By Thomas Hargrove, Scripps Howard News Service
 [Click on image for large chart]
Hargrove presentation from Health Journalism 2007: A PDF of the presentation Thomas Hargrove gave at AHCJ's annual conference. The Fatal Food project
Related links
Related stories
|
Just how good is America's network of public health departments? How ready are they in the event of a bioterrorist attack on the nation's food or water supply?
Editors at the E.W. Scripps Co. wanted to experiment with a partnership between the firm's television and newspapers to determine if a combined multimedia investigation could produce quality journalism. As a database analyst and national reporter at Scripps Howard News Service in Washington, D.C., I was assigned to identify a research project that could produce localizable news stories.
Although the historic nationwide E. coli outbreak caused by infected raw spinach was still months into the future, editors and producers in June readily accepted my suggestion to focus on food-borne illness outbreak reports collected annually by the Centers for Disease Control and Prevention. The little-known annual reports detail how many Americans are sickened or even die from food illnesses like hepatitis A, salmonella, norovirus or clostridium.
The reasoning by news managers was simple enough: Everyone eats, so concern about food safety is universal. Because no other news organization had used these particular CDC files, there was an opportunity to break new ground.
The first step was to turn written annual reports into a computer database. Using Adobe Acrobat software, Scripps Howard reporters were able to extract raw information the CDC had posted on its Internet site. It took about a week to extract and clean 6,374 records of food poisoning outbreaks that harmed 127,055 Americans reported during the five-year period from Jan. 1, 2000, through Dec. 31, 2004. The files were then transferred from an Excel spreadsheet to an SPSS file for detailed analysis.
The immediate finding was disturbing. Sixty-four percent of all outbreaks of food sickness were officially listed as from "unknown" causes. That meant most of the time, state and local health departments failed to isolate the specific bacteria or virus that is making people ill. Health investigators tend to do a better job with really large outbreaks that infect hundreds of people. But more than 50,000 Americans got sick or died in outbreaks in which the cause was undetermined.
Next, the data were aggregated by state. The total number of people who were reported sickened by contaminated food was calculated, as well as the percentage of cases that went unsolved. We also estimated the rate at which food-borne illness is detected by calculating how many people were reported sick per 100,000 population.
The variance by state was enormous. The Alabama Department of Public Health diagnosed the cause of only 5 percent of its cases. In fact, the state had not isolated the cause of any food illness outbreaks in the most recent 2 1/2 years. Florida and New Jersey were not doing much better.
"It's a real struggle. We've never identified a virus at the state level. We've always had to send viral specimens to the CDC for testing," said Alabama State Epidemiologist John Lofgren. He said many key jobs in his department have been vacant for years because of low salaries.
New Jersey officials said they, too, were hampered by a lack of local health investigators to chase the causes of food poisoning outbreaks. "It comes down to a matter of resources," said Eddy Bresnitz, New Jersey's deputy health commissioner.
But some states, such as Wisconsin, Minnesota and Hawaii, were performing very well. The Wisconsin Department of Health and Family Services diagnosed the causes of food illness 90 percent of the time, the best rate in the nation.
"This is not a magic trick. It has to do with our total system," said Wisconsin Health Secretary Helene Nelson. "We have local public health departments, we have clinicians (doctors and nurses) who are first encountering the diseases, we have a state public health division that coordinates and communicates and helps analyze with the epidemiologists, and we have really good state lab facilities. It is all those parts working together."
Perhaps even more disturbing, the data suggested that some states are virtually blind in detecting outbreaks of food illness. The state of Kentucky, for example, reported just four outbreaks over a five-year period that sickened only 35 people. Other states were detecting these illnesses at a rate 100 times greater than was the Bluegrass State.
"We had more people than that get sick just at my wedding," said Lindsey Ronay of Louisville, Ky. Doctors told her wedding guests they were the victim of the campylobacter bacteria. More than 30 people were but on antibiotics and two were hospitalized.
Kentucky health officials began an investigation after we contacted them.
"What we are reporting is just what we can document and substantiate from within our files," said Kentucky State Health Officer William Hacker. "I suspect we are underreporting what we deal with on a regular basis, just from my own experience."
Public and private laboratory technicians make most of the contagious-disease reports in Kentucky. But lab techs usually are not told if the blood, urine or stool samples they have been asked to test are part of an outbreak.
A month after we contacted Kentucky medical officials about the state's low reporting rates, authorities announced they are "enhancing" the state's Electronic Disease Surveillance Module to require people who report a communicable disease to check "yes" or "no" to questions of whether it could be part of a food- or water-related outbreak.
"We really hadn't been categorizing food- and water-borne outbreaks," said Kentucky Epidemiologist Kraig Humbaugh. "This will help us with our reporting system. It should increase the number of food outbreaks that CDC sees from us."
Database analyst Thomas Hargrove is a national reporter at Scripps Howard News Service in Washington, D.C. He has reported on the early deaths of NFL players and the increase in motorcycle crash deaths since many helmet laws were repealed .
