The non-profit group said outbreaks where the contaminated food and the contaminant were identified by public health officials declined from 41% solved in 2003 to 29% in 2012.
It said from 2009 to 2012, the average number of reported foodborne outbreaks decreased by about one-third compared to the six preceding years.
The introduction of the web-based National Outbreak Reporting System (NORS) in 2009 and reductions in public funding after 2009 could have contributed to the decrease, said the group.
CDC responds to findings
Robert Tauxe from CDC said it was clear there are substantial differences in the relative frequency of reported outbreaks from one state to another.
“Detecting, investigating, controlling and reporting outbreaks takes resources and skill in state and local health departments,” said the deputy director in the Division of Foodborne, Waterborne and Environmental Diseases, National Center for Emerging and Zoonotic Infectious Diseases.
“With the recession, many public health departments have had fewer resources, and public health staffing has been reduced. It is to be hoped that as local and state budgets recover, health departments will be able to increase their food safety capacity.”
Tauxe said one change from 2009 was a change in reporting software so outbreaks of diarrheal illness attributed to animal or person-to-person (P2P) contact could also be reported.
He added CDC thinks some states recognized some norovirus outbreaks previously reported as foodborne were really person-to-person, so as P2P outbreak reports went up, foodborne ones went down.
CSPI analyzed 10 years of state outbreaks (a total of 9,923) reported to CDC from 2003 to 2012 (see report here).
It said many states may lack adequate funding and support for public health services which could impact on the ability to prevent future foodborne outbreaks.
The 10 states in the CDC’s FoodNet surveillance program reported more outbreaks, solved more, and could identify a higher number of pathogens, than non-FoodNet states, found the report.
Boosting lab infrastructure
CSPI recommended CDC consider awarding grants to bolster lab infrastructure, Tauxe said it agreed it needs to help states increase their public health laboratory capacity.
CDC provides supplies and training so state public health labs can “fingerprint” organisms that make people sick. They are added to a national database (PulseNet) and by tracking “fingerprints” outbreaks are found.
“This means that the clinical labs (where the doctors send specimens to get the diagnosis) have to send the bacterial strain they get from that specimen on to the state public heath lab,” he said.
“Every year we distribute money to the state laboratory program through the Epi and lab capacity Building Program. With funds available we can provide limited support that way to all 50 states.”
The “fingerprinting” technology for bacteria is about 20 years old, said Tauxe.
“We are starting the jump to a new technology, based on whole genome, which gets faster and cheaper every year,” he said.
“In our pilot evaluation testing the benefit of applying the WGS method to for Listeria, we have been finding and solving more outbreaks than ever before.
“We hope that all 50 states can be using this new technology routinely for a variety of foodborne infections within five years. As resources become available we will be training and supplying support to the states (via the ELC mechanism) as we build this new network.”
Food categories and bacterial pathogens
The five food categories linked to the most solved outbreaks were produce, seafood, poultry, beef, and dairy.
Those categories made up 57% of solved outbreaks and 58% of illnesses.
Bacterial pathogens were responsible for 54% of all solved outbreaks; viruses caused 35%; chemicals and toxins caused 11%; and parasites caused less than 1%.
The bacterial pathogen most frequently identified was Salmonella spp. (19%), Clostridium spp. (11%), Bacillus cereus and E. coli spp. (6%) and Staphylococcus spp. (5%).
Information technology is a continuing challenge, according to Tauxe.
“Each state has their own data system, and over the years, CDC has constructed several different ways of transmitting different kinds of data that are not as integrated as they would be if we had designed them all at once,” he said.
“We are trying to make reporting easier and simpler for the states, for example by developing software that can automatically receive reports about outbreak investigations from their own data system without having to re-enter any data.
“Our FoodCORE program is looking at ways to make county-state data sharing work better. We are also getting better at getting our different data system to talk to each other, but there is more to do.”
Reports to CDC ranged from greater than 1,300 outbreaks in 2004 to fewer than 700 in 2009.
A high outbreak reporting rate can prevent illnesses, as it shows state and local public health officials are looking for outbreaks and are more likely to identify contaminated foods, said CSPI.
For example, Florida reported five times the number of outbreaks as Alabama when controlled for population, and Maryland reported four times the number of outbreaks as West Virginia.
CSPI said comparisons cannot be made between states due to the different ways of reporting.
Most solved outbreaks were reported from restaurants and other food establishments (40%), followed by private homes (24%) work places (8%) and catered events (6%).
Tauxe said the gap between actual and estimated foodborne illness is not a serious problem, as long as it can estimate its size.
“There will always be people who are sick, but do not visit the health care provider, choosing to tough it out at home. There will always be doctors who say “looks like a virus” and send people home without a getting a lab test done,” he said.
“It would be amazingly expensive to try to change that, and I don’t think it would make a big difference to public health. What is critical is for us to be able to measure that, and account for it, so our estimates of the total burden of illness are accurate.
“We do that by conducting population survey. The late of those was in 2006-2007. Resources permitting, we hope to be able to restart the survey in the coming years.”