Nosocomial Antibiotic Resistance in Multiple Gram-Negative Species: Experience at One Hospital with Squeezing the Resistance Balloon at Multiple Sites

Transcription

1 INVITED ARTICLE HEALTHCARE EPIDEMIOLOGY Robert A. Weinstein, Section Editor Nosocomial Antibiotic Resistance in Multiple Gram-Negative Species: Experience at One Hospital with Squeezing the Resistance Balloon at Multiple Sites James J. Rahal, Carl Urban, and Sorana Segal-Maurer Infectious Disease Section, New York Hospital Queens, Flushing, New York Increased use of antibiotics has led to the isolation of multidrug-resistant bacteria, especially in intensive care units and long-term care facilities. Resistance in specific gram-negative bacteria, including Klebsiella pneumoniae, Acinetobacter baumannii, and Pseudomonas aeruginosa, is of great concern, because a growing number of reports have documented mechanisms whereby these microorganisms have become resistant to all available antibacterial agents used in therapy. Reduction in the selection of these multidrug-resistant bacteria can be accomplished by a combination of several strategies. These include having an understanding of the genetics of both innate and acquired characteristics of bacteria; knowing resistance potentials for specific antibacterials; monitoring resistance trends in bacteria designated as problematic organisms within a particular institution on a routine basis; modifying antibiotic formularies when and where needed; creating institutional education programs; and enforcing strict infection-control practices. Strategies appropriate for primary prevention of nosocomial resistance may differ from those required for control of existing epidemic or endemic resistance. Resistance of bacteria to antibiotics first occurred within a decade after the onset of the antibiotic era. Only recently, however, has this threat received intense public scrutiny, with the United States Congress, the Institute of Medicine (Washington, DC), and the Centers for Disease Control and Prevention (Atlanta) designating it an emerging infectious disease problem [1]. Such recent attention is a consequence of bacterial evolution finally surpassing the human capacity to create new antimicrobial agents. Several bacterial species have now become resistant to all commonly used agents, primarily in the hospital setting. However, this ominous development is not universal. In fact, an increasingly wide diversity exists between the levels of antibiotic resistance in large urban medical centers and the levels in hospitals of small to moderate size that serve less densely populated areas. The reasons for this variability are not Received 7 September 2001; revised 30 October 2001; electronically published 7 January Financial support: BMA Medical Foundation, Hugaton Foundation, and the Beatrice Snyder Foundation. Reprints or correspondence: Dr. James J. Rahal, New York Hospital Queens, Main St., Flushing, NY (jjr9002@nyp.org). Clinical Infectious Diseases 2002; 34: by the Infectious Diseases Society of America. All rights reserved /2002/ $03.00 entirely clear and are probably multidimensional. Centers where bacteria have greater levels of resistance are more often teaching institutions that provide tertiary care to a population at greater risk of serious disease. Such populations are admitted more frequently from long-term care facilities, have limited preventive care, or seek medical attention at a later date. These factors lead to a longer length of hospital stay and more frequent medical and surgical complications during hospitalization, as well as increased antibiotic use. Large, active emergency medicine, trauma, transplant, or oncology facilities also contribute to an environment in which more intense antibiotic use is required and antibiotic resistance is likely to evolve. These observations are ours and are not easily documented by controlled studies of risk factors in different institutional settings. However, several studies conducted within individual institutions have demonstrated that increased use of broad-spectrum antibiotics, intubation of the respiratory, gastrointestinal, or urinary tracts, and intravascular catheterization are significant predisposing factors for the development of antibiotic resistance [2, 3]. If one assumes that the aforementioned institutional characteristics are, in some manner, related to the evolution of antibiotic resistance, it follows that potential approaches to HEALTHCARE EPIDEMIOLOGY CID 2002:34 (15 February) 499

2 the problem should vary in each institution, depending on the presence of predisposing risk factors for antibiotic resistance. Centers with few resistance risk factors and little or no experience with multidrug-resistant bacteria may need only a limited approach to prevention of drug resistance. This might include intermittent microbiologic surveillance for plasmid-mediated extended spectrum b-lactamase producing Enterobacteriaceae; vancomycin-resistant Enterococcus species; derepressed mutants among Enterobacter, Serratia, and Acinetobacter species; and carbapenem-resistant Pseudomonas aeruginosa. A more intense focus on methicillin-resistant Staphylococcus aureus may be appropriate in such settings, because this resistance mechanism is more widely distributed, as noted by the Centers for Disease Control and Prevention s Project ICARE [4]. Additional preventive measures would include education regarding potential resistance among gramnegative pathogens and enterococci, appropriate pharmacokinetic and pharmacodynamic methods, and provisions of infectious disease consultation or computer assistance to antibiotic management. Under these circumstances, an open formulary could prevent the dominant use of a single class of antibiotics and emergence of resistance to that class, a phenomenon labelled squeezing the resistance balloon by Burke [5]. In contrast to the situation described above, resistance of gram-negative bacilli to late-generation b-lactam agents emerged in many urban institutions in the late 1980s and the 1990s, often associated with a 40% 50% prevalence of methicillin resistance among isolates of S. aureus and variable vancomycin resistance in enterococci [6]. This has occurred at our center, New York Hospital Queens, a 500-bed acute care facility situated in a populous working-class urban community. Here, many patients are admitted from long-term care facilities or from an active emergency facility and trauma center. In 1988, the Infectious Disease Section was established, with its director serving as the chair of the Infection Control Committee and as the hospital epidemiologist. At that time, Acinetobacter species had become a prevalent pathogen in both medical and surgical intensive care units. It is frequently resistant to all agents except ceftazidime, imipenem, and aminoglycosides, which resulted in rapidly increasing ceftazidime use under an open formulary. In 1990, ceftazidime-resistant Klebsiella species were noted, and their origin traced back to 1989 [7, 8]. Surveillance of all ceftazidime-resistant gram-negative bacilli became a major objective of our infection control program. Duplicate isolates recovered from the same patient were eliminated, and all colonized or infected patients were placed in contact isolation. Restriction of late-generation cephalosporin use to approval by the Infectious Disease Section reduced hospitalwide ceftazidime resistance among Klebsiella species from 33% in 1989 to 13% in However, a concomitant outbreak of infection with ceftazidime- and imipenem-resistant Acinetobacter species occurred despite ongoing restriction of use of imipenem [9]. This outbreak was primarily clonal in origin and was amenable to strict infection-control procedures. Thus, restriction of both late-generation cephalosporins and imipenem, in combination with molecular epidemiology, environmental decontamination, contact precautions, and patient/staff cohorting, eliminated imipenem-resistant Acinetobacter species and reduced ceftazidime-resistant Klebsiella species by 61% in Despite implementation of the aforementioned measures, plasmid-mediated ceftazidime resistance due to extended-spectrum b-lactamase (ESBL) production by Klebsiella species enjoyed a resurgence from 1993 to 1995, when 37% of all isolates were again found to be resistant to ceftazidime. This was associated with extension of ESBL-mediated resistance in Klebsiella species to cephamycins and, in selected strains, to imipenem [10, 11]. Further study of these isolates identified a new ampc-type plasmid-mediated ESBL, designated ACT-1, with greatest activity against cephamycins and lesser activity against ceftazidime and imipenem [10]. Resistance to carbapenems was mediated by phenotypic absence of an outer-membrane porin and decreased permeability to this agent, allowing its more efficient degradation by ACT-1. Antibiotic surveillance revealed a marked increase in cefotetan use, primarily for prophylaxis and therapy of surgical and obstetric infections. This pattern had provided the selective pressure required for emergence of ACT-1 and eventual resistance of certain Klebsiella isolates to all antibiotics except polymyxin [11]. Our necessary response to this previously unidentified combination of resistance mechanisms in Klebsiella species was to withdraw the selective pressure of both cephalosporins and cephamycins by class restriction [12]. This required implementation of an intensive educational program directed at house staff, attending physicians, those in private practice, nursing staff, and other health care personnel. Administrative cooperation was essential, including Pharmacy and Therapeutics Committee antibiotic guidelines, which supported cephalosporin class restriction (with a few excepted indications, such as bacterial meningitis). Alternative antibiotics available without restriction for therapy of gram-negative infection included piperacillin-tazobactam, ampicillin-sulbactam, trimethoprim-sulfamethoxazole, doxycycline, quinolones, and aminoglycosides. Imipenem restriction was implemented after 72 h of use in intensive care units during 1996 and after 24 h subsequently. Our initial report described the results of this program after implementation in 1996, as compared with 1995 [12]. Duplicate isolates from the same patient were eliminated throughout the study. Cephalosporincephamycin use was reduced by 80%, with reduction in ceftazidime-resistant Klebsiella species by 44% throughout the hospital and by 87% in the surgical intensive care unit. In 500 CID 2002:34 (15 February) HEALTHCARE EPIDEMIOLOGY

3 addition, ESBL-producing Klebsiella species with resistance to cephamycins and imipenem were eliminated. We anticipated that restriction of cephalosporin use might increase imipenem use despite required approval of this agent by the infectious disease section after 72 h. Thus, patient-specific prescribing of imipenem (rather than pharmacy purchases or dispensing) was monitored, and a 140% increase in imipenem use was noted in 1996, with a 69% increase in imipenem resistance among P. aeruginosa. However, 85% of these imipenem-resistant isolates remained susceptible to ceftazidime, and a larger proportion of all P. aeruginosa retained such susceptibility. The increase in imipenem resistance noted in P. aeruginosa in 1996, as compared with 1995, remained constant through 2000 at a level similar to that reported by National Nosocomial Infection Surveillance [6]. This has occurred despite reinstitution of restriction of imipenem use after 24 h, rather than 72 h, and is likely the result of persistent dissemination of 1 or a few mutant clones [13]. This problem is best approached by stringent infection-control measures and use of alternative agents for therapy of P. aeruginosa infections. Selective imipenem resistance among P. aeruginosa is the result of the emergence of relatively common mutants with phenotypic D2 porin deletion [14]. The chromosomal location of this resistance mechanism is consistent with a clonal epidemiology and potential limitation of dissemination by standard infectioncontrol techniques. This mechanism contrasts with the propensity of ESBL-producing ceftazidime-resistant Klebsiella species to become resistant to all available antibiotics by a combination of plasmid-mediated and chromosomally located resistance mechanisms, yielding a complex polyclonal or clonal population. Such differences in resistance potential, resistance mechanisms, molecular epidemiology, and infection-control efficacy must be considered in the overall management of nosocomial antibiotic resistance among multiple gram-negative pathogens. With continuation of our program through 2000, the incidence of ESBL-producing ceftazidime-resistant Klebsiella species has remained at the same reduced level as that reported in 1996, with important epidemiologic changes. These include a significant increase in the proportion of ceftazidime-resistant Klebsiella isolates recovered from patients admitted from longterm care facilities and from patients chronically dependent on ventilators. This has been associated with a statistically significant reduction in recovery of ESBL-producing Klebsiella spe- Figure 1. Evolution and control of antibiotic resistance among gram-negative bacilli at New York Hospital Queens. ICU, intensive care unit. HEALTHCARE EPIDEMIOLOGY CID 2002:34 (15 February) 501

4 Table 1. Multiple techniques used for control of nosocomial gram-negative resistance at New York Hospital Queens, Surveillance of all ceftazidime- or imipenem-resistant, gram-negative bacilli ( ). Infectious disease consultation and on-site education ( ). Intensive infection control directed at pathogens exhibiting chromosomally mediated, primarily clonal, resistance ( ). Class restriction of use of cephalosporins and cephamycins, directed at pathogens exhibiting plasmid-mediated, primarily polyclonal resistance ( ). Molecular epidemiology by pulsed-field gel electrophoresis ( ). Molecular identification of resistance mechanisms ( ). cies from patients in intensive care units. As noted in other studies, long-term care facilities have become important reservoirs for the perpetuation of ESBL-mediated ceftazidime resistance in both Klebsiella species and Escherichia coli [3, 15, 16]. That such resistance adversely influences outcome among infected patients has now been demonstrated in several studies [17 19]. Our routine surveillance of all ceftazidime- and imipenemresistant gram-negative bacilli detected a clonal outbreak of infection with ceftazidime-resistant Klebsiella species in Bacteremia due to a single genotype occurred in 4 patients within 1 month [17]. Overall, 14 patients became infected, and a statistically significant increased mortality rate occurred among patients with any infection due to this strain, independent of other potential risk factors [17]. This clonal outbreak subsided with reemphasis on standard infection-control procedures. The same surveillance program uncovered an outbreak of bronchoscope transmission of imipenem-resistant P. aeruginosa that resulted from misuse of automatic endoscope reprocessor connections [20]. The identification and resolution of these outbreaks of infection emphasize the value of prospective surveillance of antibiotic-resistant pathogens and the multidimensional requirements of resistance control. A third gram-negative pathogen has lurked intermittently as a multidrug-resistant pathogen at this center since Acinetobacter baumannii, which was susceptible only to ceftazidime, imipenem, and aminoglycosides, emerged in intensive care units during the late 1980s, resulting in the following sequence: increased ceftazidime use; selection of ESBL-producing, ceftazidime-resistant Klebsiella species; increased therapeutic use of imipenem; and selection of imipenem-resistant P. aeruginosa. In contrast to the latter organism, imipenem resistance in Acinetobacter species occurs not as a selective resistant mutant among otherwise susceptible strains, but rather among strains that were previously ceftazidime resistant as a result of selection of mutants that hyperproduce ampc b-lactamase. Our restriction endonuclease studies in the early 1990s showed that such strains were strongly related to those that progressed to develop imipenem-ceftazidime-aminoglycoside resistance [9]. Resistance to imipenem in these isolates was probably due to combined high-level ampc b-lactamase production and decreased outer membrane permeability to imipenem. Fortunately, plasmid-mediated b-lactamase activity against imipenem has not yet been demonstrated among isolates of Acinetobacter species from the United States [21]. Recognition of this resistance evolution in Acinetobacter species may provide guidance in avoiding the sequential development of multidrug resistance among gram-negative pathogens, as illustrated in figure 1. Thus, detection of late generation (usually ceftazidime) resistance due to hyperproduction of chromosomal ampc b-lactamase by P. aeruginosa, or Acinetobacter, Serratia, Enterobacter, or Citrobacter species, may be an early indication of excessive use of late-generation cephalosporins. This use may be reduced by intense infection-control efforts directed at chromosomally mediated, predominantly clonal outbreaks. Selection and emergence of ESBL-producing Enterobacteriaceae may then be avoided, as well as subsequent dependence on carbapenems for therapy and emergence of carbapenem resistance [10, 22 25]. Whether substitution of cefepime for ceftazidime might abort or delay this sequence because of the greater stability of cefepime to ampc b-lactamase remains to be determined. As suggested earlier, prevention of antibiotic resistance by education regarding potential mechanisms, intermittent microbiologic surveillance, and use of appropriate infection-control procedures may obviate the necessity of restricting antibiotic use. Under these circumstances, use of an open formulary may avoid excessive use of certain agents. However, if resistance mechanisms do emerge despite implementation of such measures, multiple responses, including antibiotic control, are essential to the appropriate management of epidemic or endemic nosocomial resistance [2, 26, 27]. To suggest that any individual response is merely equivalent to squeezing a balloon at a single site oversimplifies the complexity of multiple resistance mechanisms among many pathogens [5]. This analogy implies that restriction of antibiotic use is an isolated approach to an isolated problem. Rather, it is one of several techniques required to deal with multiple potential mechanisms that may evolve simultaneously or sequentially. Clinical investigation of the efficacy of a single technique must be accomplished with all others constant. However, the results should not be interpreted as an indication for use of that technique alone. Our experience during the past decade indicates that simultaneous implementation of several infection-control measures, as recommended by a joint statement from the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America, can be effective in reduction, if not elimination, of nosocomial antibiotic resistance and its adverse effects on cost, morbidity, 502 CID 2002:34 (15 February) HEALTHCARE EPIDEMIOLOGY

5 and mortality [27]. The resistance balloon can and should be squeezed at multiple sites (table 1). References 1. 4th Decennial International Conference on Nosocomial and Healthcare-Associated Infections. March 5 9, Atlanta, Georgia: proceedings. Emerg Infect Dis 2001; 7: Kollef MH, Frazer VJ. Antibiotic resistance in the intensive care unit. Ann Intern Med 2001; 134: Wiener J, Quinn JP, Bradford PA, et al. Multiple antibiotic-resistant Klebsiella and Escherichia coli in nursing homes. JAMA 1999; 281: Fridkin SK, Steward CD, Edwards JR, et al. Surveillance of antimicrobial use and resistance in United States hospitals: Project ICARE phase 2. Clin Infect Dis 1999; 29: Burke JP. Antibiotic resistance squeezing the balloon. JAMA 1998; 280: Center for Disease Control and Prevention. National Nosocomial Infections Surveillance System, January May Available at: accessed August Meyer KS, Urban C, Eagan J, et al. Nosocomial outbreak of Klebsiella infection resistant to late generation cephalosporins. Ann Intern Med 1993; 119: Urban C, Meyer KS, Mariano N, et al. Identification of TEM-26 b- lactamase responsible for a major outbreak of ceftazidime-resistant Klebsiella pneumoniae. Antimicrob Agents Chemother 1994; 38: Go E, Urban C, Burns J, et al. Clinical and molecular epidemiology of Acinetobacter infections sensitive only to polymyxin B and sulbactam. Lancet 1994; 344: Bradford PA, Urban C, Mariano N, et al. Imipenem resistance in Klebsiella pneumoniae is associated with the combination of ACT-1, a plasmid-mediated AmpC b-lactamase and the loss of an outer membrane protein. Antimicrob Agents Chemother 1997; 41: Ahmad M, Urban C, Mariano N, et al. Clinical and molecular epidemiology associated with imipenem-resistant Klebsiella pneumoniae. Clin Infect Dis 1999; 29: Rahal JJ, Urban C, Horn D, et al. Class withdrawal of cephalosporins for control of total cephalosporin resistance in nosocomial Klebsiella. JAMA 1998; 280: Vangala K, Quale J, Landman J, et al. Imipenem-resistant (IR) Pseudomonas aeruginosa (PA) in Brooklyn, NY: prevalence and molecular epidemiology [abstract 83]. In: Program and abstracts of the Infectious Diseases Society of America (IDSA; New Orleans). Boston: IDSA, 2000: Livermore DM. Of Pseudomonas, porins, pumps and carbapenems. J Antimicrob Chemother 2001; 47: Bonomo RA. Multiple antibiotic-resistant bacteria in long-term care facilities: an emerging problem in the practice of infectious diseases. Clin Infect Dis 2000; 31; Bradford PA, Urban C, Jaiswal A, et al. SHV-7, a novel cefotaximehydrolysing b-lactamase in Escherichia coli isolates from hospitalized nursing home patients. Antimicrob Agents Chemother 1995; 39: Qavi A, Segal-Maurer S, Mariano N, et al. Outbreak of clonal ceftazidime-resistant Klebsiella pneumoniae bacteremia in a surgical intensive care unit: a case-control study [abstract 157]. In: Program and abstracts of the Infectious Diseases Society of America (IDSA; Philadelphia). Boston: IDSA, 1999: Paterson DL, Ko W-C, Von Gottberg A, et al. Outcome of cephalosporin treatment for serious infections due to apparently susceptible organisms producing extended-spectrum b-lactamases: implications for the clinical microbiology laboratory. J Clin Microbiol 2001; 39: Schiappa DA, Hayden KM, Matushek MG, et al. Ceftazidime-resistant Klebsiella pneumoniae and Escherichia coli bloodstream infection: a case-control and molecular epidemiologic investigation. J Infect Dis 1996; 174: Sorin M, Segal-Maurer S, Mariano N, et al. Nosocomial transmission of imipenem-resistant Pseudomonas aeruginosa following bronchoscopy associated with improper connection to the STERIS System 1 Processor. Infect Control Hosp Epidemiol 2001; 22: Livermore DM, Woodford B. Carbapenemases: a problem in waiting? Curr Opin Microbiol 2000; 3: Ehrhardt AF, Sanders CC, Thomson KS, et al. Emergence of resistance to imipenem in Enterobacter isolates masquerading as Klebsiella pneumoniae during therapy with imipenem/cilastatin. Clin Infect Dis 1993; 17: Lee EH, Nicolas MH, Kitzis MD, et al. Association of two resistance mechanisms in a clinical isolate of Enterobacter cloacae with high-level resistance to imipenem. Antimicrob Agents Chemother 1991; 35: Stapelton PD, Shannon KP, French GL. Carbapenem resistance in Escherichia coli associated with plasmid-determined CMY-4 b-lactamase production and loss of an outer membrane protein. Antimicrob Agents Chemother 1999; 43: Mainardi JL, Mugnier P, Coutrot A, et al. Carbapenem resistance in a clinical isolate of Citrobacter freundii. Antimicrob Agents Chemother 1997; 41: Goldmann DA, Weinstein RA, Wenzel RP, et al, for the Workshop to Prevent and Control the Emergence and Spread of Antimicrobial-Resistant Microorganisms in Hospitals. Strategies to prevent and control the emergence and spread of antimicrobial-resistant microorganisms in hospitals: a challenge to hospital leadership. JAMA 1996; 275: Shlaes DM, Gerding DN, John J Jr, et al. Society for Healthcare Epidemiology of America and Infectious Diseases Society of America Joint Committee on the Prevention of Antimicrobial Resistance guidelines for the prevention of antimicrobial resistance in hospitals. Infect Control Hosp Epidemiol 1997; 18: HEALTHCARE EPIDEMIOLOGY CID 2002:34 (15 February) 503