Trust, but verify.” Ronald Reagan’s Cold War dictum has never rung more true than in the wake of the horsemeat scandal.
As retailers and suppliers across Europe discovered horsemeat fraudulently mixed into their beef products, that lesson was abundantly clear: you really do need to know who you’re buying from, and you need to run tests to be absolutely certain your products haven’t been messed with.
But how has this lesson translated into practice, and what are the long-term implications of the horsemeat scandal on the supply chain?
The most visible change has been the introduction of speciation tests - including, obviously, for horse. Thousands of tests have been carried out since January - ABP alone completed close to 12,000 - and retailers like Tesco (which has tested more than 1,000 products) also launched “industry-leading” DNA regimes to restore consumer confidence. However, while extensive DNA tests and positive-release procedures were necessary in the immediate aftermath of ‘Horsegate’, it isn’t yet clear what level of testing is proportionate in the longer term.
A Tesco spokeswoman says tests will become “one of the permanent controls” in its supply chain, while Morrisons says it has agreed - along with the wider industry - to take part another round of testing through the FSA. “We will test all finished products and, where it adds value and is appropriate, we’ll be testing raw material as well,” adds technical director Andrew Clappen. The FSA will publish those results in June, but what happens then remains to be seen - it hasn’t decided how long testing will continue.
Cost will be a key consideration if tests are used on a permanent basis. DNA tests using polymerase chain reaction (PCR) technology cost upwards of £100, so companies can burn through a lot of money very quickly if they use them for screening. However, there are other, cheaper tests available that allow companies to test at a more manageable price (see box p41).
ABP, for example, is using Elisa-based fast tests - which can be carried out on-site rather than in a lab - to screen its raw materials and finished products. “We are told they are as accurate as Elisa, and can detect down to the 1% contamination level,” says agriculture and livestock manager Stuart Roberts. “They are good for screening, and then you can follow up with Elisa or PCR if necessary.”
But there are challenges beyond cost: speciation tests, whether cheap or expensive, can only find what you are looking for - although some broader screening tools are emerging, including a new laser gun (see box p43). Plus there is little standardisation in the testing industry, leaving lots of room for results to be challenged. “Standards are urgently needed across Europe,” says Sandra Luley, global marketing manager for applied testing at DNA technology provider Qiagen. “I have doubts some of the results we’re seeing reported are comparable.”
“Beware the false positive paradox: tests for something very rare are more likely to deliver false than genuine positives”
More complicated still is the issue of how to deal with ‘false positives’, and how and when to communicate results. Companies like Findus UK were severely criticised for delays in telling the public about positive horsemeat tests, creating an impetus for others to release results as soon as they have them.
But like all tests, DNA tests sometimes produce false positives, and people tend to underestimate the impact of those when testing for something (like horsemeat) present in only very few products, says Dr Angus Knight, principal scientist at Leatherhead Food Research. This is the false positive paradox: a statistical quirk that means a test for something very rare is more likely to deliver a false positive than a genuine positive. “If you have a test that’s accurate 99% of the time and you have an incidence of positivity of one in 1,000, you are 10 times more likely to get a false positive than a real positive,” says Knight.
That isn’t to say DNA tests aren’t useful, but it makes confirmatory tests all the more important - a point ABP would certainly agree with, after getting caught up in a false horsemeat alarm on fresh beef that went into the Asda bolognese sauce produced by Greencore (see Paul Finnerty p44). Gaining clarity on how the industry is going to use and communicate test results in the future is key, says Roberts. “It’s very easy to misuse test results and not understand at times what they mean.”
Nor do tests, on their own, make for a more secure supply chain. “Don’t get it in your head DNA testing is a panacea for everything,” warns Clappen.”It’s just one part of the whole end-to-end assurance piece.” Sainsbury’s - like Morrisons, not affected by the scandal - strikes a similar note. “It does highlight the importance of extensive supply chain assurance processes in terms of traceability, audit and analytical testing,” says Alec Kyriakides, head of product quality, safety and supplier performance.
Because the horsemeat scandal lifted the lid on a highly complex web of suppliers and traders, taking complexity out of the supply chain and bringing sourcing closer to home has been a key focus for many companies.
Laser gun is new weapon in the fight against horsemeat
Physicists at the Technical University of Berlin have developed a laser gun able to detect horsemeat in seconds.
A laser beam is shot into meat, where it is scattered and measured using Raman spectroscopy.
Molecules in different meats have a “unique fingerprint” and scatter the laser in a specific way, allowing researchers to identify what species are in a product, says Dr Heinz-Detlef Kronfeldt, who heads up the project.
The gun can be used through packaging, can test for multiple species at the same time, and can also identify how old meat is and if it’s been defrosted.
It exists only as a prototype for now, but Kronfeldt says he hopes it will be available for commercial use soon.
Cost at the moment is c. €30,000 (most of it accounted for by the spectrograph), but this could come down to €10,000
But there’s a limit to how much meat can be sourced locally (the UK isn’t self-sufficient in beef, for a start), so suggesting complexity can be removed completely is dodging the issue, believes Ishan Palit, CEO of the product services division at product safety certifier TUV Sud. “We can’t get away from complex supply chains, but we can learn how to deal with them better,” he says.
That means robust testing has to be underpinned by robust traceability, says Palit - and that’s a big challenge. Tesco says it’s currently trialling two “world-class traceability systems” for products containing proteins, which it says will “provide confidence to us and to our customers”, but in a TUV Sud survey in late 2012, just 46% in the food industry said they had complete traceability on their products.
Stringent audits and inspections can play their part, and some companies have decided in the wake of the scandal to start conducting their own audits. Retailers already do this extensively on their own suppliers, and ABP says it is also doing its own audits on raw material suppliers now, having previously relied on the EU licensing system and BRC certification alone.
But what you do after audits is as important as the audits themselves, says Doug Powell, a food scientist at Kansas State University, who has researched industry behaviour around inspections. “Companies often think passing an audit means they’ve got a clean bill of health but it’s just a snapshot. They don’t spend enough time going through the reports afterwards to follow up on any issues that were raised during the inspection.”
So when companies announce tougher checks, Powell says: “I always chuckle - those were meant to be tough before the scandal, so how do they know it’s going to work this time? The entire company culture around audits has to change.”
Preventing future scandals also requires ‘horizon scanning’. “Identifying fraud requires constant vigilance by all parties, and the sharing of information with industry by government and regulators is important in supporting our efforts to guarantee the integrity of the food supply chain,” says Kyriakides at Sainsbury’s.
Chris Sambrook, an intelligence specialist at Thames Valley Police who is writing a PhD on rural crime, adds that the industry needs to become much more professional in its approach to intelligence. “There is a huge link between counterfeiting and transnational, organised crime,” he says. “What might appear to be petty crime - a dodgy burger, for example - can often be part of a much wider net of criminal activity.”
The challenge is identifying the next big opportunity for fraud and stopping it before it is exploited - because once a fraud is under way, chances of detection and successful prosecution are low, he warns.
“If consumers start expecting virtually zero cross-contamination, where does that leave smaller processors or the local butcher and his mincer?”
To do that, the industry needs to forge closer links with formal intelligence agencies, Sambrook believes. At present these are “very, very tenuous”, but the time is right to change that. The introduction of local police and crime commissioners is “a real opportunity to influence policy direction”, he says, while the replacement in October of the Serious Organised Crime Agency with the National Crime Agency - which will have a clear focus on “third-party links” - as well as the Home Office’s new “local to global” strategy for organised crime, also offer opportunities to build closer links.
Of course, this is not a scandal in which only horsemeat has been implicated. Tests conducted in its wake also uncovered pork in some beef products and such discoveries could have equally far-reaching consequences. Most meat processing plants in the UK process more than one species, and - unlike horsemeat - pork or lamb can find its way into beef products through perfectly legitimate business practices. “There is no legal obligation to clean between red meat species,” says deputy government chemist Selvarani Elahi. “All the hygiene regulations are based on poultry.”
Even where processors do clean between species, current hygiene practices do not necessarily prevent the kind of minute cross-contamination picked up through highly sensitive DNA tests. To find out how prevalent cross-contamination between red meat species is at the moment, Elahi and her team are working on a cross-contamination project (see Fresh story p34).
The results will be of particular interest to faith groups, but the FSA is also gauging what level of cross-contamination mainstream consumers consider acceptable. At the moment, FSA guidelines deem pork at 1% and above an incident of cross-contamination, but it’s not yet clear if this will remain the threshold for non-halal and non-kosher products.
Some processors have already drawn their own conclusions and overhauled their hygiene practices - ABP says it’s doubled the number of cleaners at Dalepak to 20, introduced a new 20-point cleaning regime, and now strictly segregates species in its cold store - and manufacturers of halal products that don’t have dedicated lines will be looking closely at introducing them.
Cross-contamination risk can also be reduced significantly if processors commit to processing only one species per day, but that isn’t always an option for small-scale processors.
And here’s the rub: if consumers start expecting virtually zero cross-contamination, where does that leave smaller processors or the local butcher and his mincer? They were meant to be one of the winners of the horsemeat scandal, but many measures taken by the industry to reassure consumers require deep pockets, and therefore inherently favour big suppliers and the supermarkets.
No one wants fraudsters to pass off horsemeat as beef, and companies are right to make supply chains more secure - regardless of cost. But as the industry, regulators and, ultimately, consumers draw wider conclusions from the scandal - including on cross-contamination - balancing the need for higher levels of food authenticity with having a viable, diverse meat sector will become an important priority.
Meat species tests: the options
Polymerase chain reaction (PCR)
What is it? When people talk about ‘DNA tests’, they mean PCR. The most sensitive test, it can detect minute amounts of DNA below 1%, but precise quantification can be a challenge. Needs to be done in a lab and isn’t cheap - upwards of £100. Results take a few days to come back, although faster tests are being developed.
What is it? Tests for proteins rather than DNA. Less sensitive than PCR, but also more affordable - tens rather than hundreds of pounds. Tests need to be done in a third-party lab, so offer more cachet and precision than fast tests. Some experts warn Elisa is more suited to primary rather than processed meats.
Elisa-based fast tests
What is it? Cheap and cheerful, it’s the equivalent of a pregnancy test. Also based on proteins, tests can be run through widely available test kits, don’t require lab use and deliver results in less than an hour - through a liquid or dipstick that changes colour. The least sensitive of all three tests, useful mainly for screening.
Horsegate: Who are the winners?
- Currently reading
A new era for meat supply: Trust but verify