Traceability key to mock recall scenarios
March 26, 2015 By Peter Mitham
The key to a successful recall lies not just in the ability to recognize when something’s gone wrong and to alert consumers that they need to get potentially contaminated or defective products off their hands, but to know just how much product is affected and whether all of it is accounted for.
This makes the ability to trace product critical, even for small operations, and knowing that the systems put in place to track product and account for its whereabouts work.
Speaking to B.C. producers in 2014, Elsie Friesen of the B.C. Ministry of Agriculture said that requiring processors to regularly test recall responses through mock recalls would be part of Canada’s efforts to harmonize its food safety requirements with those of the U.S. and its new Food Safety Modernization Act.
Gary Sandberg, head of the food technology program at the B.C. Institute of Technology in Burnaby, B.C., said the Canadian Food Inspection Agency already requires federally registered processors to have a recall strategy. Now, any plant that ships product to the U.S. will be required to be federally registered and have a strategy, too.
“[If] you’re exporting to the United States, you’re going to automatically now be a federally registered plant, which means you’re under the CFIA’s jurisdiction,” Sandberg explained.
On the face of it, it’s not a bad thing. While a hazard analysis critical control points (HACCP) protocol can identify risk points and incorporate a simple response process, a proper recall strategy can be far-reaching.
“It’s like doing a disaster planning exercise,” Sandberg explained. “You definitely want to be able to move through the whole thing and, one, ensure that you can track any product which is going to have problems, and secondly, be able to pull that product back into a centralized location and deal with it accordingly.”
Putting the recall protocol to the test doesn’t mean having to go into full disaster mode and shutting the entire plant down, however. Rather, it can be as simple as trying to find out where in the process a particular lot of product is.
“You can do it … [by] looking at a lot number and saying: ‘Okay, can we actually find it?’” he said. “Then it becomes a paper-based thing where you say, okay, if we identify this food product with this lot number, and can we go through the bills of lading and say, okay, where did we ship it to, and do an inventory count and say how much do we have, how much did we manufacture, how much is not accounted for?”
Since incidents that require recalls typically don’t come to light during regular business hours, it’s important to test the ability to reach key contacts.
Sandberg served as a quality assurance manager for T.J. Lipton in Richmond, B.C., prior to becoming an instructor at BCIT. He knows first-hand the difficulties of trying to reach people and also to track product.
One of the measures that Lipton had in place to define the massive amounts of product it was producing was segmenting the production into 90-minute segments. Stock was tagged with an alpha-numeric code for each day and time. This allowed Lipton to hone in on a smaller run of product than if it could only identify a single day when the problem occurred.
“Instead of recalling maybe 5,000 or 6,000 cases, we’re maybe down to 200,” he said.
Smaller processors, or farms handling a large volume of fresh produce might find it difficult to track product from field to farm gate, but Sandberg said it’s possible to track produce by row or field, and note which runs contribute to particular processing periods. And then, produce that’s package is typically tracked by codes on the packaging that takes it to retail.
“It becomes a bit of an education process for the smaller processors, and again for the farmers, to think ‘If I did have a problem, then how do I go about controlling it?’” Sandberg said. “[But] if they’re coding their products with a date code on it and some sort of a lot number, and then the shipping documents are maintained and you’re recording date numbers and quantities … your regular inventory control system should give you a lot of the information you want.”
Sandberg also noted that grocers have inventory control systems that should dovetail with those of producers, allowing contaminated product to be identified and the public warned no matter at what point in the system a problem arises with a product.
A mock recall will put those information systems to the test. Regular testing of the systems can highlight areas that require closer attention and may identify opportunities for improving the traceability of product. Ultimately, the systems should work in the interest of all parties.
“It becomes kind of like an insurance policy – if you’re able to identify, isolate and recover that product without having to get widespread into the trade, then it’s definitely going to be helpful even if there is litigation,” Sandberg said. “You can say you do have a recall program in place, the public was notified, we did do all this due diligence to recover the product. It does work in favour of the companies involved.”
Print this page