Free access
19 July 2019

Differentiating Epidemic from Endemic or Sporadic Infectious Disease Occurrence

ABSTRACT

One important scope of work of epidemiology is the investigation of infectious diseases that cluster in time and place. Clusters of infectious disease may represent outbreaks or epidemics in which the cases share in common a point source exposure or an infectious agent in a chain of transmission pathways. Investigations of outbreaks of an illness can facilitate identification of a source, risk, or cause of the illness. However, most infectious disease episodes occur not as part of any apparent outbreaks but as sporadic infections. Multiple sporadic infections that occur steadily in time and place are referred to as endemic disease. How does one investigate sources and risk factors for sporadic or endemic infections? As part of the Microbiology Spectrum Curated Collection: Advances in Molecular Epidemiology of Infectious Diseases, this review discusses limitations of traditional approaches and advantages of molecular epidemiology approaches to investigate sporadic and endemic infections. Using specific examples, the discussions show that most sporadic infections are actually part of unrecognized outbreaks and that what appears to be endemic disease occurrence is actually comprised of multiple small outbreaks. These molecular epidemiologic investigations have unmasked modes of transmission of infectious agents not known to cause outbreaks. They have also raised questions about the traditional ways to measure incidence and assess sources of drug-resistant infections in community settings. The discoveries made by the application of molecular microbiology methods in epidemiologic investigations have led to creation of new public health intervention strategies that have not been previously considered.

INTRODUCTION

A cluster of an infectious disease that occurs in numbers in excess of what is expected for a particular time and place defines an outbreak. Multiple outbreaks of a disease in different populations or places would constitute an epidemic. An epidemic that spreads to multiple continents is a pandemic. However, if an infectious disease occurs in numbers with no major fluctuation with respect to time and place, such a pattern of disease occurrence is described as endemic. Malaria in Kenya is considered an endemic disease, but two cases of autochthonous malaria in Long Island in the United States would be called an outbreak (1). Conversely, Lyme disease in Long Island is endemic, but if it occurs in Kisumu, Kenya, it would be an epidemic. If in a 1-year period the number of malaria cases in Kenya or Lyme disease in Long Island doubles from the previous year, they would be considered epidemics according to the definition above.
The differentiation of infectious diseases into epidemic versus endemic occurrence is not always straightforward. In the temperate zone, a seasonal increase in the number of influenza cases during winter is called an epidemic, whereas the increased number of foodborne salmonellosis cases during summer is considered a seasonal fluctuation of endemic salmonellosis. Cases of salmonellosis are considered part of an epidemic or an outbreak only if they are investigated and found to have a common contaminated food exposure. However, less than 20% of the total number of salmonellosis cases annually reported to the Centers for Disease Control and Prevention (CDC) come from recognized outbreaks (2). If a case of salmonellosis is not recognized to be part of an outbreak, it is counted as a sporadic case. Many sporadic cases of salmonellosis collectively comprise endemic salmonellosis. But how many of these sporadic cases are actually part of an outbreak or epidemic? Here, the absence of well-characterized epidemiologic information is what determines salmonellosis occurrence to be endemic.
A similar diffuse description of disease occurrence is used to characterize health care-associated infections. When carbapenem-resistant (CR) Klebsiella pneumoniae infections first appeared in hospitals in New York in the early 2000s (37), they were referred to as hospital outbreaks. However, since then, these infections have become established long-term in many of the hospitals in the East Coast region of the United States, and they came to be described as endemic CR K. pneumoniae infections (8, 9). Thus, it appears that an infectious disease can occur as an epidemic in one setting or population and as sporadic or endemic in another setting or population.
There is even a mathematical model to differentiate endemic versus epidemic disease occurrence. For an infectious disease that is predominantly transmitted person to person, endemic disease occurrence is expressed as a basic reproduction number (R0) that equals 1. That is, one infected person, on the average, transmits to only one other person in a population, such that the number of cases is maintained in a steady state in that population. If the population includes a subgroup that is immune, then only the susceptible subgroup (S) can get infected and an endemic state is established when R0 × S = 1. When R0 or R0 × S becomes >1, then the disease is said to be in an epidemic state.
One negative public health consequence of describing an infectious disease occurrence as endemic is that it often gets perceived as a nonurgent problem and hence not given a high priority for intervention. The public health response to diseases described as endemic is frequently delayed or not even implemented. Acute epidemics trigger rapid responses, but if a disease is endemic, a typical response is to take time to study it to understand why it is endemic and devise control measures that are not always targeted to any specific risk factor or source identification. Furthermore, when a disease is endemic, there is an assumption or reluctant acceptance that the disease is difficult to control and that the solutions are not straightforward.
As will be demonstrated here, a molecular epidemiology approach to address this issue offers a more precise characterization of disease occurrences and will show that the differentiation of a disease occurrence as endemic or epidemic is really not meaningful. A disease occurrence is often called endemic when there is not enough information about the epidemiologic behavior of a pathogen causing the disease in question. This review will show that most so-called endemic infectious disease occurrences are actually epidemics and that rapid public health interventions against such occurrences can be made in the same way as they are done for recognized acute epidemics. Molecular epidemiology precludes excuses to describe disease occurrence as endemic.

COMMUNITY-ACQUIRED INFECTIONS

Two representative groups of community-acquired infectious diseases are discussed here to illustrate how molecular epidemiologic approaches have been used to refine our understanding of infectious disease occurrence and transmission dynamics. Foodborne illnesses are one of the most common infections acquired in the community that occur as outbreaks but more frequently as sporadic infections. On the other hand, community-acquired urinary tract infections (UTI) are the most common bacterial infectious disease of women, but outbreak occurrences of UTI in community settings are not evident. They are thus invariably considered sporadic infections. Molecular epidemiology studies of these groups of infectious disease have revealed that our previous assumptions about the transmission dynamics of these infections were incomplete. New findings made from investigations of these diseases have provided opportunities to greatly expand public health control strategies that had not been previously possible or considered.

Foodborne Diseases

Outbreaks of foodborne disease caused by bacterial pathogens are very common, and no human being is spared from such infections. In the United States, 19,119 outbreaks involving 373,531 persons, caused by bacteria, protozoa, and viruses, were reported to the CDC between 1998 and 2015 (https://wwwn.cdc.gov/foodborneoutbreaks/). Of these, 2,437 outbreaks involving 65,724 illnesses were caused by non-Typhi Salmonella, one of the most commonly reported groups of bacterial agents of foodborne illness. Another CDC surveillance system (National Notifiable Diseases Surveillance System [NNDSS; https://wwwn.cdc.gov/nndss/]) reported 51,455 cases of nontyphoidal salmonellosis in the United States in 2014 alone (10). Thus, comparing the two surveillance systems, it can be observed that cases of salmonellosis from recognized outbreaks represent a small fraction of the total number of cases reported to the CDC. The other reported cases of salmonellosis are considered sporadic cases. In fact, what is reported to the CDC in total represents a small fraction of the actual number of foodborne illnesses caused by non-Typhi Salmonella each year, which is estimated to be more than a million in the United States alone (11).
The CDC also leads an active laboratory-based surveillance system called FoodNet that reports annual changes in the number of cases of illness caused by Campylobacter, Cryptosporidium, Cyclospora, Listeria, Salmonella, Shiga toxin-producing Escherichia coli (STEC) O157 and non-O157, Shigella, Vibrio, and Yersinia spp. and cases of hemolytic-uremic syndrome (https://www.cdc.gov/foodnet/index.html). This surveillance system includes 10 sites covering about 15% of the U.S. population. In 2017, 24,484 confirmed cases, 5,677 hospitalizations, and 122 deaths from these infections were reported (https://www.cdc.gov/foodnet/index.html). Of these, Campylobacter and Salmonella were the most frequently identified pathogens, accounting for 9,421 and 7,895 cases, respectively. What this surveillance system does is show a relative change over time and place of infections caused by these pathogens. The degree of fluctuation from the expected number of cases for time and place is assessed to determine if an outbreak has occurred, and then a decision is made whether or not to initiate an investigation.
As described above, outbreaks of foodborne illnesses are confirmed as outbreaks only after the cases are investigated. Of course, most investigated cases are not linked to any recognized outbreaks, and an even greater number of cases are not even investigated. Until genotyping methods were introduced to investigate foodborne disease, cases that could not be linked to any outbreaks were counted as sporadic illnesses. Infections caused by common serotypes of Salmonella, such as Salmonella enterica serovars Typhimurium and Enteritidis, are considered endemic to most parts of the United States. Serotyping information alone in sporadic cases of salmonellosis is not very helpful in triggering an investigation of disease caused by such common serotypes unless the numbers increase substantially from what are expected for a time and place. Thus, opportunities to further investigate sporadic or endemic cases of foodborne diseases based on serotype information alone have been limited or missed.
Laboratory-based surveillance systems, such as the FoodNet, make important contributions to public health activities, because they are designed to isolate microorganisms from cases of an illness, which become available for genotyping. Salmonella, E. coli O157, other STEC organisms, Campylobacter, Listeria monocytogenes, Shigella, Vibrio cholerae, Vibrio parahaemolyticus, and Cronobacter are genotyped by pulsed-field gel electrophoresis (PFGE) by yet another CDC-led surveillance system, PulseNet (https://www.cdc.gov/pulsenet/) (see the second article of this curated collection [http://www.asmscience.org/content/journal/microbiolspec/10.1128/microbiolspec.AME-0004-2018]). In fact, it was not until the genotyping tests for foodborne pathogens became available that focused investigations of sporadic or endemic cases of these diseases became possible. These genotyping tests have expanded and refined our understanding of the pattern of occurrence and transmission of these illnesses in the last 35 years, as described below. Selected examples of applications of plasmid profile analysis, PFGE, multilocus sequence typing (MLST), and whole-genome sequencing (WGS) to investigate sporadic cases of foodborne diseases are discussed.

Salmonella enterica serotype Newport salmonellosis in New Jersey and Pennsylvania, 1981

One of the first applications of a genotyping test to investigate sporadic cases of salmonellosis involved plasmid profile analysis (see the second article of this collection [http://www.asmscience.org/content/journal/microbiolspec/10.1128/microbiolspec.AME-0004-2018]). It was used to investigate sporadic salmonellosis cases reported from New Jersey and Pennsylvania (11). In summer of 1981, an outbreak of salmonellosis was reported at a hospital in Philadelphia, PA. The outbreak involved patients as well as hospital staff. A case-control study implicated contaminated precooked roast beef as the vehicle. The precooked roast beef had been prepared at a meat processing plant in Philadelphia. Almost simultaneously, across the Delaware River in New Jersey, another outbreak of salmonellosis was reported at a wedding, which was also traced to contaminated precooked roast beef produced by the same meat processing plant.
Salmonella enterica serotype Newport and S. enterica serotype Typhimurium were isolated from patients as well as from a lot of the implicated precooked roast beef sample that was available for culture. In addition to the two outbreak sites, the same lot of meat was widely distributed to food establishments and retail stores in the two states. Therefore, a public health alert was announced in the two states to recall the implicated lot of precooked roast beef.
During the two outbreak periods, Pennsylvania and New Jersey departments of health reported an increase in the numbers of S. Newport salmonellosis cases that were not part of the two outbreaks (11). Serotype S. Newport was common and considered endemic to these states, and therefore, it was not certain that the increased numbers of these cases were caused by the precooked roast beef. However, this observation provided an unprecedented opportunity to apply a genotyping test to investigate how many of the cases of S. Newport infections that were not part of the two recognized outbreaks during this period were due to the contaminated precooked roast beef.
Plasmids were extracted from S. Newport isolates from cases in the two outbreaks, precooked roast beef, and sporadic cases from New Jersey and Pennsylvania that occurred in July to August (outbreak period) and from cases identified before and after the outbreak period. S. Newport isolates from six other states were also analyzed as geographic comparison isolates. The outbreak and meat isolates all shared two plasmids of 3.4 and 3.7 MDa, and their HaeIII endonuclease restriction fragment agarose gel electrophoresis (AGE) patterns were indistinguishable (11). Thirteen (45%) of 29 isolates from the summer of 1981 from the two states had the precooked roast beef-associated AGE pattern, while none of the isolates from the two states before the outbreak period had this pattern. Two (20%) of 10 isolates from after the outbreak period (November 1981 to January 1982) had the characteristic AGE pattern, but none of 50 isolates collected between June 1979 and February 1982 from six other states had this roast beef-associated plasmid profile.
Food intake history was obtained from 19 of 29 patients who developed S. Newport salmonellosis during July to August 1981. Seven of 10 of these patients infected with S. Newport with the characteristic roast beef-associated plasmid profile, but none of 9 infected with strains with different plasmid profiles, reported eating precooked roast beef in the 3 days before developing gastrointestinal symptoms (P = 0.003, Fisher’s exact test) (11). Food intake history was also obtained from 7 of 10 patients who developed the infection during November 1981 to January 1982; 2 were infected with S. Newport strains with the roast beef-associated plasmid profile, but none of the 7 reported eating precooked roast beef in the 3 days before their illness. One of the two patients infected with the roast beef-associated strain was a 5-month-old sibling of an older child who became ill during the summer of 1981 after eating precooked roast beef (11). The other was an 88-year-old man who had died and could not be contacted for interview about food intake.
This investigation demonstrated that the increased number of S. Newport salmonellosis cases in the two states during July to August of 1981 was largely caused by one single clonal lineage of S. Newport. It was responsible for 45% of non-outbreak-associated cases in the two states during summer of 1981. The strain was disseminated in the two states via a contaminated food product, and even after the recall of the contaminated product, cases continued to occur in the community, possibly spread by person-to-person transmission in households of the infected persons who ate the meat. Thus, in this investigation, a large proportion (45%) of “sporadic” cases of salmonellosis caused by a very common serotype was shown to be part of an unrecognized common-source outbreak. What appeared to be sporadic or endemic salmonellosis cases were indeed part of a large multistate outbreak.

Escherichia coli O157:H7 gastroenteritis in Minnesota, 1994 to 1995

In 1994 and 1995, the Minnesota Department of Health genotyped by PFGE all E. coli O157:H7 isolates submitted from the state (12). From 317 isolates subtyped, 143 distinct PFGE patterns were observed. Fifty-six (18%) of the cases were traced to 10 outbreaks, 4 of which were detected solely on the basis of the PFGE data. The five most common PFGE types were responsible for 109 (34%) of the 317 isolates, which suggested that many of the clonal lineages were not linked to recognized outbreaks. Persons infected with such strains may have been exposed to outbreak-associated vehicles outside of the recognized outbreaks (as observed with precooked roast beef exposure described above) or acquired the infection via person-to-person transmission from someone in the recognized outbreaks.
During this study time frame, 11 distinct two-week blocks were found in which the number of reported cases of E. coli O157:H7 doubled from the previous two weeks (12). In eight of these, the PFGE patterns were distinct, but in two, the increases were found to be due to strains with PFGE types suggestive of multiple simultaneous outbreaks (12). Epi-curve graphs of these cases by week of their onset and their respective isolates’ PFGE patterns revealed clustering by time of several PFGE types, indicating outbreaks.
Here, the increase in number of cases was attributed to both outbreaks as well as to infections caused by unique PFGE E. coli O157:H7 lineages. These unique lineages were considered to have caused sporadic infections, but according to the strict definition of an outbreak, cases caused by such strains may be considered to represent multiple outbreaks of one case each. As mentioned in a previous review in this curated collection (http://www.asmscience.org/content/journal/microbiolspec/10.1128/microbiolspec.AME-0001-2018), one case of cholera that occurred in Abbeville, LA, in 1978 was an outbreak because cholera had not occurred in that part of the state in over 100 years (13). Similarly, one case of gastroenteritis caused by a strain of E. coli O157:H7 with a PFGE type that had never been seen in Minnesota previously is an outbreak. It is certainly possible that such a strain could have been detected if PFGE had been done before 1994 or if the level of contamination was high in a source vehicle or if more people were exposed to such a vehicle during 1994 to 1995. Again, what is called sporadic here is a result of lack of information, and the term endemic to describe such disease occurrences is uninformative.

Salmonella Typhimurium salmonellosis in Minnesota, 1994 to 1998

In 1994, the Minnesota Department of Health began genotyping by PFGE all S. Typhimurium isolates submitted from clinical laboratories in the state (14). Of 998 isolates submitted between 1994 and 1998, 958 were genotyped by PFGE (14). Six foodborne disease outbreaks were recognized during this period involving 128 cases. Of these, PFGE was instrumental in initiating investigations in four. Seventy-nine of these cases were identified over a 15-week period in 1998. Standard food intake interviews of these cases identified two outbreaks, each with only two cases; one was associated with a wedding reception, and the other involved a child-care setting. PFGE analysis found 32 (41%) of the 79 isolates to be indistinguishable. This information allowed more focused interviews, which led to the identification of an outbreak associated with a commercial microwavable chicken product (15). This finding led to recall of the product and changes in product labeling policy. Here, unlike what happened with the precooked roast beef epidemic, the public health authorities were able to use the PFGE information related to sporadic infections to make a public health intervention.
As in many northern hemisphere regions, the number of salmonellosis cases increases during the summer in Minnesota. Over a 16-week period in the summer of 1995, 157 cases of S. Typhimurium infections were reported in the state (14). Of these, 99 (63%) were shown by PFGE to be part of three outbreaks associated with three different restaurants in the state. Two of these outbreaks were recognized by the genotype surveillance system. PFGE analysis of the isolates during this period enabled investigators to precisely quantitate the magnitude of S. Typhimurium salmonellosis cases caused by three point sources; just three lineages were responsible for two-thirds of all the reported cases during the summer of 1995! Thus, what appeared to be a seasonal increase in the number of salmonellosis cases was in fact part of several large outbreaks. Describing such an increase as a seasonal endemic fluctuation is, therefore, not meaningful. An increase in occurrence of salmonellosis in a place and time could be discovered to be an outbreak or multiple outbreaks if the causative microbes are further subtyped.

Salmonellosis in Norway, 1996 to 1999

From 1996 to 1999, the National Notification System for Infectious Diseases in Norway genotyped S. Typhimurium isolates by PFGE from 60 domestic and 42 imported cases of salmonellosis that were not known to be associated with any recognized outbreaks (16). Three subclusters comprised of identical PFGE pattern each (designated E5, F1, and G1) accounted for 47% of the 102 isolates. Most of the F1 and E5 isolates also clustered in time, while G1 isolates were more spread out (16). F1 isolates had a PFGE pattern identical to that of human S. Typhimurium isolates implicated in a nationwide outbreak in 1987, traced to contaminated chocolates in Norway (17). As shown in previous studies, a large proportion of salmonellosis cases that could not be associated with recognized outbreaks were labeled as sporadic cases, which, by a genotyping test, were shown to be comprised of several outbreaks. Here, PFGE analysis helped to quantitate the proportion of so-called sporadic cases of salmonellosis that were actually part of hidden outbreaks.

Camplylobacteriosis in New Zealand, 2005 to 2008

According to the Institute of Environmental Science and Research, campylobacter infections are one of the most common zoonotic diseases in New Zealand (https://surv.esr.cri.nz/episurv/index.php). Between March 2005 and February 2008, 502 isolates of Campylobacter jejuni were obtained from persons with gastroenteric illness in Manawatu Health District in North Island, which were genotyped by MLST (18). Fifty-one different MLST genotypes (STs) were identified, and seven of these were comprised of more than 20 cases each. These seven STs were responsible for 327 cases (65%); one of these (ST474) caused the largest number of cases: 154, or 31% of all the cases. These MLST data were used to identify risk factors for C. jejuni infections. Twenty-three STs were found to be poultry associated (including all ST474 strains), while 16 STs were associated with ruminant sources. These STs did not overlap across the two sources. Poultry sources were responsible for 350 of the cases, while ruminant sources accounted for 70. Risk factors for the other cases could not be identified (18). Children <10 years of age were more likely to be infected with ruminant-associated strains than were adults. Thus, another genotyping test—MLST—enabled investigators to identify risk factors for a large proportion of what appeared to be sporadic cases of campylobacteriosis.
Of course, it is not surprising that strains that cause outbreaks would be detected from persons who are not directly linked to any recognized outbreaks. The importance of the genotyping tests as illustrated above, however, is that they helped to (i) link sporadic cases of foodborne disease to outbreaks in which contaminated vehicles were identified, (ii) quantitate a risk factor or a vehicle of infection associated with cases that were not obviously related to recognized outbreaks, and (iii) make focused public health interventions directed at sporadic cases that would not have been possible to do otherwise.

Salmonella Typhimurium DT170 salmonellosis in Australia, 2006 to 2012

Five community outbreaks in New South Wales, Australia, that occurred between 2006 and 2012 caused by S. Typhimurium phage type DT170 were retrospectively investigated (19). Fifty-seven isolates obtained from these outbreaks were genotyped by WGS. The Salmonella isolates from these outbreaks were previously typed by multilocus variable-number tandem-repeat analysis (MLVA), and each outbreak was shown to be caused by strains with a distinct MLVA profile (19). In four of these outbreaks, the WGS analysis showed strains within each outbreak to differ by one or two single nucleotide polymorphisms (SNPs). The other outbreak occurred in a college campus in which chocolate mousse was identified as a possible contaminated source. In this outbreak, strains isolated from students and staff over a 2-day period differed by up to 12 SNPs, which suggested that the implicated food vehicle was contaminated with multiple strains of S. Typhimurium not differentiated by MLVA.
The investigators modeled reported mutation rates of S. Typhimurium to derive cutoff values for the number of SNP differences needed to determine which case should be included or excluded from an outbreak. The reported mutation rates were determined from well-characterized collections or outbreaks of S. Typhimurium from other regions of the world, which ranged from the lowest rate of 1.9 × 10−7 (20) to the intermediate rate of 3.4 × 10−7 (21) to the highest rate of 12 × 10−7 substitutions per site per year (22). The investigators assumed SNP differences to depend on the time the Salmonella organism replicates in the food (ex vivo time) and the time it replicates in the human host after infection (in vivo time) (19). They then calculated the expected maximum number of SNP differences that would occur among isolates per ex vivo/in vivo evolution time. For an ex vivo/in vivo evolution time of less than 1 month, the maximum number of SNP differences was determined to be two or four at the lowest or highest mutation rate, respectively, shown above. For ex vivo/in vivo time of up to 3 months, the SNP differences were determined to be three to nine at the lowest or highest mutation rate, respectively (19).
Using these SNP cutoff values, the investigators were able to link additional cases of S. Typhimurium DT170 infections that were not initially known to be part of the 5 outbreaks and exclude others that were not part of these outbreaks. MLVA did not provide the resolution of information needed to distinguish these isolates as being related or not to the outbreaks.
It should be cautioned, however, that even within a well-studied outbreak, the analysis of SNP differences among isolates is dependent on many factors, including the sample size of the isolates, the duration of the outbreak, the number of distinct genotypes infecting the same contaminated food vehicle, and the presence of strains that harbor genes that undergo high mutation frequency. Furthermore, SNP cutoff thresholds to rule in or rule out outbreak-associated strains determined for one bacterial species may not necessarily apply to other species. The utility of WGS for this type of analysis must be evaluated empirically with many more studies with other foodborne pathogens. We emphasize that none of the genotyping tests described in this curated collection can distinguish outbreak-associated versus unassociated strains with absolute certainty. As emphasized throughout this curated collection, just as statistical tests are used to unmask hidden associations, these genotyping tests should be regarded as yet another set of tools to assist epidemiologic investigations to make estimates and develop hypotheses that can then be further tested.
The above-described studies are a few examples of molecular epidemiology investigations that support a new epidemiologic paradigm that the notion of endemic bacterial foodborne disease is meaningless unless the etiologic agents are genotypically characterized. These studies also demonstrate that once outbreaks or clusters of cases are identified, the same approach used to investigate the outbreaks could be applied to trace sources or identify risk factors for sporadic infections. Thus, these molecular epidemiologic approaches provide opportunities to devise focused interventions targeting sporadic diseases that were not previously possible.

Extraintestinal Pathogenic Escherichia coli Infections

There are several community-onset or -acquired infectious diseases that do not obviously occur as an outbreak or epidemic—UTI (including cystitis and pyelonephritis), community-onset bloodstream infections (BSIs), bacterial pneumonia (caused by Pneumococcus, Staphylococcus aureus, or Haemophilus influenzae), and skin and soft tissue infection, to name a few common examples. Of these, UTI is the most common. Of course, UTI outbreaks in hospitals and institutional settings are frequently documented (2326), but community-acquired UTI (CA-UTI) outbreaks are rarely reported (until recently). CA-UTI are the most common bacterial infectious disease of women; the health care costs associated with this infection were estimated at $2.5 billion in 2000 in the United States (2729). CA-UTI can be complicated by pyelonephritis as well as BSIs or sepsis. While CA-UTI are caused by a variety of bacterial organisms, 75 to 85% are caused by Escherichia coli (30).
Since CA-UTI is not a reportable disease in most countries, its incidence is difficult to estimate. In one survey conducted in 2000 in the United States, about 11% of women aged 18 years and older self-reported at least one uncomplicated episode of UTI in the preceding year (29). Another recent study reported the incidence of uncomplicated recurrent UTI among women (defined as three episodes of culture-confirmed UTI per year) to be as high as 102/100,000 per year in the United States, with the highest incidence occurring in ages 18 to 34 and 55 to 64 years (31). Outbreaks of CA-UTI are not recognized, and hence, outbreaks cannot be used to indirectly estimate the incidence of this disease as was done with bacterial foodborne gastroenteritis. Thus, CA-UTI cases are practically all sporadic. Or are they?
UTI results from an infected person’s intestinal E. coli entering the bladder from the perineum. Recognized risk factors for CA-UTI are all host-related: female gender, urinary tract anatomical abnormalities, multiple sex partners, diabetes, and pregnancy. The external sources of E. coli organisms that cause CA-UTI (uropathogenic E. coli [UPEC]) are rarely sought or investigated. The E. coli needs to be first ingested orally, transit the esophagus, stomach, and small intestine, then colonize the colon and exit the rectum to colonize the perineum, and finally enter the bladder to establish an infection or cystitis. Seeding of the bladder with E. coli, especially in women, probably occurs regularly, but most such infections do not result in cystitis; a subset of these E. coli strains will succeed in causing cystitis. E. coli strains that successfully cause cystitis are referred to as extraintestinal pathogenic E. coli (ExPEC). The question is, where do the E. coli strains that colonize the intestine come from in the first place? If this question can be answered, new interventions to prevent CA-UTI could be devised. In the last 10 to 15 years, molecular epidemiology has addressed this important public health question as summarized below.

CA-UTI caused by E. coli serotype O15:H1:K52 in London, 1986 to 1987

The first suggestion that CA-UTI may occur as an epidemic was made in the United Kingdom during a study of clusters of severe community-onset infections that occurred in South London (32). Between October 1986 and October 1987, the Central Public Health Laboratory of Colindale investigated a cluster of community-acquired septicemia caused by multidrug-resistant (MDR) E. coli with an unusual drug susceptibility pattern (32). Of 34 blood isolates, 26 belonged to E. coli serotype O15:H1:K52. In the previous 17 years, only 16 of 754 cases of E. coli septicemia were caused by serogroup O15. During the same period (1986 to 1987), of 146 E. coli isolates from cases of cystitis, 135 (92%) were found to belong to serogroup O15. It was suggested that this cluster of cystitis represented a year-long outbreak of CA-UTI that was complicated by pyelonephritis, septicemia, and even meningitis (32).
Subsequently, the same serotype has appeared in clusters in Spain (33) as well as in other European countries, Africa, and the United States (3436). Strains belonging to this serotype were later shown by MLST to belong to ST393 (36).
Here, a subtyping test—serotyping—was used to show that a large proportion of cases of CA-UTI and its complications clustered in time and place. The suggestion that this was an outbreak was made after the subtyping test was performed, which is distinct from the way most outbreaks of foodborne diseases are detected. However, as the above examples of foodborne diseases also illustrate, most sporadic cases of foodborne disease, too, may be shown to be comprised of outbreaks after the subtyping tests are performed. This conclusion becomes possible only when what appear to be sporadic cases of a disease are further stratified. Below are examples of how further stratification of CA-UTI cases by genotyping UPEC led to yet another new epidemiologic paradigm that outbreaks, epidemics, and even pandemics of CA-UTI do occur.

CA-UTI in a university community caused by ST69 E. coli, 1999 to 2005

Between October 1999 and January 2000, 505 urine samples from 228 women with symptoms of UTI were consecutively collected from a California university health clinic. Of 255 UPEC isolates cultured from urine, 55 (22%) were resistant to trimethoprim-sulfamethoxazole (TMP-SMZ). Nearly half of these resistant isolates had an indistinguishable enterobacterial repetitive intergenic consensus PCR (ERIC-PCR) agarose gel electrophoretic pattern; these were designated clonal group A (CgA) (37). They were subsequently shown by MLST to belong to the ST69 complex (38). These initial sets of ST69 strains included multiple serogroups (O11, O17, O73, and O77) and were MDR. Later, additional serogroups were identified among these strains (O25b, O86, and O125ab), some of which were susceptible to all drugs tested (39, 40). Serogrouping and drug susceptibility profiles further differentiated this UPEC lineage. In addition to cystitis, ST69 strains caused pyelonephritis and bacteremia (4144).
CA-UTI TMP-SMZ-resistant UPEC isolates from patients in two other university communities were analyzed for comparison to the California university community isolates. Nearly 40% of TMP-SMZ-resistant E. coli isolates from patients with UTI at two different universities in Midwestern United States, collected around the same time period, were indistinguishable by ERIC-PCR. The isolates within each university community were more similar to each other in their PFGE patterns than across the three communities, but they were all either closely related or possibly related, according to Tenover’s criteria (37). These two universities are located 1,400 to 1,800 miles from California. This observation led investigators to suggest that three distinct outbreaks of CA-UTI caused by E. coli ST69 occurred during this period and that they may have been caused by some contaminated food product that was nationally distributed (37).
If these geographic clusters of CA-UTI caused by this ST69 clonal lineage indeed represented outbreaks, they should also cluster in time. A follow-up study analyzed UPEC isolates from the same California university clinic during the same 3.5-month period in 1999 to 2000, 2000 to 2001, 2003 to 2004, and 2004 to 2005 (45). From 1,667 patients examined at the clinic with symptoms of UTI during these four periods, 780 E. coli isolates were obtained and 584 (75%) of them were analyzed by ERIC-PCR. Only four ERIC-PCR lineages—CgA, CgC, CgH, and Cg3—accounted for 203 (35%) of the 584 genotyped isolates. CgA decreased from 26% during the first period to 5% the following period; it then increased to 13% and then decreased again to 9% in the last two periods (45). The TMP-SMZ-resistant CgC lineage was not detected in any of the periods except in the third period, when it accounted for 14% of all TMP-SMZ-resistant UPEC isolates (45). TMP-SMZ-resistant CgH appeared for the first time in period 2 and was detected again during each of the subsequent periods. Cg3 appeared for the first time in the third period (7%) and decreased to 4% the following period. Clearly, each of these lineages exhibited epidemic behavior suggestive of common-source outbreaks.
ST69 is now recognized to cause CA-UTI as well as BSIs in many regions of the United States as well as of the world (40, 4651). As intestinal pathogenic E. coli strains, such as E. coli O157:H7, have been recognized to have a food animal reservoir (cattle) and are spread by contaminated food products nationally and globally, it was not farfetched to suggest that an ExPEC lineage, such as ST69, has a reservoir in some mammalian intestine, possibly of a food animal, and is spread by a contaminated food product. To explore this suggestion, Ramchandani et al. examined 495 animal and environmental E. coli isolates belonging to serogroups O11, O17, O73, and O77, which were collected between 1965 and 2002 by the Gastroenteric Disease Center at Pennsylvania State University (52). Of these, 128 (26%) were found to be indistinguishable from CgA by ERIC-PCR; 14 of these were resistant to TMP-SMZ and were isolated from cows and turkeys. PFGE pattern (pulsotype) analysis showed that 1 of these 14 isolates, obtained from a cow in 1988, was 94% similar to one of the UPEC CgA strains isolated from a patient in California (52). Subsequent studies have shown isolation of ST69 E. coli from broiler chickens (53, 54) as well as pork (55).
Humans could serve as a reservoir of UPEC, and certainly person-to-person transmissions of UPEC, especially in households, have been documented (5658). Transmissions between humans and their companion animals also occur (5962). However, such modes of transmission would be confined within households or to restricted geographic spaces and cannot explain the widespread national or global distribution of a single lineage like ST69.
It is now recognized that the widespread clonal distribution of UPEC is not limited to ST69. MLST analyses have shown other UPEC STs that are disseminated globally that include ST10, ST73, ST95, ST127, and ST131, which, together with ST69 and ST393, have come to be termed pandemic ExPEC lineages (46). Molecular epidemiology of infections caused by these ExPEC organisms is discussed further in a separate review under a section on differentiating E. coli pathovars versus nonpathovars.
Although UPEC strains based on MLST have been shown to be shared between human and food animal sources as described above, when they are subtyped by PFGE, their electrophoretic band patterns are rarely shown to be indistinguishable. One explanation for this observation is that the PFGE genotyping test may be more discriminating than MLST and that the test is unlikely to identify multiple strains with indistinguishable pulsotypes even if the isolates were recovered from the same outbreak. Another explanation is that the human and animal food ExPEC isolates that are analyzed are not always those collected contemporaneously from the same geographic sites, and many of the isolates that are compared are convenience- rather than population-based, prospectively collected samples. More studies with well-thought-out study designs need to be conducted to document sources of UPECs (55).
Nevertheless, as the above examples illustrate, molecular epidemiologic methods applied to examine CA-UTI have made several discoveries concerning a common community-acquired infectious disease that had not been previously recognized to occur as outbreaks. These discoveries engendered a new hypothesis that CA-UTI may be foodborne, which, if clearly demonstrated, would provide an opportunity to devise novel strategies to prevent and control this important public health problem.

Community-onset BSIs in San Francisco, 2007 to 2010

Between July 2007 and September 2010, 539 Gram-negative bacterial isolates from blood cultures performed at a public general hospital in San Francisco, CA, were analyzed for drug resistance genes and genotypes (63). Of these cultures, 249 were E. coli and 220 of them were genotyped by MLST. These 220 E. coli strains were comprised of 34 distinct STs; only 5 of them (ST12, ST69, ST73, ST95, and ST131) accounted for 65% of all the genotyped E. coli isolates (63).
Note that four of these are the same pandemic lineages detected among CA-UTI patients in the California university community studies (45). This observation suggests that a large proportion of BSI cases may be preceded by episodes of CA-UTI. In fact, for more than 70% of the hospitalized patients, their first blood culture tested positive for E. coli in less than 48 h after admission. This suggests that these BSIs occurred outside the hospital before admission. In fact, 90% or more of the blood cultures tested positive in less than 48 h of admission in those infected with ST95 or ST12 complex. What appeared to be unrelated cases of BSI in hospitalized patients may represent multiple community outbreaks.
If indeed E. coli strains that cause CA-UTI infections have common sources, such as contaminated food, then E. coli strains that cause community-onset BSI may have the same sources. Such sources have been demonstrated for a small number of these infections, but more definitive studies are needed (46). Nevertheless, molecular epidemiologic studies have clearly demonstrated that sporadic infections, whether they are gastroenteritis, CA-UTI, or community-onset BSI, occur in clusters suggestive of outbreaks, which provides an opportunity to study them further to identify risk factors. This notion is discussed in more detail in a separate review.
Molecular epidemiologic methods to study sporadic infections can also be applied to assess the temporal trend of drug resistance prevalence of community-acquired infections, as discussed below.

DRUG-RESISTANT COMMUNITY-ACQUIRED GRAM-NEGATIVE BACTERIAL INFECTIONS

In February 2017, the World Health Organization (WHO) listed several groups of antimicrobial-resistant (AMR) bacterial pathogens in the priority 1 critical category. They were all Gram-negative bacteria: Acinetobacter, Pseudomonas, and members of Enterobacteriaceae (E. coli, Klebsiella, Proteus, and Serratia) (http://www.who.int/medicines/publications/global-priority-list-antibiotic-resistant-bacteria/en/). In 2013, the CDC in its report included two groups of Gram-negative bacteria among the list of three “urgent threat” pathogens—carbapenem-resistant Enterobacteriaceae, MDR Neisseria gonorrhoeae, and Clostridium difficile (64). The epidemiology of infections caused by drug-resistant Gram-negative bacteria is now a major research priority, and molecular epidemiologic applications are making new contributions to address this continuously growing public health threat.
One consistently recognized risk factor for any drug-resistant bacterial infection is, of course, the previous use of an antimicrobial agent. Exposure to an antimicrobial drug by a microbe inevitably selects for drug resistance in that microbe. Antimicrobial drug stewardship is thus based on this understanding, and therefore, the main tenet of the stewardship is to reduce and regulate antimicrobial drug use in clinical settings (65, 66). Antimicrobial drugs are widely used in human and veterinary clinical practices, as they are in food animal husbandry. In fact, worldwide, more antibiotics are used in food animal production as growth promoters or infection “prevention” than in all human clinical practices combined (6769). Thus, drug-resistant bacteria can be selected in both clinical practice and animal husbandry environments. But is selection of drug-resistant pathogens by drugs an important contributor to the prevalence of drug-resistant Gram-negative bacterial infections in community or health care settings?

Determinants of Drug Resistance in Community-Acquired Gram-Negative Bacterial Pathogens

The public health debate concerning the growing threat of drug-resistant infections often centers on whether human drug-resistant infections are caused by drug-resistant bacterial pathogens that are selected from the use of antimicrobial drugs in human clinical settings or in livestock husbandry (2, 7083). There should be no debate that outbreaks of foodborne gastroenteritis are mostly caused by drug-resistant Salmonella, Campylobacter, and STEC organisms that originate from food animal intestines. These outbreaks often trigger massive recall of implicated food products. Molecular epidemiology studies of sporadic cases of these diseases caused by drug-resistant enteric pathogens, as described above, have unmasked outbreaks, which led to food product recalls. Since antimicrobial agents are abundantly used in animal husbandry, it is conceivable that such drug-resistant enteric bacterial pathogens are frequently selected in such environments.
However, the community prevalence of drug-resistant salmonellosis caused by contaminated food products is due not to the selection of drug-resistant Salmonella but to the community distribution of the contaminated products themselves. In the above example of salmonellosis in New Jersey and Pennsylvania, 45% of sporadic cases of S. Newport salmonellosis cases during the summer of 1981 were suggested to be caused by a single lineage of S. Newport contaminating precooked roast beef (11). If this strain happened to be resistant to a drug, then the prevalence in these two states of drug-resistant S. Newport salmonellosis would have been greatly affected by this single lineage. The resistant strain may have been initially selected at a cattle farm, but the high prevalence of resistant salmonellosis in New Jersey and Pennsylvania would have been due to the community distribution of a contaminated meat product.
What about other diseases like CA-UTI or community-onset BSI? Is it the unnecessary use or overuse of antibiotics in clinical settings or in animal husbandry that affects prevalence of these drug-resistant infections? Is it both? If it is both, which source contributes more to the incidence of human drug-resistant infections? A molecular epidemiology approach to address these questions reveals that the answers have to be more nuanced than previously thought.

Assessing Intermediate and Long-Term Trends in Prevalence of Drug-Resistant UTI

In assessing prevalence of drug-resistant infections in a community setting (or even in a health care setting) such as UTI, appropriate methods for collecting clinical samples are critical. Studies based on convenience samples may overestimate the frequency of drug resistance, since such samples are likely to include isolates from patients who were not initially cultured but subsequently returned for culture because they failed the empirical treatment regimen. Accurate estimates of drug resistance prevalence need to be based on isolates obtained from prospectively and consecutively collected clinical specimens. Such estimates can then be used to assess trends—whether resistance prevalence is increasing, decreasing, or unchanging over time. Below, we discuss how a molecular epidemiology approach can be used to assess and interpret intermediate and long-term trends in prevalence of drug-resistant CA-UTI in one university community.

Change in prevalence of drug-resistant UTI at a university community, 1999 to 2005

Assessing the prevalence of drug-resistant infections in a community setting based only on bacterial species information may not provide accurate estimates and may yield misleading information about the trend of prevalence. In the CA-UTI study at the California university community described above, during October 1999 to January 2000, 22% of the 255 consecutively collected UPEC isolates were found to be TMP-SMZ resistant (37). However, nearly half of these TMP-SMZ-resistant isolates belonged to one clonal lineage, ST69. If this lineage had not been circulating at the university community during the period of the study, the prevalence of TMP-SMZ resistance would have been 11%. That is, a single clonal strain that happens to be resistant can abruptly change the prevalence of drug resistance in a community. This change in resistance prevalence had nothing to do with overuse or inappropriate use of antibiotics in the community, health care settings, or animal husbandry. Drug-resistant ST69 may have been originally selected by antimicrobial drugs in some food animal reservoir, but the 22% prevalence of TMP-SMZ resistance at the California university community was due to an epidemiologic event—introduction and circulation of this drug-resistant clone into the community that happened to occur at the time a study was being conducted.
In the multiyear study (1999 to 2005) described above, only four ST lineages—CgA (ST69), CgC (ST95), CgH, and Cg3 (ST420)—accounted for more than one-third of the UPEC isolates genotyped (45). CgA, CgC, and CgH were all MDR. In fact, MDR CgA and CgC strains accounted for 35% of 117 MDR isolates. Between periods 3 and 4, ampicillin resistance significantly decreased, from 35% to 24%, but 75% of this decrease could be attributed to the decrease of one lineage—CgC—over these two periods (45). Furthermore, none of the lineages that were initially susceptible to all drugs tested (pan-susceptible) gained resistance during this period. Thus, a large proportion of MDR UTI cases was caused by a small number of circulating genotypes, and a new selection of drug-resistant UPEC strains in this community was not detected.

Change in drug-resistant UTI in a university community, 1999 to 2017

In a long-term (17 years) follow-up study, 233 UPEC isolates were analyzed from urine samples consecutively collected from the same California university health center between September 2016 and May 2017 (84). These isolates were comprised of more than 60 distinct ST lineages. In both periods (1999 to 2000 and 2016 to 2017), the most common UPEC MLST genotypes were ST10, ST69, ST73, ST95, ST127, and ST131. In the first period, these genotypes were responsible for 125 (56%) of 225 UTI cases, while in the second period, the same genotypes caused 148 (64%) of 233 cases. These genotypes belong to pandemic ExPEC lineages reported from many other parts of the world (46).
Of 233 isolates collected during 2016 to 2017, 97 (42%) were ampicillin resistant, while 55 (24%) of 225 isolates from 1999 to 2000 were ampicillin resistant (P < 0.001). The six pandemic genotypes were responsible for 60% and 68% of all ampicillin-resistant strains in the periods 1999 to 2000 and 2016 to 2017, respectively. The frequency of ampicillin resistance increased among ST95, ST73, and ST131 strains between the two periods; they accounted for 16% of 55 ampicillin-resistant strains in 1999 to 2000 and 38% in 2016 to 2017. The overall increase in ampicillin resistance over this 17-year period was, therefore, due mostly to the increase in these genotypes. Although the ampicillin resistance frequency also increased among the nonpandemic genotypes (22% to 36%), numerically, they contributed minimally to the overall increase. This observation supports the idea that antimicrobial agents do not exert equal selective pressures on different bacterial strains to gain resistance, even after 17 years in the same community (84). The same genotypes responsible for most of the ampicillin resistance in 1999 to 2000 became even more responsible for ampicillin resistance in 2016 to 2017.
These observations demonstrate that it is the composition and the number of strains of circulating clonal lineages at any point in time that determine drug resistance prevalence in a community. Without the genotype information, misleading conclusions would have been made about the reasons for the changing prevalence of drug-resistant CA-UTI in the California university community. That is, the changing prevalence of drug resistance in a community is largely a function of transmission dynamics of clonal lineages rather than the clinical or livestock use of antibiotics. The ability of these dominant genotypes to maintain themselves in food products is what appears to account for the differences in AMR prevalence in a community. In the above-described study over a 17-year period, there was an apparent increase in ampicillin resistance of the UPEC isolates, but a large proportion of this increase was due to the higher number of just three UPEC lineages that happened to be resistant to ampicillin. Therefore, it is not sufficient to describe drug resistance prevalence in a community as increasing, decreasing, or unchanging without specifying the clonal composition of the study isolates used to make such a conclusion.
This is not to say that antimicrobial drugs are not important in determining AMR prevalence. Antimicrobial agents do select for new drug-resistant strains that subsequently get widely disseminated. If the number of such clonal lineages continues to increase in a community, then the frequency of resistance can be said to increase in that community. What molecular epidemiology analysis informs us, however, is that it is not simply the use of antimicrobial agents but also the transmission dynamics of clonal lineages that ultimately determine the resistance prevalence in a community setting. Stated in another way, just as an increase in sporadic cases of foodborne disease or CA-UTI may be comprised of one or more outbreaks, an increase in prevalence of drug resistance in a community may also be comprised of one or more outbreaks caused by bacterial lineages that happen to be drug resistant. Therefore, drug resistance prevalence in a community is ultimately determined by a mixture of outbreaks of drug-resistant and pan-susceptible microbial lineages.
This notion is not limited to community-acquired UPEC infections. In San Francisco, in the 3-year community-onset BSI study described above, only five MLST genotypes (ST12, ST69, ST73, ST95, and ST131) accounted for 71% of 119 MDR BSI isolates (63). One (ST131) singly contributed to 39% of all the MDR isolates. Between 1995 and 1998 in Salvador, Brazil, most (55%) of the meningitis cases caused by penicillin-nonsusceptible Pneumococcus were caused by serotype 14, belonging to a group of closely related BOX A1R-based PCR genotypes (85). In New York City in the 1990s, based on IS6110 analysis of Mycobacterium tuberculosis isolates, a large proportion of the newly diagnosed MDR tuberculosis cases were estimated to have been caused by just one M. tuberculosis lineage called the W strain (member of the Beijing clade) (86, 87). In fact, nearly 80% of MDR tuberculosis cases in New York City were due to transmission of a limited number of dominant genotypes of M. tuberculosis that were already MDR at the time of transmission (87). More recently, in South Africa, among 86 genotyped Mycobacterium tuberculosis isolates from patients with hospital-associated extensively drug-resistant (XDR) tuberculosis, 79 (92%) were found to belong to one spoligotype cluster, ST60 (88). Nearly all of these patients were coinfected with HIV. The ST60 cluster strains were further distinguished by IS6110 restriction fragment length polymorphism (RFLP) analysis into three groups, and one of these (called KZN) was responsible for 51 (96%) of them (88). Therefore, 51 (59%) of all 86 genotyped XDR strains belonged to one lineage: KZN.
In a separate study in KwaZulu-Natal Province, South Africa, 404 patients with XDR tuberculosis were prospectively examined to determine the proportion of cases that resulted from inadequate treatment of MDR tuberculosis that evolved into XDR tuberculosis (acquired resistance) versus those due to transmission of XDR M. tuberculosis (89). Of these participants, 77% were coinfected with HIV. A genotyping analysis of 386 participants’ M. tuberculosis isolates based on spoligotypes and IS6110 RFLP identified one large cluster (LAM4/KZN) composed of 212 (55%) participants (89). Most of these patients had never received treatment for MDR tuberculosis and therefore were deemed to have developed XDR tuberculosis from transmission of already XDR strains of M. tuberculosis instead of inadequate treatment of MDR tuberculosis. Thus, the prevalence of XDR tuberculosis in South Africa appears to be determined more by the transmission of a limited number of dominant clonal lineages of M. tuberculosis—outbreaks—rather than by acquired resistance during treatment.
Of course, inadequate treatment contributes to the selection of MDR and XDR M. tuberculosis at some point in the transmission pathways, but as highlighted in the discussion on sporadic cases of foodborne disease and CA-UTI, drug-resistant tuberculosis cases, especially those involving coinfection with HIV, are largely comprised of multiple outbreaks of clonal lineages. However, unlike the foodborne diseases, these tuberculosis cases caused by clonal M. tuberculosis lineages are not point source outbreaks but are parts of multiple person-to-person transmission pathways.

CONCLUDING REMARKS

Distinguishing epidemic and endemic occurrence of infectious diseases is one of the major concerns of public health dealt with by surveillance systems. Until molecular epidemiology methods came to be applied to study sporadic or endemic infectious diseases, population-based studies of such disease occurrences were severely limited. The examples described in this review illustrate how characterizing disease occurrence as endemic or sporadic does not paint a complete picture about transmission dynamics of infectious agents in community settings and that it hampers or stifles our ability to devise new control strategies. When seemingly unrelated infectious disease cases could be shown to belong to different subgroups or clusters based on genotypes of the infecting agents, they may be found to have epidemiologic relationships. Identification of such epidemiologic relationships provides an opportunity for intervention.
This discussion about infectious disease occurrence also highlights a debate about the origin of human drug-resistant enteric pathogens. The debate so far has focused on two major arguments: such pathogens come from overuse or inappropriate or unnecessary use of antimicrobial agents in human clinical practices, or they come from overuse or inappropriate or unnecessary use of antimicrobial agents in livestock husbandry. These arguments result from incomplete information related to bacterial strain differences, which can be unmasked by molecular microbiology tools. As the above discussion on drug-resistant infections highlights, the prevalence of drug-resistant bacterial pathogens is not only determined by antimicrobial agent use in human/veterinary clinical practices or livestock husbandry. Antimicrobial agents are necessary but not sufficient to explain the changing prevalence of drug-resistant infectious diseases in communities (and perhaps with some pathogens even in health care settings). If antibiotics exerted selective pressures on microbes equally for them to become resistant, we should observe thousands of different drug-resistant genotypes. Rather, what we observe is that a large proportion of drug-resistant infections are caused by a small set of lineages. The relative proportions of these lineages in a community are determined by epidemiologic factors, such as the geographic distribution of contaminated food products (for diseases such as gastroenteritis, CA-UTI, or community-onset BSI) or person-to-person transmission (e.g., tuberculosis), rather than by local clinical use of antimicrobial agents. However, what is still not yet explained is why the same sets of lineages become predominant nationally and globally. This issue will be explored further in another review.

Footnote

*
This article is part of a curated collection.

CURATED COLLECTION

REFERENCES

1.
Centers for Disease Control and Prevention. 2000. Probable locally acquired mosquito-transmitted Plasmodium vivax infection—Suffolk County, New York, 1999. MMWR Morb Mortal Wkly Rep 49:495–498.
2.
Cohen ML, Tauxe RV. 1986. Drug-resistant Salmonella in the United States: an epidemiologic perspective. Science 234:964–969. https://doi.org/10.1126/science.3535069.
3.
Yigit H, Queenan AM, Anderson GJ, Domenech-Sanchez A, Biddle JW, Steward CD, Alberti S, Bush K, Tenover FC. 2001. Novel carbapenem-hydrolyzing beta-lactamase, KPC-1, from a carbapenem-resistant strain of Klebsiella pneumoniae. Antimicrob Agents Chemother 45:1151–1161. https://doi.org/10.1128/AAC.45.4.1151-1161.2001.
4.
Marchaim D, Navon-Venezia S, Schwaber MJ, Carmeli Y. 2008. Isolation of imipenem-resistant Enterobacter species: emergence of KPC-2 carbapenemase, molecular characterization, epidemiology, and outcomes. Antimicrob Agents Chemother 52:1413–1418. https://doi.org/10.1128/AAC.01103-07.
5.
Bradford PA, Bratu S, Urban C, Visalli M, Mariano N, Landman D, Rahal JJ, Brooks S, Cebular S, Quale J. 2004. Emergence of carbapenem-resistant Klebsiella species possessing the class A carbapenem-hydrolyzing KPC-2 and inhibitor-resistant TEM-30 beta-lactamases in New York City. Clin Infect Dis 39:55–60. https://doi.org/10.1086/421495.
6.
Bratu S, Landman D, Haag R, Recco R, Eramo A, Alam M, Quale J. 2005. Rapid spread of carbapenem-resistant Klebsiella pneumoniae in New York City: a new threat to our antibiotic armamentarium. Arch Intern Med 165:1430–1435. https://doi.org/10.1001/archinte.165.12.1430.
7.
Bratu S, Mooty M, Nichani S, Landman D, Gullans C, Pettinato B, Karumudi U, Tolaney P, Quale J. 2005. Emergence of KPC-possessing Klebsiella pneumoniae in Brooklyn, New York: epidemiology and recommendations for detection. Antimicrob Agents Chemother 49:3018–3020. https://doi.org/10.1128/AAC.49.7.3018-3020.2005.
8.
Hujer AM, Keslar KS, Dietenberger NJ, Bethel CR, Endimiani A, Bonomo RA. 2009. Detection of SHV beta-lactamases in Gram-negative bacilli using fluorescein-labeled antibodies. BMC Microbiol 9:46. https://doi.org/10.1186/1471-2180-9-46.
9.
Gupta N, Limbago BM, Patel JB, Kallen AJ. 2011. Carbapenem-resistant Enterobacteriaceae: epidemiology and prevention. Clin Infect Dis 53:60–67. https://doi.org/10.1093/cid/cir202.
10.
Adams DA, Thomas KR, Jajosky RA, Foster L, Sharp P, Onweh DH, Schley AW, Anderson WJ, Nationally Notifiable Infectious Conditions Group. 2016. Summary of notifiable infectious diseases and conditions—United States, 2014. MMWR Morb Mortal Wkly Rep 63:1–152. https://doi.org/10.15585/mmwr.mm6354a1.
11.
Riley LW, DiFerdinando GT, Jr, DeMelfi TM, Cohen ML. 1983. Evaluation of isolated cases of salmonellosis by plasmid profile analysis: introduction and transmission of a bacterial clone by precooked roast beef. J Infect Dis 148:12–17. https://doi.org/10.1093/infdis/148.1.12.
12.
Bender JB, Hedberg CW, Besser JM, Boxrud DJ, MacDonald KL, Osterholm MT. 1997. Surveillance for Escherichia coli O157:H7 infections in Minnesota by molecular subtyping. N Engl J Med 337:388–394. https://doi.org/10.1056/NEJM199708073370604.
13.
Blake PA, Allegra DT, Snyder JD, Barrett TJ, McFarland L, Caraway CT, Feeley JC, Craig JP, Lee JV, Puhr ND, Feldman RA. 1980. Cholera—a possible endemic focus in the United States. N Engl J Med 302:305–309. https://doi.org/10.1056/NEJM198002073020601.
14.
Bender JB, Hedberg CW, Boxrud DJ, Besser JM, Wicklund JH, Smith KE, Osterholm MT. 2001. Use of molecular subtyping in surveillance for Salmonella enterica serotype typhimurium. N Engl J Med 344:189–195. https://doi.org/10.1056/NEJM200101183440305.
15.
Smith KE, Besser JM, Hedberg CW, Leano FT, Bender JB, Wicklund JH, Johnson BP, Moore KA, Osterholm MT, Investigation Team. 1999. Quinolone-resistant Campylobacter jejuni infections in Minnesota, 1992–1998. N Engl J Med 340:1525–1532. https://doi.org/10.1056/NEJM199905203402001.
16.
Heir E, Lindstedt BA, Nygård I, Vardund T, Hasseltvedt V, Kapperud G. 2002. Molecular epidemiology of Salmonella typhimurium isolates from human sporadic and outbreak cases. Epidemiol Infect 128:373–382. https://doi.org/10.1017/S0950268802007045.
17.
Kapperud G, Gustavsen S, Hellesnes I, Hansen AH, Lassen J, Hirn J, Jahkola M, Montenegro MA, Helmuth R. 1990. Outbreak of Salmonella typhimurium infection traced to contaminated chocolate and caused by a strain lacking the 60-megadalton virulence plasmid. J Clin Microbiol 28:2597–2601.
18.
Mullner P, Shadbolt T, Collins-Emerson JM, Midwinter AC, Spencer SE, Marshall J, Carter PE, Campbell DM, Wilson DJ, Hathaway S, Pirie R, French NP. 2010. Molecular and spatial epidemiology of human campylobacteriosis: source association and genotype-related risk factors. Epidemiol Infect 138:1372–1383. https://doi.org/10.1017/S0950268809991579.
19.
Octavia S, Wang Q, Tanaka MM, Kaur S, Sintchenko V, Lan R. 2015. Delineating community outbreaks of Salmonella enterica serovar Typhimurium by use of whole-genome sequencing: insights into genomic variability within an outbreak. J Clin Microbiol 53:1063–1071. https://doi.org/10.1128/JCM.03235-14.
20.
Okoro CK, Kingsley RA, Connor TR, Harris SR, Parry CM, Al-Mashhadani MN, Kariuki S, Msefula CL, Gordon MA, de Pinna E, Wain J, Heyderman RS, Obaro S, Alonso PL, Mandomando I, MacLennan CA, Tapia MD, Levine MM, Tennant SM, Parkhill J, Dougan G. 2012. Intracontinental spread of human invasive Salmonella Typhimurium pathovariants in sub-Saharan Africa. Nat Genet 44:1215–1221. https://doi.org/10.1038/ng.2423.
21.
Mather AE, Reid SW, Maskell DJ, Parkhill J, Fookes MC, Harris SR, Brown DJ, Coia JE, Mulvey MR, Gilmour MW, Petrovska L, de Pinna E, Kuroda M, Akiba M, Izumiya H, Connor TR, Suchard MA, Lemey P, Mellor DJ, Haydon DT, Thomson NR. 2013. Distinguishable epidemics of multidrug-resistant Salmonella Typhimurium DT104 in different hosts. Science 341:1514–1517. https://doi.org/10.1126/science.1240578.
22.
Hawkey J, Edwards DJ, Dimovski K, Hiley L, Billman-Jacobe H, Hogg G, Holt KE. 2013. Evidence of microevolution of Salmonella Typhimurium during a series of egg-associated outbreaks linked to a single chicken farm. BMC Genomics 14:800. https://doi.org/10.1186/1471-2164-14-800.
23.
Tullus K, Hörlin K, Svenson SB, Källenius G. 1984. Epidemic outbreaks of acute pyelonephritis caused by nosocomial spread of P fimbriated Escherichia coli in children. J Infect Dis 150:728–736. https://doi.org/10.1093/infdis/150.5.728.
24.
Okuda T, Endo N, Osada Y, Zen-Yoji H. 1984. Outbreak of nosocomial urinary tract infections caused by Serratia marcescens. J Clin Microbiol 20:691–695.
25.
Su LH, Ou JT, Leu HS, Chiang PC, Chiu YP, Chia JH, Kuo AJ, Chiu CH, Chu C, Wu TL, Sun CF, Riley TV, Chang BJ, Infection Control Group. 2003. Extended epidemic of nosocomial urinary tract infections caused by Serratia marcescens. J Clin Microbiol 41:4726–4732. https://doi.org/10.1128/JCM.41.10.4726-4732.2003.
26.
Ikram R, Psutka R, Carter A, Priest P. 2015. An outbreak of multi-drug resistant Escherichia coli urinary tract infection in an elderly population: a case-control study of risk factors. BMC Infect Dis 15:224. https://doi.org/10.1186/s12879-015-0974-0.
27.
Foxman B, Brown P. 2003. Epidemiology of urinary tract infections: transmission and risk factors, incidence, and costs. Infect Dis Clin North Am 17:227–241. https://doi.org/10.1016/S0891-5520(03)00005-9.
28.
Gupta K, Hooton TM, Naber KG, Wullt B, Colgan R, Miller LG, Moran GJ, Nicolle LE, Raz R, Schaeffer AJ, Soper DE, Infectious Diseases Society of America, European Society for Microbiology and Infectious Diseases. 2011. International clinical practice guidelines for the treatment of acute uncomplicated cystitis and pyelonephritis in women: a 2010 update by the Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases. Clin Infect Dis 52:e103–e120. https://doi.org/10.1093/cid/ciq257.
29.
Foxman B, Barlow R, D’Arcy H, Gillespie B, Sobel JD. 2000. Urinary tract infection: self-reported incidence and associated costs. Ann Epidemiol 10:509–515. https://doi.org/10.1016/S1047-2797(00)00072-7.
30.
Naber KG, Schito G, Botto H, Palou J, Mazzei T. 2008. Surveillance study in Europe and Brazil on clinical aspects and antimicrobial resistance epidemiology in females with cystitis (ARESC): implications for empiric therapy. Eur Urol 54:1164–1175. https://doi.org/10.1016/j.eururo.2008.05.010.
31.
Suskind AM, Saigal CS, Hanley JM, Lai J, Setodji CM, Clemens JQ, Urologic Diseases of America Project. 2016. Incidence and management of uncomplicated recurrent urinary tract infections in a national sample of women in the United States. Urology 90:50–55. https://doi.org/10.1016/j.urology.2015.11.051.
32.
Phillips I, Eykyn S, King A, Gransden WR, Rowe B, Frost JA, Gross RJ. 1988. Epidemic multiresistant Escherichia coli infection in West Lambeth Health District. Lancet 1:1038–1041. https://doi.org/10.1016/S0140-6736(88)91853-3.
33.
Prats G, Mirelis B, Llovet T, Muñoz C, Miró E, Navarro F. 2000. Antibiotic resistance trends in enteropathogenic bacteria isolated in 1985–1987 and 1995–1998 in Barcelona. Antimicrob Agents Chemother 44:1140–1145. https://doi.org/10.1128/AAC.44.5.1140-1145.2000.
34.
Johnson JR, Stell AL, O’Bryan TT, Kuskowski M, Nowicki B, Johnson C, Maslow JN, Kaul A, Kavle J, Prats G. 2002. Global molecular epidemiology of the O15:K52:H1 extraintestinal pathogenic Escherichia coli clonal group: evidence of distribution beyond Europe. J Clin Microbiol 40:1913–1923. https://doi.org/10.1128/JCM.40.6.1913-1923.2002.
35.
Cagnacci S, Gualco L, Debbia E, Schito GC, Marchese A. 2008. European emergence of ciprofloxacin-resistant Escherichia coli clonal groups O25:H4-ST 131 and O15:K52:H1 causing community-acquired uncomplicated cystitis. J Clin Microbiol 46:2605–2612. https://doi.org/10.1128/JCM.00640-08.
36.
Olesen B, Scheutz F, Menard M, Skov MN, Kolmos HJ, Kuskowski MA, Johnson JR. 2009. Three-decade epidemiological analysis of Escherichia coli O15:K52:H1. J Clin Microbiol 47:1857–1862. https://doi.org/10.1128/JCM.00230-09.
37.
Manges AR, Johnson JR, Foxman B, O’Bryan TT, Fullerton KE, Riley LW. 2001. Widespread distribution of urinary tract infections caused by a multidrug-resistant Escherichia coli clonal group. N Engl J Med 345:1007–1013. https://doi.org/10.1056/NEJMoa011265.
38.
Tartof SY, Solberg OD, Manges AR, Riley LW. 2005. Analysis of a uropathogenic Escherichia coli clonal group by multilocus sequence typing. J Clin Microbiol 43:5860–5864. https://doi.org/10.1128/JCM.43.12.5860-5864.2005.
39.
Colomer-Lluch M, Mora A, López C, Mamani R, Dahbi G, Marzoa J, Herrera A, Viso S, Blanco JE, Blanco M, Alonso MP, Jofre J, Muniesa M, Blanco J. 2013. Detection of quinolone-resistant Escherichia coli isolates belonging to clonal groups O25b:H4-B2-ST131 and O25b:H4-D-ST69 in raw sewage and river water in Barcelona, Spain. J Antimicrob Chemother 68:758–765. https://doi.org/10.1093/jac/dks477.
40.
Skjøt-Rasmussen L, Olsen SS, Jakobsen L, Ejrnaes K, Scheutz F, Lundgren B, Frimodt-Møller N, Hammerum AM. 2013. Escherichia coli clonal group A causing bacteraemia of urinary tract origin. Clin Microbiol Infect 19:656–661. https://doi.org/10.1111/j.1469-0691.2012.03961.x.
41.
Adams-Sapper S, Sergeevna-Selezneva J, Tartof S, Raphael E, Diep BA, Perdreau-Remington F, Riley LW. 2012. Globally dispersed mobile drug-resistance genes in gram-negative bacterial isolates from patients with bloodstream infections in a US urban general hospital. J Med Microbiol 61:968–974. https://doi.org/10.1099/jmm.0.041970-0.
42.
Ajiboye RM, Solberg OD, Lee BM, Raphael E, Debroy C, Riley LW. 2009. Global spread of mobile antimicrobial drug resistance determinants in human and animal Escherichia coli and Salmonella strains causing community-acquired infections. Clin Infect Dis 49:365–371. https://doi.org/10.1086/600301.
43.
Manges AR, Perdreau-Remington F, Solberg O, Riley LW. 2006. Multidrug-resistant Escherichia coli clonal groups causing community-acquired bloodstream infections. J Infect 53:25–29. https://doi.org/10.1016/j.jinf.2005.09.012.
44.
Manges AR, Dietrich PS, Riley LW. 2004. Multidrug-resistant Escherichia coli clonal groups causing community-acquired pyelonephritis. Clin Infect Dis 38:329–334. https://doi.org/10.1086/380640.
45.
Smith SP, Manges AR, Riley LW. 2008. Temporal changes in the prevalence of community-acquired antimicrobial-resistant urinary tract infection affected by Escherichia coli clonal group composition. Clin Infect Dis 46:689–695. https://doi.org/10.1086/527386.
46.
Riley LW. 2014. Pandemic lineages of extraintestinal pathogenic Escherichia coli. Clin Microbiol Infect 20:380–390. https://doi.org/10.1111/1469-0691.12646.
47.
Johnson JR, Manges AR, O’Bryan TT, Riley LW. 2002. A disseminated multidrug-resistant clonal group of uropathogenic Escherichia coli in pyelonephritis. Lancet 359:2249–2251. https://doi.org/10.1016/S0140-6736(02)09264-4.
48.
Dias RC, Marangoni DV, Smith SP, Alves EM, Pellegrino FL, Riley LW, Moreira BM. 2009. Clonal composition of Escherichia coli causing community-acquired urinary tract infections in the State of Rio de Janeiro, Brazil. Microb Drug Resist 15:303–308. https://doi.org/10.1089/mdr.2009.0067.
49.
Johnson JR, Murray AC, Kuskowski MA, Schubert S, Prère MF, Picard B, Colodner R, Raz R, Trans-Global Initiative for Antimicrobial Resistance Initiative (TIARA) Investigators. 2005. Distribution and characteristics of Escherichia coli clonal group A. Emerg Infect Dis 11:141–145. https://doi.org/10.3201/eid1101.040418.
50.
Johnson JR, Menard M, Johnston B, Kuskowski MA, Nichol K, Zhanel GG. 2009. Epidemic clonal groups of Escherichia coli as a cause of antimicrobial-resistant urinary tract infections in Canada, 2002 to 2004. Antimicrob Agents Chemother 53:2733–2739. https://doi.org/10.1128/AAC.00297-09.
51.
Johnson JR, Menard ME, Lauderdale TL, Kosmidis C, Gordon D, Collignon P, Maslow JN, Andrasević AT, Kuskowski MA, Trans-Global Initiative for Antimicrobial Resistance Analysis Investigators. 2011. Global distribution and epidemiologic associations of Escherichia coli clonal group A, 1998–2007. Emerg Infect Dis 17:2001–2009. https://doi.org/10.3201/eid1711.110488.
52.
Ramchandani M, Manges AR, DebRoy C, Smith SP, Johnson JR, Riley LW. 2005. Possible animal origin of human-associated, multidrug-resistant, uropathogenic Escherichia coli. Clin Infect Dis 40:251–257. https://doi.org/10.1086/426819.
53.
Agersø Y, Jensen JD, Hasman H, Pedersen K. 2014. Spread of extended spectrum cephalosporinase-producing Escherichia coli clones and plasmids from parent animals to broilers and to broiler meat in a production without use of cephalosporins. Foodborne Pathog Dis 11:740–746. https://doi.org/10.1089/fpd.2014.1742.
54.
Johnson JR, Porter SB, Johnston B, Thuras P, Clock S, Crupain M, Rangan U. 2017. Extraintestinal pathogenic and antimicrobial-resistant Escherichia coli, including sequence type 131 (ST131), from retail chicken breasts in the United States in 2013. Appl Environ Microbiol 83:e02956-16. https://doi.org/10.1128/AEM.02956-16.
55.
Vincent C, Boerlin P, Daignault D, Dozois CM, Dutil L, Galanakis C, Reid-Smith RJ, Tellier PP, Tellis PA, Ziebell K, Manges AR. 2010. Food reservoir for Escherichia coli causing urinary tract infections. Emerg Infect Dis 16:88–95. https://doi.org/10.3201/eid1601.091118.
56.
Ender PT, Gajanana D, Johnston B, Clabots C, Tamarkin FJ, Johnson JR. 2009. Transmission of an extended-spectrum-beta-lactamase-producing Escherichia coli (sequence type ST131) strain between a father and daughter resulting in septic shock and emphysematous pyelonephritis. J Clin Microbiol 47:3780–3782. https://doi.org/10.1128/JCM.01361-09.
57.
Johnson JR, Owens K, Gajewski A, Clabots C. 2008. Escherichia coli colonization patterns among human household members and pets, with attention to acute urinary tract infection. J Infect Dis 197:218–224. https://doi.org/10.1086/524844.
58.
Johnson JR, Clabots C. 2006. Sharing of virulent Escherichia coli clones among household members of a woman with acute cystitis. Clin Infect Dis 43:e101–e108. https://doi.org/10.1086/508541.
59.
Ewers C, Bethe A, Wieler LH, Guenther S, Stamm I, Kopp PA, Grobbel M. 2011. Companion animals: a relevant source of extended-spectrum β-lactamase-producing fluoroquinolone-resistant Citrobacter freundii. Int J Antimicrob Agents 37:86–87. https://doi.org/10.1016/j.ijantimicag.2010.09.007.
60.
Platell JL, Cobbold RN, Johnson JR, Heisig A, Heisig P, Clabots C, Kuskowski MA, Trott DJ. 2011. Commonality among fluoroquinolone-resistant sequence type ST131 extraintestinal Escherichia coli isolates from humans and companion animals in Australia. Antimicrob Agents Chemother 55:3782–3787. https://doi.org/10.1128/AAC.00306-11.
61.
Bert F, Panhard X, Johnson J, Lecuyer H, Moreau R, Le Grand J, Johnston B, Sinègre M, Valla D, Nicolas-Chanoine MH. 2008. Genetic background of Escherichia coli isolates from patients with spontaneous bacterial peritonitis: relationship with host factors and prognosis. Clin Microbiol Infect 14:1034–1040. https://doi.org/10.1111/j.1469-0691.2008.02088.x.
62.
Johnson JR, Miller S, Johnston B, Clabots C, Debroy C. 2009. Sharing of Escherichia coli sequence type ST131 and other multidrug-resistant and urovirulent E. coli strains among dogs and cats within a household. J Clin Microbiol 47:3721–3725. https://doi.org/10.1128/JCM.01581-09.
63.
Adams-Sapper S, Diep BA, Perdreau-Remington F, Riley LW. 2013. Clonal composition and community clustering of drug-susceptible and -resistant Escherichia coli isolates from bloodstream infections. Antimicrob Agents Chemother 57:490–497. https://doi.org/10.1128/AAC.01025-12.
64.
Centers for Disease Control and Prevention. 2013. Antibiotic Resistance Threats in the United States, 2013. Centers for Disease Control and Prevention, Atlanta, GA.
65.
Barlam TF, Cosgrove SE, Abbo LM, MacDougall C, Schuetz AN, Septimus EJ, Srinivasan A, Dellit TH, Falck-Ytter YT, Fishman NO, Hamilton CW, Jenkins TC, Lipsett PA, Malani PN, May LS, Moran GJ, Neuhauser MM, Newland JG, Ohl CA, Samore MH, Seo SK, Trivedi KK. 2016. Implementing an antibiotic stewardship program: guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis 62:e51–e77. https://doi.org/10.1093/cid/ciw118.
66.
Barlam TF, Cosgrove SE, Abbo LM, MacDougall C, Schuetz AN, Septimus EJ, Srinivasan A, Dellit TH, Falck-Ytter YT, Fishman NO, Hamilton CW, Jenkins TC, Lipsett PA, Malani PN, May LS, Moran GJ, Neuhauser MM, Newland JG, Ohl CA, Samore MH, Seo SK, Trivedi KK. 2016. Executive summary: implementing an antibiotic stewardship program: guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis 62:1197–1202. https://doi.org/10.1093/cid/ciw217.
67.
Food and Drug Administration. 2010. CVM Updates—CVM Reports on Antimicrobials Sold or Distributed for Food-Producing Animals. Food and Drug Administration, Silver Spring, MD.
68.
Van Boeckel TP, Brower C, Gilbert M, Grenfell BT, Levin SA, Robinson TP, Teillant A, Laxminarayan R. 2015. Global trends in antimicrobial use in food animals. Proc Natl Acad Sci U S A 112:5649–5654. https://doi.org/10.1073/pnas.1503141112.
69.
Krishnasamy V, Otte J, Silbergeld E. 2015. Antimicrobial use in Chinese swine and broiler poultry production. Antimicrob Resist Infect Control 4:17. https://doi.org/10.1186/s13756-015-0050-y.
70.
Cherubin CE. 1981. Antibiotic resistance of Salmonella in Europe and the United States. Rev Infect Dis 3:1105–1126. https://doi.org/10.1093/clinids/3.6.1105.
71.
Cherubin CE, Timoney JF, Sierra MF, Ma P, Marr J, Shin S. 1980. A sudden decline in ampicillin resistance in Salmonella typhimurium. JAMA 243:439–442. https://doi.org/10.1001/jama.1980.03300310027017.
72.
Angulo FJ, Johnson KR, Tauxe RV, Cohen ML. 2000. Origins and consequences of antimicrobial-resistant nontyphoidal Salmonella: implications for the use of fluoroquinolones in food animals. Microb Drug Resist 6:77–83. https://doi.org/10.1089/mdr.2000.6.77.
73.
US Congress Office of Technology Assessment. 1995. Impacts of Antibiotic-Resistant Bacteria. US Government Printing Office, Washington, DC.
74.
Smith DL, Harris AD, Johnson JA, Silbergeld EK, Morris JG, Jr. 2002. Animal antibiotic use has an early but important impact on the emergence of antibiotic resistance in human commensal bacteria. Proc Natl Acad Sci U S A 99:6434–6439. https://doi.org/10.1073/pnas.082188899.
75.
Sarmah AK, Meyer MT, Boxall AB. 2006. A global perspective on the use, sales, exposure pathways, occurrence, fate and effects of veterinary antibiotics (VAs) in the environment. Chemosphere 65:725–759. https://doi.org/10.1016/j.chemosphere.2006.03.026.
76.
Phillips I, Casewell M, Cox T, De Groot B, Friis C, Jones R, Nightingale C, Preston R, Waddell J. 2004. Does the use of antibiotics in food animals pose a risk to human health? A critical review of published data. J Antimicrob Chemother 53:28–52. https://doi.org/10.1093/jac/dkg483.
77.
Soulsby EJ. 2005. Resistance to antimicrobials in humans and animals. BMJ 331:1219–1220. https://doi.org/10.1136/bmj.331.7527.1219.
78.
Spellberg B, Guidos R, Gilbert D, Bradley J, Boucher HW, Scheld WM, Bartlett JG, Edwards J, Jr, Infectious Diseases Society of America. 2008. The epidemic of antibiotic-resistant infections: a call to action for the medical community from the Infectious Diseases Society of America. Clin Infect Dis 46:155–164. https://doi.org/10.1086/524891.
79.
Aitken SL, Dilworth TJ, Heil EL, Nailor MD. 2016. Agricultural applications for antimicrobials. A danger to human health: an official position statement of the Society of Infectious Diseases Pharmacists. Pharmacotherapy 36:422–432. https://doi.org/10.1002/phar.1737.
80.
Hu Y, Cheng H. 2016. Health risk from veterinary antimicrobial use in China’s food animal production and its reduction. Environ Pollut 219:993–997. https://doi.org/10.1016/j.envpol.2016.04.099.
81.
Cerniglia CE, Pineiro SA, Kotarski SF. 2016. An update discussion on the current assessment of the safety of veterinary antimicrobial drug residues in food with regard to their impact on the human intestinal microbiome. Drug Test Anal 8:539–548. https://doi.org/10.1002/dta.2024.
82.
Cabello FC, Godfrey HP. 2016. Even therapeutic antimicrobial use in animal husbandry may generate environmental hazards to human health. Environ Microbiol 18:311–313. https://doi.org/10.1111/1462-2920.13247.
83.
Aarestrup FM. 2015. The livestock reservoir for antimicrobial resistance: a personal view on changing patterns of risks, effects of interventions and the way forward. Philos Trans R Soc Lond B Biol Sci 370:20140085. https://doi.org/10.1098/rstb.2014.0085.
84.
Yamaji R, Rubin J, Thys E, Friedman CR, Riley LW. 2018. Persistent pandemic lineages of uropathogenic Escherichia coli in a college community—1999–2017. J Clin Microbiol 56:e01834-17. https://doi.org/10.1128/JCM.01834-17.
85.
Ko AI, Reis JN, Coppola SJ, Gouveia EL, Cordeiro SM, Lôbo TS, Pinheiro RM, Salgado K, Ribeiro Dourado CM, Tavares-Neto J, Rocha H, Galvão Reis M, Johnson WD Jr, Riley LW. 2000. Clonally related penicillin-nonsusceptible Streptococcus pneumoniae serotype 14 from cases of meningitis in Salvador, Brazil. Clin Infect Dis 30:78–86. https://doi.org/10.1086/313619.
86.
Frieden TR, Sherman LF, Maw KL, Fujiwara PI, Crawford JT, Nivin B, Sharp V, Hewlett D Jr, Brudney K, Alland D, Kreisworth BN. 1996. A multi-institutional outbreak of highly drug-resistant tuberculosis: epidemiology and clinical outcomes. JAMA 276:1229–1235. https://doi.org/10.1001/jama.1996.03540150031027.
87.
Friedman CR, Stoeckle MY, Kreiswirth BN, Johnson WD Jr, Manoach SM, Berger J, Sathianathan K, Hafner A, Riley LW. 1995. Transmission of multidrug-resistant tuberculosis in a large urban setting. Am J Respir Crit Care Med 152:355–359. https://doi.org/10.1164/ajrccm.152.1.7599845.
88.
Gandhi NR, Weissman D, Moodley P, Ramathal M, Elson I, Kreiswirth BN, Mathema B, Shashkina E, Rothenberg R, Moll AP, Friedland G, Sturm AW, Shah NS. 2013. Nosocomial transmission of extensively drug-resistant tuberculosis in a rural hospital in South Africa. J Infect Dis 207:9–17. https://doi.org/10.1093/infdis/jis631.
89.
Shah NS, Auld SC, Brust JC, Mathema B, Ismail N, Moodley P, Mlisana K, Allana S, Campbell A, Mthiyane T, Morris N, Mpangase P, van der Meulen H, Omar SV, Brown TS, Narechania A, Shaskina E, Kapwata T, Kreiswirth B, Gandhi NR. 2017. Transmission of extensively drug-resistant tuberculosis in South Africa. N Engl J Med 376:243–253. https://doi.org/10.1056/NEJMoa1604544.

Information & Contributors

Information

Published In

cover image Microbiology Spectrum
Microbiology Spectrum
Volume 7Number 430 August 2019
eLocator: 10.1128/microbiolspec.ame-0007-2019
Editors: Michael Sadowsky, BioTechnology Institute, University of Minnesota, St. Paul, MN, Ronald E. Blanton, Center for Global Health & Diseases, Case Western Reserve University, Cleveland, OH

History

Received: 13 March 2019
Returned for modification: 1 April 2019
Published online: 19 July 2019

Contributors

Author

Lee W. Riley
Division of Infectious Diseases and Vaccinology, School of Public Health, University of California, Berkeley, CA 94720

Editors

Michael Sadowsky
BioTechnology Institute, University of Minnesota, St. Paul, MN
Ronald E. Blanton
Center for Global Health & Diseases, Case Western Reserve University, Cleveland, OH

Notes

Correspondence: Lee W. Riley, [email protected]

Metrics & Citations

Metrics

Note: There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.

Citation counts come from the Crossref Cited by service.

Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media

Figures

Media

Tables

Share

Share

Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy