63%. These 1992 and 1998 percentages represent the only data available that allow a comparison between a baseline of pre-HACCP preventive controls and the initiation of preventive controls as a result of the HACCP program.

For HACCP plan development, in 1998 the percentages of firms with a plan that succeeded in drafting specific components of their plans averaged 69% on a component-by-component basis. The percentage of these processors that succeeded in drafting all components of their plans correctly was 31%, however, demonstrating that the majority of processors with a plan had mastered some, if not most, components of their plans but were not yet successful in drafting all components. In 1999, the percentage of firms with a plan that drafted all components successfully improved to 44% while the average success rate on individual elements improved to 79%.

For implementation of the activities that are listed in a HACCP plan, the 1998 percentages of all firms with a plan that succeeded in implementing specific elements of HACCP averaged 74% on an element-by-element basis, although these percentages were lower for certain industry segments, as will be discussed below. In 1999, for all firms with a plan, the average success rate on individual elements improved to about 80%. As with plan development, the percentages of firms that succeeded on individual implementation elements were consistently higher than the percentage of firms that succeeded on everything (47% of all firms inspected in 1998 and 54% of all firms inspected in 1999), i.e., adequate plan and implementation of all program elements, again demonstrating the rigor of the grading system. It was not easy, at least through the first two years of the program, for processors to score "100%" on all elements, even though they were able to succeed on many of the individual elements.

The data on the program element relating to the taking of corrective actions when HACCP controls do not function properly suggest that industry performance on this element should be the subject of closer FDA scrutiny. The percentages of firms that either took adequate corrective action when a problem was found or did not need to take a corrective action were actually higher (81% in 1998 and 86% in 1999) than the average rate of success for individual program elements overall (74% in 1998 and 80% in 1999). Nonetheless, because appropriate corrective actions are vital to keeping potentially unsafe products out of commerce, it is important that performance on this element be very good.

It is possible that the success rate for corrective actions was actually higher than recorded. In the early years of the program many firms established HACCP controls that were not really needed for safety, declined to take corrective actions when these controls were not fully followed, and may have been downgraded by FDA inspectors as a consequence. Nonetheless, it is worth placing special emphasis on this program element during inspections and other opportunities for feedback to industry, at least to obtain a clearer sense for what is occurring.

This evaluation also arrays the data on successful development of HACCP plans and system implementation by eight categories of food safety hazard: pathogens, parasites, ciguatera, histamine, environmental chemicals, aquaculture drugs, unapproved food additives, and physical hazards. In 1999, from 92% to 99% of processors with a HACCP plan had no deficiencies in their preventive controls for six of the eight hazards. These processors included those for which active controls for the specific hazard in question were needed and those who had accurately concluded that, for them, the hazard was not reasonably likely to occur.

For two of the hazards, however, pathogens and histamine formation in certain species of finfish, the percentages were noticeably lower: 67% for pathogens and 83% for histamine. These percentages do not mean that many seafood products are contaminated with either pathogens or histamine, but it does indicate that the successful implementation of preventive controls for these hazards has lagged behind those for other hazards. In many respects, the controls for these hazards can be more complex than for other hazards.

For purposes of this evaluation, FDA subdivided the types of processors for which pathogens are a relevant hazard into firms that make: (1) cooked, ready-to-eat products; (2) smoked products; (3) cured products; (4) dried products; and (5) breaded and stuffed products. Cured and dried products comprise only a tiny market niche. Processors of breaded and stuffed products are showing notably higher rates of successful implementation than are processors of the other four categories. Consequently, FDA is in the process of focusing special regulatory attention on processors of cooked, ready-to-eat and smoked products in addition to processors of products for which histamine is a hazard. These latter three categories of processors comprise approximately 56% of all processors that needed and had HACCP plans in 1999.

To date, program focusing has included the following:

  • A special emphasis on regulatory follow-up in cases where processors of cooked, ready-to-eat products, smoked products, or histamine-forming products have not adequately responded to FDA demands to correct significant HACCP deficiencies.
  • Improvements in preventive control strategies for these hazards, to be integrated into the next edition of FDA's seafood HACCP guidelines for processors, the "Fish and Fishery Products Hazards and Controls Guide."
  • The development of a new training course for industry and regulators that focuses on preventive measures for these hazards.
  • The development of a new training course for regulators only that places emphasis on how to inspect processors of these products.
  • The development of a seafood HACCP certification program for FDA inspectors in which training and auditing occurs during inspections of processors of these products.

Initiatives now in the planning stage include:

  • The development of criteria for good aquaculture practices that could be integrated into HACCP controls. Aquaculture ponds, especially those in developing nations, can be the recipients of run-off from nearby agricultural and other activities that can include animal and human waste. A recent FDA survey of salmonella on commercial seafood found an overall low frequency of occurrence; however, the highest frequency of occurrence was for aquaculture products.
  • The development of good fishing practices criteria for fishing vessels that catch species that can form histamine if not handled properly. These criteria would be integrated into HACCP controls that would serve as incentives to processors to obtain fish only from vessels that follow the criteria.

In addition to requiring processors to operate HACCP systems, FDA's program regulations require processors to monitor eight areas of sanitation that are fundamental to safety and record the results of that monitoring. Because sanitation tends to involve conditions throughout a food processing plant and not just a small number of critical control points, the regulations allow processors to perform this monitoring and recordkeeping as part of their HACCP systems or to do it separately. These requirements for sanitation monitoring and recordkeeping are in addition to maintaining sanitary conditions and practices. Proper sanitation practices in commercial food processing establishments were already required in FDA regulations that pre-date the HACCP regulations for seafood.

The logic behind the new sanitation monitoring and recordkeeping requirements was twofold. First, good sanitation is generally regarded as an important prerequisite to HACCP. Secondly, the seafood industry had experienced chronic sanitation problems that had been resistant to correction. It was expected that progress could finally be made if processors had to monitor and record their own sanitation practices rather than rely solely on feedback from regulatory inspections.

Data from the first two years of the program show the first measurable improvement in sanitation practices that FDA had seen in years, especially when compared to baseline data from the 1992 survey. For all firms inspected, the percentage of firms with no significant sanitation violations increased from 48% in 1998 to 56% in 1999. While this increase might not appear significant at first glance, it should be understood that perfect sanitary conditions and practices are nearly impossible to maintain, and that improvements often require changes in human habits as well as improvements in equipment.

For the types of firms for which 1992 data are available, improvement on specific aspects of sanitation was relatively dramatic. Average percentages of firms that adequately maintained the condition of all their food contact surfaces went from 49% in 1992 to 75% in 1998 and 87% in 1999.

For sanitation monitoring and recordkeeping, the percentages show similar, but not identical trends. For all firms, overall sanitation monitoring increased by 75% from 1998 to 1999 (from 44% to 77%). Where monitoring for specific aspects of sanitation could be tracked back to 1992, there are no significant differences in the percentages between 1992 and 1998, but relatively large improvements between 1998 and 1999. Sanitation recordkeeping also showed steady progress but, unlike monitoring, the most significant improvement occurred between 1992 and 1998. This observation is not surprising because sanitation recordkeeping generally was not practiced before the effective date of the program.

The FDA seafood HACCP program applies to foreign products exported to the United States in addition to domestic products. Over half of all seafood consumed in this country is of foreign origin from nearly 160 countries. It is impossible for FDA to inspect the many thousands of foreign processors that ship to the United States. The agency's traditional strategy for imports has been to examine and sample selected products at ports-of-entry.

Recently, FDA has begun augmenting this approach with specially targeted inspections overseas to determine compliance with U.S. requirements. For seafood HACCP, this targeting has been directed toward nations that are unlikely to be operating their own relatively advanced HACCP systems. In 1999, FDA visited 37 processors in Viet Nam, Ecuador, The Philippines, and Taiwan. As expected, the success rates on program elements for these processors tended to lag somewhat behind their U.S. counterparts, but they all were attempting to operate HACCP systems in accordance with U.S. requirements. The FDA inspection teams devoted a considerable portion of each country visit to education, both for government regulatory officials and industry, in order to promote and encourage progress.

The FDA seafood HACCP regulations also impose new responsibilities on importers of foreign seafood products into the United States. These importers must take "affirmative steps" to provide at least threshold evidence that their foreign suppliers are meeting U.S. requirements. FDA inspectors are now inspecting importers at their places of business to determine whether the importers are meeting their "affirmative steps" obligations. The results of these inspections are being stored in a new data system developed for the seafood HACCP program, but because the programming for this data system was still ongoing as of this evaluation, these data will not be available until next year's program evaluation. A preliminary "hand" review of information from some inspections indicates that the biggest problem with importers during the first two years of the program was just getting started with meeting the requirements. Importers who attempted the "affirmative steps" requirements tended to be successful.

I. EVALUATING AGAINST A NEW STANDARD: GENERAL CONSIDERATIONS

The second year of full operations under the FDA's seafood Hazard Analysis Critical Control Point (HACCP) program ended in December, 1999. This analysis provides a first internal evaluation of the program through the second year. It focuses on the progress that the industry has made in meeting the new requirements of the program. A follow-on evaluation will examine how FDA has implemented the program.

HACCP is a system of preventive controls for food safety based on seven internationally recognized principles. It was applied to food originally to ensure that food for astronauts was safe. The application of scientifically proven preventive controls provided a level of confidence that otherwise would have required extreme end-product testing to generate.

The seven principles of HACCP are:

  1. Hazard analysis: determining what food safety hazards are reasonably likely to occur in a given situation in the absence of preventive controls.
  2. Critical control points: identifying places or operations in a processing system where control can be applied to prevent or reduce the occurrence of the hazard, (e.g., the processing cook step, where pathogens can be destroyed).
  3. Critical limits: establishing maximum or minimum operating limits of the critical control point (e.g., the processing cook time and temperature combination that ensures the destruction of pathogens).
  4. Monitoring: conducting regular inspections of the critical control points to ensure that they are operating within their critical limits.
  5. Recordkeeping: recording the results of monitoring and other HACCP functions.
  6. Corrective actions: taking the appropriate actions when monitoring reveals that the critical limits are not being adhered to at a critical control point, in other words, when something goes wrong in the HACCP system.
  7. Verification: taking periodic steps to ensure that a HACCP system is working as it should.

Classic HACCP systems also include the development of "HACCP plans," which describe the HACCP system for a particular processing operation. HACCP plans typically include the results of the hazard analysis, list the critical control points and critical limits, and describe how the remaining principles will be carried out.

Regulatory seafood HACCP is a tool that FDA uses to help implement the food safety provisions of the Federal Food, Drug, and Cosmetic Act (FD&C Act) as well as certain provisions of the Public Health Service Act. The primary provision of law is section 402(a)(4) of the FD&C Act, which essentially states that food is adulterated if it is prepared, packed or held under conditions whereby it may have been rendered injurious to health. Under this provision, conditions are objectionable if they do not sufficiently protect against contamination by potentially unsafe substances in amounts that cause the food to be adulterated.

Because HACCP is aimed at conditions of production, it is not a tool for deciding whether the presence of a contaminant, or some amount of it, causes food to be adulterated. In many cases, that kind of decision involves an independent, risk assessment-type judgement on the merits by a public health agency. Once such a decision has been made, however, HACCP can be a vehicle for achieving compliance through preventive controls exercised by the industry and enforced by regulatory authority.

Primarily, regulatory HACCP for seafood raises the standard for conditions of production by requiring that seafood processors practice preventive controls for safety in a systematic, focused, and demonstrable way. FDA's pre-existing regulations for food processors, "Current Good Manufacturing Practice in Manufacturing, Packing, or Holding Human Food" at 21 CFR 110, contain provisions for a wide range of practices to prevent food contamination, involving both specific processes and controls (e.g., pasteurization) and sanitary practices (e.g., cleaning of food contact surfaces), but they do not provide a framework for identifying hazards likely to occur based on a hazard analysis or for focusing on the relatively few points in a process most critical to controlling those hazards through daily monitoring and recordkeeping, as HACCP does. By requiring such a framework, a mandatory HACCP system for seafood should increase the margin of safety over that which previously existed.

FDA's seafood program was the first national regulatory program to operate on the basis of HACCP principles. The impetus for mandating a HACCP regulatory program for seafood processors began in the late 1980's. While there was no public health crisis associated with seafood -- a National Academy of Sciences report in 1991 stated that most seafood is unlikely to cause illness to the consumer -- but there was an erosion of confidence in seafood safety, caused by a number of factors. Public concerns about pollution in the marine environment and growing public interest in food safety lead to concerns about the adequacy of existing regulatory controls for seafood.

Moreover, before HACCP, an understanding of food safety hazards and controls was not generally a prerequisite to the commercial processing of seafood. Adequate assurance that processors fully understood safety was lacking. An FDA survey of the highest risk segments of the industry in the early 1990's showed that preventive controls were often not employed. As the National Academy of Sciences pointed out, there are clearly food safety risks associated with seafood and some illnesses do occur.

The concerns generated by these factors are especially understandable in light of the complexity and diversity presented by seafood species, origins, and processing conditions. Unlike beef and poultry, seafood processors and regulators must take into account a broad range of hazards, none of which stands out as causing a single, overarching, problem. There are over 150 commonly eaten species that come from a wide variety of habitats all around the world and are subject to different potential hazards. Seafood processing varies from the simple to the complex, and utilizes a wide variety of food processing techniques and technologies, many of which have potential hazards associated with them.

The FDA seafood safety program identifies two categories of hazards. The first is inherent hazards associated with the particular species of fish. Inherent hazards may be environmental, such as pathogens from the harvest area, usually associated with raw molluscan shellfish, or caused by other factors, such as scombrotoxin, a problem in several dozen species of fish that can form a toxin if time/temperature abused after harvest. Scombrotoxin can cause an allergic-type reaction in humans and is second leading reported cause of illness. Other inherent hazards include natural toxins produced by marine algae such as that which causes red tide, chemical contaminants, and marine parasites.

The second category of hazard is process-related. It includes pathogens introduced in the processing environment, improper use of food and color additives, and physical hazards such as metal fragments.

In developing its program, FDA was able to draw on a substantial amount of work that had already been done to apply HACCP to seafood processing to control these hazards. Program design was heavily influenced by an industry-government seafood HACCP design project (what HACCP would look like for different seafood species/products) in the late 1980's that had been run by the National Marine Fisheries Service of the Department of Commerce with the seafood industry, as well as a joint seafood HACCP pilot project that FDA had conducted with the National Marine Fisheries Service in the early 1990's and other seafood HACCP pilot projects that had been solely operated by industry. FDA also modeled aspects of the program on its longstanding, successful regulatory program for low acid canned foods, which included some HACCP-type principles.

FDA published a notice of proposed rulemaking in January, 1994 and final rules for the program in December, 1995. The final rules included a two year grace period to allow industry to prepare and implement HACCP systems and to allow FDA time to train inspectors and others and to develop an enforcement strategy. At the end of this grace period, the program became effective for FDA's entire seafood processing inventory. The agency significantly increased the frequency of its inspections of seafood processors and FDA inspectors were given initial responsibility for determining the adequacy of processors' HACCP systems. FDA also began inspecting importers of seafood for the first time, regardless of whether they processed seafood in addition to importing it. Importers have specific responsibilities under the program to ensure that the products they import meet U.S. requirements.

It was not realistic to expect that industry would be able to achieve this higher standard for processing controls immediately. For that reason, FDA opted for a "staged" strategy based on the premise that mastering science-based hazard analysis and preventive controls is an educational process, inevitably involving problem-solving. Thus, inspections during the first year, which in most cases provided the first direct FDA feedback to processors on their HACCP systems, were primarily designed to be advisory and educational in nature. Regulatory action was not contemplated except for problems that required immediate correction to protect the public health. Second year inspections, on the other hand, were to result in warnings of possible regulatory action ("warning" letters) in the absence of at least significant, good faith progress either before or immediately after the inspection.

II. DOMESTIC INDUSTRY PROGRESS: FOCUSING ON PROGRAM ELEMENTS

a. Background

FDA's seafood HACCP regulations apply to all domestic seafood processors in interstate commerce. The regulations essentially define a "processor" as one who commercially handles, stores, prepares, heads, eviscerates, shucks, freezes, changes into different market forms, manufactures, preserves, packs, labels or unloads or holds at dockside (21 CFR 123(k) and (l)). The regulations exempt harvesters and retailers, primarily for practical reasons, and the FD&C Act exempts common carriers.

FDA's official establishment inventory of processors of fish and fishery products, exclusive of processors of raw molluscan shellfish (see Section IV), contains about 3,600 domestic firms that meet the regulatory definition of processor. Although the regulations do not differentiate among processors, the most common categories are manufacturers, repackers, and warehouses. These establishments vary in size from operations with small volumes to relatively large plants with multiple lines, products and species. However, the majority of plants are small, independently owned operations.

b. Compliance Measurements Before and After HACCP

Before the seafood HACCP program, it was unusual for FDA to initiate regulatory action against a seafood processor solely on the basis of conditions in the plant. More typically, plant conditions would cause FDA to sample products to determine whether the products were contaminated so as to be adulterated under the law. In the absence of such contamination, FDA would advise the plant to upgrade its processing conditions, but the plant was not usually regarded as being out of compliance. Compliance rates based on inspection and sampling results were typically over 90%.

As indicated in the introduction to this report, FDA's seafood HACCP requirements mandate practices designed to prevent contamination from occurring in the first place. Thus, HACCP deficiencies could cause processors to be out of compliance even in the absence of contamination in product samples. Moreover, HACCP deficiencies included conditions that could be reasonably expected to cause contamination eventually, not just conditions requiring urgent correction.

Thus, after the effective date of the program, "compliance" figures relating to HACCP requirements were measuring progress toward a new and much more rigorous regulatory standard, from which processors were generally starting from scratch. Initial post-HACCP compliance figures were expected to be much lower than traditional compliance statistics, then to rise steadily through a problem solving process. The measurements that showed over 90% compliance against the traditional standard have remained essentially unchanged since the implementation of HACCP.

While pre-HACCP compliance provides a foundation for the program, it does not really provide a basis from which to measure progress toward meeting the new HACCP standard. This is because, as indicated earlier, lack of compliance under the traditional standard was often measured by whether contamination had actually occurred, rather than by whether preventive controls were adequate.

c. Comparison to 1992

The only data available for comparing pre-HACCP preventive conditions against post-HACCP preventive conditions were generated from a survey that FDA conducted in 1992 of preventive practices in three segments of the seafood industry. The three segments included processors that made: (1) cooked, ready-to-eat products; (2) stuffed, breaded products; and (3) products involving species that can form scombrotoxin if mishandled (i.e., time/temperature abuse) after harvest. These three were chosen for the survey because of the importance of preventive controls to ensure safety. Cooked, ready-to-eat products, for example, will not typically receive a cook step by the consumer that would likely kill any remaining pathogens. Stuffed, breaded products, if mishandled during processing, can accumulate a bacteria-related toxin that cannot be removed through cooking by the consumer. Similarly, scombrotoxin cannot be removed through cooking. Moreover, consumers cannot always detect the decomposition that accompanies scombrotoxin formation.

Table Number 1, which is designed to show the level of progress that the domestic industry is making toward full implementation of HACCP, includes data from the 1992 survey in addition to data from the first two years of HACCP implementation (1998 and 1999). Because of the nature of the information collected in 1992, stemming in part from the fact that there was no mandatory HACCP system at that time, the data from that survey can be compared against data on implementation of some, but not all of the key regulatory requirements in today's program. Nonetheless, the survey data show that industry adoption of preventive controls has progressed substantially since 1992 as a result of the seafood HACCP program.

d. HACCP Plans: Overall

The data in Table 1 appear as percentages of processors. The first entry provides the combined percentages of processors that attempted to draft HACCP plans when they needed plans and those that legitimately did not need plans. The second and third entries divide these processors into those that did not need plans and those that produced adequate plans when needed. The second entry shows that about 30 percent of processors inspected (many of them warehouses and processors of non-scombroid finfish) had no hazards associated with their products and did not need plans. The FDA seafood HACCP program requires processors to create plans when there is at least one food safety hazard associated with their products that is reasonably likely to occur.

Because there was nothing similar to a plan requirement in 1992, processors generally did not have HACCP plans at that time. Consequently, the 1992 survey did not attempt to capture data on plans. Nonetheless, progress is clearly visible from 1998 to 1999.

As the third entry shows, in 1999 there was a 10% increase from the previous year in firms that correctly recognized that they had food safety hazards to control via HACCP and attempted to develop HACCP plans documenting how that control would occur. Moreover, the fourth entry shows a corresponding 13% increase in firms with plans that were now fully appropriate for their circumstances. Successful plan development is a major hurdle for processors to overcome because it reflects how well they fully understand HACCP principles as well as the science and technology of seafood hazards and controls, and how to apply this information to their own day-to-day operations. Consequently, it is to be expected in the early years that only some of those plans would be fully adequate.

Moreover, the application of HACCP principles can change significantly over a year as a processor shifts from one type of seasonal operation to another, with different hazards and controls. It was not unusual for FDA to critique a processor's plan for one type of product during the 1998 inspection, but see a plan from that processor covering another type of product in 1999. The plan for the first product might now be adequate as a result of agency feedback in 1998, but the plan for the second product might need further upgrading. Thus, the processor would still be regarded as deficient in plan development, even though the plan for the first product had been successfully upgraded.

e. HACCP Plans: Specific Components

Entries 5-9 in Table 1 provide percentages of firms with plans that got individual components of their HACCP plan correct. These components are identification of hazards, identification of critical control points, establishment of critical limits, establishment of monitoring procedures and establishment of procedures for taking corrective actions. Because written HACCP plans generally did not exist in 1992, no comparable data exist from the 1992 survey. The 1998 - 1999 data range from 66% - 76% in 1998 and gain an average of 8% in 1999. It is worth noting that these percentages are roughly twice as high as those for processors that got all components of their plans correct, thus demonstrating the challenge of "scoring 100%" on the development of an entire plan.

Two anomalies appear in the 1998-99 data. First, the fact that no data are available on identification of critical control points for 1998 reflects a problem in the data collection system for that year that was corrected in 1999. Consequently, it is impossible to chart progress on that item during the first two years. The second involves an apparent one percent decrease in industry performance from 1998 to 1999 in the establishment of critical limits. To some extent, this apparent decrease probably really reflects an improved understanding of appropriate critical limits by FDA personnel in the second year of inspections as a result of additional training and guidance and the application of that understanding in the evaluation of processors' critical limits. In other words, FDA probably "graded" more stringently in 1999 than in 1998.

The apparent decrease may also reflect the fact that the adequacy of critical limits has tended to be an area of some controversy between industry and FDA. The controversial nature of critical limits in the early years of the program reveals that a pre-HACCP consensus has been lacking on key aspects of food safety control. For example, in areas such as time/temperature abuse for ready-to-eat products, e.g., the amount of time that a product can safely undergo processing outside of refrigeration, critical limits were never worked out or agreed upon collectively among FDA, academia and industry. In that respect, the HACCP program is helping to surface those kinds of issues and serving as a catalyst for resolution. The collection of time/temperature guidance for controlling pathogen growth and toxin formation in seafood now in FDA guidelines ("The Fish and Fishery Products Hazards and Controls Guide," 2nd Edition, Table A-2, page 234) reflects a significant step forward in food safety in this area.

Where FDA believes that a processor's critical limit is inadequate but the firm is equally sure that its critical limit is sufficient to control a hazard, FDA will still typically grade a firm as deficient in this area until the controversy is resolved. There may be limited exceptions to this practice to allow for the completion of a scientific study that will resolve the matter. But in any event, whether processors are making progress on this element -- and the extent to which industry and government are making progress in resolving remaining scientific issues associated with critical limits -- should be revealed in future years.

f. Classroom Training and OJT

Entry 10 on Table 1 indicates the percentages of processors that met the program training requirement. Processors must utilize the services of a trained individual, or an individual with appropriate on-the-job experience, when carrying out certain key HACCP functions. This is one of the few program provisions that goes beyond simply requiring a functional HACCP system, in that it is intended to help the processor eventually succeed. Recipients of the training have overwhelmingly reported that it was essential to their understanding and ability to develop a HACCP system (as reported in the April, 2000 Sea Grant survey entitled "National Seafood Industry HACCP Implementation Survey Report").

The seafood HACCP training requirement is based on a similar provision in the highly effective FDA program for controlling botulism in acidified foods and low acid canned foods. As occurred with the botulism control program, the seafood training requirement has served as a catalyst for the development of private sector training programs for industry in HACCP, i.e., the identification of hazards and the establishment of focused preventive controls. Courses of this nature are valuable for their own sake, even if HACCP were not required, and could not be provided solely by FDA.

In 1992, few such courses existed, so the survey did not even attempt to obtain data on training. The 1998 - 1999 data show that about 85% of the inspected industry met the training requirement, so the program has had a significant impact in exposing the industry to basic food safety control principles. The 85% figure includes both firms that need full HACCP systems and firms that correctly conclude, on the basis of hazard analysis, that they do not have food safety hazards to control now. The training is designed to help them both to make that assessment initially and to periodically reassess it as circumstances warrant.

There was no notable change in the percentage between 1998 and 1999. On the other hand, it is already relatively high, certainly much higher than the percentage of firms that have successfully implemented the program. This difference between training and successful implementation suggests that, as important as training may be, it cannot fully substitute for experience over time, including the problem solving that is inevitably a part of it.

g. HACCP System Implementation: Overall

The percentages of firms that have successfully implemented the program, including both HACCP plan development and implementation of the plan, were 47% in 1998 and 54% in 1999 (entry 11). It is important to remember that these figures represent only the firms that scored "100%" on all the program elements discussed above and all HACCP implementation elements, such as monitoring and recordkeeping. While these numbers are interesting -- and show a positive trend -- they may not be as important as the percentages of processors that succeeded on the individual elements, since overall perfection will probably always be difficult to attain. The percentages on the individual elements tend to be significantly higher than they are for firms that were essentially perfect on all significant aspects of the program.

As a tangential matter, an apparent incongruity in the data is the fact that both the 1998 and 1999 percentages for overall successful implementation are higher than the percentages in the fourth entry in Table 1 for firms with adequate HACCP plans (31% and 44%), even though it is harder to get both plan and implementation right than to get just the plan right. The overall successful implementation percentages are higher, however, because they include those processors that do not have hazards and thus do not need plans. For these processors, successful implementation involved hazard analysis that correctly indicated no current hazards. For obvious reasons, these firms were not included in the percentages of firms with adequate plans.

h. HACCP System Implementation: Specific Elements

The next entries in Table 1 involve specific implementation elements. The first of these (entry 12) is effective monitoring of critical control points to determine whether they are functioning within their critical limits. Data exist for all processors with a plan, then for manufacturers of ready-to-eat products (entry 13), stuffed/breaded products (entry 14), and species that can form scombrotoxin if time/temperature abused (entry 15). The latter three allow comparison to data from the 1992 survey.

For all processors with a plan, the percentages of processors that were effectively monitoring were 72% in 1998, rising to 78% in 1999. Note that a slightly higher percentage of processors were effectively monitoring than had identified effective monitoring procedures in their plans (the eighth entry in Table 1). There are two possible explanations for this difference. First, if a processor's plan contained an inadequate monitoring scheme but the processor was implementing it as written, FDA has graded only the plan as deficient but has not also graded the actual monitoring as deficient (because the processor has shown a willingness and ability to monitor in accordance with its plan). The real issue in that situation is what constitutes adequate monitoring for the particular control in question.

Second, it is possible to operate HACCP-type preventive controls, at least for awhile, without committing these controls to writing through HACCP plans or monitoring records, and some of the data may be reflective of that kind of situation. Clearly, it is more important to practice preventive control than to produce paper about it, if a choice had to be made. Generally speaking, paper deficiencies are not as critical as implementation deficiencies. However, the agency remains convinced that recording HACCP systems and outcomes is essential in the long run as a means of maintaining stable, properly functioning programs.

The follow-on subdivision of processors (ready-to-eat, stuffed/breaded, scombroid) shows large increases from 1992 in the number of processors that are monitoring for hazards. (The 1992 data also demonstrate that, at least on some scale, HACCP-type preventive controls can occur in the absence of paper that memorializes those controls.) However, the percentages of ready-to-eat and breaded/stuffed processors that are effectively monitoring are notably lower than for all processors. The issues raised by this finding are discussed below in Section III.

The next four entries (16-19) track the percentages of firms that effectively recorded the results of their monitoring. Again, the data are broken down into all processors with a plan (entry 16), followed by manufacturers of ready-to-eat products (entry 17), stuffed/breaded products (entry 18), and scombroid species (entry 19).

For those firms for which data are available, the 1992 figures for recordkeeping were lower than they were for monitoring since, in the absence of a recordkeeping requirement, recording the results of monitoring is less likely to occur than the monitoring itself. As for monitoring, however, the figures show very large improvements from 1992 to 1998 -- in this case, from near zero to a range of 56% to 77%. Moreover, the 1998 recordkeeping percentages are roughly the same as for 1998 monitoring. Unlike monitoring, however, the 1999 percentages for recordkeeping remain roughly flat, and even slightly lower, from 1998 and only increase for one type of processor.

There are two likely reasons why the data did not show the consistent increases between 1998 and 1999 were recorded for other elements. First, FDA has been "grading" recordkeeping strictly, so that one recordkeeping error or omission can cause this element to be graded as a deficiency, even though a processor's recordkeeping is otherwise adequate. For most critical control points, recordkeeping must occur several times every day so an error or omission at some point might be difficult to avoid over the long term. Whether FDA should revise its grading system for recordkeeping (e.g., to give partial credit when most records are adequate) is a question that the agency is considering.

Second, in many cases it appears that recordkeeping had not yet caught up with significant advances in controls. For example, the HACCP program has served as a catalyst for new time/temperature controls for the processing of scombroid species and ready-to-eat products, but recordkeeping has not kept pace. The lag is most apparent in the data that compares adequate monitoring by all processors (78%) to adequate recordkeeping (69%) and adequate monitoring for pathogens by ready-to-eat processors (62%) with adequate recordkeeping by the same processors (55%).

The sequential implementation that the program is experiencing, with monitoring starting out ahead of recordkeeping, is not surprising. Also, as processors start to adjust to new and more rigorous monitoring regimes, FDA inspectors will focus more than they have previously on recordkeeping and will find mistakes that went previously ungraded. All of this is to be expected in a maturing program. Recordkeeping is an area where more years of data are needed in order for the situation to clarify.

The data in entry 20, "Adequate corrective actions taken - all firms with a plan" are reflective of several complex issues that are worth addressing. The percentages -- 81% for 1998 and 86% for 1999 -- were relatively high for HACCP implementation.

It is possible, moreover, that the percentages were actually higher than recorded. That's because some firms were graded as deficient for failure to take corrective actions when they exceeded critical limits that were not really germane to safety. Some processors have established critical limits that exceed what is necessary to achieve a safe product. When those critical limits are exceeded, but firms choose not to act, FDA inspectors are faced with a situation that is not always easy to sort out: an apparent HACCP violation even though the product is not really adulterated or likely to be adulterated.

Because FDA regulatory HACCP is a tool to implement food safety provisions in the FD&C Act, agency policy is to address failures to take corrective action only when a processor exceeds a critical limit germane to safety. Nonetheless, FDA inspectors often have erred on the side of caution by counting any failure to take corrective action as a deficiency, even when safety is not an issue. FDA continues to counsel its inspectors on how to distinguish when a corrective action is needed for purposes of regulatory HACCP from when it is not, but this situation demonstrates how a complex program has placed new burdens on inspectors in addition to the regulated industry that will take time to master.

Even though the actual numbers are probably higher than recorded, there is some cause for concern. The data do not tell us what percentage of firms needed to take corrective action or, of those that did, what percentage did so effectively. If, for example, only a minority of firms needed to take corrective action, the data suggest that most of them did not do so effectively. Because corrective action can be a last defense against the entry of adulterated product into interstate commerce, ineffective corrective action can be a significant problem. On the other hand, if most firms had to take a corrective action, the percentages are more reassuring. In any event, the data indicate that, as a matter of prudence, FDA should focus more of its attention on the adequacy of corrective actions.

The corrective action data are also reflective of the difficult choices the agency has had to make on the kinds of information that it will obtain from HACCP inspections. While the types of new data that FDA obtains from its HACCP inspections are highly detailed relative to other types of inspections, they will never be enough to answer every question that could be asked about this program. FDA originally obtained more data about the program than it does now, but realized that it had to cut back so as not to overburden either its own inspectors or those of States that were also attempting to implement HACCP programs under contract or partnership with FDA.

The next entry (entry 21) records adequate corrective action records and, again, includes firms that took corrective actions and those that did not need to do so. Not surprisingly, for 1998 the percentage of those that kept adequate records was nearly 10% lower than for firms that took adequate corrective actions. However, for 1999, the percentage of adequate recordkeeping caught up with that for adequate corrective actions. This is a positive development that shows that adequate recordkeeping need not lag behind the action being recorded when processors understand the importance of the record. In this case, processors may be realizing that it is in their self interest to maintain records showing how they successfully solved a problem that otherwise could have resulted in some liability.

i. Sanitation Controls

The remaining entries in Table 1 provide information on sanitation controls. The FDA seafood HACCP regulations require processors to monitor eight areas of sanitation that are fundamental to safety and to record the results of that monitoring. Because sanitation tends to involve conditions throughout a food processing plant and not just to a small number of critical control points, the regulations allow processors the choice of whether to include this monitoring and recordkeeping in their HACCP systems or to do it separately.

Proper sanitation is already required for food generally in FDA's regulations entitled "Current Good Manufacturing Practice in Manufacturing, Packing, or Holding Human Food" at 21 CFR 110 (GMPs). These regulations list a wide range of sanitation practices for processors to follow, but do not specifically require HACCP-type monitoring and recordkeeping by the processor's management to ensure that the practices are being followed throughout the plant. The seafood HACCP regulations add this monitoring and recordkeeping feature for many of the practices listed in the GMPs.

There were essentially two reasons for this addition. First, sanitation is regarded as a fundamental prerequisite to HACCP. It is unlikely that HACCP can succeed if attempted under unsanitary conditions. Second, the U.S. seafood industry has historically suffered from poor sanitation and FDA had made little progress in improving sanitation generally over the years. The agency therefore concluded that new requirements were warranted to improve this situation.

The first entry (entry 22) documents the adequacy of all significant sanitation controls, including both the pre-existing GMP requirements and the new monitoring and recordkeeping requirements, for all firms inspected in 1998 and 1999. No comparison could be made to 1992 because that survey did not cover all firms. The percentages -- 21 and 34 -- are by far the lowest in the Table, and confirm that this area has been a chronic problem. These percentages reflect those that scored virtually "100%" on all of the many sanitation requirements in the GMPs and, in addition, on the new sanitation monitoring and recordkeeping requirements. FDA recognizes that this is not easy to do. Sanitation is not an area that lends itself to perfection, even among the best firms. More realistically, sanitation deficiencies regularly occur and are corrected.

FDA does regard the 13% percent improvement between 1998 and 1999 as a highly positive development. An improvement of that magnitude in one year on sanitation in the seafood industry is unprecedented, and strongly indicates that the new requirements and emphasis on sanitation are having an effect. It remains to be seen whether progress will continue at this rate.

The next entry (entry 23) shows the percentages of firms that maintained adequate sanitary standards, regardless of whether they monitored and kept records. Clearly, good sanitation is the purpose behind monitoring and recordkeeping, so it is worth screening out the latter activities in order to obtain a state of the industry for actual sanitation. Those percentages doubled and show an equivalent increase from 1998 to 1999. It is not possible to know whether this increase reflects firms that also adopted the new monitoring and recordkeeping requirements for sanitation, but it is a reasonable assumption since these requirements appear to be serving as a catalyst for improvements in sanitation.

The next two entries (24-5) provide data, again without monitoring and recordkeeping, on adequate condition of food contact surfaces for processors of ready-to-eat products and for processors of stuffed/breaded products. This break-out is provided because data exist on it for these processors from the 1992 survey. Because sanitation conditions can have a greater effect on the safety of ready-to-eat products than for products that will be cooked after processing, FDA has regarded improvements in conditions as an important program objective. Similarly, sanitation is important for the safety of stuffed/breaded products because they are handled extensively during processing and, if mishandled, could contain a toxin that cannot be removed by cooking.

The data show 23% to 29% improvements in these sectors between 1992 and the first year of the HACCP program. Again, improvements like these have been rarely seen in the seafood industry. Moreover, the improvements continued at rates of 9% and 14% between 1998 and 1999.

It is also worth noting that the percentages for adequacy of food contact surfaces in 1998 and 1999 were roughly 30% higher than for no significant GMP violations, reaffirming that processors are tending to succeed with many individual aspects of the program even when they are not achieving 100% for all aspects.

The next five entries (26-30) provide percentages of firms that succeeded in implementing the program's sanitation monitoring requirements. For all firms inspected, this element experienced the highest percentage of increase (33%) from 1998 to 1999 of any program element tracked by Table 1. If the premise behind the monitoring requirement is correct -- that internal monitoring by a plant of its own sanitation practices will lead to improvements in those practices -- this is an important development.

It is worth noting, however, that the 1999 percentage of 77% for sanitation monitoring is nearly 20% higher than for no significant sanitation violations in the same year (56%), a difference that, at first glance, would appear to belie the premise. The likely explanation is that, as stated previously, perfect sanitation is so difficult to achieve that, even with adequate monitoring, there will likely always be some sanitation deficiencies, even when sanitation is being internally monitored. It is not surprising, therefore, that monitoring percentages have exceeded perfect sanitation. Monitoring should help correct deficiencies quickly when they are found in addition to improving the sanitation percentages in the long run.

Entries 27-30 involve practices (i.e., monitoring food contact surfaces and hand washing facilities) for which data from 1992 are available. Interestingly, there was no meaningful change in this monitoring between 1992 and 1998 like there was for HACCP-type monitoring of preventive controls for safety, but the increases from 1998 were significant and raise these percentages to among the highest on the chart.

The next three entries (31-33) provide percentages of firms that maintained adequate sanitation monitoring records. Like those for HACCP monitoring records, there were notable increases from 1992 to 1998. Unlike HACCP monitoring records, however, the increases continued into 1999. This is a positive development, although it is not yet clear why processors showed improvement in this area between 1998 and 1999 while the data remained flat for HACCP monitoring records during the same time period. FDA expects this question to clarify over time.

The last entries (34-35) relate to corrective actions taken and recorded for sanitation deficiencies. These percentages closely track the corrective action percentages for HACCP and the same types of considerations apply.

Table #1
Domestic Industry Progress - Focusing on Program Elements
Adoption of HACCP and Sanitation Preventive Controls by Regulatory Provision (1998, 1999) with Comparison to 1992 Preventive Control Survey
Mandatory HACCP/Sanitation Provision1992 (%)1998 (%)1999 (%)
1HACCP plan present when needed or HACCP plan not needed – all firms inspectedND7984
2No HACCP plan needed - all firms inspectedND3230
3HACCP plan present when needed - all firms that need a planND6878
4Adequate HACCP plans - all firms with a planND3144
5Adequate identification of hazards in plan - all firms with planND6683
6Adequate identification of Critical Control Points in plan – all firms with planNDND86
7Adequate identification of Critical Limits in plan - all firms with planND6766
8Adequate identification of monitoring procedures in plan – all firms with planND6776
9Adequate identification of corrective actions in plan - all firms with planND7683
10Meet training requirement - all firms inspectedND8485
11Satisfactory HACCP system, including plan (when needed) and implementation – all firms inspectedND4754
12Adequate implementation of HACCP monitoring procedures - all firms with a planND7278
13Adequate monitoring of pathogen hazard – all cooked ready-to-eat manufacturers115362
14Adequate monitoring of pathogen hazard – all stuffed/breaded manufacturers315969
15Adequate monitoring of scombroid hazard – all scombroid manufacturers137779
16Adequate HACCP monitoring records - all firms with a planND7169
17Adequate records for pathogen hazard - all cooked ready-to-eat manufacturers95655
18Adequate records for pathogen hazard - all stuffed/breaded manufacturers85869
19Adequate records for scombroid hazard - all scombroid manufacturers07774
20Adequate corrective actions taken after critical limit deviations or no corrective actions needed - all firms with a planND8186
21Adequate corrective action records when corrective actions taken or no corrective actions needed - all firms with a planND7288
22Adequate sanitation controls (GMPs and sanitation monitoring & record keeping) – all firms inspectedND2134
23No significant GMP violations - all firms inspectedND4856
24Adequate condition of food contact surfaces - all cooked ready-to-eat manufacturers527584
25Adequate condition of food contact surfaces - all stuffed/breaded manufacturers457490
26Adequate HACCP-type sanitation monitoring - all firms inspectedND4477
27Adequate monitoring of food contact surfaces - all cooked ready-to-eat manufacturers656986
28Adequate monitoring of food contact surfaces - all stuffed/breaded manufacturers717390
29Adequate monitoring of hand washing facilities - all cooked ready-to-eat manufacturers656988
30Adequate monitoring of hand washing facilities - all stuffed/breaded manufacturers716988
31Adequate sanitation monitoring records – all firms inspectedND4358
32Adequate sanitation monitoring records - all cooked ready-to-eat manufacturers125362
33Adequate sanitation monitoring records - all stuffed/breaded manufacturers244361
34Adequate sanitation corrections taken when sanitation deficiencies noted or no sanitation corrections needed - all firms inspectedNDND94
35Adequate sanitation correction records when sanitation corrections taken or no sanitation corrections needed - all firms inspectedND7289
Note: ND = No data available

 

III. DOMESTIC INDUSTRY PROGRESS - FOCUSING ON SPECIFIC HAZARDS

Table Number 1 provides information of both the status of the industry as a whole and on the progress it is making year-to-year in addressing the most significant elements of the HACCP and sanitation monitoring programs. Among other things, this information helps FDA identify program elements (e.g. identification of critical limits, implementation of monitoring procedures) that are the source of compliance problems generally so that the agency may direct additional regulatory attention toward those aspects of the program.

Table Number 2 shows the same data on success in achieving the significant elements of the program, but arrayed on a hazard-by-hazard basis. The information provided by this chart helps the agency identify industry sectors toward which it should direct improvement efforts. Many hazards are associated with specific seafood species or processing strategies.

a. First Column of Table Number 2

The first column in Table Number 2 lists, for eight categories of food safety hazards, percentages of seafood processors with a HACCP plan that had adequate preventive controls in place (i.e., adequate HACCP plans and complete implementation). In 1999, between 92% and 99% of the processors with a HACCP plan had no deficiencies in their preventive controls for six of the eight hazards. The exceptions were processors for which pathogens and histamine were relevant hazards. In 1999, 67% of the processors had no deficiencies in their preventive controls for pathogens and 83% had no deficiencies for histamine. It is worth noting, however, that, for both of these hazards, the percent of processors with adequate preventive controls increased from 1998 to 1999, as was the case for all of the hazards. Trend analysis is more fully discussed in the preceding section.

For purposes of calculating the percentages in this column, it is important to note that many products do not present hazards for pathogens or histamine, and so preventive controls for these hazards in these products would not be necessary. Processors of such products are counted in the first column of Table #2 as having no deficiencies in their preventive controls for these hazards. This method of counting applies to all eight hazards in the first column. Limitations in the data system make it difficult to separate out products for which active control for a specific hazard is needed from those for which the hazard does not normally exist. Even so, the data still show that most processors are in good shape for any given hazard. They also provide a basis for comparing implementation among hazards (i.e. identifying those hazards that processors essentially have little problem with and those for which implementation is lagging).

Five segments of the seafood industry were selected for assessment of pathogen control, processors that make cooked ready-to-eat seafood, smoked seafood, cured seafood (e.g. pickled, salted), dried seafood, and breaded and stuffed seafood). . The pathogen hazard is generally relevant to processors in these five segments. The segment of the industry that processes species of fish that can form scombrotoxin when time/temperature abused, regardless of the market form of the product (e.g. fillets, canned product) is included in the "histamine" assessment and the scombroid hazard is generally relevant to processors of scombroid species. Additionally, relevant industry segments were selected for each of the other hazards listed in the table. This focus on industry segments minimizes the effect, exhibited by the data in the first column, where processors for which the hazard is not relevant are counted as having no control problem with that hazard.

b. Second Column and Beyond of Table Number 2

For the industry segments associated with the pathogen hazard, the breaded and stuffed seafood segment has higher percentages, indicative of success, for most program elements than do the other four segments. The preventive controls necessary to control the pathogen hazard for these products are much less complex than those necessary for the other four categories. The slight drops from 1998 to 1999 in the reported percentages of processors of breaded and stuffed products that correctly identified the pathogen hazard as significant for their product and that successfully identified appropriate critical limits, are likely the result of improved understanding by the regulatory workforce of the relevance of the hazard to these products, and, thus, to stricter "grading" in 1999. Nonetheless, compliance by this group of processors is such that a refocusing of efforts towards them does not appear warranted.

In the remaining four industry segments, relatively low percentages of the processors correctly identified the pathogen hazard as significant for their product in 1998. This lack of insight represents a significant conceptual deficiency that a HACCP system is designed to reveal and eventually correct, as evidenced by the increases in 1999 for these four segments. However, consistent with Table Number 1, the percentage of processors in these segments that properly identified critical limits in their HACCP plan fell significantly from 1998 to 1999. This pattern is likely accounted for by: 1) processors that had been cited in 1998 for not having identified the pathogen hazard, corrected this aspect of their HACCP plan, only to fall short on the details of the control of the hazard, especially the most scientifically demanding aspect of the plan, the critical limits; and 2) further training of regulators caused them to be more demanding of the scientific adequacy of processors’ critical limits in 1998 than they were in 1999.

The relatively low percentages of processors that identified appropriate critical limits, especially for cooked ready-to-eat seafood and smoked seafood, can be attributed to initial resistance by processors to adopt critical limits equivalent to those recommended by FDA (especially time/temperature limits for the processing of cooked ready-to-eat seafood and water phase salt limits for smoked seafood). This resistance was likely the result of: 1) scientific dispute of the scientific necessity of FDA’s recommended limits; and 2) the need for substantial change in operations in order to accommodate the recommended limits (e.g., improved chilling procedures for cooked ready-to-eat seafood, changes in brining or dry salting procedures for smoked seafood).

The percent of processors that maintained proper monitoring records also dropped significantly from 1998 to 1999, in three of the four categories. This pattern, which was also revealed in Table Number 1, is likely accounted for by: 1) the previously described increase in the regulators’ expectations for the critical limits resulted in a linked increase in expectations for monitoring (i.e., frequency and nature of monitoring) from 1998 to 1999; and 2) the placing of more emphasis by regulators on record review as other elements of the preventive control system mature. A thorough review of records is more meaningful when a processor has identified credible critical limits and monitoring procedures. Nonetheless, experience from the decades-old low acid canned food program has shown that more processors tend to demonstrate a level of noncompliance with record keeping requirements than they do for other program elements, and that these problems tend to be persistent. In both programs, recordkeeping requirements tend to be more prescriptive than other program elements, and include a number of relatively minor provisions, making full compliance all the more challenging.

Generally, the other attributes of the HACCP system improved, some significantly, from 1998 to 1999 for these four categories of products. Particularly important are the improvements in taking corrective actions and documenting those actions. As discussed in Section II of this evaluation, further improvements to processors’ critical limits will likely increase these percentages even more, because most corrective action deficiencies appear to relate to the setting of unnecessarily restrictive critical limits. An example of an overly restrictive critical limit is a "maximum cooler temperature of 40° F," when product is held on ice in a cooler. In this case, an elevated cooler temperature (i.e. above 40° F) for a short period of time does not jeopardize the safety of the iced product. The processor usually recognizes that fact and opts not to take a corrective action, at the expense of being adversely "graded" by the regulator for not doing so. A more meaningful critical limit would be "product always fully covered with ice." This critical limit avoids the problems caused by periodic small fluctuations that occur in production coolers during working hours. Investigators have recently been trained to recommend these kinds of HACCP plan improvements and to avoid downgrading for failure to take corrective actions that have no genuine bearing on safety.

With respect to the control of histamine in scombroid species, the percentages of processors that identified appropriate critical limits and that maintained proper monitoring records were relatively low, even in 1999. This phenomenon is likely the result of the same factors attributed to the relatively low percentages for cooked ready-to-eat seafood and smoked seafood processors, as well as to the industry generally.

The percentages in column 2 and beyond in Table #2 for the other hazards (e.g., parasites in cured seafood, aquaculture drugs in aquacultured fish), are significantly higher than for pathogens or histamine. The percentages range from 79% to 100%, with most 90% or greater. Generally, the control strategies for these other hazards are less complex than the strategies for pathogens and histamine.

When the percentages for specific industry segment/hazard combinations, shown in columns 2 and beyond of Table #2, are compared to the corresponding percentages for the seafood industry as a whole for all hazards, as shown in columns 2 and 3 of Table #1 (i.e., entry numbers 5-9, 12, 16, 20, and 21), an interesting pattern emerges. The overall industry percentages (Table #1) are nearly the same or only slightly higher than percentages for the cooked, ready-to-eat seafood, smoked seafood, and scombroid species segments (Table #2), which were identified previously as the poorest performing segments. This similarity in the data can be explained in part by the fact that these three segments comprise approximately 56% of the number of processors inspected that needed and had HACCP plans in place. In other words, the data from these three industry segments dominate the data in Table #1 for processors with a HACCP plan.

Additionally, the data in Table Number 1 relate to processors that have satisfactory program elements for all hazards that are relevant to their operations, while the data in Table #2 relate to processors that have satisfactory preventive controls for only one selected hazard. Many industry segments have multiple hazards that require preventive controls (e.g. parasites, ciguatoxin, and environmental chemical hazards in nonscombroid species of fish). The more demanding nature of the criteria in Table Number 1 also helps account for the small distinction between the overall industry (Table #1) and the lowest performing industry segment/hazard combinations (in Table #2).

c. Program Refocusing

The relatively low percentages for adequate pathogen and histamine controls indicate that these are areas where FDA should focus additional efforts to foster compliance with the regulations. For pathogens, the focus should be placed on cooked, ready-to-eat and smoked products. As stated previously, breaded and stuffed products tend to "grade" higher and do not appear to warrant increased attention at this time. In addition, the other two product categories (i.e. cured and dried seafood) comprise very small proportions of the seafood industry as a whole, and generally present significantly lower risk of illness. Refocusing will be considered toward these products if improvements are not forthcoming.

In fact, the Agency has already begun to refocus in this way. In response to comments from stakeholders, including trade associations, academicians, and federal and state regulators, FDA developed a better understanding of viable control strategies for these hazards and is in the process of integrating control strategy improvements into the third edition of FDA's "Fish and Fishery Products Hazards and Controls Guide." Providing guidance that is more readily understandable and that can be more easily incorporated into processors’ current activities is likely to facilitate rapid adoption by the industry.

FDA worked with the Seafood HACCP Alliance, a coalition of federal and state regulators, academicians, and industry representatives, to develop and present a training course for regulators and industry members called the Seafood HACCP Encore Course. The course emphasized the preventive measures necessary (1) for the control of pathogen hazards in smoked fish and cooked ready-to-eat products and (2) for the control of histamine in scombroid species. The training included the most recent thinking on the subjects.

These points have been re-emphasized in a training course designed by FDA specifically for federal and state regulators called the Seafood HACCP Regulators Training Program - The Sequel, a follow-up to the original training course provided to regulators at the start of the HACCP program implementation effort. Additionally, in an on-going effort to implement a certification program for FDA seafood inspectors, smoked fish, cooked ready-to-eat seafood, and scombroid fish processors are being selected for training and auditing of seafood certification auditor candidates. All auditors receive specialized in-plant training in these commodity areas and are then required to pass two certification audits in processing plants that manufacture these products. These auditors then serve as trainers for other federal and state inspectors in their districts and as auditors for FDA certified seafood inspector candidates in other districts. The goal is to develop a well-trained inspection workforce that is better equipped to assist and assess the industry, especially in those areas where implementation has lagged.

In a complimentary effort, FDA has placed special emphasis on regulatory follow-up on cases where processors of smoked fish, cooked ready-to-eat seafood, or scombroid species have not made a good-faith effort and reasonable progress in the implementation of controls for the pathogen and histamine hazards. To date, one processor of smoked fish have been placed under permanent injunction for failure to implement proper controls for Clostridium botulinum, resulting in a court-ordered HACCP program and a second injunction proceeding is in progress. Additionally, court-ordered HACCP controls for C. botulinum, Listeria monocytogenes, and enteric pathogens have been added to consent decrees resulting from several pre-existing permanent injunctions of smoked fish and cooked ready-to-eat seafood processors.

FDA has also directed its field offices to place top priority on the inspection of processors of scombroid species, cooked ready-to-eat seafood, and smoked seafood where implementation problems were encountered in FY2000. These firms will be reinspected as necessary during FY2001 to ensure that they make expedited progress in addressing any areas of noncompliance.

FDA is in the process of developing a good fishing practices guidance document, which will emphasize the control of histamine formation on-board the harvest vessel. It is expected that this effort will mesh with the Agency's current efforts to ensure that processors that obtain scombroid species directly from the harvest vessel insist on documentation of proper on-board handling as a condition of purchase. The Agency plans to include fishing trade associations in both the development and efforts toward adoption of fishing vessel guidelines. Establishing working relationships with these organizations will constitute a new initiative for FDA

Finally, FDA is currently involved in the planning stages of a project to develop criteria for good aquaculture practices that could be integrated into HACCP controls. Aquaculture ponds, especially those in developing nations, can be the recipients of run-off from nearby agricultural and industrial activities that can include animal and human waste. A recent FDA survey of salmonella on commercial seafood found an overall low frequency of occurrence; however, the highest frequency of occurrence was for aquaculture products.

Many of these refocusing efforts are aimed at improvement in the identification of critical limits for cooked ready-to-eat seafood, smoked seafood, and scombroid species products. With respect to improvement in the maintenance of monitoring records, the Agency will first attempt to reassess the manner in which it is collecting data on this point. It is important to be able to separate deficiencies relating to minor or infrequent record keeping errors from more significant and chronic errors. At present, the data system does not permit such separation.

Table #2
Domestic Industry Progress - Focusing on Specific Hazards
Adoption of HACCP preventive controls by hazard (1998, 1999)
HazardAdequate Plans and Implement (Overall)Hazards Properly IdentifiedCCP(s) Properly IdentifiedCL(s) Properly IdentifiedMonitor Properly IdentifiedCA Properly IdentifiedMonitoring Properly ImplementProper Monitoring RecordsCA Properly ImplementProper CA Records
'98'99'98'99'98'99'98'99'98'99'98'99'98'99'98'99'98'99'98'99
PathogensAll firms with a plan6267                  
Cooked ready-to-eat firms only  7580 837260707581877275786881867690
Smoked seafood firms only  7483 816254676972757374706684847090
Cured seafood firms only  7073 827364677973888085776780917388
Dried seafood firms only  7275 757267786283837279677172877292
Breaded/stuffed seafood firms only  9288 898079798180878381808178927692
ParasitesAll firms with a plan9299                  
Cured seafood firms only  77100 1009010097989710093989396979887100
Nonscombroid finfish species firms only  9099 10093999499949996999299999992100
CiguateraAll firms with a plan9799                  
Nonscombroid finfish species firms only  9797 9996979598969895989297981009399
HistamineAll firms with a plan7683                  
Scombroid species firms only  6889 906067657872826976676879843284
Environmental chemicalsAll firms with a plan8994                  
Aquacultured fish firms  7891 969091849192918698968994988898
Crustacean firms  9699 999598919895999296939597999399
Nonscombroid finfish species firms  9195 989396969596979695929297958994
Aquacultured drugsAll firms with a plan9397                  
Aquacultured fish firms  6793 968689829390917896908788988296
Food additivesAll firms with a plan8992                  
Crustacean firms  9399 968989859690978499828691958793
Physical hazardsAll firms with a plan8994                  
Mixed fishery products firms  9394 94909097961009693969092100989098
Breaded/stuffed products firms  8997 999498939590969596929394988797

 

IV. DOMESTIC SHELLFISH INDUSTRY

Tables Number 1 and 2 do not contain data from processors of raw molluscan shellfish (i.e. oysters, clams, mussels). That is because these processors are inspected by State shellfish control authorities that do not submit the results of their inspections into the National Seafood HACCP Inspection Database. Under a Federal/State cooperative arrangement for the regulation of the raw molluscan shellfish industry known as the National Shellfish Sanitation Program (NSSP), state regulatory authorities inspect these processors and FDA audits each State regulatory program. As part of this audit, each year, FDA accompanies state inspectors on a number of their inspections. A total of 230 accompanied inspections were performed in 1999. This number represents more than 10% of the total number of these kinds of processors nationwide. During these inspections, the FDA shellfish specialists completed an inspection form that collects data similar to that contained in the National Seafood HACCP Inspection Database. Where direct comparisons can be made between the two databases, the data are presented in Table Number 3.

Like Table Number 1, the data in Table Number 3 are organized in a way that helps FDA identify program elements that may be the source of compliance problems generally so that regulatory attention might be directed toward those aspects of the program. Unlike Table Number 1, the data in Table Number 3 cover inspections conducted in 1999 only. For reasons described below, this is the first year for which data on HACCP performance exist for the molluscan shellfish industry. Instead of a comparison between years, a comparison is made between processors of molluscan shellfish, cooked ready-to-eat seafood, and scombroid species. Cooked ready-to-eat and scombroid species were selected because they, like molluscan shellfish, are identified by FDA as posing a relatively high risk, for which HACCP programs are likely to be relatively complex.

The rates of successful implementation for the processors of raw molluscan shellfish in the sample for the various program elements are generally comparable to the rates for processors of cooked ready-to-eat and scombroid species. Nonetheless, some noteworthy differences exist. On average, molluscan shellfish processors performed better in the development of their HACCP plans and in the taking of corrective action than did the processors of the other two products. On the other hand, on average, they performed somewhat worse in the implementation of their HACCP plans than did the other processors.

The superior performance of molluscan shellfish processors in plan development can likely be accounted for by the detailed nature of the NSSP, that these processors are obliged to follow. In particular, the NSSP Model Ordinance, which contains many of the requirements for the shellfish program that the participants in the Federal/State cooperative have agreed to follow, provides many of the HACCP plan elements for the processors, including mandatory critical limits. Also noteworthy is the fact that molluscan shellfish processors are inspected by State Shellfish Control Authorities between two and four times per year (depending upon the nature of their business). It is likely that this high level of regulatory contact also accelerated their plan development process.

The somewhat lower performance of the molluscan shellfish processors in plan implementation can likely be accounted for by the one to two year delay in the implementation of the Seafood HACCP Regulation in the molluscan shellfish industry relative to the rest of the seafood industry. Successful implementation tends to lag behind plan development for many processors. Thus, it is not surprising that the raw molluscan shellfish industry is slightly behind other industry segments in HACCP implementation.

This lag was anticipated and unavoidable given the mechanics of the program. The Interstate Shellfish Sanitation Conference meets once a year to adopt new requirements into the program. Many features of the HACCP program for raw molluscan shellfish have had to be adopted sequentially at these annual meetings. The sequential features of the program include: 1) adoption by the ISSC of new, mandatory HACCP provisions (i.e. Model Ordinance changes); 2) development and adoption by the ISSC of standardized inspection forms and evaluation criteria tailored to raw molluscan shellfish processing; and 3) standardization of FDA shellfish specialists, followed by standardization of State shellfish control authority personnel in how to perform HACCP-based inspections in a consistent manner. Consequently, the time frame for phasing in the program takes a few years longer than is needed for FDA to implement it unilaterally for the remainder of the seafood industry.

No explanation is readily identifiable for the superior performance of the molluscan shellfish processors in taking corrective action, relative to the other processors. Either fewer corrective actions were needed or they were taken better or with greater frequency when they were needed. In either case, the performance of the industry on this regulatory provision was commendable. On the other hand, follow-up training for FDA Shellfish Specialists and State Shellfish Control Authority personnel may be necessary to strengthen the state regulatory response to inadequate monitoring and record keeping.

Table #3
Molluscan Shellfish Data
Comparison of Molluscan Shellfish, Cooked Ready-to-Eat and Scombroid Species Processors Adoption of HACCP Preventative Controls by Regulatory Provision (1999 data)
Mandatory HACCP/Sanitation ProvisionMolluscan ShellfishCooked, Ready-to-EatScombroid Sp.
Adequate identification of hazards in plan – all firms with a plan858089
Adequate identification of Critical Control Points in plan - all firms with a plan798390
Adequate identification of Critical Limits in plan – all firms with a plan756067
Adequate identification of monitoring procedures in plan – all firms with a plan847578
Adequate implementation of HACCP monitoring procedures – all firms with plan707576
Adequate HACCP monitoring records – all firms with a plan606868
Adequate corrective actions taken after critical limit deviations or no corrective actions needed – all firms with a plan938684

 

V. Progress of Foreign Firms That Export Seafood to the United States

FDA’s Seafood HACCP program applies to imported as well as domestically produced seafood. The majority of all seafood consumed in the United States is imported from a total of about 159 countries. A number of these countries have advanced regulatory systems for seafood, while others do not. There are estimated to be about thirteen thousand foreign processors that export to the US. FDA’s traditional strategy is to review entries of products being offered for import into the United States, and on that basis, to select products for physical examination at ports-of-entry. The examination is primarily directed toward determining whether the product contains substances that would cause it to be adulterated under US law. This examination would not necessarily reveal whether the products were produced under HACCP-type preventive controls in the country of origin. And, while it could be inferred that adulterated products were not processed under fully functioning HACCP systems, the specific HACCP deficiency probably could not be identified without more information.

Consequently, FDA has developed new strategies to augment port-of-entry examination. The first strategy is a requirement in the seafood HACCP regulations that U.S. importers take "affirmative steps" to help ensure that imported seafood products have been processed in accordance with the new U.S. HACCP requirements. Affirmative steps may be basic, threshold indicators of compliance, such as obtaining copies of foreign processors’ HACCP plans and records. Importers are not expected to be surrogate regulatory agencies. Nonetheless, the "affirmative steps" requirement was novel and, FDA anticipated that it would take some time to be fully understood and implemented by the entire importing community. FDA now inspects importers at their places of business to determine whether they are meeting their "affirmative steps" obligation.

The second strategy involves a significant increase in foreign regulatory inspections by FDA inspectors. So far, these inspections have been directed towards developing countries that are major exporters of seafood. FDA chose developing countries because they are less likely to be providing advanced, HACCP-based regulatory feedback to their own processors than are highly developed countries. Countries with advanced regulatory systems for seafood are generally implementing and enforcing HACCP-type systems for their products. The findings so far (see below) have generally paralleled those for domestic inspections. Most processors are implementing HACCP but have not perfected it yet. For that reason, education both for the industry and the local regulatory agencies is a major component of these inspection activities.

In 1999, FDA sent inspection teams to four foreign countries that are major importers of seafood to the United States. The four visited were Vietnam, Ecuador, The Philippines and Taiwan. The trips had several purposes. First, FDA wished to engage in direct, in-plant compliance/enforcement inspections of foreign processors that are subject to US HACCP requirements. Second, FDA sought to develop contacts and relationships with the competent inspection authorities and industries in these countries. Third, FDA wished to provide training and educational opportunities to interested persons in both the industry and government.

During these visits, a total of 37 seafood processors were inspected for compliance with the HACCP regulation and 3 industry seminars were held, as well as numerous meetings with governmental authorities. The results of the inspections are compared to the results of domestic inspections for the same year in Table 4 below.

Comparison of Foreign and Domestic Firms for HACCP Implementation

Although foreign seafood processors are subject to the same regulatory requirements as domestic firms, FDA expected that HACCP implementation in developing countries would lag somewhat behind that in the United States. There were three reasons for this expectation. First, it is more difficult for processors in developing countries to obtain necessary information in the form of training and guidance in order to help them understand the principles and requirements involved. Second, as previously stated, the firm may not be receiving regulatory feedback on HACCP principles from local regulatory authorities. Third, FDA’s HACCP program may not be consistent with local requirements. FDA’s strategy to help overcome these disadvantages is to continue its foreign inspection program and to continue working with local regulatory officials to upgrade their understanding and capability. There is no good substitute for domestic regulatory control and regular, consistent feedback on the adequacy of HACCP systems. On the plus side, foreign processors that ship to the United States from developing countries are clearly making an effort to learn the new requirements and comply with them.

Table #4
Comparison of Compliance of Foreign and Domestic Seafood Producers
Mandatory HACCP Provision1999 Foreign (%)1999 Domestic (%)
HACCP plan present when needed or no HACCP plan needed - all firms inspected10084
No HACCP plan needed – all firms inspected030
HACCP plan present when needed – all firms that needed a plan10078
Adequate HACCP plans – all firms with a plan3344
Adequate identification of hazards in plan – all firms with a plan5083
Adequate identification of Critical Control Points in plan – all firms with a plan5086
Adequate identification of Critical Limits in plan – all firms with a plan5366
Adequate identification of monitoring procedures in plan - all firms with a plan3676
Adequate identification of corrective actions in plan - all firms with a plan4783
Meet training requirement - all firms inspected10085
Satisfactory HACCP system, including plan (when needed) and implementation – all firms inspected3354
Adequate implementation of HACCP monitoring procedures – all firms with a plan3978

As expected, in the aggregate, the foreign firms inspected in these countries are behind their US counterparts in most categories, although they did better in some. All foreign firms visited in 1999 had received the required HACCP training and had at least attempted to draft written HACCP plans. Clearly, there was an understanding by these processors that a HACCP plan was required if they had a significant food safety hazard to control and that HACCP training was also required. Also, because FDA selected producers of relatively high risk products to inspect, all of these firms did, in fact, have hazards associated with their products and did need a plan. These findings must be tempered by the fact that, unlike inspections in the United States, these inspections were announced in advance. Nonetheless, if foreign inspections by FDA serve as an impetus for compliance, that is reason enough to continue performing them, if only on the limited scale that resources allow.

VI. Progress of Importers of Foreign Origin Seafood

As mentioned above, importers of seafood into the United States must comply with several special requirements of the seafood HACCP regulation that are designed to ensure that the products they offer for entry are processed under an acceptable HACCP system. One option for providing such assurance is for importers to purchase products from countries with which FDA has a bilateral agreement based on either the equivalence of the foreign seafood safety system or its ability to ensure compliance with U.S. requirements. However, since no such agreements presently exist, importers are required to formulate product specifications and take one or more "affirmative steps" to help ensure that the product offered for entry has been processed under HACCP controls. They must also keep and make available to FDA certain records that document compliance with these requirements. These records are examined during unannounced inspections of importers by FDA inspectors at the importer's places of business.

Before the seafood HACCP regulations FDA did not directly impose requirements on importers or conduct unannounced inspections of importers, so there was no need to maintain an establishment inventory for importer of fish and fishery products. Since the advent of seafood HACCP, the agency has been working to develop such an inventory. FDA's best estimate of the number of seafood importers is 8500 firms. During 1998 and 1999 the agency performed 908 and 665 inspections of these businesses, respectively, for compliance with the importer requirements in the seafood HACCP regulation. For each inspection, the FDA inspector must fill out a data form to provide detailed information about the inspection into the national seafood HACCP database. This database is still under development as of the writing of this evaluation and the computer programming necessary to array the results of importer inspections is not complete. Consequently, this evaluation can only provide limited information derived from the results of 26 importer inspections referred to the FDA Office of Seafood to determine whether regulatory action (i.e., a "warning letter") is warranted against the importer. This information focuses, therefore, on importers that were deficient in their implementation of the importer requirements.

A qualitative examination of the data from these noncompliant importers reveals importers experienced problems with all three major provisions of the importer requirements: preparation of product specifications; taking affirmative steps; and keeping appropriate records. However, it also reveals that these importers were reasonably successful at implementing the regulatory requirements when they attempted to do so.

The most important hurdle to overcome for compliance appears to be in educating importers on their responsibilities and ensuring that they take these responsibilities seriously. A strategy to provide importers with additional information about HACCP requirements, coupled with proper enforcement, to ensure compliance, may be appropriate.