Car time used to be the best thinking time for me. While living in Guelph and working on my MSc I’d drive semi-weekly to the greenhouse vegetable capital of North America, Leamington, Ontario (that’s in Canada) to do some on-farm food safety work with the industry. I’d throw some tunes on, rock out and try to get stuff straight in my head. On the drive I’d dream up my next activity or op-ed idea; between Tim Horton’s stops and refinement conversations (I had just got my first cell phone) some salient ideas might have developed.
Today I spent 15 hours in the car driving from Port Hope, Ontario (also in Canada) to Raleigh, NC with Dani and 21-month old Jack in the vehicle (which is now a family friendly minivan equipped with a DVD player). Elmo and The Wiggles DVDs (4 and 6 viewings today, respectively) have replaced rock-out-friendly tunes and phone conversations have been replaced by pointing out buses, cows, planes and boats to Jack.
Regardless, I still had some time to think about some stuff.
Coverage surrounding our food safety infosheet evaluation paper last week has been pretty decent, with pick-up from Scientific American, AP and USA Today (which Doug has already mentioned) as well as a few blogs. The focus of the paper has been represented pretty well, but there have been a few things worth clarifying and addressing.
From Scientific American:
– And recent research by food safety specialist Ben Chapman of North Carolina State University found that meals prepared in commercial kitchens have been involved in up to 70 percent of food poisoning.
Our team didn’t look at tracing where contamination happens and how many meals have led to outbreaks in the study, although we reported in the intro that up to 70% of outbreaks have been traced to meals outside of the home. We grabbed that estimate from a few sources including a combination of CDC outbreak line listing summaries and Ontario, Canada outbreak statistics.
A limitation of the data is that it is an estimate derived only from confirmed outbreaks, which are usually reported by year. We saw estimates as low as 14% (Lee, M., and D. Middleton. 2003. Enteric illness in Ontario, Canada, from 1997 to 2001. J. Food Prot. 66:953–961. ) and as high as 83% for one year. We settled on up to 70% after looking at all the papers and eliminating/combining sources with assumptions and averaging outbreaks data out over a 5-year timeframe.
Casey and Doug tackled this question in a much more succinct way in a Foodborne Pathogens and Disease paper last year (Where Does Foodborne Illness Happen—in the Home, at Foodservice, or Elsewhere—and Does It Matter?): Current surveillance systems focus on the place where food is consumed rather than the point where food is contaminated. Rather than focusing on the location of consumption—and blaming consumers and others—analysis of the steps leading to foodborne illness should center on the causes of contamination in a complex farm-to-fork food safety system.
Colin Caywood at Marler Clark’s Food Poisoning Journal used the paper to make the point that cross-contamination in restaurant kitchens can magnify the impact of an outbreak:
With that in mind, a recent article by Nicole Norfleet caught my attention for its insight into the way that outbreaks such as Subway’s can be made exponentially worse by poor food safety practices at the restaurant…. Among the risky behaviors cited were workers using aprons and other garments to dry hands, as well as using the same utensils and surfaces to prepare both raw and cooked foods.
I definitely agree with cross-contamination making things worse, but I’m not sure if the current Subway-linked outbreak, where illnesses have been associated with food at 46 outlets, making it appear to be a common supplier source is the best example. Several foodhandlers testing positive for Salmonella serotype Hvittingfoss is beginning to look like the 2009 Fat Duck outbreak.
From The AP story:
Joan McGlockton, a food policy representative for the National Restaurant Association, was cited as saying that while the study is disconcerting, the association doesn’t feel it is representative of the entire restaurant industry.
Yeah, I agree and the study wasn’t built to allow for generalizations. Our aim was to evaluate the efficacy of the infosheets as a behavior-changing intervention. While we were also able to gain some data that can be used in risk assessments, it has to be used carefully with realistic assumptions, because it’s the only video data set out there. Based on the time, effort and resources committed by the company we worked with, what we saw might represent the best practices out there. But maybe it doesn’t. We’re both guessing and if the NRA has some behavior data that we can compare our findings to that would be great.
I’m back in NC and starting a couple of new projects with the NC State team (Audrey Kreske and Allison Smathers) over the next week, further measuring what people actually do when it comes to food safety. More on these as data starts rolling in.