War of numbers
by Steve Bjerklie
KANSAS CITY -- In several recent public appearances, executives with the American Meat Institute, including president and chief executive officer J. Patrick Boyle, have stated that the prevalence of E. coli O157:H7 in beef products is going down, and they’ve used data from the Food Safety and Inspection Service and the Centers for Disease Control and Prevention to support the claim. Using test data from the F.S.I.S., the A.M.I. has in particular emphasized that from 2000 through 2008, E. coli prevalence in ground beef dropped 45%, suggesting the beef industry’s various efforts to control the pathogen have been effective.
But Barbara Kowalcyk, director of food safety for the Center for Foodborne Illness, Research & Prevention in Grove City, Pa., and a doctoral student in molecular epidemiology and environmental health, said the A.M.I. is misusing the data to paint a rosy picture for consumers that’s dishonest. In a strongly worded statement released last month, in which she called Mr. Boyle’s recent comments about E. coli reduction in beef “inappropriate and misleading,” Ms. Kowalcyk wrote: “U.S.D.A.’s E. coli O157:H7 microbiological testing program is strictly regulatory and was not statistically designed to estimate the prevalence of E. coli O157:H7 in raw ground beef. Different establishments are sampled each year. Further, the methods used to select establishments and to conduct the microbial testing have changed over the years. As a result, it is inappropriate to make year-to-year comparisons. Several sources, including U.S.D.A. itself, have noted the limitations of the data obtained from U.S.D.A.’s Verification Testing Programs."
“I don’t think it’s possible to draw any conclusions about the prevalence of E. coli from these data,” she said. “This is not a small issue, this is a big one.”
Jim Hodges, executive vice-president of the A.M.I. and director of the American Meat Institute Foundation, which supports scientific research, said he doesn’t necessarily dispute Ms. Kowalcyk’s statements.
“She takes a pure view of how these numbers are derived,” he said. “Fundamentally, I don’t disagree with her. But the fact is, everyone uses these numbers to look at trends – F.S.I.S. does it, C.D.C. does it, everyone does it. If we wanted to be perfectly accurate, we’d say there has been a significant reduction in E. coli, but people want to know how much, so we use percentages to show a trend.
“I guarantee you trends are important to everyone,” he added. “And that’s all we’re trying to communicate. The point is trends, not actual numbers.”
Ms. Kowalcyk isn’t buying it. She said it isn’t accurate to compare selected year-to-year data from tests that were never intended to point toward a trend. She said the test data collected in 2008 came from a different group of beef plants than the test data from eight years earlier. The side-by-side comparison does show a 45% drop, she admitted, but likens such a comparison to comparing someone who weighed 300 pounds in 2000 to someone else who weighed 150 pounds in 2008 and drawing the conclusion that people in general have experienced a 50% drop in weight.
“If you want to do a trend analysis, let’s be honest about it,” she said. “The misuse of statistics is what gives statistics a bad name,” adding that U.S.D.A.’s own Web site misapplies E. coli test data to prove trends that don’t exist.
She also said the A.M.I.’s recent claims that C.D.C. data show that E. coli infections have decreased 44%from 2000 through 2008 are also misleading and ignore more recent trends.
“In actuality, C.D.C does not compare individual years of FoodNet data (i.e. 2008 versus 2000). Rather, C.D.C. compares the data for a given year to a composite of the 1996-1998 FoodNet data and to a composite of the preceding three years, which in this case would be 2005-2007 FoodNet data. This is done to account for changes in the number of FoodNet sites and changes in the size of the population. It is true that, when comparing the 2008 FoodNet data to the 1996-1998 composite, E. coli O157:H7 infections have decreased 25%," she said last month. “However, when comparing the 2008 FoodNet data to the 2005-2007 composite, E. coli O157:H7 infections have not changed significantly.”
Mr. Hodges responded: "Our numbers probably over-estimate the prevalence of E. coli. They’re regulatory samples and tend to be biased toward where the problems are. But they’re the best data we’ve got. To not use them is a complete disservice to the public."
Ms. Kowalcyk and Mr. Boyle appeared together in October on a segment of CNN’s "Larry King Live" interview program in which beef safety was the focus, following a New York Times report that uncovered serious lapses and gaps in the industry’s efforts to control E. coli.
Ms. Kowalcyk herself has a direct connection to the issue: her two-and-a-half year-old son Kevin died in 2001 from O157:H7 poisoning that apparently came from adulterated ground beef that was later subject to a recall.
“It very much concerns me when the government or industry puts out information that can give the public a false sense of security,” she said. FSM
The author is a contributor to Meat&Poultry Magazine, a sister-publication to Sosland Publishing’s The Food Safety Monitor.