Preventing foodborne disease a ‘farm to fork issue’

Some restaurants in the United States have begun serving raw chicken. Chicken sashimi, chicken tartare — dishes that might make an infectious disease physician queasy with thoughts of Salmonella poisoning — are being eaten by customers in the U.S.

“That’s Russian roulette,” Timothy F. Jones, MD, state epidemiologist for the Tennessee Department of Health, said in an interview.

Likewise, Kirk Smith, DVM, MS, PhD, manager of the foodborne, waterborne, vector-borne and zoonotic diseases section at the Minnesota Department of Health, said raw chicken, no matter how it is sourced or prepared, is too dangerous to consume.

“Pathogens are common on raw chicken,” Smith said. “If that becomes more of a widespread practice, I guarantee you we will see outbreaks.”

Source: CDC

Each year, around 17% of people living in the U.S. contract a foodborne disease, causing 128,000 hospitalizations and 3,000 deaths, according to CDC estimates. These grim statistics tell only part of the story. Getting a complete picture of foodborne illness in the U.S. is difficult.

In search of the latest clinically relevant information, Infectious Diseases in Children spoke with several leading experts about what is causing foodborne disease in the U.S., how infections are being diagnosed and tracked and how the behavior of patients and physicians is having an impact.

“We live in a microbial world, and I don’t think we will ever have totally sterile food,” Robert V. Tauxe, MD, MPH, director of the CDC’s Division of Foodborne, Waterborne and Environmental Diseases, said in an interview.

‘We’re at an interesting point’

According to data from the CDC’s FoodNet surveillance network, which tracks seven bacterial and two parasitic causes of enteric disease, Campylobacter was the leading cause of laboratory-diagnosed bacterial foodborne illness in the U.S. last year, followed by Salmonella, Shigella, Shiga toxin-producing Escherichia coli (STEC), Cryptosporidium, Yersinia, Vibrio, Listeria and Cyclospora. Overall, out of 31 known pathogens that cause foodborne illness, norovirus is by far the most common, accounting for more than half of all infections, according to the CDC.

Established in 1995, FoodNet monitors sites in 10 states, covering about 15% of the U.S. population. According to Tauxe, data from the program have revealed overall progress over the past 2 decades against some bacterial pathogens, including the deadly E. coli O157:H7 strain, Campylobacter and Listeria, all of which have become less common because of important improvements in meat safety. However, the incidence of these infections has not continued to decrease but instead leveled off recently. Meanwhile, infections caused by Salmonella have remained steady at the same levels as 20 years ago. According to Tauxe, surveillance of these pathogens is now being affected by changes in how infections are diagnosed and reported.

“We’re at an interesting point now where both the nature of the tests that are being ordered and possibly physicians’ behavior may be changing,” Tauxe said.

CIDTs impact patient care and public health

Physicians are increasingly using culture-independent diagnostic tests (CIDT) to rapidly identify foodborne illnesses. Results that might take days to get from a culture can be had in hours. CIDTs are comprehensive, too, testing for many pathogens at once, including parasites. They can even detect viruses like norovirus, which cannot be cultured.

“It’s easier, it’s faster, so the advantages to the clinician are great. You can get the results the same day,” Jones said.

But there are disadvantages. Used alone, CIDTs tell only part of the story, leaving out critical information for foodborne surveillance. They do not yield pathogen isolates for molecular subtyping, making it more difficult to link infections to outbreaks, including multistate outbreaks like the recent spate of Salmonella infections in at least a dozen states that were linked to imported Maradol papayas from Mexico. Multistate outbreaks are being detected more frequently, chiefly due to a larger proportion of isolates that are subtyped — a trend that will start to go backward with increasing use of CIDTs, according to Smith.

Missing data from CIDT-only results means that reported outbreaks might be larger than they seem; some may even be missed entirely.

“CIDTs are great for physicians and patients, but we lose a really important public health component, which may not impact the individual patient but has a really large impact on public health,” Elaine Scallan, PhD, director of the Rocky Mountain Public Health Training Center, co-director of the Colorado Integrated Food Safety Center of Excellence and associate professor in the department of epidemiology at the Colorado School of Public Health, said in an interview.

CIDTs also do not detect antimicrobial resistance, making it more difficult to monitor trends. Yet, for the first time, national prevalence data for 2016 collected by FoodNet and published by the CDC included some infections that were identified only with CIDTs, leaving important gaps in data. The inclusion of CIDT-only results raised the incidence rates for six of the most frequently reported bacterial foodborne illnesses, including Campylobacter. Culture-confirmed Campylobacter infections decreased significantly compared with the average number of cases in the previous 3 years, but the use of CIDT results led to a “slight but not insignificant” increase in incidence, Tauxe and colleagues wrote in the report.

Other issues exist. CIDTs can sometimes produce false-positive results, leading to skewed estimates. Interpreting data that include CIDT-only results makes it difficult to know if an uptick in cases is due to increased testing, varying test sensitivity, an actual increase in infections or a combination of all three. Studies have raised questions about the accuracy of some tests, including one published in 2016 in the Journal of Clinical Microbiology that demonstrated highly variable sensitivity, specificity and positive predictive value of stool antigen CIDTs for Campylobacter.

Smith said better data are needed to learn if physicians are ordering CIDTs at a higher rate than they were ordering cultures in the past because it may give the perception that the incidence of foodborne illness is increasing when in fact it is stable, or even decreasing.

“CIDTs have introduced this huge wrinkle because we don’t know how physician testing practices are changing,” he said. “Are they testing basically the same proportion of patients that they see? Are they testing a greater proportion? That can have a huge impact on what’s identified and reported to public health, which makes our job in measuring foodborne diseases incidence a lot tougher.”

As a practical solution to these problems, some states, including Tennessee and Minnesota, have passed laws requiring that clinical or commercial laboratories that receive a positive CIDT result send the original specimen to the state lab for culturing and additional testing. This saves clinical labs from covering additional costs associated with more testing.

“It’s increased work for the state lab, but it’s in our interest,” Jones said. “We have pretty good compliance. By the time things get shipped and mailed, bugs can die on the way and you can lose some information, but by and large it’s working pretty well.”

Ordering follow-up cultures can reveal information missed by CIDTs. However, according to Tauxe, only around 60% of positive CIDT results are followed by culture of the same positive specimen — a so-called reflex culture. The percentage is higher for some pathogens, running to almost 90% for STEC, for example. Tauxe said physicians can play a role in improving those numbers.

“I think there’s a really important role for the clinician, especially [the] infectious disease clinician, in encouraging laboratories to do the reflex cultures and, when they get a positive back from the lab by a PCR test, encouraging their colleagues to ask up-front that the same specimen be cultured either in that lab or sent to the state lab for culture,” he said.

This is perhaps easier said than done. A physician’s main job is treating their patient, and CIDTs provide faster diagnoses, Jones said. Plus, insurance companies do not necessarily want to cover the extra costs that come with ordering more tests.

“In the health care system that we have, they don’t like to do multiple tests that get at the same thing,” Jones said.

He said instituting protocols to make such reflex testing automatic, without a specific order, may be worth considering in some situations.

Whole-genome sequencing

Although CIDTs can make it more difficult to detect outbreaks of foodborne illness, when isolates are available, whole-genome sequencing (WGS) has given state and federal labs the capability to identify them more easily. According to the CDC, WGS reveals the complete DNA makeup of an organism, allowing health officials to define the scope of foodborne outbreaks, identifying them faster and with more accuracy. The agency began using the technology in 2013 to quickly detect outbreaks of Listeria with as few as two cases.

Some state labs also have begun using WGS, but the method is still years away from being feasible for clinical labs, which may have less of a need to use the technology than state or federal labs.

“When you’ve got a hospital laboratory that’s trying to help a doctor save a patient, some of the really high-tech stuff isn’t necessary. That’s more a public health value,” Jones said.

According to Smith, the CDC has provided money for expensive sequencers and reagents to build up the capacity for state labs to perform the tests. In addition to identifying more and smaller outbreaks, the process allows public health officials to better classify them. Whereas pulsed-field gel electrophoresis, the longtime CDC standard for detecting foodborne disease, might show a multistate outbreak of several hundred cases across the country, WGS might identify it not as one big outbreak but as many smaller outbreaks or as one outbreak embedded in many otherwise unrelated cases.

“That focus on what’s more likely to be truly related to a common source has really helped save public health resources from trying to put together 200 cases that aren’t related,” Smith said.

Eventually, Jones said WGS may replace both cultures and CIDTs, ultimately giving experts a clearer picture of what is buried in an organism’s genes, including detailed subtyping and antibiotic susceptibility. Currently, for example, WGS can detect the presence of many genes that are known to be associated with antimicrobial resistance, but it is not able to identify previously unknown mechanisms that may emerge, Jones said. That still requires culturing.

WGS has gotten cheaper, but hurdles for wide implementation remain, including dealing with the volume of data that it provides.

“In the not-too-distant future, it’s going to be the test that gets done,” Jones said.

The impact of diet

Although methods of detecting infections and discovering outbreaks have improved, it is difficult to know if changes in diet are affecting foodborne illness in the U.S.

“The data are surprisingly poor,” Jones said.

There are some constants. Most Salmonella infections are still associated with poultry, such as undercooked chicken. But there also have been changes. STEC — including its most common form, E. coli 0157 — was once thought to be primarily associated with beef but is now being linked to other foods.

“We’re starting to see more and more nonmeat-associated outbreaks — leafy greens, unpasteurized apple juice and all kinds of things,” Jones said.

According to both Jones and Smith, there is a long-term trend of Americans eating more produce, which can lead to more infections. In fact, Smith said there has been an increase in foodborne outbreaks linked to produce.

“But we certainly can’t argue with the benefits in terms of chronic disease like cardiovascular disease that are gained from eating produce,” he said. “Ideally, we would like to eat a lot more produce and not have outbreaks.”

On the other end of the dietary spectrum, Smith said there has been an increase in outbreaks linked to processed foods like microwavable meals, snack foods and meal replacement powders eaten by consumers looking for fast, convenient ways to satiate their appetites.

Leafy greens, peanut butter, alfalfa sprouts and flour have been linked to outbreaks, but whether they are “new” causes of foodborne infection or are just being recognized because of better testing is unclear.

“The list of outbreaks that CDC is involved in investigating where the cause is a new vehicle that hasn’t caused an outbreak in the past is getting longer and longer,” Jones said.

According to a CDC report, foodborne outbreaks associated with organic food have grown more common in recent years, coinciding with an increase in organic food production and consumption. But the CDC said assessing the risk for infection from organic food is not possible because foodborne outbreak surveillance does not systematically collect data on the method of food production.

“We rarely have good information in outbreaks about whether the food involved was organic or not. It’s just not determined, people don’t remember or it’s hard to figure out,” Tauxe said.

However, experts seem to agree that it is unlikely that organic food is more or less safe for consumers from a microbiological standpoint.

“I get asked that a lot. I would love to see data and I’ve seen none,” Jones said.

The standards that apply to the way organic food is produced have other health benefits. For example, organic foods are less likely to contain antibiotic or pesticide residue.

“My subjective impression is that we’re certainly not seeing an increase or any substantial number of outbreaks associated with organic foods,” Jones said. “If we’re going to start to see a difference, it’s probably going to have more to do with the food preparation than actual food contamination. And whether or not vegetarians or people who like organic food cook more safely, I have no idea.”

‘A farm to fork issue’

There is a scenario in which Jones said he would consider eating raw chicken.

“Maybe if we started to irradiate food, then sure, I might think about it,” he said. “But now, the proportion of chicken that you can culture pathogens from is huge.”

Irradiated food is hit with ionizing radiation to eliminate pathogens, control insects, preserve or sterilize food or delay sprouting or ripening of produce. The FDA, which regulates the sources of radiation that are used to irradiate food, said it has been evaluating the process for several decades and has found it to be safe. NASA astronauts eat irradiated food in space to avoid getting sick.

The FDA has approved a variety of foods for irradiation in the U.S., including many commonly consumed meats and produce, but irradiated food remains rare and there are perhaps more practical ways to reduce foodborne illness, including government regulation and personal behavior.

The Food Safety Modernization Act (FSMA), signed into law by President Barack Obama in 2011, was “the most sweeping legislation we’ve had in food safety in this country in 100 years,” Smith said. It gave the FDA the power to order mandatory recalls of products and makes food production in the U.S. and the importation of food from other countries subject to science-based minimum standards of safety. Under the act, the FDA has rolled out numerous rules over the past several years, intending to be more proactive in preventing foodborne outbreaks. In July, the agency gave 43 states a record $30.9 million for risk-based programs. Overall, however, implementation of the legislation has been slow, and a government watchdog raised concerns recently that the FDA has not done enough under guidelines set by the FSMA to protect the food supply. According to a review by the HHS’ Office of Inspector General, the FDA failed to conduct a follow-up inspection within 1 year in half of cases in which a food facility was found to have a significant violation. It did no follow-up inspection at all in 17% of these cases, according to the report.

“I don’t think it’s ever really been funded to the extent that it needs to be to do all the things that it wants to do,” Scallan said. “Some of the things that it is asking the FDA to do, like inspection and enforcement, is hard without being funded considering how many producers and processors there are in the country. A lot of the ideas about being more proactive and risk-orientated were spot-on, but I don’t think they have the resources to do that.”

There are interventions that can be implemented closer to the table. Norovirus, which the CDC estimates causes nearly 5.5 million infections each year, is highly contagious and can be transmitted from the feces of food workers to ready-to-eat food items if the employee has not washed his or her hands properly. There are billions of virus particles per gram of stool.

“There’s a reason why it’s the most common foodborne pathogen in this country,” Smith said. “Turnover in the food worker industry makes it hard to keep up on training. So, we know what the problems are with norovirus, we just have not been able to get interventions implemented enough to have a big impact on that.”

Bacterial illnesses are a different problem. Whereas humans are norovirus’ only host, bacterial infections are both a production and a household issue.

“Even if 40% of chicken has Campylobacter on it, if the consumer cooks it thoroughly and doesn’t cross-contaminate, there won’t be any disease,” Smith said. “It’s truly a farm to fork issue, and I really think we need better efforts both in production and processing, and in consumer handling.”

Jones said physicians could also make a difference by educating patients on the importance of safe food preparation.

“Now, is that a physician’s priority and do they have time to do it? No,” he said. “That’s probably not a fair thing to ask of them, but ultimately it could change the trend in the U.S.” – by Gerard Gallagher

• References:
• CDC. Burden of foodborne illness: findings. 2016. https://www.cdc.gov/foodborneburden/2011-foodborne-estimates.html.
• CDC. Whole genome sequencing (WGS). 2016. https://www.cdc.gov/pulsenet/pathogens/wgs.html.
• Fitzgerald C, et al. J Clin Microbiol. 2016;doi:10.1128/JCM.01925-15.
• FDA. Food irradiation: what you need to know. 2016. https://www.fda.gov/food/resourcesforyou/consumers/ucm261680.htm.
• Harvey RR, et al. J Food Prot. 2016;doi:10.4315/0362-028X.JFP-16-204.
• Marder EP, et al. MMWR Morb Mortal Wkly Rep. 2017;doi:10.15585/mmwr.mm6615a1.
• Newsweek. Chicken sashimi is the other pink meat. 2013. http://www.newsweek.com/2013/07/17/chicken-sashimi-other-pink-meat-237734.html.
• Office of Inspector General. Challenges Remain in FDA’s Inspections of Domestic Food Facilities. https://oig.hhs.gov/oei/reports/oei-02-14-00420.asp. Accessed September 28, 2017.

• For more information:
• Timothy F. Jones, MD, can be reached at Tim.F.Jones@tn.gov.
• Elaine Scallan, PhD, can be reached at Elaine.Scallan@ucdenver.edu.
• Kirk Smith, DVM, MS, PhD, can be reached at Kirk.Smith@state.mn.us.
• Robert V. Tauxe, MD, MPH, can be reached at media@cdc.gov.

Disclosures: Jones, Scallan, Smith and Tauxe report no relevant financial disclosures.