Nosocomial outbreak of genetically related IMP-1 β-lactamase-producing Klebsiella pneumoniae in a general hospital in Japan

International Journal of Antimicrobial Agents 29 (2007) 306–310

Nosocomial outbreak of genetically related IMP-1 ␤-lactamase-producing Klebsiella pneumoniae in a general hospital in Japan Shinako Fukigai a , Jimena Alba b , Soichiro Kimura b , Toshie Iida a , Noriko Nishikura a , Yoshikazu Ishii b,∗ , Keizo Yamaguchi b b

a Division of Microbiology, Department of Clinical Laboratory, Hanyu General Hospital, Saitama, Japan Department of Microbiology and Infectious Diseases, Toho University Faculty of Medicine, 5-21-16 Omori-nishi, Ota-ku, Tokyo 1438540, Japan

Received 16 July 2006; accepted 28 October 2006

Abstract Gram-negative bacteria with acquired metallo-␤-lactamase (MBL) resistance are being increasingly described worldwide. Here we report the first case of an outbreak by a cluster of genetically related strains of Klebsiella pneumoniae producing the IMP-1 MBL. Six isolates of K. pneumoniae with a ceftazidime minimum inhibitory concentration ≥64 ␮g/mL were collected between February 2003 and June 2004 in Hanyu General Hospital, Saitama, Japan. These isolates were analysed to establish the mechanism of resistance. The zone of inhibition of these isolates using ceftazidime or imipenem disks on Mueller–Hinton agar containing dipicolinic acid was much larger than on Mueller–Hinton agar without dipicolinic acid. Polymerase chain reaction and DNA sequencing confirmed that the isolates contained blaIMP-1 as well as intI1 as a class I integrase gene. Pulsed-field gel electrophoresis was performed, showing that five of the six isolates were related. This outbreak was controlled by restrained and careful use of antibiotics as well as strict hygiene practices. © 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. Keywords: Klebsiella pneumoniae; Metallo-␤-lactamase; IMP-1; Nosocomial infection

1. Introduction Klebsiella pneumoniae is one of the major nosocomial pathogens causing intra-abdominal infections, urinary tract infections and primary bacteraemia [1]. These organisms have been susceptible to ␤-lactam antibiotics except penicillins, but in the past 2 decades they have developed resistance to expanded-spectrum ␤-lactams, largely owing to the production of extended-spectrum ␤-lactamases (ESBLs) [2]. Carbapenems are stable to most ␤-lactamases produced by Gram-negative bacteria. However, the number of carbapenem-resistant Gram-negative isolates is increasing. Carbapenem resistance in Gram-negative bacteria may be due to ␤-lactamase production [3,4], outer membrane imper∗ Corresponding author. Tel.: +81 3 3762 4151x2396; fax: +81 3 5493 5415. E-mail address: [email protected] (Y. Ishii).

meability or activation of efflux pumps. However, the most frequent mechanism is the production of ␤-lactamases [5,6]. Carbapenem hydrolysing activity is shown not only by class B metallo-␤-lactamases (MBLs) but also by class A and class D enzymes [7,8]. IMP-1 was the first acquired MBL to be described in the world [9]. Plasmid-acquired class B ␤-lactamases are classified into three main molecular groups: IMP-type, VIM-type and SPM-type enzymes [3,4]. Gram-negative bacilli producing IMP-type and VIM-type MBLs have been increasingly reported in Europe, Asia, and North and South America [4]. Recently, IMP-1 and VIM-2 ␤-lactamases were described as the predominant MBLs in Japan [10]. MBLs in K. pneumoniae have occurred sporadically. IMP1 ␤-lactamase has been reported in K. pneumoniae from Japan [10], Singapore [11] and Brazil [12], and IMP-8 has been described in Taiwan [13]. VIM-1 and VIM-4 MBLs have recently been described in Greece and Italy [14,15]. Here we

0924-8579/$ – see front matter © 2006 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved. doi:10.1016/j.ijantimicag.2006.10.011

S. Fukigai et al. / International Journal of Antimicrobial Agents 29 (2007) 306–310

report the first cluster of genetically related IMP-1 MBLproducing K. pneumoniae isolates in a hospital in Japan.

2. Method 2.1. Bacterial strains From February 2003 to June 2004, 456 clinical isolates of K. pneumoniae were obtained from patients in Hanyu General Hospital, Saitama, Japan. Nine (2.0%) of the 456 isolates had a ceftazidime minimum inhibitory concentration (MIC) ≥64 ␮g/mL. Six of these nine K. pneumoniae isolates were confirmed as MBL-producing strains by a screening test (see below). Details of the six MBL-producing K. pneumoniae are shown in Table 1. 2.2. Identification of bacterial strains Identification of all K. pneumoniae isolates was performed using the BD PhoenixTM identification system (Becton Dickinson Co., Tokyo, Japan) using NMIC/ID-23 panels (Becton Dickinson Co.).

307

2.4. Screening of MBL production Production of MBLs was screened by the disk approximation test [17]. Briefly, Mueller–Hinton agar plates (Becton Dickinson Co.) containing 200 ␮g/mL dipicolinic acid (DPA) (Sigma–Aldrich Japan K.K.) as a final concentration were prepared for screening. Mueller–Hinton agar plates in the absence of DPA were also prepared [17]. DPA is a chelating agent with a similar activity to ethylenediaminetetraacetic acid (EDTA) or sodium 2-mercaptoacetic acid; moreover, DPA has no inhibitory activity for bacterial cell growth. Bacterial suspensions were prepared as described above. The test was performed by placing two commercially supplied Kirby–Bauer disks, each containing 30 ␮g of ceftazidime or 10 ␮g of imipenem (Eiken Chemical Co. Ltd., Tokyo, Japan) on Mueller–Hinton agar plates in the presence or absence of DPA. The plates were then incubated at 35 ◦ C for 16–18 h. If the inhibition zones around the disk of 32 mm for ceftazidime or 20 mm for imipenem on Mueller–Hinton agar with DPA were bigger than those on the Mueller–Hinton agar plate without DPA (no inhibition zone), the strain was considered a MBL-producer.

2.3. Antibiotic susceptibility testing

2.5. Polymerase chain reaction (PCR) and DNA sequencing analysis

MICs were determined following the guidelines of the Clinical and Laboratory Standards Institute [16], using the broth microdilution method with cation-adjusted Mueller–Hinton broth (Difco, Detroit, MI). Bacterial suspensions were prepared as described above. Organisms were inoculated at ca. 5 × 104 cells per well. The MIC was defined as the lowest concentration preventing visible growth after incubation for 18 h at 35 ◦ C. Amikacin sulfate and minocycline hydrochloride were purchased from Sigma–Aldrich Japan K.K. (Tokyo, Japan). Levofloxacin was a gift from Daiichi Pharmaceutical Co. Ltd. (Tokyo Japan). Piperacillin was a gift from Taisho Toyama Pharmaceutical Co. Ltd. (Tokyo, Japan). Tazobactam was a gift from Taiho Pharmaceutical Co. Ltd. (Tokyo, Japan). Imipenem was a gift from Banyu Pharmaceutical Co. Ltd. (Tokyo, Japan). Ceftazidime was a gift from GlaxoSmithKline (Tokyo, Japan). Aztreonam was a gift from Eisai Co. Ltd. (Tokyo, Japan). Quality control for antibiotic susceptibility testing was performed using the following reference strains: Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 [16].

The MBL gene cassettes and integrons were detected by PCR using previously described specific primer sets for blaIMP , blaVIM , blaSPM , aacA4, IntI1, IntI2 and IntI3 [18]. blaIMP-1 was confirmed using positive control strains producing IMP-1 ␤-lactamase. PCR was performed in a GeneAmp PCR system 2400 thermal cycler (Applied Biosystems Japan, Tokyo, Japan). The thermocycle protocol used was: initial denaturation step at 94 ◦ C for 2 min, followed by 25 cycles of denaturation at 94 ◦ C for 30 s, annealing at 55 ◦ C for 30 s, elongation at 72 ◦ C for 90 s, plus a final extension step at 72 ◦ C for 7 min. PCR fragments were determined by the Agilent 2100 bioanalyser (Agilent Technologies, Palo Alto, CA). Twelve of the 16 wells were used for experimental samples, one was used for a kit-supplied molecular weight ladder and three were used for loading the gel–dye mix. DNA 7500 LabChips were prepared and loaded with samples as recommended by the manufacturer. Briefly, microchannels were filled by pipetting 9 ␮L of gel–dye mix into the appropriate well and then forcing the mix into the microchannels by applying pressure to the well via a 1 mL syringe.

Table 1 Background of studied Klebsiella pneumoniae isolates Strain

Isolation date

Specimen

Patient age (years)

Sex

Pre-existing disease

Ward

TUM2220 TUM2221 TUM2223 TUM2224 TUM2225 TUM2226

27 February 2003 16 June 2003 28 June 2003 6 October 2003 23 January 2004 24 June 2004

Sputum Sputum Urine Sputum Urine Vaginal swab

70 50 80 58 83 81

Male Male Female Male Female Female

Cerebral infarction Intracerebral bleeding Cerebral infarction Subarachnoid haemorrhage Hepatic cirrhosis, pyelonephritis Hepatic cirrhosis

4th 3rd 3rd 3rd Outpatient 2nd

308

S. Fukigai et al. / International Journal of Antimicrobial Agents 29 (2007) 306–310

Chips were immediately inserted into the bioanalyser and processed. All experiments were performed using Agilent Biosizing software (version B.01.02). The PCR-amplified DNA segment of the blaIMP-1 gene was purified with QIAquick PCR purification kit (Qiagen Inc., Tokyo, Japan), prepared with the ABI Prism Big Dye Terminator version 3.1 cycle sequencing ready reaction kit (Applied Biosystems) and sequenced with the ABI Prism 310 Genetic Analyzer (Applied Biosystems), using sequence specific primers for blaIMP and aacA4. For sequencing of blaIMP , the following PCR primer set was used: CGGATGAAGGCACGAAC (forward) and AAGCAGACTTGACCTGA (reverse), as reported by Iyobe et al. [19]. A similarity search for the deduced amino acid sequences was performed using the BLAST program at the DNA Data Bank of Japan (Shizuoka, Japan). 2.6. Pulsed-field gel electrophoresis (PFGE) analysis PFGE profiles of genomic DNA were analysed after digestion with the restriction enzyme XbaI. DNA fragments were electrophoresced in 1% agarose gels in 0.5× Tris–borate–EDTA buffer in the presence of 100 ␮M of thiourea (Sigma–Aldrich, Japan K.K.). Bacteriophage ␭ concatemers (Bio-Rad Laboratories Inc., Hercules, CA) were used as DNA size markers. The gel was run with a linear increase in switching times (from 4 s to 70 s) over a period of 26 h, a 120◦ switch angle and a gradient of 6.0 V/cm and 14 ◦ C. The gel was stained with ethidium bromide and photographed with ultraviolet illumination. A dendrogram of PFGE profiles was generated by Fingerprinting II (Bio-Rad Laboratories, Inc.) using the Dice coefficient and clustering by unweighted pair group method using average linkages (UPGMA).

3. Results 3.1. Antibiotic susceptibility testing and results of screening for MBL production Table 2 shows the antibiotic susceptibilities of the K. pneumoniae isolates. All isolates were resistant to all tested

Table 3 Polymerase chain reaction results for blaIMP-1 , aacA4 and integrase genes among the studied Klebsiella pneumoniae isolates Strain

blaIMP-1

aacA4

intI1

TUM2220 TUM2221 TUM2223 TUM2224 TUM2225 TUM2226

+ + + + + +

− − − − − −

+ + + + + +

␤-lactams including carbapenems such as imipenem and meropenem. The tested ␤-lactamase inhibitors also had no effect on these isolates. These antibiotic susceptibility data strongly suggested that the strains produce a MBL. On the other hand, all K. pneumoniae isolates were susceptible to minocycline, amikacin and levofloxacin. Production of a MBL was confirmed using imipenem and ceftazidime disks applied to Mueller–Hinton agar plates with or without DPA. For all isolates, a larger inhibition zone was observed around the antibiotic disks on the Mueller–Hinton agar plate containing DPA compared with those on the plate without this compound (Table 2). 3.2. Antibiotic resistance gene and integrase gene The blaIMP gene was found in all isolates by PCR analysis (Table 3). blaVIM or blaSPM was not detected by PCR. Also, intI1 was found in all isolates, however the other integrase genes were not observed in these strains (Table 3). The presence of blaIMP-1 was confirmed by DNA sequencing analysis. The results of PCR analysis suggested that blaIMP was encoded by a class 1 integron. Most class 1 integrons also encode the aacA4 gene [4], therefore PCR analysis for the aacA4 gene was performed. However, the aacA4 gene encoding 6 -N-acetyltransferase (AAC(6 )-Ib) was not detected in any isolate by PCR analysis (Table 3). All isolates were susceptible to amikacin (Table 2), confirming that these isolates do not produce AAC(6 )-Ib, an enzyme that typically modifies amikacin [20]. The detailed integron structure will be confirmed in the near future.

Table 2 Minimum inhibitory concentrations (MICs) of Klebsiella pneumoniae strains Drug

Piperacillin Piperacillin/tazobactam Ceftazidime Imipenem Aztreonam Amikacin Minocycline Levofloxacin Metallo-␤-lactamasea a

MIC (␮g/mL) TUM2220

TUM2221

TUM2223

TUM2224

TUM2225

TUM2226

32 16/4 128 32 0.125 0.5 16 0.5 +

8 8/4 >128 64 0.125 0.25 0.5 ≤0.063 +

>128 >128/4 >128 4 0.125 0.5 0.5 ≤0.063 +

16 8/4 >128 64 0.125 0.5 0.25 ≤0.063 +

>128 >128/4 >128 8 0.125 0.5 16 ≤0.063 +

>128 >128/4 64 4 ≤0.063 0.5 0.5 ≤0.063 +

Metallo-␤-lactamase was detected using dipicolinic acid.

S. Fukigai et al. / International Journal of Antimicrobial Agents 29 (2007) 306–310

309

Fig. 1. Separation of six Klebsiella pneumoniae isolates based on pulsed-field gel electrophoresis. Dendrogram generated by Dice coefficient with clustering by unweighted pair group method using average linkages (UPGMA). Five isolates had >82% similarity, suggesting that they are genetically related.

3.3. PFGE analysis The chromosomes of these MBL-producing K. pneumoniae isolates were digested by restriction enzyme XbaI (Fig. 1). The resulting gel showed that five isolates shared ≥82.8% similarity. The similarity of K. pneumoniae TUM2226 with the other isolates was 35.7–44.4%.

4. Discussion Use of DPA to detect MBL in P. aeruginosa was first described by Kimura et al. [17]. They also reported that DPA did not destroy ampicillin, ceftazidime, aztreonam, imipenem or meropenem. By contrast, imipenem and meropenem are hydrolysed by 2-mercaptopropionic acid. We observed a growth inhibition effect on some tested strains by EDTA, however DPA does not have this effect on P. aeruginosa or K. pneumoniae. DPA is also convenient to use as it is autoclavable and is therefore very suitable for the detection of MBLs. To date, infections by MBL-producing K. pneumoniae have been reported from Taiwan (IMP-8) [13], Singapore (IMP-1) [11], Greece (VIM-1) [21] and France (VIM-1) [22]. IMP-1 ␤-lactamase is the most common plasmid-acquired MBL in Japan [10]. These enzymes can hydrolyse all ␤-lactams except for aztreonam. The MIC of aztreonam for all isolates was ≤0.125 ␮g/mL (Table 2). Fortunately, most K. pneumoniae isolates do not produce class C ␤-lactamases. Accordingly, aztreonam is still a therapeutic choice for K. pneumoniae infections in general. All isolates were susceptible to amikacin, making it an alternative antibiotic for these infections. Four patients received administration of meropenem (average 9.5 days) before the isolation of MBL-producing K. pneumoniae. Other patients received cefmetazole or cefepime before isolation of a MBL-producer. These data suggest that administration of ␤-lactam antibiotics was one of the risk factors for selecting MBL-producing isolates in these patients. All MBL-producing K. pneumoniae infections were cleared by administration of minocycline or levofloxacin. Yamaguchi et al. [23] reported that the isolation

frequency of levofloxacin-resistant K. pneumoniae was 0.8% in Japan in 2002. Furthermore, Hoban et al. [24] reported that the isolation frequency of levofloxacin-susceptible K. pneumoniae was 97.5%. Thus, levofloxacin is potentially a useful antibiotic for imipenem-resistant K. pneumoniae infection. Of 456 K. pneumoniae clinical specimens isolated in Hanyu General Hospital from February 2003 to June 2004, six strains (1.3%) appeared to produce MBL (Table 2). Six of these patients had indwelling urinary catheters. The average period of hospitalisation was 101.2 days (61–151 days). Moreover, PFGE results suggested that five isolates were identical, however, K. pneumoniae TUM2226 did not share the same origin as the others (Fig. 1). Clinical data and the results of PFGE strongly suggested that five of the MBLproducing K. pneumoniae infections in this hospital were caused by a device-related or healthcare-associated infection. Isolation of MBL-producing K. pneumoniae was reduced after reinforcing standard procedures and taking appropriate precautions such as hand hygiene, no-touch technique, barrier nursing and reduction in the number of people staying in a hospital ward including staff, patients and visitors. In November 2003, the clinicians of Hanyu General Hospital also agreed with the Infection Control Committee to restrict the use of ceftazidime, meropenem, ciprofloxacin, pazufloxacin, vancomycin and teicoplanin. Some reports show that the restriction of antibiotic use is efficient in reducing antimicrobial resistance [25,26]. Our data also suggest that restriction of carbapenem use was effective in decreasing the selection of carbapenem-resistant bacteria. Two MBL-producing K. pneumoniae strains were isolated from two patients after the provisions for reducing MBLproducing K. pneumoniae infection (Table 1). All people complied with the infection control procedures in this hospital. The PFGE banding pattern of K. pneumoniae TUM2226 was different from the other MBL-producing K. pneumoniae isolates. On the other hand, K. pneumoniae TUM2225 was isolated from an outpatient. This isolate shared ≥82.8% similarity with TUM2220, TUM2221, TUM2223 and TUM2224. This patient had stayed on the same ward (3rd) in Hanyu Gen-

310

S. Fukigai et al. / International Journal of Antimicrobial Agents 29 (2007) 306–310

eral Hospital with the other patients with MBL-producing K. pneumoniae. MBL-producing K. pneumoniae was isolated from patients in this ward several times between 1997 and 2004. However, since June 2004, ceftazidime-resistant K. pneumoniae have not been isolated from this ward. In summary, this is the first report of an outbreak by genetically related MBL-producing K. pneumoniae. Aztreonam, amikacin or levofloxacin are effective antibiotics for the presented K. pneumoniae infections. Moreover, restrained and careful use of antibiotics as well as strict hygiene practices are critical to prevent the emergence and spread of MBLproducing K. pneumoniae.

[11]

[12]

[13]

[14]

[15]

Acknowledgments This work was supported by a grant from the Ministry of Health, Labor, and Welfare of Japan (H15-Shinko-09, H15Shinko-10 and H18-Shinko-11) and Toho University Project Research Grants (18-7). J. Alba was supported by the Japan Health Science Foundation. We thank K. Nakashima, R. Shimatsu and R. Shibuya for their technical assistance. We also thank Becton Dickinson for placing the PhoenixTM instrument at our disposal during this study and for providing the consumables.

[16]

[17]

[18]

[19]

References [20] [1] Podschun R, Ullmann U. Klebsiella spp. as nosocomial pathogens: epidemiology, taxonomy, typing methods, and pathogenicity factors. Clin Microbiol Rev 1998;11:589–603. [2] Paterson DL, Ko WC, Von Gottberg A, et al. International prospective study of Klebsiella pneumoniae bacteremia: implications of extendedspectrum ␤-lactamase production in nosocomial infections. Ann Int Med 2004;140:26–32. [3] Nordmann P, Poirel L. Emerging carbapenemases in Gram-negative aerobes. Clin Microbiol Infect 2002;8:321–31. [4] Walsh TR, Toleman MA, Poirel L, Nordmann P. Metallo-␤-lactamases: the quiet before the storm? Clin Microbiol Rev 2005;18:306–25. [5] Nikaido H. Prevention of drug access to bacterial targets: permeability barriers and active efflux. Science 1994;264:382–8. [6] Piddock LJ. Clinically relevant chromosomally encoded multidrug resistance efflux pumps in bacteria. Clin Microbiol Rev 2006;19: 382–402. [7] Jacoby GA, Munoz-Price LS. The new ␤-lactamases. N Engl J Med 2005;352:380–91. [8] Walther-Rasmussen J, Hoiby N. OXA-type carbapenemases. J Antimicrob Chemother 2006;57:373–83. [9] Watanabe M, Iyobe S, Inoue M, Mitsuhashi S. Transferable imipenem resistance in Pseudomonas aeruginosa. Antimicrob Agents Chemother 1991;35:147–51. [10] Shibata N, Doi Y, Yamane K, et al. PCR typing of genetic determinants for metallo-␤-lactamases and integrases carried by gram-negative

[21]

[22]

[23]

[24]

[25]

[26]

bacteria isolated in Japan, with focus on the class 3 integron. J Clin Microbiol 2003;41:5407–13. Koh TH, Babini GS, Woodford N, Sng LH, Hall LM, Livermore DM. Carbapenem-hydrolysing IMP-1 ␤-lactamase in Klebsiella pneumoniae from Singapore. Lancet 1999;353:2162. Lincopan N, McCulloch JA, Reinert C, Cassettari VC, Gales AC, Mamizuka EM. First isolation of metallo-␤-lactamase-producing multiresistant Klebsiella pneumoniae from a patient in Brazil. J Clin Microbiol 2005;43:516–9. Yan JJ, Ko WC, Tsai SH, Wu HM, Wu JJ. Outbreak of infection with multidrug-resistant Klebsiella pneumoniae carrying blaIMP-8 in a university medical center in Taiwan. J Clin Microbiol 2001;39:4433–9. Giakkoupi P, Xanthaki A, Kanelopoulou M, et al. VIM-1 metallo-␤lactamase-producing Klebsiella pneumoniae strains in Greek hospitals. J Clin Microbiol 2003;41:3893–6. Luzzaro F, Docquier JD, Colinon C, et al. Emergence in Klebsiella pneumoniae and Enterobacter cloacae clinical isolates of the VIM-4 metallo-␤-lactamase encoded by a conjugative plasmid. Antimicrob Agents Chemother 2004;48:648–50. National Committee for Clinical Laboratory Standards. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 6th ed. Approved standard. M7-A5. Wayne, PA: NCCLS, 2003. Kimura S, Ishii Y, Yamaguchi K. Evaluation of dipicolinic acid for detection of IMP- or VIM-type metallo-␤-lactamase-producing Pseudomonas aeruginosa clinical isolates. Diagn Microbiol Infect Dis 2005;53:241–4. Kimura S, Alba J, Shiroto K, et al. Clonal diversity of metallobeta-lactamase-possessing Pseudomonas aeruginosa in geographically diverse regions of Japan. J Clin Microbiol 2005;43:458–61. Iyobe S, Kusadokoro H, Ozaki J, et al. Amino acid substitutions in a variant of IMP-1 metallo-␤-lactamase. Antimicrob Agents Chemother 2000;44:2023–7. Shaw KJ, Cramer CA, Rizzo M, et al. Isolation, characterization, and DNA sequence analysis of an AAC(6 )-II gene from Pseudomonas aeruginosa. Antimicrob Agents Chemother 1989;33:2052–62. Ikonomidis A, Tokatlidou D, Kristo I, et al. Outbreaks in distinct regions due to a single Klebsiella pneumoniae clone carrying a blaVIM-1 metallo-␤-lactamase gene. J Clin Microbiol 2005;43:5344–7. Kassis-Chikhani N, Decre D, Gautier V, et al. First outbreak of multidrug-resistant Klebsiella pneumoniae carrying blaVIM-1 and blaSHV-5 in a French University hospital. J Antimicrob Chemother 2006;57:142–5. Yamaguchi K, Ohno A; Levofloxacin Surveillance Group. Investigation of the susceptibility trends in Japan to fluoroquinolones and other antimicrobial agents in a nationwide collection of clinical isolates: a longitudinal analysis from 1994 to 2002. Diagn Microbiol Infect Dis 2005;52:135–43. Hoban DJ, Biedenbach DJ, Mutnick AH, Jones RN. Pathogen of occurrence and susceptibility patterns associated with pneumonia in hospitalized patients in North America: results of the SENTRY Antimicrobial Surveillance Study (2000). Diagn Microbiol Infect Dis 2003;45:279–85. Brahmi N, Blel Y, Kouraichi N, et al. Impact of ceftazidime restriction on gram-negative bacterial resistance in an intensive care unit. J Infect Chemother 2006;12:190–4. Girouard S, Levine G, Goodrich K, et al. Infection control programs at children’s hospitals: a description of structures and processes. Am J Infect Control 2001;29:145–51.