Antibiotics can have ecological effects that impact the efficacy of other antimicrobial agents or facilitate the development of secondary infections [1,2]. When antibiotics are administered, particularly when they are overused or misused, they change the environment and the biome, which in turn can lead to the selection or development of bacterial strains resistant to a wide range of antibiotic agents, extending beyond the particular antibiotic or antibiotic class initially administered. Certain antibiotic agents also change the normal bacterial flora or environment within the gastrointestinal tract, which in turn can promote the colonization and overgrowth of particular bacteria (e.g., Clostridium difficile), and increase the risk of gastrointestinal infections associated with these bacteria. Antibiotic usage can also have an impact on skin and mucosa colonization (such as for methicillin-resistant Staphylococcus aureus) with significantly increased risk of subsequent infections. These forms of ‘collateral damage’ associated with antibiotic use are important considerations when deciding how best to use antibiotics to prevent or treat infections in the hospital (and community) setting.
This review looks at some of the ecological effects of antibiotics used in the hospital and their potential for collateral damage of the nosocomial environment. Collateral damage is becoming an increasing problem due to the increasing severity of illness in hospitalized patients and the increasing use of broad-spectrum antibiotics . The ultimate goal is to understand how to better use antibiotics to optimize their beneficial effects, while minimizing risk of collateral damage, in other words, to improve antibiotic stewardship within hospitals and other institutions.
Clostridium difficile-associated disease
The increasing prevalence of C. difficile infection (CDI), including diarrhea and colitis, is a classic example of how use of certain antibiotics can disrupt bacterial flora of the gut and cause a ‘superinfection’ associated with potentially dire consequences. CDI also provides an illustrative case of how improved antibiotic stewardship and infection control measures can be used to prevent the occurrence of problematic bacterial infections.
CDI is the primary cause of nosocomial gastrointestinal disorders; its incidence and severity continue to rise among hospitalized patients in the United States and other countries [3–5]. Zilberberg et al.  reported an approximate doubling in incidence of adult CDI hospitalizations in the United States from 5.5 cases per 10 000 population in 2000 to 11.2 in 2005, and in the CDI-related age-adjusted case-fatality rate from 1.2% in 2000 to 2.2% in 2005. Particularly high rates are observed in patients of at least 65 years of age (Fig. 1). The incidence of infant CDI hospitalizations has also nearly doubled in the United States, from 2.8 cases per 10 000 in 2000 to 5.1 cases in 2005 . Increased mortality and case-fatality rates may be due to emergence of more virulent strains of C. difficile. Redelings et al.  reported a dramatic increase in mortality from CDI in the United States from 5.7 per million population in 1999 to 23.7 per million in 2004. According to the United States Centers for Disease Control and Prevention (CDC), hospital-acquired CDI accounts for approximately 9000 deaths annually, and another 3000 deaths postdischarge, whereas nursing home-onset cases account for 16 500 deaths annually .
Patients with extended stays in a hospital or long-term care facility often become colonized with C. difficile, although only a subset develop symptomatic infection or CDI . Clinical manifestations of CDI range from self-limited mild diarrheal illness to a fulminant life-threatening colitis . CDI has been associated with increased healthcare costs and longer hospital stays . Based on a sampling of United States hospital discharge abstracts nationwide, Ricciardi et al.  reported an increase in the rate of C. difficile colitis from 261 cases per 100 000 discharged patients in 1993 to 546 per 100 000 in 2003. In a study by Kyne et al. , hospitalized patients with C. difficile-associated diarrhea had an extra length of stay of 3.6 days and 54% higher adjusted hospital costs compared with patients without C. difficile-associated diarrhea.
Antibiotics as a risk factor for Clostridium difficile infection
Antibiotic therapy is the most important risk factor for CDI development. In fact, at least 90% of CDI cases are associated with concurrent or prior antibiotic administration [3,4,12], with most beginning 3–7 days after initial exposure . Other risk factors include age more than 65 years, severe underlying disease, neoplastic chemotherapy, length of hospital stay, hypoalbuminemia, and nasogastric intubation [4,13–15]. Antibiotics promote development of CDI by disrupting the normal flora of the colon, enabling C. difficile colonization and overgrowth [3,4]. Symptomatic disease, as opposed to asymptomatic carriage, CDI is associated with the acquisition of a toxigenic strain of C. difficile , together with failure of the host/patient to mount an anamnestic toxin A/B immunoglobulin G antibody response .
More than 40 000 bacterial species reside in the human gastrointestinal tract , and it was thought that these bacteria lived a solitary existence, living, growing, reproducing, and dying in relative isolation, without real communication or interaction with other commensal microorganisms [18,19]. However, more recent research on ‘quorum sensing’ indicates bacteria have complex chemical signaling systems that enable intraspecies and interspecies communication through secretion of molecules that impact the behavior, growth, and sometimes virulence of other gastrointestinal inhabitants [18–20]. More specifically, organisms release small amounts of autoinducers or transcription activators, and their concentration increases as cell density increases, until a minimal threshold concentration triggers a shift in microbial gene expression, which is associated with metabolic changes that can alter bacterial competence and virulence expression. These changes include production of proteases, biofilm formation, antibiotic formation, changes in motility, and sporulation.
The microbial gastrointestinal population is diverse, becoming increasingly more complicated down toward the colon, where there is a predominance of anaerobic bacteria [21,22]. The important point is that whenever administering an antibiotic that is excreted in the gut, there is an effect on the fecal flora, with the potential for subsequent changes and developments through various factors involved in intraspecies and interspecies communication.
Although a wide range of antibiotics have been associated with increased risk of CDI, broad-spectrum antibiotics such as third-generation cephalosporins, fluoroquinolones, ampicillin, and clindamycin have been most strongly linked [3,4,14,15,23]. All fluoroquinolones are strongly associated with increased risk of CDI [23,24]. A hypervirulent strain of C. difficile– known by various interchangeable designations as NAP1 (North American pulse-field gel electrophoresis type 1 by CDC), as ribotype 027, or BI by REA typing – is resistant to fluoroquinolones and has been associated with increasing severity of CDI [23,25]. In a study by Loo et al. , severe CDI was observed in 16.7% of patients with NAP1 C. difficile strains, characterized by binary toxin genes and a partial deletion of the tcdC gene , and in none of the patients with isolates lacking both binary toxin genes and the partial deletion of the tcdC gene. The study also demonstrated that incidence and mortality attributed to CDI are generally age-dependent, regardless of the predominant C. difficile strain involved (Table 1).
Preventing Clostridium difficile infection
Various approaches, separate or combined, are available to prevent the onset, recurrence, or persistence of CDI. It is important to understand that colonization with C. difficile and subsequent development of CDI depends not only on disruption of the normal gastrointestinal flora (e.g., through antibiotic use), but also on the acquisition of a toxin-producing strain of C. difficile. Both are necessary events for the occurrence of C. difficile colonization, overgrowth, and CDI development . Furthermore, ineffective host immunity is required for C. difficile colonization and overgrowth to lead to symptomatic disease and the greater likelihood of recurrence.
Because C. difficile is often transmitted from patient to patient within the hospital environment by hospital personnel or acquired through contact with pathogen on hospital furniture or equipment, infection control measures are also key to controlling the spread of C. difficile within the institutional setting. Such infection control measures include bleach for environmental disinfection, regular effective hand washing, and various barrier or contact precautions, when necessary, such as placement in private rooms, patient cohorting, and donning of gowns and gloves by hospital personnel [3,4,13]. C. difficile are spore-producing bacteria, and the spores survive routine environmental cleaning with detergents and hand hygiene with alcohol-based gels ; enhanced cleaning techniques are required to control spread. Practice recommendations for the prevention of C. difficile infections have been developed by the Society for Healthcare Epidemiology of America (SHEA) and Infectious Diseases Society of America (IDSA), and include antimicrobial stewardship to reduce the risk of C. difficile infection, including various disinfection and barrier methods to prevent the patient from being exposed to C. difficile . Diluted household bleach is recommended for environmental contamination in an outbreak setting or one of hyperendemicity, and soap and water for hand hygiene, rather than alcohol hand hygiene products.
Effective infection control measures have been shown to reduce CDI incidence compared to rates prior to implementation, either when used by themselves or as part of a multipronged intervention strategy that includes improved antibiotic stewardship [23,28]. As reported in the 2005 study by Loo et al. , implementation of major infection control measures for a multiinstitutional outbreak of CDI in June 2004 was associated with a decrease in CDI incidence, from 22.5 cases per 1000 admissions for January to June 2004, to 13.6 cases per 1000 for July to December 2004, and 12.4 cases per 1000 for January to June 2005. A recent systematic review and meta-analysis on the prevention of endemic healthcare-associated C. difficile infection suggested that antimicrobial stewardship, glove use, hand hygiene, and disposable thermometers should be routinely used for the prevention of CDI .
Other potential approaches for prevention include stopping unnecessary antibiotic use (improved antibiotic stewardship), as well as a number of nonantibiotic strategies, such as colonizing patients with nontoxigenic strains of C. difficile to boost host immunity prior to subsequent exposure to a toxin-producing strain; restoration of the normal flora or use of ‘probiotics’ to confer health benefits to the host; and use of vaccines to boost host immunity [30–33]. A recent systematic review and meta-analysis on the prevention of endemic healthcare-associated C. difficile infection suggested that environmental cleaning and probiotics merit further study in the prevention of CDI . However, their analysis is problematic as they include both primary and secondary CDI prevention trials using probiotics and make conclusions about the utility of probiotics in general. Not included in that analysis was a recent study that showed a proprietary Lactobacillus probiotic significantly reduced the occurrence of antibiotic-associated diarrhea .
Studies have shown that improved antibiotic stewardship or prescribing practices, including reducing the use of broad-spectrum antibiotics most strongly associated with CDI risk, reduces the incidence of CDI [34–40]. In addition, a recent review by McFarland  of probiotics for prevention of CDI indicated that most studies to date seem to support such an approach (Fig. 2), but a variety of different probiotic strains and concentrations have been studied (making comparison across studies difficult), and the studies have typically been small. Quality control of probiotic products is also generally poor in the current environment. Hence, although promising, no firm conclusions or recommendations can be made at this time about the use of probiotics for CDI prevention. Further research is warranted. Other approaches for CDI prevention mentioned above are also currently experimental and also warrant further study.
Antibiotic use and resistance
Another form of collateral damage associated with antibiotic use is the selection for or induction of antibiotic-resistant pathogens. Studies have pointed to a linkage between antibiotic use and antibiotic resistance [42,43]. Antibiotic resistance is important because infections caused by these pathogens are associated with increased morbidity and mortality, longer hospital stays, and increased healthcare costs . Moreover, if further antibiotic resistance is not curtailed, clinicians may face multidrug-resistant (MDR) pathogens or ‘superbugs’ that are resistant to all currently available agents, such as the newly described NDM-1 Escherichia coli . Given that few new agents with novel mechanisms of action are becoming available and only few are expected in the foreseeable future [46,47], such a development could have catastrophic consequences.
Although some organisms are inherently resistant to particular antibiotics, most resistant strains emerge due to the selective pressure created through antibiotic use. Mechanisms of acquired resistance typically involve gene mutations resulting in altered porin channels limiting penetration of the antibiotic into the bacterial cell (where the target site exists); production of enzymes that inactivate the antibiotic; alteration of the target site for the antibiotic; or formation of efflux pumps extruding the antibiotic from the cell interior before it can act on its target . Characteristics conferring resistance can be acquired due to spontaneous gene mutations that enable the bacterium to survive the bacteriostatic or bacteriolytic effects of an antibiotic . The bacterium possessing this characteristic then becomes ‘selected,’ in the Darwinian sense of ‘survival of the fittest,’ and when it reproduces/divides, it passes on this ‘survival gene’ to its progeny. Bacteria can also become resistant by acquiring new genetic material from other innately resistant bacteria. This ‘horizontal transfer’ may occur within a given species or between different species. The existence of mobile integrons with gene cassettes containing genes conferring resistance to multiple drug classes increases the potential for dissemination of multidrug resistance within and among divergent bacterial pathogens [49,50].
How resistance is determined: susceptibility breakpoints
The clinical susceptibility or resistance of bacteria to an antibiotic can be described along a continuum, with greater or lesser certainty of therapeutic success when using that antibiotic [51–53]. According to the European Union Committee on Antimicrobial Susceptibility Testing (EUCAST), a microorganism is defined as susceptible by in-vitro study when achieving a minimum inhibitory concentration (MIC) level (μg/ml) that is associated with a high likelihood of therapeutic success . This means that normal or standard antibiotic dosing should be associated with high likelihood of success when dealing with a susceptible organism, that is, one that is not intrinsically resistant and that has not acquired characteristics reducing its susceptibility to the antibiotic. A microorganism is defined as having an intermediate MIC when associated with a drug level (μg/ml) that is associated with an uncertain therapeutic effect . Infections caused by these bacteria might be successfully treated when a higher dose of the drug is used or when high levels of the drug are concentrated in the infected body site. A microorganism is defined as resistant when the MIC (μg/ml) concentration of the antimicrobial is associated with a high likelihood of therapeutic failure . Resistant bacteria are typically unaffected by the drug, regardless of the dosage or administration schedule used.
Changing breakpoints: Clinical and Laboratory Standards Institute 2010
A MIC breakpoint is a drug concentration that discriminates between susceptible, intermediate, and resistant bacterial strains [51,53]. In the United States, the Clinical and Laboratory Standards Institute (CLSI) is the primary organization that develops antibiotic and organism-specific susceptibility breakpoints, as well as provides recommendations on susceptibility testing and how to apply susceptibility breakpoints [51,53]. The CLSI document (M100) providing breakpoints is updated regularly. The US Food and Drug Administration (FDA) also sets breakpoints for an antibiotic at the time of its approval, but these criteria are not updated on a regular basis. The EUCAST in Europe provides a similar function to the CLSI in the United States.
The breakpoints described by the CLSI, FDA, and EUCAST may be discordant, complicating clinical decision-making. Moreover, because the CLSI (for example) periodically updates its guidelines, breakpoints for susceptible, intermediate, and resistant organisms are often a ‘moving target’. In the latest M100 document from CLSI (M100-S20), the breakpoints were revised for several cephalosporins (cefazolin, cefotaxime, ceftazidime, ceftizoxime, and ceftriaxone) and the monobactam aztreonam for Enterobacteriaceae, in large part due to the increasing prevalence of extended-spectrum β-lactamase (ESBL)-producing Enterobacteriaceae. In addition, there is consideration in changing the breakpoints for Pseudomonas aeruginosa. To the clinician, this might mean that isolates that would be susceptible by today's breakpoints might be resistant by the criteria of the new breakpoint, without any change in MIC level or acquisition of any new plasmids or resistance genes. To avoid confusion, it is important that clinicians remain knowledgeable about the latest changes in susceptibility breakpoints and the data upon which they are based. Effective antimicrobial stewardship should include educating all parties involved with patient care (and communication between the different parties) as to changes in susceptibility breakpoints and proper interpretation of local laboratory data. For example, local antibiograms based on the prior year may now be ‘dated,’ that is, based on prior susceptibility breakpoints. In that scenario, clinical decision needs to be based on the actual MIC (or other relevant ‘hard’ measure), rather than the ‘S,’ ‘I,’ or ‘R’ that may appear on the laboratory report.
Bacterial resistance of particular concern includes methicillin-resistant S. aureus (MRSA), vancomycin-resistant enterococci (VRE), P. aeruginosa, Acinetobacter species (e.g., Acinetobacter baumannii), Klebsiella pneumoniae, and E. coli [54,55]. One concern is the potential for selection of these drug-resistant organisms or the development of colonization and infection with these MDR species due to antibiotic use.
Care in choosing antibiotic therapy requires attention to local susceptibility patterns and the potential selection for antibiotic-resistant species. Before designating a certain class of drugs as the hospital ‘workhorse,’ there needs to be consideration of local resistance patterns and the possibility of collateral damage associated with the drug class' preferential use. In fact, no antibiotic therapy should be considered an institutional workhorse without consideration of local susceptibility/resistance patterns, the potential for emergence of resistance, and all other possible adverse consequences.
Examples of selective pressure affecting antimicrobial resistance
Macrolides are frequently used to treat penicillin-allergic patients with Streptococcus pyogenes infections. During the 1990s, there was a resurgence of severe forms of S. pyogenes infections worldwide . In Finland, a marked increase in S. pyogenes erythromycin resistance occurred, rising from 5% in 1988 to 13% in 1990 . This rise in resistance was studied by examining local health records and was ‘significantly correlated’ with the levels of erythromycin use . Due to wide publicity of this problem, Finish physicians were ‘educated about the use of alternative drugs’, which resulted in a marked reduction in erythromycin use and subsequent ‘significant decline in the frequency of erythromycin resistance’ among S. pyogenes isolates from throat and other clinical samples from 1993 to 1996 .
Another example was reported by Rahal et al.  who noted a rampant ceftazidime-resistant Klebsiella outbreak and implemented a hospital-wide formulary restriction on the use of cephalosporin. Within 1 year, there was an 80.1% reduction in cephalosporin use, which was accompanied by a 44% reduction in the incidence of ceftazidime-resistant Klebsiella (P < 0.01). Unfortunately, this restriction led to an increased use of imipenem and a resultant increase in imipenem-resistant P. aeruginosa isolations (P < 0.01). These documented examples highlight and reinforce our intuitive feelings that there is a direct effect of antimicrobial usage and clinical resistance. One can consider extrapolating these examples to other clinical infections and other drugs.
A review of collateral damage by Paterson  highlighted the linkage between use of third-generation cephalosporins and subsequent infection with VRE, ESBL-producing K. pneumoniae, β-lactam-resistant Acinetobacter species, and CDI. Use of fluoroquinolones has also been linked to the development of MRSA, fluoroquinolone-resistant P. aeruginosa, and – more recently – CDI . On the basis of these findings, the author concluded continued use of third-generation cephalosporins and fluoroquinolones in hospitals as workhorse antibiotic therapy should be monitored and discouraged.
The remainder of this article examines the potential for selection or development of resistance when administering antibiotic therapy, in the specific case of treatment of intraabdominal infections and focuses on the effect of expanded use of a carbapenem, ertapenem, and also the possible impact of secondary antibiotic use on the emergence of resistance.
Antibiotic treatment of intraabdominal infections selects for resistance
A recent preclinical study using mice examined the differential effects of subcutaneous treatment with imipenem/cilastatin, ertapenem, piperacillin/tazobactam, or ceftriaxone on persistence of intestinal colonization by two ESBL-producing strains of K. pneumoniae (P62 and P10045) . Ceftriaxone, to which the ESBL-producing strains were resistant, promoted intestinal overgrowth in the experimental model (compared with saline controls), whereas ertapenem suppressed colonization, and imipenem/cilastatin neither promoted nor suppressed colonization. Although both strains were susceptible to ertapenem and imipenem/cilastatin, only ertapenem is excreted in large amounts into the gastrointestinal tract, which probably accounts for the differences in activity in this study. Interestingly, the impact of piperacillin/tazobactam on persistence of colonization was strain dependent. If the more susceptible P62 K. pneumoniae strain was involved, then piperacillin/tazobactam suppressed colonization, whereas if the resistant P10045 strain was isolated, then overgrowth occurred. Consequently, one cannot assume that all patients with ESBL isolates can be empirically treated with piperacillin/tazobactam and that knowing the MIC of a specific ESBL strain is important in the selection of an antimicrobial therapeutic agent.
Similar results were observed in the Optimizing Intra-Abdominal Surgery with Invanz studies (OASIS-I and OASIS-II), which were randomized, open-label, active-comparator clinical trials comparing ertapenem with piperacillin/tazobactam (OASIS-I) or ceftriaxone/metronidazole (OASIS-II) [62,63] for the treatment of complicated intraabdominal infections in patients requiring surgery. In the overall OASIS-I analysis, similar percentages of patients experienced a favorable clinical response with ertapenem (94%) or piperacillin/tazobactam (90%) . However, subanalyses for both studies revealed different rates of bowel colonization with resistant Enterobacteriaceae for patients treated with ertapenem, piperacillin/tazobactam, or ceftriaxone/metronidazole .
Enterobacteriaceae-resistant species were recovered from 12.2% of piperacillin/tazobactam recipients at the end of therapy (vs. 0.6% at baseline; P < 0.001) in OASIS-I, and from 17.1% of ceftriaxone/metronidazole recipients at therapy end (vs. 2.6% at baseline, P < 0.001) in OASIS-II. Conversely, resistant Enterobacteriaceae were recovered from only 0.6% and 0.5% of ertapenem recipients in OASIS-I and OASIS-II, respectively, significantly less than the percentage with the respective comparator (P < 0.001 for each study). These data suggest that emergence of resistant Enterobacteriaceae is less likely with ertapenem than with piperacillin/tazobactam or ceftriaxone/metronidazole therapy in patients with complicated intraabdominal infections who are awaiting surgery. With respect to colonization with an ESBL-producing E. coli or Klebsiella species, ceftriaxone/metronidazole treatment was associated with a significant increase in rate compared with baseline (9.3 vs. 2.1%, P < 0.001) in OASIS-II, whereas no ertapenem recipient was colonized with an ESBL producer at the end of therapy in either study. Ertapenem is active against ESBL-producing bacteria, whereas ceftriaxone/metronidazole is not, suggesting the latter treatment selected for ESBL producers. Imipenem-resistant P. aeruginosa were uncommon in all treatment groups.
Recent reports have documented the occurrence of ertapenem-resistant Enterobacteriaceae [64–68]. Lartigue et al.  reported a case of a 50-year-old compromised woman who was treated with imipenem for 10 days and subsequently had an ertapenem-resistant E. coli isolated from a peritoneal lavage culture. This isolate had both a CTX-M β-lactamase and a deficiency in porin OmpC. Garcia-Fernandez et al.  reported the isolation of two carbapenem-resistant, ESBL-producing strains of K. pneumoniae in a Roman hospital with an additional lack of OmpK35 and OmpK36 porins and suggested similar populations may be ‘hidden within populations of susceptible bacteria’. Yan et al.  studied 9722 E. coli isolates from a Taiwanese hospital from 1999 to 2007 and found ertapenem-resistant isolates, many genetically unrelated, in 66 patients. As ertapenem was not introduced into that hospital in 2007, the authors suggested the increased resistance from 0.1% in 1999 to 1.7% in 2007 was due to the use of antibiotics other than ertapenem. Still, the authors voiced concern that the selective pressure of increased ertapenem use might accelerate the process. As suggested by Doumith et al. , ertapenem resistance requires the combination of both an ESBL-producing pathogen and a change in permeability caused by a loss of porin channels, which is relatively uncommon and which did not spread nationally in the United Kingdom. Efflux pumps were not implicated in the development of ertapenem resistance.
A retrospective subanalysis comparing bowel colonization with VRE after ertapenem, piperacillin/tazobactam, or ceftriaxone/metronidazole treatment in OASIS-I and OASIS-II demonstrated generally low and comparable rates in all treatment groups, with acquisition rates at the end of therapy or 2 weeks poststudy compared with baseline ranging from 0 to 3.5% . These data suggest there is minimal risk of colonization with VRE after treatment with any of these agents. Nonetheless, a small number of patients did acquire VRE, and hospital transmission is not only related to antibiotic usage and low hand hygiene compliance, but may be complicated by environmental colonization on fomite surfaces, thus facilitating spread. This highlights not only the importance of attention to hand hygiene and infection control practices in the hospital environment, but also the need for more effective environmental cleaning agents and practices. Resistant fecal pathogens are frequently transmitted within the hospital on the hands of personnel who have failed to use proper precautions.
The potential for disruption of indigenous gut flora and colonization with resistant bacteria probably occurs with any antibiotic that is excreted in significant concentrations into the gastrointestinal tract. A 2006 study by Nord et al.  evaluated the effect of daily administration of the newer antibiotic tigecycline (10-day treatment) on normal oropharyngeal and intestinal microflora in 13 healthy adult patients. Substantial fecal but minimal salivary concentrations of the drug were observed on day 8. Correspondingly, tigecycline had a minor effect on the oropharyngeal microflora, but a significant effect on intestinal microflora. In particular, there was a significant reduction in the numbers of enterococci and E. coli in the intestinal microflora at day 8 (P < 0.05), together with an increase in other enterobacteria (and yeasts). There was also a significant reduction in lactobacilli and bifidobacteria (P < 0.05), two common Gram-positive inhabitants of the human gastrointestinal tract. In addition, two K. pneumoniae strains and five Enterobacter cloacae strains resistant to tigecycline emerged on day 8 of treatment. Tigecycline had no impact on Bacteroides, and no C. difficile strains were isolated.
Effects of expanded clinical use of ertapenem
Rising rates of infection caused by ESBL-producing bacteria have necessitated the increased usage of carbapenems, given that these agents continue to exhibit activity against ESBL-producing organisms. Hence, carbapenems are now considered the drugs of choice for serious ESBL infections [71,72]. Unlike antipseudomonal (group 2) carbapenems, ertapenem is a nonpseudomonal (group 1) carbapenem with activity against ESBL-producing bacteria, but not P. aeruginosa and nonfermenters . Ertapenem use has increased under circumstances wherein coverage of ESBL-producing bacteria, but not P. aeruginosa, is desired. However, there has been concern that ertapenem, as a nonpseudomonal carbapenem, will cause the emergence of P. aeruginosa strains resistant to the pseudomonal carbapenems (imipenem, meropenem, and doripenem).
A number of recent studies have examined the effect of ertapenem use on the hospital ecology, particularly the effect on the susceptibilities of P. aeruginosa to pseudomonal carbapenems and have failed to observe any decrease in susceptibility related to ertapenem use [2,74–78]. For example, a retrospective analysis by myself and colleagues of hospital susceptibility data for June 2002 to December 2005 from a 344-bed community teaching hospital in Santa Monica, California, failed to show any decrease in the susceptibility of P. aeruginosa or other gram-negative bacteria to imipenem or several other antibiotics after ertapenem was introduced to the hospital formulary . In fact, there was evidence for improved susceptibility of P. aeruginosa to imipenem coincident with ertapenem use.
Ertapenem was added to the formulary after an audit showed surgeons were commonly prescribing ampicillin/sulbactam, and 40% of hospital E. coli were resistant to ampicillin/sulbactam . Following a 9-month introductory period, an autosubstitution policy was instituted whereby clinicians ordering ampicillin/sulbactam for their patients received ertapenem instead. Ertapenem use rose steadily from 0 defined daily dosages per 1000 patient days (DDD) at baseline, to 35 DDD during the introductory period, and a general range of 36–48 DDD after institution of the autosubstitution policy. Coincidentally, ampicillin/sulbactam use dramatically decreased, and a less dramatic decrease in imipenem usage was observed. Of particular note, P. aeruginosa susceptibility to imipenem improved from 67% just prior to adding ertapenem to the formulary, to a median susceptibility (interquartile range) of 88% (82–95%) after institution of the autosubstitution policy. Moreover, subsequent analyses demonstrated an increase of 0.38% for every unit decrease in the monthly DDD of imipenem (P = 0.008). These data suggest P. aeruginosa susceptibility to imipenem improved due to decreased selective pressure consequent to decreased use of the antipseudomonal agent.
Additionally, P. aeruginosa susceptibilities for levofloxacin, cefepime, and piperacillin/tazobactam tended to increase during the study period (Fig. 3) . Increased ertapenem usage had either minor or no changes in the susceptibilities of E. coli, Proteus mirabilis, Klebsiella species, and Enterobacter cloacae to the various antibiotic agents studied, including imipenem, although there was a slight decrease in E. coli susceptibility to levofloxacin. The prevalence of ESBL-producing E. coli and Klebsiella species remained generally constant following ertapenem addition to the hospital formulary and autosubstitution. The study has a number of limitations, including the limitation of group-level studies, the issue of using proportion vs. incidence as the main outcome, and the limitation of a single-institution study.
A comprehensive multicenter assessment of the impact of ertapenem use on P. aeruginosa susceptibility to antipseudomonal carbapenems was performed in the recent Ertapenem Utilization and Resistance Emergence among Collateral Antimicrobials (EURECA) study . Using sophisticated methods and data analysis, and a 6-year evaluation period, EURECA examined the impact of ertapenem adoption on the emergence of resistance among P. aeruginosa and other important nosocomial bacterial pathogens at 25 anonymous teaching and community hospitals, geographically distributed throughout the United States. The 6-year evaluation included the 3 years prior to and the 3 years after adoption of ertapenem at each of the study hospitals.
No significant change in P. aeruginosa susceptibility to antipseudomonal carbapenems was detected in the 25 hospitals during the 6-year study period, although the ertapenem use density ratio increased during this period, whereas that of other carbapenems remained essentially constant (Fig. 4) . Hence, no relationship was established between rate of ertapenem use and change in P. aeruginosa carbapenem susceptibility over the 6-year evaluation period. These results are similar to those reported in a recent published report by Pakyz et al.  and generally consistent with the findings from the retrospective review by Goldstein et al.  just discussed.
Impact of secondary antibiotic use on resistance
As mentioned earlier, studies have pointed to an association between antibiotic use and antibiotic resistance [42,43]. This is easy to conceive when focusing on the development of resistance to the original drug administered or to other members of the same drug class. However, misuse or overuse of a particular drug can also promote certain pathogens to develop resistance to drugs from unrelated classes, that is, cross-resistance or co-resistance. That is, not only primary antibiotic use, but also secondary antibiotic use may be associated with collateral damage in terms of emergent resistance.
This latter concept was highlighted in a recent study by Bosso et al.  examining the impact of imipenem, ceftriaxone, cefepime, piperacillin/tazobactam, or ciprofloxacin on susceptibility rates of P. aeruginosa, E. coli, K. pneumoniae, and E. cloacae to these various antibiotics. Interestingly, trends for decreases in organism susceptibility over time were not statistically associated with the primary drug, for example, for susceptibility to imipenem with imipenem use. But trends for decreased susceptibility were associated with secondary drug use. In particular, piperacillin/tazobactam use was associated with decreased susceptibility of E. coli to cefepime. Other studies have also reported cases in which susceptibility of a pathogen(s) to a drug of one particular class was altered (either increased or decreased) by prior use of an antibiotic from another class.
Results such as these are important because they suggest a resistance problem may not be adequately addressed by simply altering the utilization of the primary antibiotic. The mechanisms accounting for development of co-resistance between members of different antibiotic classes are often unclear, but might involve development of multidrug efflux pumps  or the involvement of mobile integrons with gene cassettes conferring multidrug resistance [49,50].
Inappropriate or overuse of antibiotics – including selection that does not give sufficient attention to the local susceptibility or resistance patterns to the agent selected – can have adverse consequences that are harmful to the patient, the hospital ecology, and society. These adverse consequences have been described as ‘collateral damage’ and may include, for example, ‘superinfection’ secondary to disruption of bacterial flora of the gastrointestinal tract, selection or induction of drug-resistant organisms, and unwanted colonization or infection with MDR organisms. In addition, as the report by Rahal et al. illustrated , restriction of one class of antibiotics (because of concern of rising resistance to that drug class) can result in increased resistance to drugs replacing the first class.
When deciding whether to administer antibiotic therapy, and selecting among available agents, attention should be paid to the potential collateral damage associated with antibiotic use, as well as potential benefits. With better antimicrobial stewardship, the collateral benefits of antibiotic therapy can be enhanced, while minimizing the collateral damage.
This supplement was supported through an educational grant from Merck & Co., Inc.
The author declares no conflicts of interest.
1 Paterson DL. ‘Collateral damage’ from cephalosporin or quinolone antibiotic therapy. Clin Infect Dis 2004; 38(Suppl 4):S341–S345.
2 Weber DJ. Collateral damage and what the future might hold. The need to balance prudent antibiotic utilization and stewardship with effective patient management. Int J Infect Dis 2006; 10:S17–S24.
3 Riddle DJ, Dubberke ER. Clostridium difficile infection in the intensive care unit. Infect Dis Clin North Am 2009; 23:727–743.
4 Sunenshine RH, McDonald LC. Clostridium difficile-associated disease: new challenges from an established pathogen. Cleve Clin J Med 2006; 73:187–197.
5 Vaishnavi C. Established and potential risk factors for Clostridum difficile infection. Indian J Med Microbiol 2009; 27:289–300.
6 Zilberberg MD, Shorr AF, Kollef MH. Increase in adult Clostridium difficile-related hospitalizations and case-fatality rate, United States, 2000–2005. Emerg Infect Dis 2008; 14:929–931.
7 Zilberberg MD, Shorr AF, Kollef MH. Increase in Clostridium difficile-related hospitalizations among infants in the United States, 2000–2005. Pediatr Infect Dis J 2008; 27:1111–1113.
8 Redelings MD, Sorvillo F, Mascola L. Increase in Clostridium difficile-related mortality rates, United States, 1999–2004. Emerg Infect Dis 2007; 13:1417–1419.
10 Kyne L, Hamel MB, Polavaram R, Kelly CP. Healthcare costs and mortality associated with nosocomial diarrhea due to Clostridium difficile. Clin Infect Dis 2002; 34:346–353.
11 Ricciardi R, Rothenberger DA, Madoff RD, Baxter NN. Increasing prevalence and severity of Clostridium difficile colitis in hospitalized patients in the United States. Arch Surg 2007; 142:624–631, discussion 631.
12 Marra AR, Edmond MB, Wenzel RP, Bearman GM. Hospital-acquired Clostridium difficile-associated disease in the intensive care unit setting: epidemiology, clinical course and outcome. BMC Infect Dis 2007; 7:42.
13 Barbut F, Petit JC. Epidemiology of Clostridium difficile-associated infections. Clin Microbiol Infect 2001; 7:405–410.
14 Bignardi GE. Risk factors for Clostridium difficile infection. J Hosp Infect 1998; 40:1–15.
15 Dubberke ER, Reske KA, Yan Y, et al
. Clostridium difficile-associated disease in a setting of endemicity: identification of novel risk factors. Clin Infect Dis 2007; 45:1543–1549.
16 Kyne L, Warny M, Qamar A, Kelly CP. Asymptomatic carriage of Clostridium difficile and serum levels of IgG antibody against toxin A. N Engl J Med 2000; 342:390–397.
17 Frank DN, Pace NR. Gastrointestinal microbiology enters the metagenomics era. Curr Opin Gastroenterol 2008; 24:4–10.
18 Federle MJ, Bassler BL. Interspecies communication in bacteria. J Clin Invest 2003; 112:1291–1299.
19 Williams P, Winzer K, Chan WC, Camara M. Look who's talking: communication and quorum sensing in the bacterial world. Philos Trans R Soc Lond B Biol Sci 2007; 362:1119–1134.
20 Shank EA, Kolter R. New developments in microbial interspecies signaling. Curr Opin Microbiol 2009; 12:205–214.
21 Edmiston CE, Walker AP. Microbiology of intraabdominal infections. Infect Dis Clin Pract 1996; 5:S15–S19.
22 Finegold SM. Anaerobic bacteria in human disease. New York: Academic Press Inc; 1977.
23 Loo VG, Poirier L, Miller MA, et al
. A predominantly clonal multiinstitutional outbreak of Clostridium difficile-associated diarrhea with high morbidity and mortality. N Engl J Med 2005; 353:2442–2449.
24 Gaynes R, Rimland D, Killum E, et al
. Outbreak of Clostridium difficile infection in a long-term care facility: association with gatifloxacin use. Clin Infect Dis 2004; 38:640–645.
25 McDonald LC, Killgore GE, Thompson A, et al
. An epidemic, toxin gene-variant strain of Clostridium difficile. N Engl J Med 2005; 353:2433–2441.
26 Gerding DN, Muto CA, Owens RC Jr. Measures to control and prevent Clostridium difficile infection. Clin Infect Dis 2008; 46(Suppl 1):S43–S49.
27 Cohen SH, Gerding DN, Johnson S, et al
. Clinical practice guidelines for Clostridium difficile infection in adults: 2010 update by the Society for Healthcare Epidemiology of America (SHEA) and the Infectious Diseases Society of America (IDSA). Infect Control Hosp Epidemiol 2010; 31:431–455.
28 Weiss K, Boisvert A, Chagnon M, et al
. Multipronged intervention strategy to control an outbreak of Clostridium difficile infection (CDI) and its impact on the rates of CDI from 2002 to 2007. Infect Control Hosp Epidemiol 2009; 30:156–162.
29 Hsu J, Abad C, Dinh M, Safdar N. Prevention of endemic healthcare-associated Clostridium difficile infection: reviewing the evidence. Am J Gastroenterol 2010; 105:2327–2339.
30 Bauer MP, van Dissel JT. Alternative strategies for Clostridium difficile infection. Int J Antimicrob Agents 2009; 33(Suppl 1):S51–S56.
31 Gao XW, Mubasher M, Fang CY, et al
. Dose-response efficacy of a proprietary probiotic formula of Lactobacillus acidophilus CL1285 and Lactobacillus casei LBC80R for antibiotic-associated diarrhea and Clostridium difficile-associated diarrhea prophylaxis in adult patients. Am J Gastroenterol 2010; 105:1636–1641.
32 Merrigan MM, Sambol SP, Johnson S, Gerding DN. New approach to the management of Clostridium difficile infection: colonisation with nontoxigenic C. difficile during daily ampicillin or ceftriaxone administration. Int J Antimicrob Agents 2009; 33(Suppl 1):S46–S50.
33 Trejo FM, Perez PF, De Antoni GL. Co-culture with potentially probiotic microorganisms antagonises virulence factors of Clostridium difficile in vitro. Antonie Van Leeuwenhoek 2010; 98:19–29.
34 Climo MW, Israel DS, Wong ES, et al
. Hospital-wide restriction of clindamycin: effect on the incidence of Clostridium difficile-associated diarrhea and cost. Ann Intern Med 1998; 128:989–995.
35 Davey P, Brown E, Fenelon L, et al.
Interventions to improve antibiotic prescribing practices for hospital inpatients. Cochrane Database Syst Rev 2005:CD003543.
36 Fowler S, Webber A, Cooper BS, et al
. Successful use of feedback to improve antibiotic prescribing and reduce Clostridium difficile infection: a controlled interrupted time series. J Antimicrob Chemother 2007; 59:990–995.
37 Khan R, Cheesbrough J. Impact of changes in antibiotic policy on Clostridium difficile-associated diarrhoea (CDAD) over a five-year period in a district general hospital. J Hosp Infect 2003; 54:104–108.
38 Ludlam H, Brown N, Sule O, et al
. An antibiotic policy associated with reduced risk of Clostridium difficile-associated diarrhoea. Age Ageing 1999; 28:578–580.
39 O'Connor KA, Kingston M, O'Donovan M, et al
. Antibiotic prescribing policy and Clostridium difficile diarrhoea. QJM 2004; 97:423–429.
40 Valiquette L, Cossette B, Garant MP, et al
. Impact of a reduction in the use of high-risk antibiotics on the course of an epidemic of Clostridium difficile-associated disease caused by the hypervirulent NAP1/027 strain. Clin Infect Dis 2007; 45(Suppl 2):S112–S121.
41 McFarland LV. Evidence-based review of probiotics for antibiotic-associated diarrhea and Clostridium difficile infections. Anaerobe 2009; 15:274–280.
42 Dellit TH, Owens RC, McGowan JE Jr, et al
. Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America guidelines for developing an institutional program to enhance antimicrobial stewardship. Clin Infect Dis 2007; 44:159–177.
43 Tacconelli E. Antimicrobial use: risk driver of multidrug resistant microorganisms in healthcare settings. Curr Opin Infect Dis 2009; 22:352–358.
44 Maragakis LL, Perencevich EN, Cosgrove SE. Clinical and economic burden of antimicrobial resistance. Expert Rev Anti Infect Ther 2008; 6:751–763.
45 Kumarasamy KK, Toleman MA, Walsh TR, et al
. Emergence of a new antibiotic resistance mechanism in India, Pakistan, and the UK: a molecular, biological, and epidemiological study. Lancet Infect Dis 2010; 10:597–602.
46 Bad bugs, no drugs: as antibiotic R&D stagnates, a public health crisis brews. Alexandria, VA: Infectious Diseases Society of America; 2004.
47 Spellberg B, Guidos R, Gilbert D, et al
. The epidemic of antibiotic-resistant infections: a call to action for the medical community from the Infectious Diseases Society of America. Clin Infect Dis 2008; 46:155–164.
48 Tenover FC. Mechanisms of antimicrobial resistance in bacteria. Am J Med 2006; 119:S3–S10, discussion S62–S70.
49 Bennett PM. Plasmid encoded antibiotic resistance: acquisition and transfer of antibiotic resistance genes in bacteria. Br J Pharmacol 2008; 153(Suppl 1):S347–S357.
50 Mazel D. Integrons: agents of bacterial evolution. Nat Rev Microbiol 2006; 4:608–620.
51 Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing; 20th informational supplement. CSLI document M100-S20. Wayne, PA: CLSI; 2010.
53 Turnidge J, Paterson DL. Setting and revising antibacterial susceptibility breakpoints. Clin Microbiol Rev 2007; 20:391–408, table of contents.
54 Rice LB. Federal funding for the study of antimicrobial resistance in nosocomial pathogens: no ESKAPE. J Infect Dis 2008; 197:1079–1081.
55 Rice LB. The clinical consequences of antimicrobial resistance. Curr Opin Microbiol 2009; 12:476–481.
56 Bronze MS, Dale JB. The reemergence of serious group A streptococcal infections and acute rheumatic fever. Am J Med Sci 1996; 311:41–54.
57 Seppala H, Nissinen A, Jarvinen H, et al
. Resistance to erythromycin in group A streptococci. N Engl J Med 1992; 326:292–297.
58 Seppala H, Klaukka T, Lehtonen R, et al
. Outpatient use of erythromycin: link to increased erythromycin resistance in group A streptococci. Clin Infect Dis 1995; 21:1378–1385.
59 Seppala H, Klaukka T, Vuopio-Varkila J, et al
. The effect of changes in the consumption of macrolide antibiotics on erythromycin resistance in group A streptococci in Finland. Finnish Study Group for Antimicrobial Resistance. N Engl J Med 1997; 337:441–446.
60 Rahal JJ, Urban C, Horn D, et al
. Class restriction of cephalosporin use to control total cephalosporin resistance in nosocomial Klebsiella. JAMA 1998; 280:1233–1237.
61 Pultz MJ, Donskey CJ. Effects of imipenem-cilastatin, ertapenem, piperacillin-tazobactam, and ceftriaxone treatments on persistence of intestinal colonization by extended-spectrum-beta-lactamase-producing Klebsiella pneumoniae strains in mice. Antimicrob Agents Chemother 2007; 51:3044–3045.
62 Dela Pena AS, Asperger W, Kockerling F, et al
. Efficacy and safety of ertapenem versus piperacillin-tazobactam for the treatment of intra-abdominal infections requiring surgical intervention. J Gastrointest Surg 2006; 10:567–574.
63 Dinubile MJ, Friedland I, Chan CY, et al
. Bowel colonization with resistant gram-negative bacilli after antimicrobial therapy of intra-abdominal infections: observations from two randomized comparative clinical trials of ertapenem therapy. Eur J Clin Microbiol Infect Dis 2005; 24:443–449.
64 Chen PL, Yan JJ, Wu CJ, et al.
Salvage therapy with tigecycline for recurrent infection caused by ertapenem-resistant extended-spectrum beta-lactamase-producing Klebsiella pneumoniae. Diagn Microbiol Infect Dis 2010; 68:312–314.
65 Doumith M, Ellington MJ, Livermore DM, Woodford N. Molecular mechanisms disrupting porin expression in ertapenem-resistant Klebsiella and Enterobacter spp. clinical isolates from the UK. J Antimicrob Chemother 2009; 63:659–667.
66 Garcia-Fernandez A, Miriagou V, Papagiannitsis CC, et al
. An ertapenem-resistant extended-spectrum-beta-lactamase-producing Klebsiella pneumoniae clone carries a novel OmpK36 porin variant. Antimicrob Agents Chemother 2010; 54:4178–4184.
67 Lartigue MF, Poirel L, Poyart C, et al
. Ertapenem resistance of Escherichia coli. Emerg Infect Dis 2007; 13:315–317.
68 Yan JJ, Wu JJ, Lee CC, et al.
Prevalence and characteristics of ertapenem-nonsusceptible Escherichia coli
in a Taiwanese university hospital, 1999 to 2007. Eur J Clin Microbiol Infect Dis 2010; 29:1417–1425.
69 DiNubile MJ, Friedland IR, Chan CY, et al
. Bowel colonization with vancomycin-resistant enterococci after antimicrobial therapy for intra-abdominal infections: observations from 2 randomized comparative clinical trials of ertapenem therapy. Diagn Microbiol Infect Dis 2007; 58:491–494.
70 Nord CE, Sillerstrom E, Wahlund E. Effect of tigecycline on normal oropharyngeal and intestinal microflora. Antimicrob Agents Chemother 2006; 50:3375–3380.
71 Falagas ME, Karageorgopoulos DE. Extended-spectrum beta-lactamase-producing organisms. J Hosp Infect 2009; 73:345–354.
72 Rodriguez-Bano J, Pascual A. Clinical significance of extended-spectrum beta-lactamases. Expert Rev Anti Infect Ther 2008; 6:671–683.
73 Keating GM, Perry CM. Ertapenem: a review of its use in the treatment of bacterial infections. Drugs 2005; 65:2151–2178.
74 Crank CW, Hota B, Segreti J, Stroger JHJ. Effect of ertapenem utilization on Pseudomonas aeruginosa
susceptibility to imipenem [abstract no.285]. 44th Annual Meeting of IDSA; 12–15 October 2006; Toronto.
75 Eagye KJ, Nicolau DP. Absence of association between use of ertapenem and change in antipseudomonal carbapenem susceptibility rates in 25 hospitals. Infect Control Hosp Epidemiol 2010; 31:485–490.
76 Goff DA, Mangino JE. Ertapenem: no effect on aerobic gram-negative susceptibilities to imipenem. J Infect 2008; 57:123–127.
77 Goldstein EJ, Citron DM, Peraino V, et al
. Introduction of ertapenem into a hospital formulary: effect on antimicrobial usage and improved in vitro susceptibility of Pseudomonas aeruginosa. Antimicrob Agents Chemother 2009; 53:5122–5126.
78 Lima AL, Oliveira PR, Paula AP, et al
. The impact of ertapenem use on the susceptibility of Pseudomonas aeruginosa to imipenem: a hospital case study. Infect Control Hosp Epidemiol 2009; 30:487–490.
79 Pakyz AL, Oinonen M, Polk RE. Relationship of carbapenem restriction in 22 university teaching hospitals to carbapenem use and carbapenem-resistant Pseudomonas aeruginosa. Antimicrob Agents Chemother 2009; 53:1983–1986.
80 Bosso JA, Mauldin PD, Salgado CD. The association between antibiotic use and resistance: the role of secondary antibiotics. Eur J Clin Microbiol Infect Dis 2010 [Epub ahead of print].
81 Aeschlimann JR. The role of multidrug efflux pumps in the antibiotic resistance of Pseudomonas aeruginosa and other gram-negative bacteria. Insights from the Society of Infectious Diseases Pharmacists. Pharmacotherapy 2003; 23:916–924.