Antimicrobial Use and Risk Analysis

Effective food safety systems integrate science and risk analysis—risk assessment, risk management, and risk communication—at all levels. In risk assessment, scientific data are used to identify, characterize, and measure hazards; assess exposure; and characterize risks.

Resistance gene transfer is a complex scientific phenomenon that takes place within a larger macrobiologic ecosystem and social system. It is critical that this bigger picture be recognized when potential policy changes are considered. A thorough risk assessment can provide a framework for the needed “big picture” view of a problem, its sources, and the consequences of proposed policy changes. Observance of similar resistance genes in food animals and humans does not explain the causal pathway or flow of genetic information; a solid epidemiologic method is essential to gain an understanding of the system.

When making a decision about a “risky” new technology, the negative impact is only part of the equation. Thus, it is also important that the benefits of risk management options be evaluated. The current U.S. regulatory framework is geared toward protecting the public from risk without consideration of benefits. Regulators must reject any practice that appears to produce any apparent risk unless a demonstrated higher risk would occur with rejection of the practice in question.

Some evidence is accumulating, especially in the poultry industry, that there are significant human health benefits from antibiotic use in food animals to prevent or control disease. Subclinical disease levels of birds at slaughter significantly impact carcass contamination with pathogens such as Salmonella and Campylobacter. Antibiotic use in food animals reduces levels of subclinical disease, and has been shown to have a human health benefit that more than compensates for the risk of antibiotic use. More specifically to this point, investigators studying the pros and cons of using virginiamycin to reduce bacterial illness in poultry showed, using conservative estimates, that at least 40,000 illnessdays/ year in humans are prevented by the continued use of the antibiotic in the flocks. Similar results have been reported for other antibiotics used in poultry.

To inform the public debate on use of antimicrobials in food animals, investigators developed a model for assessing annual farm-to-patient risk of foodborne transmission of antibiotic resistance genes to the U.S. population that might occur through uses of tylosin and tilmicosin. Tylosin is used in poultry, swine, and cattle to prevent or control disease and enhance growth performance. Tilmicosin is used to treat and control respiratory disease in cattle. The investigators estimated that the risk of illness treatment failure resulting from tylosin- or tilmicosin-resistant Campylobacter or Enterococcus faecium infection associated with poultry, pork, or beef consumption is very low (< 1 in 10 million). Their analysis suggests that policies regarding antibiotic use in food animals should be developed on a case-by-case basis. Additionally, the potential benefits of antibiotic use, such as more uniform food animal quality, better evisceration, and reduced levels of pathogen (Salmonella spp., Campylobacter spp.) carcass contamination should also be considered.

The ban in Denmark of antibiotics (avoparcin, bacitracin, spiramycin, tylosin, and virginiamycin) for growth promotion resulted in increased intestinal disease in animals and increased therapeutic use of antibiotics. While this action led to decreased prevalence of resistance to some antibiotics among isolates from some animals, resistance prevalence increased for other antibiotics, bacterial strains, and animals. In addition, resistance of isolates from ill humans increased from 18% to 46% during a five year period after the ban. The Danish experience is instructive for showing that thorough risk assessments should be used to guide selection of risk management actions so that unintended consequences are avoided or minimized.

The complexity of the antibiotic resistance issue precludes simple solutions. Resistance proclivity varies with the antimicrobial, bacterium, and usage pattern. Therefore, sweeping risk management measures that are proposed for a certain classification of use (non-therapeutic, growth promotion, and routine disease prevention, for example) can be draconian and without predictable results. The most effective way to address the complexity and totality of the farm-to-food-tofailure chain is to base decisions on results of risk assessment. Conducting science-based risk assessments for specific product uses and tracking those bacteria that may become resistant as a result of those uses would provide insight into what mitigation interventions would be most effective. Further, given the data requirements and uniqueness of interactions between bacteria and antibiotics, analyses should be done on specific bacterium– antimicrobial pairs.