We don’t know for certain that the outbreak is linked to produce, but that is a reasonable supposition. If it is, it is a catastrophic problem that has caused human suffering beyond any outbreak previously known to be related to produce. As of this writing, there were 24 reported deaths, nearly 650 victims of HUS and around 2,400 known to be sick.
Many of the on- and off-the-record public statements of authorities have been incorrect or unsupported. Although E. coli 0104 has shown up in the US previously, this strain of bacteria is not one of the so-called “Big Six” non-0157 variants of E. coli that are increasingly being studied (E. coli strains 026, 045, 0103, 0111, 0121, and 0145 are the “Big Six”). So we actually have very little information on how common 0104 actually is.
Indeed, Germany’s reaction to blame, and Spain’s response to defend its farmers, was so quick and strong that one senses this was not simply a scientific dispute about the origin of a pathogen. The fact that such a localized German outbreak was instantly blamed by German authorities on a Mediterranean country was received in Spain as a slap in the face. It was seen as Teutonic arrogance assuming that there existed lower standards in Spain and the Mediterranean. In contrast, a German producer or processor is assumed to be world-class.
The thing we really don’t know is whether this particular variant of E. coli is exceedingly virulent or whether this is a very, very large outbreak.
In general, the idea that it was Spanish cucumbers didn’t make sense from a produce distribution standpoint. The outbreak is too large to just be an episodic incident — a bird flying over a field — and Spanish cucumbers are sold over a widely dispersed area. So a large scale outbreak would typically cover a larger area. Think of an outbreak in the US that is affecting only people in Ohio but is supposedly due to Chilean Grapes — it doesn’t really make sense.
A local producer of sprouts seems to fit the geography, but the numbers are much, much too large. We’ve been to Germany, and it is not as if sprouts are the national food. The largest E. coli outbreak ever was a 1996 radish sprout outbreak in Japan, but that is how the Japanese eat, not the Germans. If we figure 2,000 known sick and, say 60 unreported sicknesses for each one reported, we are talking about 120,000 people getting sick from eating sprouts from one producer — that would be exceptional.
Based on the geography and the scale of the outbreak, the most likely produce source, if there is one, would have to be a regional fresh-cut processor.
One very unfortunate part of this episode is that public health epidemiologists have come across like the Keystone Cops. They are under pressure, of course, to tell consumers how to protect themselves, to restore confidence in other parts of the food supply and to respond to political demands.
Politically it seems more difficult in Europe to establish a case epidemiologically, as there seems to be an expectation that any finding will be confirmed through testing. This is an unreasonable expectation in produce, where shelf-life can be short and testing products extensively today just doesn’t tell much about the state of product a month ago.
The German apology to Spain was odd as Spain’s point was not that Spanish epidemiologists had studied the data and vindicated Spanish produce; it was that none of the variant of E. coli that got people sick had been found in Spain, which was irrelevant. In the face of such irrelevant arguments, German public health officials folded, which shows that they either bent to political pressure — and thus public health is hopelessly compromised — or that they never had much faith in their epidemiological studies to begin with.
His concern, of course, aside from helping people right now, is that such “first it is this” “then it is that” and “maybe it is this” creates a boy-who-cries-wolf effect and undermines public confidence in the whole public health infrastructure.
We will see how the situation develops, but there are some valuable lessons that are already evident:
1. The enemy is always changing.
The industry has funded much research since the Spinach Crisis, but pathogens evolve. We need to run fast just to stay in place. All our testing panels, including those used by FSIS, are now out of date. Just imagine the horror if next year there was an outbreak in the US related to this strain. Although Secretary Of Agriculture Tom Vilsack is going around saying he is “reasonably confident” that US consumers won’t face the type of E. coli outbreak that is currently hitting Europe, one can only say that such a statement is, at best, political.
2. Genetics are going to be a game-changer.
The fact that this is a new variant of E. coli was quickly established by a genetics lab in China. Due to the Pundit Poppa’s illness, we have done a deep dive into genetics. We asked Raphael Lehrer, a Harvard Ph.D. who is the founder and Chief Scientific Officer of GeneKey — a company we are utilizing as part of our efforts to help the Pundit Poppa—what he thought of this use of the technology:
There are two things that I think:
1) This is a good use of the current technologies. E. coli has an approximate base-pair genome of 5.4 million while humans have approximately 3 billion base-pairs. If it took the Chinese 3 days to sequence, they must have done a number of comparator sequences (i.e., other strains of E. coli), which is very helpful as well.
2) The second thing is that infectious diseases are similar to cancer in that both mutate quickly (due to fast reproduction) and therefore benefit from natural selection.
However, infectious diseases are harder in that when someone new gets infected, they start with already mutated superbug. When a new person gets cancer, that person starts with the unmutated ordinary cancer. I conclude from this that in many ways, cancer is an easier problem to solve than infectious diseases. (Especially those that mutate extra-fast like HIV and Hep C.)
Of course, infectious disease has its reasons for being easier too: 1) you can distinguish what type of bug it is via the microscope; 2) other organisms had already developed defenses against infectious diseases (e.g., penicillin), while it’s not as easy to turn to the natural world for cancer.
With the advent/acceleration of genomics, we are beginning to solve problem #1. Over the next 10-20 years, there should be a revolution in cancer... much like the 80s for heart disease and the 90s for HIV, an era of breakthrough progress. #2 is gradual, but we’ve come a long way in the last 50 years or so in making compounds.
I believe that there is still a lot of untapped potential in the current drugs, and in combinations of drugs. Unlocking this will bring more clarity to the areas of unmet need (or need that is “unmeetable” with current agents), but pharma is already pretty good at that, so while it will probably change the emphasis, the current portfolios will still be mostly relevant.
So genetic analysis can be an important tool, but the quick mutation of pathogens makes this a very difficult battle to win.
3. Claims of European superiority on food safety should be taken with a grain of salt.
Robert Stovicek, Ph.D., President at Primus Labs, wrote us back in 2007 cautioning that we should be skeptical of claims that European food safety standards were producing better results than those in the US. He put it this way:
The value of comparing the success or failure of various national or regional auditing schemes is questionable. No foreign regulatory agency is working with a system as sophisticated or efficient as the CDC’s PulseNet.
What will be identified as an “outbreak” (two or more individuals infected by a common source) in the US will more than likely go undetected in other nations. We’re not being measured using the same tools. And certainly under the current US tort system the ramifications of failure are not the same.
His point, of course, was that without identical surveillance systems, cross-country comparisons are very dubious.
4. We need to focus on treatment.
Back in 2007, we urged the produce industry to focus some of its efforts not just on reducing the incidence of food safety problems but, also, on reducing the impact of food safety problems by researching more effective treatments and cures. By all accounts, the German medical establishment has been overwhelmed by the task of treating people. Indeed, one wonders if some of the deaths being attributed to the pathogen aren’t just as much attributable to lousy medical care. There seems to be confusion about the effectiveness of antibiotics and a lack of protocol to deal with such an outbreak.
Obviously, this is a public health matter and the relevant authorities need to address this in training and resource provision. Still, it is wrong for the industry to just walk away and say this is someone else’s problem. A level of 100% safety is difficult, perhaps impossible to achieve, so sometimes fresh produce will cause illness. We should be at the forefront of supporting research on these illnesses that are often foodborne and we should support efforts to make sure hospitals and doctors are trained and equipped to deal with such outbreaks.
5. We are obviously not prepared for a bioterrorism attack.
Der Spiegel had this chilling quote:
Helge Karch, the director of the RKI’s EHEC consulting laboratory at the Münster University Hospital in western Germany, has devoted almost his entire life as a researcher to EHEC bacteria. “But I’ve never encountered something like this,” he says....
On Tuesday, the German newspaper Süddeutsche Zeitung reported that Karch had discovered that the O104:H4 bacteria responsible for the current outbreak is a so-called chimera that contains genetic material from various E. coli bacteria. It also contains DNA sequences from plague bacteria, which makes it particularly pathogenic. There is no risk, however, that it could cause a form of plague, Karch emphasized in remarks to the newspaper.
Dr. Karch is well-respected, so it is reassuring to hear that this won’t bring back the plague. We also have no reason to believe this outbreak has anything to do with terrorism. But what if it did? We’ve mentioned previously the Rajneeshi’s intentional release of pathogens on salad bars in April, 2007 to achieve political goals.
Even a moderately successful terrorism effort could result in many multiples of the number of people sickened by this outbreak. Clearly Germany isn’t prepared. We don’t have much confidence that the rest of the West is in any better shape.
6. We need better epidemiology.
Epidemiology is as much art as science, and it is clear that epidemiologists keep having a problem with the same thing: The confounding of ingredients.
The Spinach crisis was relatively straight-forward as, although spinach can be eaten with other ingredients — spring mix, bacon, cheese, etc. — it is often eaten alone and as a retail-driven outbreak, it was purchased alone. In contrast, the salmonella Saintpaul outbreak was difficult because salsa often looks like just tomatoes but can contain many other ingredients. Equally the Taco Bell outbreak that was blamed first on green onions, though, ultimately, it was deemed to be lettuce, was difficult because of the multitude of ingredients associated with what people said they ate.
So one obvious problem is that surveys that show overconsumption of lettuce, tomatoes and cucumbers — are just saying salad, and it could be a protein on the salad or any of many ingredients that either wasn’t asked about or that consumers forgot they ate.
But the issue goes beyond that. In the early phases of this current outbreak, there were numerous stories hypothesizing that the pathogen was particularly aggressive toward adult women. We had talked down a reporter who interviewed the Pundit with a similar hypothesis at the start of the Spinach Crisis. In both cases, the over-representation of women among those ill could be more easily explained by female preferences for salad than by the theory of sex-preference aggressiveness among pathogens.
It is also possible, though, that over-indexing for salad consumption can be proxy for other behaviors: eating more seafood, doing Yoga, drinking bottled water — whatever. There doesn’t seem to be an easy way to sort this out. If we don’t find one, however, we will continue to struggle with mis-identified foodborne illness outbreaks.
Robert A LaBudde, Ph.D., PAS, Dpl. ACAFS, who runs a technical software and consultancy organization for the food industry in the areas of product formulation, quality assurance and food safety, and who we previously quoted here, made the point clear in a post geared toward food safety experts:
There is an obvious lesson to be learned from the German E. coli O104 and the USA Salmonella St. Paul outbreaks: The quick and dirty statistical analysis used by epidemiologists to compare odds ratios based on retrospective attack rates is flawed in concept.
Firstly, they look at scores of odds ratios to find connections. This “multiple comparisons” situation naturally leads to many false positives. Instead, they should adjust their threshhold level (i.e., level of significance) to take into account the number of total comparisons made, so that the familywise error is controlled at 0.05, not the individual tests.
Secondly, although it may be “teaching your grandmother to suck eggs”, epidemiologists need to understand the concept of confounding. If a mixture ingredient is in the common source (e.g., salads, salsa), then the principal constituents of that ingredient would be confounders. In both outbreaks this included tomatoes. In the German one, it included lettuce and cucumbers.
Thirdly, the retrospective nature of the studies biases the results. Tomatoes and cucumbers are easy to remember from a meal, as well as whether one ate a lettuce-based salad or not. Remembering that the salad was sprinkled with sprouts, or that one had added sprouts along with other things in small quantities from a salad bar, may be much more difficult to remember. This will substantially increase the false positives due to confounders.
Fourthly, the fact the German outbreak primarily affected adult women and not children or the elderly was consistent with salad ingredients, which were already suspect. Once salad ingredients were suspect, a re-interviewing of the subjects should have been done to improve accuracy on minor components, such as sprouts, peppers and spring onions that have been implicated more often in similar outbreaks.
Fifthly, the investigating agency should publicly release the attack rate tables for the outbreak. For some reason, the agencies involved don’t believe in transparency when indicting food sources.
All-in-all, the methodology used in investigating the two outbreaks was flawed. Announcements of implicated sources were made on insufficient proof, the level of scrutiny and investigation was weak, and the reasons opaque. The nature of the outbreak was clearly North German based, yet the first announcements targeted non-German food sources.
If the data were made public, we would not have these investigation fiascoes.
7. The focus on traceability should be moved to epidemiology.
James Gorny, PhD, at the time Executive Director, Postharvest Technology Research & Information Center at UC Davis but currently the Vice President Quality Assurance & Technology at the FDA, wrote us a prescient letter that we incorporated in a piece titled Tomato/Salmonella Situation Cries For Improved Epidemiology. One has to give him kudos for seeing the issue more clearly than most.
The industry has focused enormous resources on traceability, partly because this is an area where the industry can do something, but it has been proven time and again that traceability is of little value if the epidemiology isn’t right.
In fact, traceability for all large retailers is already so good that almost always the epidemiology is what slows the investigation down, not the traceability.
8. Industry involvement in epidemiology is absolutely essential
It is all fine and dandy to survey consumers on what they ate but before public health authorities point a finger at specific items, they really need access to industry experts.
Typically, to minimize competitive issues, public health needs to have the ability to pull together some old lions in the field and get the benefit of their expertise. Does it make sense to say that so many people ate organic bean sprouts? Does it make sense to say that these cucumbers would have only gone to Hamburg and environs? These types of reality checks can help public health make sensible decisions.
Following the Salmonella Saintpaul crisis, there were many cries for the US government to provide compensation to tomato farmers, especially because the tremendous financial losses suffered by those in the tomato industry were caused by a governmental error.
Here at the Pundit we did an interview with Professor Richard Epstein, a noted legal authority, and he basically nixed any legal claim. We titled the piece, With FDA/CDC Protected By Sovereign Immunity, Compensation For Losses Looks Bleak Says Professor Richard Epstein. The industry had no more success gathering political support to create a compensation fund. Yet in Europe, Spain immediately put the continent on notice that it would likely seek compensation for its farmers for losses suffered due to Germany’s accusation.
In fact, it seems to be quite widely assumed that compensation will be given. The European Union Farm Commission has proposed a $220 million fund to assist farmers.
However, Spain says this is insufficient and wants more.
Now ultimately producers have to charge consumers for all the costs of production. This includes the cost of dumping product when people won’t buy. Now with Europe taking this cost and socializing it, one would imagine it could be seen as an unfair trade practice — US producers have to cover their losses during outbreaks; the Europeans don’t — this would seem to offer the Europeans a competitive advantage. Of course, rather than filing with the WTO, many American producers would rather see the situation as a precedent.
One issue is that compensating growers is just part of the story. What about wholesalers, traders, retailers, restaurant? Freshfel, the European trade association, made just such a plea in a position paper. But it seems unlikely that compensation would go so far, partly because of the difficulty in making calculations.
Freshfel is also asking for a fund to rebuild confidence in fresh produce:
Freshfel requested Commissioner Ciolos to consider in addition significant sums for promoting the fresh fruit and vegetable category in a relatively short term. These funds, fully financed by the EU, should amount to 50-55 million EURO, which should enable to launch a proper campaign across the EU restoring consumer confidence.
Such issues were discussed in the US after the spinach crisis but never went anywhere. The decision was made to focus efforts on research through the Center For Produce Safety and the California Leafy Greens Marketing Agreement — partly because it was clear no public entity was going to give the US produce industry 50-55 million Euro — 73-81 million in US dollars — to restore confidence and partly because the sense was that the best way to restore confidence was to act in such a way that the problem wouldn’t recur.
In any case, this is clearly a question to be analyzed. If we wish growers to sell produce inexpensively, we can’t burden them with enormous costs. The European solution is to collectivize the costs. America’s solution is to mostly keep the costs in the private sector. Of course, now that the US bailed out the banks and other financial institutions, perhaps the attitude would be different in the future.
People are dropping dead from eating food and we are too cowardly to suggest irradiation. This is bizarre. The failure of public health authorities to demand the use of irradiation is evidence of a politically influenced establishment that is not, in fact, putting public health first.
It is clear that the industry needs a kill step and this is the one available. NPR isn't exactly a bastion of right-wing thinking but even it is running pieces with titles such as Irradiation Underused To Fight E. Coli in Foods.
At the very least, the industry should move ahead and offer irradiated product as a choice. For nursing homes, hospitals, etc. — anyone who serves immune-compromised folks — this is a no brainer.
Opponents of irradiation make vague objections about the long-term consequences of irradiation. Yet 24 people are dead and hundreds severely ill with many never likely to make a full recovery. These people did not have to die. The opponents of irradiation need to be held accountable, and public health authorities need to speak clearly in favor of irradiation.
The whole matter is not easy. It is obvious that public health authorities cause enormous economic damage by finger-pointing without a strong case. It is also true that with such a lethal pathogen, there is a great urgency to get out what information we can, as early as possible, so as to allow consumers to make their own decisions.
We would say the best solution is to maximize transparency. In today’s world, there are brilliant people everywhere who would be anxious to chip in and try to help resolve the epidemiological riddle. The great challenge of management is how to reach out and tap into this broader community. For most part the public health bureaucracies on both sides of the Atlantic have tried to keep control of the information flow, often claiming “privacy” as a cover, in order to sustain their own importance and their claim on public funding. It would, indeed, be embarrassing for these organizations if some freelance epidemiologist in Minnesota, Oregon, Wisconsin or South Africa, for that matter, cracks the case while it is befuddling massive public health agencies. Of course, the public cares not a whit for this and wants the case solved as quickly and inexpensively as possible.
It has to start by sharing information contemporaneously with receiving it — not in a report two years later. This is something the European authorities and the FDA ought both to take to heart.