GAPs/GMPs And HACCP Plans
Jim Prevor’s Perishable Pundit, November 17, 2006
At the end of October, United Fresh Produce Association distributed to its members a take on the spinach crisis by United’s President/CEO Tom Stenzel. You can read it here.
It was an excellent summary of the crisis by someone in the thick of it. Tom had used portions of the paper at numerous presentations, including the Spinach Town Hall Meeting at PMA, which we analyzed here.
One part of the letter struck the Pundit, and so we asked Tom for clarification:
Hi Tom —
Wanted to see if I could trouble you to clarify one point:
In your letter to the industry, you said, “Let’s all understand one fact — regulators have stated publicly that if today’s Good Agricultural Practices and Good Manufacturing Practices had been followed in this case, the outbreak would not have occurred.”
We’ve asked this question many, many times of both FDA and the state agencies in California. They have always been unwilling to identify any specific practice that was in violation of the GAP or GMP documents. They have indicated “areas of concern” but when I questioned them, these areas of concern seemed less like violations of the documents than evidence that some additional steps are needed. So, for example, the commodity-specific guidelines might say something like, “Make sure the water is suitable for the purpose intended” and the grower may have put in a HACCP plan that required testing the water annually to ascertain that.
Well, since it didn’t prevent the outbreak, you could argue they didn’t test it frequently enough — but that is with 20-20 hindsight.
Similarly with fences that had holes in them, wells that weren’t covered, etc. — these all were really questions, more than violations. How often should a fence be walked to check for holes or wells be checked that the covers are on tight? Once a year? Once a quarter? Once a month? Once an hour?
So, two things: since you explain that they said this publicly, can you point me to where the regulatory authorities said this? And, second, can you help me understand better what kind of violations were committed?
All the Best —
Tom was kind enough to send a thorough response:
Your question raises a most important point about the current discussion of GAPs. What level of specificity and measurement are most appropriate? While I believe most in the industry and government would now agree it is time to be more specific, that does not imply that current GAPs are weak. From a food safety/HACCP approach, general guidelines may in fact be more strenuous in an environment in which strongly motivated and scientifically expert people are implementing those guidelines.
Problem is we’ve been learning recently that not all understand the rigor that is called for by a statement, such as “make sure water is of suitable quality.” In a food processing plant, experts are likely to be extremely rigorous in setting up restrictions, safeguards, testing protocols, etc., to ensure that they do just that. But in some environments, we’ve come to understand that sophistication is not there, and instead we need to add specific measurable criteria about what water to test, how often, with what action levels, etc.
This is about making sure that every user clearly understands what is expected, and every observer has a precise measuring device to determine if there is compliance. In a regulatory sense, you’d call this a “command and control” approach where one outlines exactly what someone must do, or must not do, to comply with the rules. A true HACCP model would identify the risks and require that each operator study, understand, and manage their own unique risks.
So, now what about GAPs and were GAPs “violated.” First, FDA’s 1998 Guidance and industry’s more recent Lettuce and Leafy Greens document clearly identify all of the major risk factors that must be controlled. If someone is controlling all those risks adequately, this type of contamination would not have occurred. FDA officials have made these general comments in meetings and on the press calls, although this is indeed a generality. I don’t hear them either saying “broken fences are a violation of GAPs (although not controlling animals certainly is).”
But the real issue is that a responsible party adequately controlling all risks identified in the GAPs should have prevented it — that’s back to the HACCP approach vs. command and control. There is nothing about this outbreak to indicate it was caused by an extraordinary unknown risk factor not already warned about in GAPs. Now, there’s good news — the system failed so terribly only on one day, with raw material going through one processing plant, but fail it did on that day. Can we say where GAPs/GMPs were not complied with on that ranch or in the plant — only time will tell.
Try this analogy to the HACCP approach — It is like a warning, “Don’t drive too fast or you might have a wreck.” But, what is too fast? That depends on the conditions — rain, fog, etc. — but the actual speed is your determination. That’s why you write your own HACCP plan based on the unique circumstances of risk you are trying to control. So, it’s impossible to look at the current GAPs and say someone exceeded the posted speed limit of 55 mph.
Another example: the Nunes Company found a level of generic E. coli in a reservoir that made their company’s experts nervous. They took action. They implemented the current GAPs perfectly — consider the suitability of water for its intended use, and they made their best judgment that perhaps that water shouldn’t have been used.
We can second-guess them too about whether they should or shouldn’t have taken that action, but we know they didn’t have an outbreak. Now, if this were another company without the expertise or the same decision-maker, would a bright line action level for generic E coli in surface water be better? Until now, that’s been a scientific question for each operator — but one of the lessons learned in this outbreak is that we had better be more specific, so standards of practice can be more uniform.
Your question on fences is a good one. It begs for a rule, such as “check the fences once a week, month, or year” as the standard, and then we can all know that the operator “complied” with the rules. But, GAPs today would tell us that an operation that harvests a field with fences broken down and animal tracks in the field “did not check the fences enough” to adequately control the risks. On these types of measurements, we’re likely moving to a pre-harvest field GAP assessment of some sort.
It is unlikely that we can ever prevent 100% of a pathogen from getting into a field, but our controls need to be focused on leaving that product behind and not letting it get into commerce. Just like harvest guidelines today that instruct lettuce cutters to leave behind any heads that look at all possibly suspect, we will likely be doing stronger pre-harvest assessments of things likes fences and animal tracks.
You are also right that everything here is colored by 20-20 hindsight. In the real world, if you have an outbreak associated with your products, you did not do enough. That’s the law, and there will be penalty to pay. The converse is tough to prove though — if you didn’t have an outbreak, did you do enough to prevent it or just get lucky? Companies will likely not know definitely, but this is not an acceptable roulette game to play on a company by company basis. Therefore, the industry will simply have to find a way to take the strong scientific intent behind the current GAPs and put that into measurement criteria to address the risks we know.
Hope this helps. You could really have a more cogent discussion of all this with Jim [Gorny] or Dave [Gombas], I suspect, and we’d be glad to set up a meeting with you anytime to discuss industry food safety issues.
It helps a lot, Tom, and we appreciate the effort laying this all out for us.
Certainly it seems unarguable that some people have not been doing their HACCP plans that well. As we’ve written before, many of these look like photocopied documents from other operations rather than real plans specifically drawn up after assessing the risks of one particular operation.
Minimum standards will help and certainly make it easier for regulators to judge compliance. So we can say that all water sources must be tested at least every X days.
But if the problem is faulty HACCP plans, maybe a solution is to require professional expertise in these plans. In most jurisdictions, if you are going to build a house it doesn’t matter if the homeowner is safety conscious or not because those home plans must be stamped by both a licensed architect and a structural engineer.
Maybe every grower should be required to have a HACCP plan developed and reviewed on a set schedule by a professional. And that professional, like a C.P.A., should have liability for the adequacy of the plan.
This would put teeth in the HACCP plans.
Still, perhaps we should also be looking for a “safe harbor” provision in any ultimate plan. To take the position that any outbreak proves the inadequacy of a HACCP plan, and thus the culpability of growers or processors seems unreasonable.
Every day of the week, autos are sold that we know, for a fact, will kill people. And, generally speaking, auto manufacturers are not culpable.
Planes fly, though we know some will crash.
A grower or processor who follows the legal minimums required and has a HACCP plan drawn up in good faith by licensed professionals and who implements the plan diligently should not be liable if even these best efforts don’t happen to prevent an outbreak.