No Change in Risk for Antibiotic-Resistant Salmonellosis From Beef, United States, 2002–2010

Solenne Costard; Jane G. Pouzou; Keith E. Belk; Paul S. Morley; John W. Schmidt; Tommy L. Wheeler; Terrance M. Arthur; Francisco J. Zagmutt


Emerging Infectious Diseases. 2020;26(9):2108-2117. 

In This Article

Abstract and Introduction


Restricting antibiotic use in food production animals is a target for reducing antimicrobial drug–resistant infections in humans. We used US surveillance data to estimate the probability of antibiotic-resistant nontyphoidal salmonellosis per meal made with beef during 2002–2010. Applying data for nontyphoidal Salmonella in raised-without-antibiotics cattle, we tested the effect of removing antibiotic use from all beef cattle production. We found an average of 1.2 (95% credible interval 0.6–4.2) antibiotic-resistant nontyphoidal salmonellosis cases per 1 million beef meals made with beef initially contaminated with antibiotic-resistant nontyphoidal Salmonella at slaughter or retail and 0.031 (95% credible interval 0.00018–0.14) cases per 1 million meals irrespective of beef contamination status. Neither outcome showed sustained change except for increases in 2003 and 2009 (>98% confidence) when larger or more outbreaks occurred. Switching all beef production to a raised-without-antibiotics system may not have a significant effect on antibiotic-resistant nontyphoidal salmonellosis (94.3% confidence).


Increased antimicrobial resistance (AMR), or antibiotic resistance, has resulted in initiatives to reduce the use of antibiotics in food production animals,[1,2] but quantification of the public health effects of decreasing antibiotic use in livestock remains limited.[3,4] Reduction of antibiotic use in livestock can lower resistance prevalence (i.e., proportion of pathogens with resistance) in animals,[4] but some studies show that pathogen prevalence may be higher in livestock raised without antibiotics.[5] Because transmission of foodborne pathogens is proportional to the prevalence of pathogens in the food source,[6] quantifying the change in human antibiotic-resistant foodborne illnesses resulting from reduced antibiotic use in livestock is vital.

In the United States, the most common bacterial cause of foodborne illness is nontyphoidal Salmonella (NTS), which leads to >1 million foodborne illnesses and 20,000 hospitalizations per year.[7] Antibiotic-resistant NTS is among the top 18 AMR threats in the United States,[8] causing 100,000 infections annually. The Centers for Disease Control and Prevention National Antimicrobial Resistance Monitoring System (NARMS) tracks resistance to 25 antibiotics in patient samples positive for isolates such as NTS,[9] including the clinically relevant antibiotics ciprofloxacin and ceftriaxone.

Multiple assessments of human AMR risk from meats have been performed.[10–14] However, most focused on only 1 class of antibiotic,[10,11] had limited or no longitudinal data,[14] or were not based on nationwide surveillance at the animal source.[11] Quantitative assessments of AMR risk with a more comprehensive resistance definition,[15] such as resistance to any class, or to ≥3 classes, that use representative, longitudinal data, are critical to defining the risks and benefits from policy with regard to antibiotic use in livestock.[3] Surveillance studies of antibiotic use and AMR in humans and livestock can be used to generate estimates of risk based on empirical data and can show the results of long-term conditions or systematic changes over time.

Our objective with this study was to use beef as a model to quantify trends in the longitudinal relationship human NTS infections and antibiotic-resistant NTS in meats. We also used the estimates to predict change in antibiotic-resistant salmonellosis resulting from hypothetical scenarios of antibiotic restriction in beef cattle.