White Paper on Human Illness Caused by Salmonella from All Food and Non-Food Vectors

This white paper examined current methods for surveillance, and the collection of epidemiological information related to foodborne outbreaks caused by Salmonella spp. were analyzed to assess their effectiveness and limitations. Information on illness caused by salmonellae was critically evaluated with research projects and modifications of epidemiological data collection needed to close gaps in our understanding of these issues were suggested.

Funded in part by The Beef Checkoff
Co-funded by the Pork Checkoff

Download final report

Objectives

The goals for this research were to:

1. Summarize all historical and epidemiological data on the relationships between food and non-food vehicles and vectors, and human illnesses caused by Salmonella
2. Develop an historical timeline outlining major events related to Salmonella emergence, epidemiology, surveillance, regulation, and industry initiatives to control salmonellae
3. Describe surveillance strategies, regulations, and industry initiatives to control this pathogen
4. Analyze effectiveness and limitations of current epidemiological data collection strategies for food attribution data
5. Identify gaps in understanding illness caused by Salmonella and propose research and possible procedural changes for epidemiological studies in order to close these gaps

Conclusions

Over the past ten years since the review by Mead on foodborne illness and death in the U.S., coverage of the active surveillance system for foodborne infections has increased from approximately 5% to 15% of the U.S. population (315) Incidence of salmonellosis cases/100,000 persons was estimated at 16.0 (13.7) in 1996 (1997), the baseline period, and, according to the most recent figures, was14.9 (14.8) in 2007 (2006). (http://www.cdc.gov/foodnet/reports.htm). Hospitalization rates for salmonellosis were 21-22% in 1996-97 and were somewhat higher, 26%, in 2004 (the last year with complete published data). However, the case-fatality rate was similar in 1997 (0.589%) and in 2004 (0.584%). Overall, the picture of salmonellosis in the U.S. has not changed greatly in the past decade.

However, there have been some changes in important serotypes and vehicles of infection. S. Typhimurium has declined as a percentage of serotyped Salmonella isolates from FoodNet sites from 29-32% in 1997-1998 to 16-19% in 2006-2007. S. Enteritidis isolates accounted for about 16-17% of serotypes identified during this decade although there was some yearly variation. S. Newport was an infrequent isolate in 1997-1998 (about 3.5%) but since 1999 has averaged about 11% of identified Salmonella isolates.(http://www.cdc.gov/foodnet/reports.htm)

Extrapolating from data provided by FoodNet to the U.S. population can be problematic because FoodNet sites are not representative of the whole population. Data from peer-reviewed articles in scientific journals are also not truly representative because only a small fraction of reported outbreaks, usually those which affect many people or are unusual or interesting in some way, are subsequently described in scientific literature.This may skew perceptions on the importance of certain serotypes or the frequency of different vehicles of infection.

Nevertheless there are some trends in the data and some contrasts between the U.S. and other countries that are of interest and suggest the need for further research and interventions.

Serotypes

• S. Enteritidis is by far the most common serotype in Europe and Japan while S. Typhimurium is the most common in Australia. In the U.S., S. Enteritidis and S. Typhimurium appear to have a similar prevalence. The reason that S. Enteritidis is uncommon in Australia is  apparently due to the fact that it has not become established in chickens in that country. Are some serotypes better adapted for surviving in  our food and agricultural production and processing systems?
• Case fatality rates for different serotypes vary 100-fold wth S. Choleraesuis and S. Dublin being significantly more virulent than many other serotypes. Investigation of factors that increase pathogenicity of certain serotypes may yield information useful in preventing or treating infections.
• Salmonellosis outbreaks traced to mung beans were first recorded in 2000. All mung bean outbreaks in the U.S. were caused by rare phage types of S. Enteritidis. In some of the incidents, mung bean seeds were known to have been imported from China. Did these rare phage types originate in China or were they previously unrecognized phage types already present in North America?

Antibiotic Resistance.

• This remains a concern as data from NARMS indicates that prevalence of antibiotic sensitivity in Salmonella isolated from turkeys, swine, and cattle has decreased between 1999 and 2006 while that of chicken isolates has remained about the same. More than 65% of isolates from turkeys and swine are resistant to at least one antibiotic while 30-40% of isolates from chickens and cattle exhibit some level of resistance. Resistance to multiple antibiotics has increased over time.
• S. Typhimurium DT104, resistant to 5 or more antibiotics, appears to have decreased in prevalence in recent years in the U.S. But there are increasing reports of significant antimicrobial resistance in S. Newport and S. Paratyphi.  Additionally, antibiotic- resistant strains of salmonellae are being disseminated internationally along with global trade in food.
• There is some evidence that antibiotic-resistant strains are more virulent. This should be examined more closely.

Vehicles of Infection

• Outbreaks and cases of many foodborne pathogens have increased as people consume more fresh fruits and vegetables and production and processing of these commodities has become more concentrated.  Some serotypes are more often associated with these outbreaks than others. How are they adapted for attaching to and surviving on plant surfaces?
• Importation of foods from underdeveloped countries carries the risk that sanitary practices may not be adequate to produce safe and healthy foods.
• Some European countries have undertaken well organized, nationwide programs to control Salmonella in poultry and swine. Some of their practices may be useful in improving U.S. production systems.

Deliverable

This paper effectively summarized the outbreak data and illness cases attributed to Salmonella from both food and non-food sources.  It also identified knowledge gaps which can be used to direct future research.