THE APPLICATION OF HACCP AFTER THE PILOT PROGRAM.
The U.S. Food and Drug Administration (FDA) is considering whether to propose regulations under existing legislative authority that would require food manufacturers to provide food safety controls based upon the seven principles of Hazard Analysis and Critical Control Point (HACCP) published by the National Advisory Committee on Microbiological Criteria for Foods (NACMCF). FDA has concluded that its determination on whether to adopt HACCP regulations will be facilitated by the operation of a HACCP Pilot Program. FDA is conducting a pilot program with a number of volunteers from the food manufacturing industry who are using HACCP to control food hazards.
Purpose of HACCP Pilot Program
The pilot program is intended to provide information that food science professionals can use in determining whether HACCP should be expanded beyond seafood as a food safety regulatory program. The information is being gathered from firms that produce several different types of food products and from firms which control a variety of potential food hazards. The pilot program is also intended to provide FDA with additional experience in working with the audit type inspection necessary for verifying a HACCP program. The pilot program is continuing at several additional firms which are not covered in this report.
The pilot program is limited to manufacturers of foods other than seafood because FDA and the National Marine Fisheries Service have previously conducted a joint seafood HACCP pilot program with volunteer domestic seafood processors and retailers. This experience was valuable in determining FDA's course of action toward seafood HACCP. Retail food operations also are not included in the pilot program. However, HACCP principles have been incorporated into the Food Code, and FDA is actively cooperating with state food agencies to facilitate the adoption of the Food Code and HACCP at retail establishments.
The Role of FDA in the HACCP Pilot Program
FDA's role in the HACCP Pilot Program is to verify that the firm's HACCP program is effective and is being followed. Verification involves on-site evaluations of the firm's HACCP system as documented by the firm's HACCP records. The focus of the evaluations is to determine whether food hazards that are reasonably likely to occur are being properly controlled at critical control points (CCPs). HACCP enhances food safety by requiring a detailed analysis of hazards affecting the product, and rigorous auditing of records to ensure that controls are being applied and any deviations are being corrected. This is the same type of systematic approach that has been successfully used by industry in scheduled processing of low-acid canned foods. During the pilot program, FDA assessed whether the firms were applying acceptable process controls and were producing safe products on a continuing basis. All the provisions, powers, and requirements under existing food regulations also continued to apply to the firms.
The Role of Industry Participants in the HACCP Pilot Program
The HACCP Pilot is a voluntary, non-regulatory program. Each industry participant agreed to (1) provide management support for HACCP; (2) develop, implement, and maintain a HACCP plan; (3) maintain records of HACCP controls; (4) verify that the HACCP plan is being implemented; and (5) train staff at all levels in their areas of HACCP responsibility. Data voluntarily submitted to FDA under the pilot program are considered confidential.
The Role of Trade Associations and States in the HACCP Pilot Program
Industry trade associations play an important role in HACCP and FDA sought input and cooperation from these groups in addressing issues of how HACCP can be effectively applied to the food manufacturing industry. One firm used the technical expertise of an industry trade association in conjunction with a state university research center. These organizations became partners in the pilot program at the firm where they worked closely on the hazard analysis and HACCP plan development. They continued to be involved during the implementation of the HACCP program.
State authorities, who share regulatory oversight responsibilities for the food manufacturing industry, participated with FDA in one of the pilot programs.
Pilot Program Implementation Protocol
FDA is conducting the pilot program based on the procedures outlined in the "Invitation to Participate in a Voluntary HACCP Pilot Program for the Food Industry," published in the Federal Register on August 4, 1994. The procedures include a voluntary agreement between FDA and the firm, a HACCP plan review, and on-site evaluations by an FDA site evaluation team. Details regarding this protocol were provided in the first interim report.
First Interim Report
In June 1996, FDA issued a first interim report that provided observations and comments on the HACCP Pilot Program. The report included information on aspects of the program that worked well, areas that presented challenges, and issues that needed to be addressed to decide how to proceed with a HACCP program. The primary focus of the first report was on the firm's experience with developing a HACCP program and beginning its implementation.
Firms Completing the HACCP Pilot Program
There are six (6) companies that have completed the voluntary pilot program. They represent a wide range of foods products, hazards, and manufacturing processes. Some of these food products undergo further processing before reaching consumers, while others are produced for final consumer marketing. The number of food safety hazards that were reasonably likely to occur that the firms designated in their HACCP plans as requiring HACCP controls ranged from one (1) to seven (7). Chemical, microbial, and physical hazards were represented.
Second Interim Report
Observations and comments from the six firms that have completed the pilot program are provided in this second interim report. These firms are Alto Dairy, Campbell Soup Company, ConAgra, EarthGrains (formally Campbell-Taggart, Inc.), Pillsbury, and Ralston Foods. Information for this report was obtained from the on-site evaluation reports prepared during the course of the pilot program and the close-out reports prepared jointly by the firm and FDA at the completion of the pilots. This second report provides updated information on the implementation of the firms' HACCP programs including an analyses of the pilot participants' experiences, and their final assessment of the benefits and challenges of HACCP. The information is presented in manner which maintains the confidentiality of the firms.
Organization of Second Interim Report --- The report is organized in three parts. The first part includes information on aspects of the pilot program not covered in the first interim report. This information was primarily developed at the conclusion of the pilot program at each firm and covers validation and verification activities, costs and benefits of HACCP, and ways to measure the effectiveness of HACCP.
The second part provides information that updates aspects of the pilot program covered in the first report. This report concludes with a discussion of the continuing implementation of HACCP at the firms after completion of the pilot program, and the potential roles FDA can play as HACCP continues to be adopted by the food industry.
At the conclusion of the pilot program at each firm, FDA discussed the firms' experiences with evaluating their own HACCP programs, assessing the HACCP costs and benefits, and measuring the effectiveness of HACCP. The information provided by the firms is discussed in this section of the report.
1. VALIDATION AND VERIFICATION
The firms developed evaluation procedures to determine the adequacy of their HACCP system. These procedures became components of their HACCP plans and served to distinguished these HACCP plans from many types of quality assurance (QA) programs. There are two principle types of evaluations: (1) studies to determine the validity of the HACCP plan, and (2) audits to verify that the plan is being properly implemented. Each is important to the overall effectiveness of HACCP.
Validation
Validation of the HACCP plan is determining whether the food safety hazards that are reasonably likely to occur have been identified, and whether the control measures in the HACCP plan are capable of controlling these hazards. Validation evaluations were conducted by the firms during the initial implementation phase when the HACCP plans were being checked to ensure they were functioning as intended. During this phase, the firms assessed the adequacy of the CCPs, critical limits, monitoring, record-keeping procedures, and corrective actions designated in the HACCP plan. The firms reported that several adjustments were made in the HACCP plans as a result of these checks. This is discussed in more detail later in this report in the section on development of the HACCP plan. The amount of time it took to complete this initial validation phase and to fine tune the HACCP plans varied. Two firms reported it took 12 months.
Once a HACCP plan is adopted and implemented, NACMCF recommends that the plan be revalidated on an "as needed" basis. Revalidation may be needed because of a system failure or a significant product, process, or packaging change. The type of changes reported by the firms that might trigger a revalidation include changes in raw materials or the source of raw materials, product formulation, finished product distribution systems, or the intended use or target consumers of the finished product. During the pilot, HACCP plan modifications were routinely made as a result of such changes.
Validation and revalidation also apply to the prerequisite program components of the HACCP system. Examples of changes made to the prerequisite programs during the pilot program are presented later in the report in the section on prerequisite programs.
Verification
The second type of evaluation is auditing the HACCP plan to verify that the HACCP system is functioning according to its design. During the pilot program the firms demonstrated that properly applied verification audit procedures provided the built-in safeguards that can be effectively substituted for routine end-product sampling. That is, the firms were able to rely on frequent reviews of their HACCP plan, CCP monitoring records, and corrective action records showing that procedures are being implemented to ensure that the final product is safe rather than relying upon end-product sampling results.
Three types of verification audits were performed during the pilot program. Information regarding these audits is provided in Table 1.
Daily HACCP Record Review --- The first type is the daily review of HACCP system records to ensure that standard operating procedures (SOPs) are being followed, deviations are detected, and corrective actions are being taken as required. Each firm conducted this type of verification audit, although some records of bulk product processing needed only weekly review.
Audit of Record Review --- A second type of verification audit is the periodic but frequent evaluation of the HACCP records to ensure that the daily reviews are being done and appropriate follow-up corrective actions are being taken if a critical limit deviation occurs. Table 1 shows that these audits were performed weekly, monthly or, in one case, quarterly by five of the six firms.
Audit of System Implementation --- The third type of verification audit is the in-depth evaluation of the HACCP system to ensure that the prerequisite programs and HACCP plan are being implemented as designed. Each firm conducted at least one of these audits during the pilot program. In some cases, the audits were conducted over a one-year period with different operational departments being audited in different quarters. In other cases, third parties hired by customers or corporate level auditors performed comprehensive audits in addition to the in-house audits. One firm used ISO 1011 auditing guidelines to conduct both its monthly and annual verification audits. Another firm found that audits were needed at least annually because of the turnover of employees and changes in employee practices.
TABLE 1. Types of Verification Audits Performed During Pilot Program
| Firms | Type of Audit | Frequency | Changes Made |
| A |
|
|
|
| B |
|
|
|
| C |
|
|
|
| D |
|
|
|
| E |
|
|
|
| F |
|
|
|
Verification Records
Verification audits rely to a large extent on records, and the quality of an audit will normally depend upon the quality of the records. The firms found that their HACCP systems did not necessarily require more records but did require more focused records that documented specific activities. As was reported in Table 1, some firms made changes in their record-keeping procedures after their initial audits in order to sharpen the content and focus of their records. FDA asked the firms which records were useful to the firms' verification audits. The information provided is summarized in Table 2.
Consumer Complaints --- Three firms identified consumer complaints pertaining to safety as playing a useful role in helping to validate and verify their HACCP plans. The other firms reviewed their consumer complaints but did not consider them to be a useful verification tool because they were too infrequent and imprecise, or they related to quality rather than safety issues. Two firms reported receiving consumer complaints during the pilot. One firm said that the number of complaints was reduced three-fold indicating greater customer satisfaction. The other firm noted that consumer complaints helped them decide to add physical hazard detectors and reduce the number of trucking companies used for outgoing shipment in order to improve its control of quality parameters and safety hazards. Another firm cautioned that relying on consumer complaints places the firm in a position of reacting to a problem, perhaps months after the particular lot was manufactured, rather than preventing its occurrence.
TABLE 2. Types of HACCP System Records Reported to be Important to a Firm's Verification Activities
| Type of HACCP System Record Used for Verification | Periods of Record Retention | Firm | |||||
| A | B | C | D | E | F | ||
| CCP monitoring records | Shelf life, 1 1/2x-2x shelf life, 1 yr, 2 yr | X | X | X | X | X | X |
| CCP deviation and corrective action records | Shelf life, 1 1/2x-2x shelf life, 1 yr, 2 yr | X | X | X | X | X | X |
| Sanitation cleaning records | 1 yr | X | X | X | X | X | |
| Incoming ingredient control records | 1 yr | X | X | ||||
| Lot numbers and traceability records | 1 yr | X | |||||
| Consumer complaints | 1 yr | X | X | X | |||
| Calibration records | 1 yr | X | X | X | X | X | X |
| Microbial tests | 1 yr | X | X | X | X | X | X |
| Staff and employee training records | 1 yr | X | |||||
2. COSTS AND BENEFITS
At the conclusion of the pilot program, the firms were asked to provide FDA with any available information on the costs and the benefits the firms attributed to their HACCP program. Most of the information provided was in a narrative form. FDA found that trying to quantify this type of information is difficult because, as one firm advised, "most of the costs related to the HACCP program revolve around the cost of time and people". These costs are not easily calculated. Further, most of the benefits were intangible and also not easily measured.
Costs
The costs experienced by firms in developing and implementing a HACCP program depended on several factors. Costs were associated with the following:
The impact, or at least the perception of the impact, of these potential costs varied significantly among different firms. The information provided by the firms relative to factors affecting costs is summarized in Table 3.
Benefits
The principal benefits from HACCP reported by the firms are (1) more effective and efficient operations, (2) a higher level of confidence in the safety of the product, and (3) greater customer satisfaction. The firms attributed these benefits to HACCP because:
Improvement in employee performance was perhaps the most significant benefit from HACCP expressed to FDA by the firms. One firm reported that "due to increased HACCP awareness, employees have been instrumental in designing new processes/procedures for monitoring and control". This firm gave as an example, a processing step that was changed, based on an employee suggestion, to reduce the likelihood of occurrence of a physical hazard. The information provided by the firms relative to benefits is also summarized in Table 3.
TABLE 3. Factors Affecting Costs and Benefits of the HACCP Pilot Program
| Firms | Costs | Benefits |
| A |
|
|
| B |
|
|
| C |
|
|
| D |
|
|
| E |
|
|
| F |
|
|
3. MEASURING THE EFFECTIVENESS OF HACCP
Measuring the effectiveness of HACCP is an important consideration if FDA is to develop information concerning advantages of adopting a HACCP-based control program. Both the food industry and the regulatory agencies share the same goal of ensuring a safe food supply. Nevertheless, food processors and regulators may have different perspectives as to how effectiveness should be measured because each may place different values on the various benefits that HACCP provides. Another challenge is to identify measures of effectiveness that are objective and direct, and measures that have a baseline against which an assessment of effectiveness can be made and/or a change can be calculated.
Effectiveness of HACCP from the Regulatory Perspective
FDA's goals regarding HACCP are stated in the Advance Notice of Proposed Rulemaking (ANPR), published in the Federal Register on August 4, 1994. These goals are to:
Information from the pilot program suggests that it may be feasible to objectively measure whether HACCP will accomplish these goals. Although it was not within the scope of the pilot program to conduct such an assessment, FDA did analyze what was learned from the pilot program about measuring effectiveness.
Making the Food Supply Safer --- The data provided by the firm's verification audits and FDA's quarterly evaluations showed that deviations from critical limits were detected, appropriate corrective actions were taken, and product produced under the HACCP plan at each firm met all safety criteria established in the HACCP plan when it left the processor. FDA did not attempt, however, to measure whether the product leaving the firms was "safer". FDA did not have objective baseline data on the application of safety controls and the safety of the product prior to the pilot program.
Using Resources More Efficiently --- The pilot program showed that HACCP evaluations do provide a more effective use of regulatory resources. FDA was able to review HACCP records verifying that standard procedures had been implemented and corrective actions taken. These data showed that the food being produced and marketed during the entire one-year period of the pilot program was safe. Under the current GMP-based inspection approach, FDA and the states rely upon observations of controls being implemented on the day of the inspection, the analytical results from any in-process and end-product samples that FDA may collect, and any consumer complaints submitted to FDA to make an assessment of product safety. Normally, these data only show whether preventive control and corrective actions were taken during the short period of the inspection and the period of production represented by any samples collected.
Reassuring the Consumer --- Reassuring and educating the consumer about the safety of the food supply or, at least, the safety of product leaving the firms, was beyond the scope of the pilot program. Informing consumers about the benefits of producing products under HACCP preventive controls will need to be addressed in a separate and specific program of consumer education.
Enhancing Industry's Role in Problem Prevention --- The firms reported that HACCP created a need for additional training on protecting against food safety hazards, and provided an incentive to strengthen their prerequisite programs in order to catch and prevent problems before they come under the scope of the HACCP plan. This included taking a closer look at potential hazards in their incoming ingredients and materials. Examples of changes the firms made to strengthen these controls to better prevent problems from occurring are provided later in the report in Table 5. They also reported they spent more effort determining the cause of a deviation from a critical limit under their HACCP plan so the problem could be fixed and the deviation prevented from reoccurring. FDA was able to confirm from its HACCP evaluations that the firms did place increased emphasis on their incoming ingredient controls and other prerequisite program controls, that corrections were made, and that there were no reoccurring deviations from critical limits.
Effectiveness of HACCP From the Industry Perspective
FDA asked the firms what was learned from the pilot program that could be used to measure the effectiveness of HACCP. The firms identified several key benefits that they attributed to the effectiveness of HACCP. Each firm stated that to some degree HACCP provided 'more' of one or more of these benefits than was formerly provided by a GMP control program alone. These key benefits or outputs are:
Objective Measurements --- The firms reported that they used two types of objective and direct measurements to assess these key benefits or outputs.
Level of Customer Dissatisfaction --- The first measurement was the change in the level of consumer dissatisfaction with the product. This was derived by measuring the change in numbers of consumer complaints before and after HACCP. However, the firms also advised FDA that the vast majority of consumer complaints did not relate to food safety, and thus there were few complaints that are relevant to the HACCP program. Another related type of measurement used was an assessment of customer satisfaction. One firm used customer initiated, third-party verification audits as a measure of the level of customer satisfaction.Effectiveness of Controls --- The second approach used by the firms was measuring the change in the occurrence of product safety problems. This was derived by calculating the amount of product produced or the number of production runs in which product deviated from critical limits before and after HACCP. Deviations also can be measured in terms of how often a product safety incident report is prepared because a CCP is out of control. Two firms reported that they saw the amount of product discarded reduced after they instituted HACCP.
Subjective Measurements --- In most instances, the firms used more subjective measures to conclude that HACCP was an improvement over previously used QA procedures. For example, managers reported that employees had a better understanding of food safety, they were more diligent in responding to deviations promptly, and they maintained tighter production controls including finding additional ways to provide more efficient operations. Managers also reported intangible benefits, such as an increased level of confidence in the safety of their product, because they knew any deviations from production controls would be promptly detected and appropriate corrective actions would be taken.
Two important aspect of examining how best to implement HACCP program for the food manufacturing industry is determining how firms integrated HACCP with existing QA programs, and examining whether the full application of the principles upon which HACCP are based are, in fact, essential to its proper application. These issues were examined in the first interim report on the pilot program and are revisited in this section. The intent is to supplement the information previously provided and not to repeat what has already been observed.
1. QUALITY ASSURANCE AND HACCP SYSTEMS
HACCP is a food safety assurance program that shares common characteristics with recognized QA programs such as Total Quality Management (TQM) and ISO 9000. These common characteristics include: (1) identifying the safety attributes of the product the customer requires, (2) establishing points of control, (3) monitoring these control points, and (4) verifying that the program is being implemented as designed. The firms with QA and/or sanitation control programs in place incorporated HACCP into these programs. They reported that HACCP was not a major departure from what was done before and required refinements more than substantial changes.
Confusion can develop over the integration of HACCP with pre-existing QA systems because, on a practical basis, it may be difficult to separate controls of safety hazards from controls that deal with quality and productivity factors. The consumer wants safe food that is of good quality and is provided at a low cost. A firm's system of controls to ensure safety is, in many respects, identical to systems that ensure quality and efficient production. As a result, most firms found it was a challenge to separate safety hazards that were to be controlled at CCPs from quality factors that were to be controlled at other types of control points. This is discussed in more detail later in this report.
2. HACCP PLAN DEVELOPMENT STEPS
The HACCP plan of each pilot participant was specific to a processing facility and a food product. In some cases, the HACCP plan grouped types of food products and production methods together. In each instance, the firms determined the product and scope of the HACCP plan at the outset, including the point in the food production chain where the plan began and where the plan ended. Once the scope of the plan was determined, pilot participants took several other preparatory steps including: (1) establishing the HACCP team, (2) providing HACCP training, and (3) preparing and verifying a process flow diagram.
Establishing a HACCP Team
Each firm assembled a HACCP team consisting of individuals with specific knowledge of and expertise in the process and product. Membership on the teams consisted of representatives of management and operational units, such as production, sanitation, quality control, and laboratory. In most instances, corporate representatives played an important role in the initial stage of developing the hazard analysis and providing scientific input for the HACCP plan. The operating unit personnel, however, were responsible for developing the technical aspects of the HACCP plan and implementing the plan. Two firms had plant managers on their HACCP teams, and all firms had at least one department manager on the team. The teams were multi-disciplinary and included employees familiar with the variability and limitations of the daily operations.
Two firms used outside consultants who were knowledgeable in the potential hazards associated with the product and process. One firm reported that the consultants "played a large role in the plan development, in developing a safety system concept and model programs, and in serving as a liaison with regulatory officials". The other firm reported that the outside consultant "did not add value to the system".
The HACCP teams at the firms reflected the strong commitment of each firm's management to HACCP. The firms reported that management support was essential for various reasons: (1) providing adequate time and resources for the team to meet, including time for regular meetings to discuss any problems, better ways of doing their jobs, etc.; (2) providing training for the HACCP team members and special training on specific HACCP related topics for all employees; (3) providing capital resources for control measures and monitoring equipment; and (4) supporting auditing, root cause analysis, and corrective actions.
Conducting HACCP Training
At the beginning of the pilot program, each firm agreed to provide HACCP training to key members of the HACCP teams. Firms reported that this training of employees was instrumental to the firm's success with HACCP. The training helped develop a positive attitude about food safety, and it helped empower employees. In-house training and standardized materials provided in training manuals also helped ensure that a common approach was used for HACCP throughout the corporation.
The information obtained on the type of training provided and the role that training played in the success of the pilot program are summarized in Table 4. These data show that all firms trained at least two members of their HACCP team. These people, in turn, trained other team members. Most firms relied more on corporate and in-house training than training from trade and professional associations. Line employees usually were provided with an overview of HACCP, and CCP operators received hands-on training in monitoring and record-keeping. Two firms stressed the importance of repetitive and continuous training for line employees. One firm provided HACCP training to all line employees because the employees are cross-trained in a variety of jobs, and everyone is responsible for the plan's success.
Process Flow Diagrams
Each pilot participant used a process flow diagram to provide a clear and simple outline of the steps involved in the process being controlled by the HACCP plan. The flow diagrams were limited in scope to those steps in the processes that were under the direct control of the pilot participants. Some flow diagrams also included steps in the food chain that occurred both before and after the processing steps. Some firms reported that their initial flow diagram was designed by plant engineers and needed to be simplified so the diagram and descriptive words were clear and could more easily be used by HACCP team members. In general, the process flow diagrams provided an important basis for the hazard analysis.
Firms noted the need to review their flow diagram with on-line supervisors to help verify, reformat, and revise it as needed. Further, procedures were needed to communicate to appropriate managers changes in product specifications and ingredients, processing, or equipment that might necessitate a change in the flow diagram. Each firm modified their equipment and/or process during the pilot, and five firms changed their process flow diagrams. One firm changed its flow diagram because a processing system was relocated. This company reported that its flow diagram can easily be updated because it is computerized.
In four cases, firms also changed their HACCP plans as a result of changes in operations. One firm reported that changes were made because new products and equipment were added, existing equipment was modified, and equipment was relocated on the production line. They had to work with several drafts before they developed an accurate flow diagram. This firm now reviews the flow diagram on a quarterly basis.
TABLE 4. Contribution of HACCP Training to Success of Pilot Program
| Firm | Number Trained | Firms' Training Experience | |
| Outside | In-house | ||
| A | 2 |
| |
| B | 1 | 6 |
|
| C | 2 |
| |
| D | 6 |
| |
| E | 3 | 5 |
|
| F | 3 | 3 |
|
3. PREREQUISITE PROGRAMS
Each firm built its HACCP plan upon a strong foundation of prerequisite programs. These programs establish the procedures used to document and control operational conditions. The NACMCF guidelines explain that prerequisite programs are established and managed separately from HACCP plans. At the same time they may overlap, and portions of the prerequisite programs may be incorporated into the HACCP plan.
QA Programs --- All firms had QA programs, GMP programs, or TQM systems that included food safety, regulatory, and quality controls before they became participants in the pilot program. Most firms did not use the term 'prerequisite programs' to describe these programs.
Prerequisite Programs --- Once the firms began developing their HACCP programs, they had little difficulty deciding what comprised their prerequisite programs. The list included:
Control Points --- Most firms followed the NACMCF guidelines and called the steps or points in their operation and process at which these prerequisite program controls are applied 'control points.' A few companies used other terms including quality control points, process control points, and regulatory control points.
The firms also had some activities which were important to HACCP but were not preventive in nature and, thus, were not controls per se. Firms considered these activities to be part of their HACCP system, and for some firms, also part of their prerequisite programs. They included:
Strengthening Prerequisite Programs --- Over the course of the pilot program, firms made several changes to improve their prerequisite program controls. These changes provided increased assurance that the hazards were being adequately control at 'control points' and, therefore, would not need to be controlled at CCP's. Examples of these changes are provided in Table 5. By the end of the pilot program, each firm had a comprehensive set of prerequisite programs that provided a strong foundation for the implementation of its HACCP plan. In several instances, having these prerequisite programs in place led to a reduction in the number of CCPs in the firm's HACCP plan.
TABLE 5. Improvements Made to Prerequisite Programs During Pilot Program to Reduce Number of Hazards Controlled at CCP's.
| Firms | Changes Made to Prerequisite Program |
| A |
|
| B |
|
| C |
|
| D |
|
| E |
|
| F |
|
Good Manufacturing Practices (GMPs)
FDA and the firms discussed the type of control measures that should be included in prerequisite programs and the relationship of these prerequisite program controls to the CCPs in the HACCP plan. Most of the control measures firms included in their prerequisite programs pertain to GMPs. GMPs include procedures to prevent adulteration caused by the following:
Food hazards --- biological, chemical, and physical properties including any poisonous or deleterious substances which may render the food injurious to health and cause it to be adulterated within the meaning of Sec. 402(a)(1), (2) and (4) of the Food, Drug, and Cosmetic Act (the Act); and any color additive causing it to be adulterated within the meaning of Sec. 402(c) of the Act;
Non-safety contaminants --- substances that would cause a food to consist in whole or part of any filthy, putrid, or decomposed substance and make it unfit for food and cause it to be adulterated within the meaning of Sec. 402(a)(3), (4), (5), (6) and (7) of the Act; and
Quality and economic defects including misbranding --- properties, substances, or other factors that affect the quality and salability of the food product in the market place, and include factors that would cause the food to be adulterated within the meaning of Sec. 402(b) and (d) of the Act; and factors that would cause the food to be misbranded within the meaning of Sec. 403 of the Act.
All firms had in place or developed SOPs to manage GMPs. Before starting the pilot, FDA conducted an initial review of operations to ensure that these SOPs were being properly implemented. In addition, most firms used outside GMP auditors, or had customers who conducted audits to verify compliance with GMPs.
Incoming Ingredient and Material Controls
The firms reported that when they conducted their a hazard analysis, they found that an important means of controlling potential hazards is to use raw ingredients and materials that are free of hazards. If the raw ingredients are free of hazards, the only hazards that need to be controlled by a CCP are those that may arise from within the processing operation itself. These are often only physical hazards.
All the firms either had controls in place for incoming ingredient or developed additional control measures during the pilot program.
Table 6 summarizes the control measures used by the firms. These control measures, in affect, extend control of hazards backward in the food production chain to the primary producers and suppliers.
Suppliers With HACCP Programs --- The firms noted that the most effective control measure is a HACCP program used by a producer or supplier. Table 6 shows that three of the six firms required suppliers to have HACCP programs. This allowed these firms to conclude in their hazard analysis that the potential hazards that could occur in the raw ingredient were being prevented and were unlikely to occur. These potential hazards did not need to be controlled by a CCP under their HACCP plan. Instead, the firms were able to monitor these supplier HACCP programs under their prerequisite programs.
Suppliers Without HACCP Programs --- Incoming ingredient and material controls that fell short of HACCP-type controls provided less assurance that potential hazards were unlikely to occur. In these instances, the firms had to make a case-by-case determination whether a reasonable likelihood existed that a hazard would occur in the particular type of raw ingredient or material. The firms based this determination upon experience, illness data, scientific reports, and other information they collected. For several firms, this type of information had not been collected when the hazard analysis was first conducted at the beginning of the pilot program. This is a principal reason why some changes were made in the number of CCPs during the course of the pilot program which will be discussed in more detail later in this report.
Incoming Controls Detect Problems --- During the period of the pilot program, several instances occurred where the incoming material controls detected problems. One instance involved a supplier that was required to have a HACCP plan to control a problem that was detected. The firm reported that:
There have been three recent incidents of foreign material contamination in shipments from a supplier. The foreign material found was not specifically named in the plant's hazard analysis, but the material was due to the negligence of the supplier. Although the supplier's hazard analysis identified several foreign materials, the items found are not likely to reoccur, and would not normally have been identified when conducting a hazard analysis. The supplier's controls for the foreign materials were part of its prerequisite programs. Finding the hazards started a sequence of events that included holding the product. The supplier's HACCP program was evaluated to determine overall effectiveness. The findings also triggered an audit of the supplier to ensure that safety controls were in place.
Thus, the firm took steps to ensure that the supplier was effectively implementing their prerequisite programs to prevent situations that might develop into more substantial problems at the processing level.
TABLE 6. Types of Incoming Ingredient and Material Controls Used by Firms*
| Type of Controls | Firms | |||||
| A | B | C | D | E | F | |
| Receipt of raw materials only from known/approved sources of supply | X | X | X | X | X | X |
| Letters of guaranty required based upon product specifications | X | X | X | X | ||
| Certificates of Analysis for selected components including periodic chemical analyses | X | X | X | X | X | X |
| HACCP plan required for classification as 'select' or 'certified' supplier | X | X | X | |||
| Inspection of ingredients and shipping containers for evidence of insanitary conditions | X | X | X | X | X | X |
| Shipping containers required to include record of prior use and cleaning | X | X | X | |||
| Bulk ingredients put through screens and magnets prior to acceptance | X | X | X | |||
| Ingredient containers marked with date of arrival | X | X | ||||
| Separate receiving area provided near storage area for incoming | X | X | X | |||
| In-house laboratory samples analyzed to confirm acceptability; samples retained | X | X | X | X | X | X |
| Environmental growing conditions monitored relative to safety of raw ingredients | X | X | ||||
| Quick tests of indicator parameters used to assess acceptability of raw ingredients | X | X | X | |||
| Seals required on closed or bulk containers | X | X | X | X | X | |
| Parameters established on time after harvest for use of raw ingredients | X | |||||
| Screening of new vendors included initial sample analysis by outside laboratory and random sampling for six months | X | |||||
| Vendor audits required to be performed by firm or third-parties | X | X | X | X | ||
| Problem report generated if there is a deviation from SOPs | X | X | X | X | ||
* Types of controls used at each firm depends, in part, upon the type of incoming ingredients and materials, and upon other in-plant controls.
Hazards Controlled Through the Prerequisite Programs
The firms reported that properly designed and implemented prerequisite programs can reduce the likelihood of occurrence of many potential hazards. This helps limit the number of hazards that need to be controlled under the HACCP plan. One firm noted that:
The controls that are used in prerequisite programs are a separate program from the HACCP program. The HACCP plan only addresses CCPs. We are organizing a QA manual which will include all quality control points, process control points, food safety control points, and regulatory control points. CCPs will remain in a separate HACCP manual.
Another firm reported that it investigates any instances where hazards get through its prerequisite programs because the intent is to control hazards by prerequisite programs.
Decision Trees --- Five of the six firms said that they used the NACMCF decision tree or a comparable decision tree. The decision tree helped identify those significant hazards that did not need to be controlled at CCPs but did need to be controlled under the firms' prerequisite programs. However, these five firms also reported that they did not rely solely on the NACMCF decision tree. One participant said that the NACMCF guideline by itself is "too general...there is a lot of confusion...it is interpreted differently by different people". Instead, this firm supplemented the NACMCF guideline with other internal guides.
Controlling Hazards at a CP vs a CCP --- Firms needed flexibility to be able to include some prerequisite program controls in their HACCP plans. No explicit criteria arose during the pilot program for determining when a control measure should be managed under a prerequisite program or elevated to the HACCP plan. In some instances, the same type of hazards were controlled under the prerequisite programs at one firm and under the HACCP plan at another firm. Either approach appeared to be effective. In most instances, however, firms tried to maximize the use of their prerequisite programs to control hazards in order to avoid the more arduous requirements of a CCP. Exceptions arose where two firms elected to include some prerequisite program controls in their HACCP plan because it involved a quality or economic factor of great importance to the firm.
Prerequisite Program Records
Each firm confirm that the prerequisite program controls were being implemented as set forth in the firm's SOPs by the use of monitoring and correction records. The prerequisite program records differed from the HACCP plan monitoring records in that they normally documented that a procedure was followed as scheduled rather than documenting that a critical limit was continuously met. The prerequisite program records were signed and dated by the employee performing the operation.
The monitoring records helped the firms identify problems and correct them in a timely manner. The records of corrective action were also important because they documented the correction action taken when there was a failure to implement a control, especially when that failure might have reasonably resulted in a food hazard being introduced into the product.
4. HAZARD ANALYSIS
All the firms, except for two that had pre-existing HACCP programs, reported that the most challenging aspect of developing a HACCP plan was conducting the hazard analysis. In large part, this was because the hazard analyses provided by most of the participants at the beginning of the pilot program did not include a comprehensive analysis and rationale. As a result, FDA asked these firms to provide additional documentation of the hazards analyzed; the potential significance of the hazards; and how the hazards were to be controlled, eliminated, or reduced to acceptable levels. Several firms determined that they needed to develop additional data on their ingredients and operations in order to better assess the significance of certain potential hazards, especially their likelihood of occurrence.
At the conclusion of the pilot program, most firms reported that preparing a written hazard analysis, including preparing a list of control measures, helped them to (1) conduct a more scientific analysis rather than just a qualitative one, and (2) communicate to employees the public health significance of the hazards that were being controlled. However, one firm, which had a pre-existing HACCP program, commented that in situations where a firm produces a product with no history of significant hazards, creating a documented hazard analysis created significant added work of little value.
Critical Control Points
The participants confirmed that the information provided by the hazard analysis is essential for the HACCP team to determine which steps and which control measures in the process are CCPs. Although a CCP decision tree was used by five of six of the firms to determine if a particular step or measure met the criteria of a CCP, it was used only as a tool. More importantly, firms relied on their experience, expert knowledge, operational and scientific data, and customer experience to make the final determination of which control measures were to be CCPs.
Changes in the Number of CCP's --- FDA tracked whether the firms changed their CCPs during the pilot program. Most changes were made by the firms during the preliminary stages of the pilot program when the firms and FDA were discussing whether the firms' HACCP plans were consistent with the NACMCF guidelines. During this period, one firm, for example, reduced the number of CCP's from 80 to 16 and then from 16 to 2 before starting the pilot program.
Any changes made to the number of CCP's after the pilot started were at the complete control and discretion of the firms. FDA discussed any modifications to the HACCP plans with the firms during the quarterly pilot site evaluations. Table 7 summarizes the changes made to the number of CCP's by the firms during the course of the pilot program. The data show that five of six firms changed the number of CCPs controlled. The sixth firm would have changed the number if they were operating under a regulatory program, but elected not to since it was a pilot program. Three of the five firms reduced the number of CCPs. Two firms increased the number of CCPs when additional data indicated that hazards not being controlled under the HACCP plan were likely to occur.
Developing Additional Information --- In each instance the changes in number of CCPs resulted because the firms developing additional data on the likelihood of occurrence of potential hazards. These more comprehensive hazard analyses conducted by the firms increased their awareness of the significance of some potential hazards and caused them to change their control methods. Consumer and customer complaints were also useful to some firms to assess the significance of potential hazards.
A 'Seasoned' HACCP Plan --- By the end of the pilot program, each firm had a HACCP plan well designed to meet the needs of the firm. In general, these more seasoned HACCP plans evolved after the firm had developed an accurate process flow diagram, identified all potential hazards affecting their product and operation, and developed sufficient data to assess the likelihood of occurrence of these hazards. The firms also had determined which hazards are best controlled by the prerequisite programs and by the HACCP plan.
TABLE 7. Changes Made to the Number of CCPs During the Pilot Program.
| Firms | # CCPs Start |
# CCPs End |
Comments Made by Firm |
| A | 3 | 3-4 | The number of CCPs varied with different packaging lines. A CCP was added to control physical hazards. The firm also expanded the scope of two (2) CCPs to cover additional control measures controlling the same hazard. |
| B | 4 | 7 | Controls for foreign materials and allergens were elevated to CCPs when additional data developed for the hazard analysis showed that these hazards were reasonably likely to occur. |
| C | 8 | (0) (1) |
The number of CCPs decreased over the course of the pilot program from eight to two and then to zero at the end of the pilot at the firm's discretion. The firm's analysis of the likelihood of occurrence, severity of the injury, and liability convinced them that all of the hazards could be controlled as CPs under the prerequisite programs. FDA disagreed and felt there should be at least one (1) CCP. |
| D | 16 | (16) (1) |
All but one (1) CCP controls non-safety, GMP-related hazards. Under the pilot the hazards are controlled under the HACCP plan at the firm's discretion because the controls for non-safety factors are treated the same by the firm as the controls for the safety hazards. Under a regulatory scheme, the firm would have one(1) CCP. |
| E | 8 | 7 | A CCP was eliminated when studies showed that a prior control for an essential quality factor was effective in controlling for the safety hazard as well. |
| F | 2 | 1 | This firm began with 80 CCP's, but the number decreased to two (2) at the start of the pilot when a more comprehensive hazard analysis was made. During the pilot it was determined by the firm that one (1) CCP could be adequately controlled under the prerequisite programs. |
Processing Control Points
Five of the six firms had processes that raised an issue concerning the role processing quality controls should play in determining appropriate CCPs. This issue concerned whether a hazard that was reasonably likely to occur needed to be controlled at a CCP if there was a processing step essential to the quality and marketability of the product which also had the affect of controlling the hazard.
Example of a Bakery Product --- One example encountered during the pilot involved a bakery product. The hazard analysis determined that it is reasonably likely that the raw ingredients may contain microbial hazards. During processing, the product is baked in order to produce a product of a quality that can be marketed. Studies conducted of the internal temperatures during baking showed that the baking processing step is lethal to any microbial hazard. The studies also showed that any product that was incompletely baked to such an extent that the microbial hazards might survive would be too doughy on the inside to undergo a further processing step and would not be suitable for marketing. At issue was whether the baking step should be considered a CCP because it controls a safety hazard or whether it can reasonably be considered an essential processing step that controls a quality attribute. The firm determined that in this instance the baking step could be reasonably characterized as a quality or processing control point since any under-cooked product would be destroyed as it underwent further processing.
Example Involving Physical Hazards --- The same question arose concerning physical hazards. One example involved milled flour which was being produced as a raw ingredient. Unmilled wheat is reasonably likely to contain physical hazards such as stones. The milling process involves a series of air purifiers, rollers, and sifters that, in addition to producing a marketable flour product, will remove or crush any physical hazards to the point where they are no longer hazards. The firm determined that these essential milling steps could be reasonably characterized as processing control points since any wheat containing physical hazards that did not go through the rollers and shifters could not be made into flour. This wheat would need to be reprocessed or discarded.
5. HACCP PLANS
NACMCF explains that the HACCP plan is "the written document which...delineates the procedures to be followed to assure the control of a specific process or procedure." The HACCP plans initially submitted to FDA by the pilot participants, in most instances, provided only summary information. Most firms did not provide any supporting documentation including the rationale for selecting CCPs or validation studies establishing that the control measure(s) and critical limit(s) were effective.
These initial submissions highlighted that there is not always a clear understanding or agreement on what constitutes a HACCP plan. In some cases, the summary plan was the only document submitted because the firm had just developed its first HACCP plan and had not prepared supporting documentation of the analysis that had been done. In other cases, the HACCP plan was a collection of documents taken from the firm's larger QA program. In these cases the firms had not always separated the HACCP program out from their QA program. Another firm reported that supporting data may be kept at corporate headquarters where authorities know how to interpret the data rather than at individual facilities.
Without supporting documentation of the hazard analysis and CCP determination, FDA was unable to review and assess the adequacy of the HACCP plan submitted by the firms prior to the start of the pilot. Instead, FDA used the initial visit to the firm to review operations, interview HACCP team members, and indicate the type of supporting information needed to determine that the HACCP plan provided necessary controls. The firms agreed and, during the course of the pilot program, supporting data and information on the rationale for decisions were provided to FDA upon request. Most documentation pertained to decisions about what control measures would be used and whether hazards needed to be controlled through the prerequisite programs or the HACCP plan.
Model or Standardized HACCP Plans
One issue considered during the pilot program was the usefulness of standard formats and model HACCP plans. Standardized approaches often can help ensure that the information needed for a HACCP plan is provided and that the information is presented in an easily understood manner. For example, FDA found in reviewing the firms' HACCP plans that it was difficult to understand why certain CCPs were selected unless information from the hazard analysis and the analysis which determined the CCPs was presented in a clear and logical manner.
The experience of the firms with standard formats and model plans is summarized in Table 8. This table shows that four of six firms reported they used a standard format, usually the one provided by NACMCF, and found the standardized format to be helpful. The other two firms used a summary table for their HACCP plan that was similar to the NACMCF summary chart.
Only two of the firms used model plans. The majority of firms advised FDA that they did not consider model plans to be worthwhile. Five of six firms also did not find pre-approval of HACCP plans to be useful. The firms stated that a HACCP plan is changed too often during its initial implementation to make pre-approval practical, and that the pre-approval process would be too time consuming.
TABLE 8. Information on HACCP Plan Development
| Firms | Standard Format Used | Generic Plan Used | Pre-approval of Plans Useful | Time to Develop Plan |
| A | Yes, format based on NACMCF was beneficial | No generic plan was used | No, development of HACCP plans is an ongoing process | 9-12 months to develop and another 12 months to fine tune |
| B | Yes, format based on NACMCF was used | No generic plan was used, nor are they considered to be applicable | No, not necessary since assistance available from trade associations and private sources | 3-6 months following an orientation and weekly training for 3 months |
| C | No, NACMCF format was not helpful | Yes, model plans and information from other firms were helpful | No, would not work because program would evolve and change from the submitted plan | 2 years unless there are assigned employees and experts |
| D | Yes, NACMCF based format was used but was modified | No generic plan was used, nor are they considered to be worthwhile | No, not advisable because it would take too long, and on-site familiarity is needed | No estimate given - pre-existing plan in place |
| E | Yes, NACMCF format was helpful | Yes, model plan developed by industry used | No, pre-approval is not necessary because models are available, and too many changes occur | 6 - 12 months, but dependent upon many factors |
| F | No standardized format was used | No generic plan was used because not as detailed as is needed | Yes, it would be ideal and should include hazard analysis | 5 months to develop and train plus 12 months to implement |
6. CRITICAL LIMITS
The firms established critical limits to distinguish between safe and unsafe operating conditions at a CCP. The factors considered by the firms in setting their critical limits included (1) detection limits of their equipment, (2) experience from conducting risk assessments, (3) production specifications, and (4) existing regulatory limits. The firms used regulatory limits where they were applicable or set limits based upon the scientific and technical literature. One firm relied upon historical data and experience of experts in setting processing/operating limits for indicators of natural toxins in incoming ingredients to ensure that the finished product met critical limits.
Four of the six firms did not encounter problems in establishing their critical limits, but two did. One firm had problems establishing indicator parameters to trigger increased surveillance of incoming ingredients that were subject to intermittent contamination. A second firm had problems establishing limits to effectively monitor temperatures in transportation vehicles. The critical limits were modified at three firms during the pilot; in two instances to make them consistent with regulatory limits.
7. MONITORING
The firms reported that they were able to establish monitoring procedures to track operations, assess whether a CCP is under control, and produce an accurate record for future use in verification. Monitoring frequencies were based on factors such as operations feasibility, prior experience and judgment, manufacturer recommendations, and corporate policies. One firm reported that:
Monitoring frequency was determined by the record log of retests on the metal detector. The data identified the frequency of metal occurrences which aided in determining the frequency of monitoring.
Three firms adjusted the frequency of their monitoring based upon their experiences during the pilot program. They reported that auditing results were used to improve monitoring frequency. In each case, the frequency was increased in order to decrease the amount of product put at risk.
Microbiological Testing --- Microbiological testing was not used by the firms to monitor a CCP because of the time-consuming nature of the analytical methods. Two firms did use chemical tests to monitor CCPs. In both instances, rapid analytical methods provided timely test results. In addition, at least one firm reported that its historical microbial test data verified that pathogenic microbial hazards are of such low occurrence that they do not warrant control through HACCP. However, some microbial and chemical testing was conducted by several firms. The principal uses of testing are to:
In general, the firms confirmed that chemical and microbial testing play an essential role in validating and verifying the effectiveness of HACCP systems. Chemical and microbial testing helps ensure that an ineffective HACCP plan will not be implemented and continue to be used until a complaint or a foodborne outbreak reveals that the system is ineffective.
8. CORRECTIVE ACTIONS
Each firm developed specific corrective actions for each CCP as part of their HACCP plans. Although the firms designed their HACCP plans to prevent deviations from critical limits from occurring, ideal circumstances did not always prevail. All six firms experienced at least one deviation from a critical limit at a CCP or a failure in monitoring equipment that required follow-up corrective action. Three of the six firms experienced at least one non-conformity with procedures to implement their HACCP plan. When these events occurred, two assessments were made by the firms.
Product Assessment --- First, an assessment was made concerning the immediate action to be taken with any product affected by the deviations or non-conformities. Each firm's corrective action plan specified the action to be taken and the employee responsible for that action for all anticipated deviations. In the case of deviations that occurred during the pilot program, the monitoring and corrective actions were effectively implemented at each firm, and all affected product was properly handled.
HACCP Plan Assessment --- The second assessment concerned the integrity of the HACCP plans. When a non-conformity occurred, each firm evaluated whether the HACCP plan was properly designed to control the hazard encountered. In all instances, the deviations occurred when a physical hazard was detected in the product. These deviations were of the type expected and the integrity of the plan was not in doubt. However, Table 9 shows that there was one instance when non-conformities were detected that were more serious, and the firm modified the HACCP plan to provide for better record-keeping and verification.
TABLE 9. Experience with Non-Conformities with HACCP Plan Procedures
| Firms | Non-Conformities Detected | Changes in HACCP Plan? | Product Recall Plan Available/Used? |
| A |
|
Control measures modified to ensure record-keeping and verification procedures followed | Yes, and used once during the pilot on a problem not related to food safety |
| B |
|
None required -- Action plan was effective | Yes, but not used during pilot |
| C |
|
None required -- Action plan was effective | Yes, but not needed for HACCP because corrective actions are effective |
| D | None | None required -- Action plan was effective | Yes, but not used during pilot |
| E | None | None required -- Action plan was effective | Yes, but not used during pilot |
| F | None | None required -- Action plan was effective | Yes, but not used during pilot |
9. POINTS OF DISCUSSION AND INTERPRETATION
Over the course of the pilot program, there were many discussions by FDA and the firms on issues and interpretations of the NACMCF guidelines. At the conclusion of the pilot program, the participants were asked to identify any issues of disagreement that arose. Few disagreements were reported by the firms, but the information provided offers an insight into the types of issues that might arise between a regulator and a firm over the firm's HACCP program and the regulator's evaluation procedures. The potential issues are as follows:
Relatively few disagreements occurred because reliance was placed upon the NACMCF guidelines. Different interpretations were resolved through discussion and reliance on scientific information.
1. THE FIRMS
The final question FDA asked the firms was what plans they had regarding the application of HACCP to their operations after the conclusion of the pilot program. Five of six firms reported that they intended to continue with HACCP. In addition, several firms reported they intended to use the lessons learned through the pilot program for adopting HACCP throughout the company. The pattern for expansion that emerged from the pilot was to:
One firm reported that its HACCP operation will be used as a model by its trade association and its state who also participated in the pilot with the firm. Four firms requested that FDA continue to work with them as they expand HACCP to other facilities. The continued cooperation will include training, HACCP plan reviews, and evaluations.
2. THE STATES
The state that participated in the pilot program reported that it intend to continue to conduct HACCP evaluations as the opportunity arises. This state made a valuable contribution to the pilot program. The state supervisor and inspectors who were trained by FDA were highly competent members of the evaluation team. They contributed practical knowledge about the firm and the food safety and processing issues associated with the firm. They brought technical expertise as well as field experience to the site team.
Working with the state also highlighted the need for cooperation and coordination in harmonizing federal and state laws that apply to the food processing operations. As firms continue adopting HACCP programs and establishing critical limits for their CCPs, state and federal requirements need to be consistent. Further, where firms have model plans that are to be adopted by multiple facilities in different states, there needs to be consistency in requirements between states.
3. FDA
FDA used a team approach to conduct the HACCP pilot program evaluations. Most FDA field investigators and many state food inspectors have had training and have experience in conducting audit-type inspections. Many FDA investigators have received specific training in conducting HACCP audits, both in the context of the pilot program and FDA's Seafood HACCP program. Investigators were able to provide valuable assistance to the firms in developing their HACCP programs. FDA found this approach to be beneficial.
The firms had differing opinions as to the role FDA should play in HACCP after the pilots. These opinions included:
CCP -- Critical Control Point
CP -- Control Point
FDA -- Food and Drug Administration
GMP -- Good Manufacturing Practice
HACCP -- Hazard Analysis and Critical Control Point
NACMCF -- National Advisory Committee on Microbiological Criteria for Foods
QA -- Quality Assurance
SOP -- Standard Operating Procedures
TQM -- Total Quality Management