Decades of CDC surveillance data analyzed in this study illustrate that Salmonella Dublin more often causes bloodstream infections, hospitalizations (with longer hospital stays), and deaths than other Salmonella serotypes. Our findings support previous descriptions of Salmonella Dublin as a cattle-adapted serotype (10).

In the past decade, more than half of Salmonella Dublin infections have been resistant to >7 antimicrobial drug classes, and clinical outcomes have been more severe. The proportion of Salmonella Dublin isolates that were resistant was ≈2.7 times greater during 2005–2013 than during 1996–2004; isolates from the later period were also resistant to more antimicrobial drug classes. Multidrug resistance probably has direct clinical implications because bloodstream infections that require antimicrobial therapy tend to develop in patients with Salmonella Dublin infections (3). Most Salmonella Dublin isolates were resistant to third-generation cephalosporins (including ceftriaxone), which are often the treatment of choice for children with bloodstream infections because of the contraindication for fluoroquinolone use in children.

Clinical severity of Salmonella Dublin infections, as measured by the proportion of hospitalizations and deaths, also increased between 1996‒2004 and 2005‒2013. Our study did not directly measure the association between antimicrobial resistance and clinical severity of Salmonella Dublin infection by linking isolate data to outcome data. Nevertheless, by comparing both measures over the 2 periods, we showed that, for Salmonella Dublin infections, antimicrobial drug resistance and clinical severity increased in parallel. In addition to the older age of patients and concurrent conditions often associated with Salmonella Dublin infections (7,11), we hypothesize that the multidrug resistance profile has led to the higher rates of treatment failure, prolonged hospitalizations, and higher mortality rates observed in our study.

Virulence factors may also contribute, particularly those factors located on resistance plasmids that are co-selected for when antimicrobial drugs are used in cattle. Salmonella Dublin has been described as having a serotype-specific virulence-associated plasmid that is associated with invasive infection and remains stable through multiple generations of nonselective bacterial passage (12). Additional analyses, with use of whole-genome sequencing, particularly methods like those developed by Pacific Biosciences (Menlo Park, CA, USA) to use long-sequence reads and facilitate plasmid analysis, would enable investigation into the respective contributions of virulence factors and resistance mechanisms.

The recently observed increase in human infections with Salmonella Dublin resistant to ceftriaxone and nalidixic acid probably resulted, in part, from the agricultural use of comparable antimicrobial drugs in animals. Over the past 15 years, ceftriaxone resistance among Salmonella Dublin isolates from FSIS-PR/HACCP (Pathogen Reduction/Hazard Analysis and Critical Control Point) samples from cattle increased from 0 to 86% (13). Davis et al. determined that among Salmonella Dublin isolates from cattle, resistance to the third-generation cephalosporin ceftazidime increased over a 5-year period; they suggested that antimicrobial resistance in Salmonella Dublin is probably driven by antimicrobial drug use in cattle without influence of antimicrobial drug use in humans (14). Berge et al. also observed increasing resistance to third-generation cephalosporins and fluoroquinolones in calves in California during 1998–2002 (15). These findings demonstrate that antimicrobial stewardship and judicious use programs are essential for maintaining the efficacy of drugs used in human and veterinary medicine.

Our data indicate that the incidence of Salmonella Dublin infections has increased while the incidence of other Salmonella infections has remained mostly stable (Figure 1). The peak in overall Salmonella infections that occurred in the mid-1980s was driven by a nationwide outbreak of Salmonella Typhimurium (16). The simultaneous spike in Salmonella Dublin resulted largely from consumption of raw milk (7,17), particularly from a large California dairy (18). The dairy producer promoted its raw milk as having health benefits (19), and many persons with compromised immune systems (e.g., young, elderly, or HIV-positive) became ill (20). As a result of a public health investigation (17), production and sales were halted, and FDA banned the interstate sale of raw milk in 1987 (21). A sharp decline in Salmonella Dublin infections soon followed.

Although additional data on food histories and the role of the environment will help elucidate the sources of human infections, the risk for Salmonella Dublin infection among humans is probably still caused, in part, by consumption of raw milk and beef. Raw dairy products have been linked to numerous Salmonella Dublin illnesses in the United States (17,22) and abroad (23,24). Incidence of Salmonella Dublin infections was 3 times higher in states that allow the sale of raw milk or permit cow shares than in states where raw milk sales are illegal. US surveillance data from PulseNet, the USDA Agricultural Marketing Service (25), and FDOSS also indicate that Salmonella Dublin has been isolated from ground beef and boneless beef products and has been associated with outbreak-associated illnesses from beef products. Salmonella Dublin has also been found in beef cattle and calves (26,27).

In our study, the higher proportion of Salmonella Dublin infections among men than women may be partially attributable to consumption patterns. Although the 2006–2007 FoodNet Population Survey found no differences by sex for consumption of raw milk or cheese items (28), numerous studies have found that men consume more beef and more undercooked beef than women (28,29). Occupational exposure to cattle may also contribute to the increased frequency of infection among men.

Most Salmonella Dublin infections continue to be reported from California (Figure 2), but illnesses have occurred nationwide. They are probably associated with an ongoing outbreak of Salmonella Dublin infections among US dairy and beef cattle. A 2014 dairy study conducted by the USDA National Animal Health Monitoring System found antibodies directed against Salmonella Dublin lipopolysaccharide O-antigens in 8% of bulk tank milk samples (30). Of operations in participating western states (California, Colorado, Idaho, Texas, and Washington), 52% were positive, compared with 2.8% of operations in eastern states (Kentucky, Michigan, Minnesota, Missouri, New York, Ohio, Pennsylvania, Vermont, Virginia, Wisconsin). In 2013, the Animal Health Diagnostic Center at Cornell University (Ithaca, NY, USA) issued an animal health advisory, warning cattle owners about an increase in MDR Salmonella Dublin infections among cattle in the northeastern United States, treatment difficulties associated with these infections, the potential for long-term environmental contamination, and the dangers (including death) that these infections pose to animals and humans (31).

Changes in the geographic distribution of Salmonella Dublin infections in cattle probably explain the similar geographic spread among humans. Historically, Salmonella Dublin in cattle was associated with the western United States and was not discovered in cattle east of the Rocky Mountains until 1968 (32). Salmonella Dublin continued to spread by transport of animals and their products and can now be found in cattle populations throughout the contiguous United States (26).

In Denmark, in response to the specific threat to human and animal health posed by Salmonella Dublin infections, in 2006, the Danish government passed legislation intended to eradicate this serotype. Their policy actions included heightened surveillance for cattle and abattoirs, voluntary interventions to reduce environmental contamination and disease spread within infected herds, economic sanctions for producers who do not control Salmonella Dublin in their herds, and closing of infected herds to live-animal trade (33,34). In the United States, precedent for the successful eradication of other host-adapted Salmonella serotypes in production animals has been set by use of vaccines and improved management practices. An example is the USDA National Poultry Improvement Plan, which has successfully eradicated Salmonella Gallinarum and Pullorum from domestic commercial poultry (35,36). Efforts are under way to decrease the burden of Salmonella Dublin among cattle. An oral modified-live Salmonella Dublin vaccine has been evaluated for use in calves; however, this vaccine has not been effective for reducing the incidence of disease, and research into finding an effective vaccine continues (37).

Interventions developed for the Denmark cattle and US poultry industries may not be completely applicable to the US cattle industry because of regulatory and production differences. For example, in Denmark, to control Salmonella Dublin infections, trade restrictions are applied to farms with affected herds, and in the United States, biosecurity procedures for poultry producers generally enable tighter environmental control than do those for cattle producers. However, judicious use of antimicrobial drugs in cattle, coupled with improved specific husbandry and management practices on the farm, could decrease antimicrobial-resistant Salmonella Dublin infection in cattle. In 2012, FDA prohibited certain extralabel uses of cephalosporins in chickens, turkeys, cattle, and swine (38). This new prohibition has the potential to slow the spread of cephalosporin resistance among food animals and is a valuable step toward protecting the effectiveness of current antimicrobial drugs. Nevertheless, other extralabel uses of cephalosporin drugs are still permitted.

Salmonella Dublin is a cattle-adapted Salmonella serotype that causes severe and antimicrobial drug–resistant infections in humans and cattle, and its incidence is on the rise. Reducing Salmonella Dublin carriage by cattle could benefit animal and human health. Unlike most other Salmonella infections in food animals, Salmonella Dublin can cause high mortality rates, particularly among calves, and heavy economic burdens for producers (39). It is well established that use of antimicrobial agents is a major driving force for the global surge in antimicrobial resistance. Food animal management practices, including veterinary use of antimicrobial drugs, may be contributing to the increasing resistance in Salmonella Dublin and to Salmonella Dublin–associated illness and death among humans (15). Therefore, careful evaluation of management practices and judicious use of antimicrobial drugs in cattle is critical for the control of antimicrobial drug–resistant Salmonella Dublin infections in cattle and humans. The 2016 FDA Veterinary Feed Directive aims to eliminate the use for food production purposes (i.e., growth promotion and feed efficiency) of antimicrobial drugs that are considered medically important in humans and seeks to bring all remaining therapeutic use of antimicrobial agents in feed and water under the oversight of licensed veterinarians (40). Agricultural and public health authorities will need to engage in ongoing, meaningful collaborations to reduce inappropriate antimicrobial use in food-producing animals to protect human and animal health.