Received 21 June 2002/Returned for modification 23 July 2002/Accepted 24 September 2002

Transcription

1 JOURNAL OF CLINICAL MICROBIOLOGY, Dec. 2002, p Vol. 40, No /02/$ DOI: /JCM Copyright 2002, American Society for Microbiology. All Rights Reserved. Comparison of a Repetitive Extragenic Palindromic Sequence-Based PCR Method and Clinical and Microbiological Methods for Determining Strain Sources in Cases of Nosocomial Acinetobacter baumannii Bacteremia David Martín-Lozano, 1 * José Miguel Cisneros, 1 Berta Becerril, 2 Lucila Cuberos, 2 Trinidad Prados, 2 Carlos Ortíz-Leyba, 3 Elías Cañas, 1 and Jerónimo Pachón 1 Infectious Diseases Service 1, Microbiology Service, 2 and Intensive Care Unit, 3 Virgen del Rocío University Hospitals, Seville, Spain Received 21 June 2002/Returned for modification 23 July 2002/Accepted 24 September 2002 Using a repetitive extragenic palindromic PCR (REP-PCR), we genotypically characterized strains causing nosocomial Acinetobacter baumannii infections and analyzed the source of bacteremia in 67 patients from an institution in which infections by this bacterium were endemic. Six different genotypes were found, including 21, 27, 3, 9, 3, and 4 strains. The probable source of bacteremia, according to clinical and/or microbiological criteria, was known in 42 patients (63%): respiratory tract (n 19), surgical sites (n 12), intravascular catheters (n 5), burns (n 3), and urinary tract (n 3). The definite source of bacteremia, according to REP- PCR, could be established in 30 (71%) out of the 42 patients with strains from blood and other sites; in these cases clinical and microbiological criteria for the source of bacteremia were thus confirmed. In the remaining 12 patients (29%) the probable source was refuted by the REP-PCR method. The definite sources of bacteremia according to genotype were as follows: respiratory tract in 13 patients (31%), surgical sites in 8 (19%), intravascular catheters in 4 (9%), burns in 3 (7%), and urinary tract in 2 (5%). A comparison of strains from blood cultures and other sites with regard to their REP-PCR and antimicrobial resistance profiles was also made. Taking the REP-PCR as the gold standard, the positive predictive value of antibiotype was 77% and the negative predictive value was 42%. In summary, the utility of the diagnosis of the source of nosocomial A. baumannii bacteremia using clinical and/or microbiological criteria, including antibiotyping, is limited, as demonstrated by REP-PCR. Downloaded from Acinetobacter baumannii has been reported with increasing frequency as a significant pathogen involved in outbreaks of nosocomial infection. A. baumannii may cause a great number of clinical conditions, including pneumonia, bacteremia, urinary tract infections, wound infections, endocarditis, and meningitis. Treatment can be complicated by emergence of isolates with multiple-drug resistance (2, 5). The characteristics of A. baumannii bacteremia are widely described in the literature (2, 5, 21, 22, 26). The source of bacteremia is of epidemiological and prognostic interest (5, 21, 22, 26). According to recent studies, the source of A. baumannii bacteremia is unknown in 43 to 50% of the cases and, when the source can be established, the primary focus of infection is the respiratory tract, followed by intravascular catheters (5, 21, 22, 26). The diagnosis of the source of bacteremia is established by assessing clinical and, whenever possible, microbiological data. However, this method may not be completely appropriate in hospitals in which infections by A. baumannii are endemic, where the patients with bacteremia may be colonized by several different strains. In these cases, additional studies are necessary to determine the strain responsible for the infection. Methods such as antibiotyping have been used by different authors (6, 8, 24), but it seems to be poorly discriminatory for epidemiological purposes (9, 10, 13, 18, 22). The usefulness of molecular techniques has already been demonstrated in epidemiological studies of infections due to A. baumannii. The DNA restriction fragment length polymorphism analysis determined by pulsed-field gel electrophoresis is considered the gold standard method (19, 20) but is complex to perform, requires a considerable amount of time, and uses expensive reagents. Contrary to this method, several authors (3, 4) have shown that repetitive extragenic palindromic PCR (REP-PCR) is a simple, rapid, and relatively low-cost method, useful in the characterization of nosocomial outbreaks of A. baumannii infections, being comparable to pulsed-field gel electrophoresis in regard to discriminatory power and reproducibility. Moreover, REP-PCR has a higher power of discrimination and reproducibility than do other PCR-based fingerprinting systems (3, 11, 17, 23, 25). In the present work we carried out a genotype study of A. baumannii strains, isolated from bacteremic patients, using REP-PCR, with two objectives: first, to know the clonal distribution of bacteremia, and second, to compare the results obtained with the conventional clinical and/or microbiological diagnosis to those from the molecular identification. on November 18, 2018 by guest * Corresponding author. Mailing address: Servicio de Enfermedades Infecciosas, Hospitales Universitarios Virgen del Rocío, Avda. Manuel Siurot s/n, 41013, Seville, Spain. Phone: Fax: jpachon@hvr.sas.junta-andalucia.es. MATERIALS AND METHODS Setting. The Virgen del Rocío University Hospitals, Seville, Spain, is a 1,700- bed tertiary and teaching center. Approximately 58,000 patients are admitted each year, and there are two adult intensive care units (ICU), a severe-burn unit, and an active program of solid-organ and bone marrow transplantation. 4571

2 4572 MARTÍN-LOZANO ET AL. J. CLIN. MICROBIOL. Population study. The study was carried out using the strains isolated from the case patients of a prospective case control study designed to assess the attributable mortality of nosocomial A. baumannii bacteremia (J. M. Cisneros, D. Martín, B. Becerril, et al., Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. K-1713, 2000), during the period from April 1997 to March Only true A. baumannii bacteremia were included, defined as an episode of bacteremia in a patient with clinical evidence of infection. The following data from each patient with A. baumannii bacteremia were recorded: gender, age, hospitalization ward (ICU, medical or surgical), nosocomial procedures (intravascular catheter, urinary tract catheter, mechanical ventilation, surgery, nasogastric tube, parenteral nutrition, and antimicrobial treatment), and score of Acute Physiology and Chronic Health Evaluation II (14). All cases were monitored until the death of the patient or discharge from the hospital. Nosocomial acquisition and the probable source of bacteremia were defined by standard clinical and microbiological criteria (7, 12). All data refer to the first episode of A. baumannii bacteremia in each patient. Microbiological and genotypic study. Blood cultures were requested by the physician in charge of each patient and processed using the BACTEC NR-860 system (Becton Dickinson, Cockeysville, Md.). When available, samples from the probable source of bacteremia were collected. Identification of A. baumannii was carried out using MicroScan (Baxter H.C., West Sacramento, Calif.), the API 20NE System (Bio-Mérieux, Marcy l Etoile, France), and the temperature growth test (44 C). Susceptibility to antimicrobial agents (ampicillin, ampicillinsulbactam, aztreonam, imipenem, cefotaxime, ceftazidime, gentamicin, amikacin, tetracycline, ciprofloxacin, and trimethoprim-sulfamethoxazole) was determined using the MicroScan system (Neg Breakpoint Combo Panel 2I) and the E test for colistin, following the criteria from the National Committee for Clinical Laboratory Standards (15). A. baumannii isolates were stored at 80 C in brucella broth containing 20% glycerol. The genotypic study was developed as described by other authors, using whole cells (23). Briefly, strains were grown on blood agar plates and were incubated overnight at 37 C to isolate single colonies. After incubation, four or five discrete colonies of each strain were resuspended in 50 l of sterile distilled water in a 0.5-ml Eppendorf tube. To lyse the cells and extract DNA, the tubes were heated for 10 min at 95 C, were then cooled on ice, and were centrifuged in a microcentrifuge (Microfuge Lite; Beckman) at 6,000 rpm for 60 s to remove the cell debris. These crude DNA extracts were kept on ice for immediate use. Portions (5 l) of these extracts were used in 50- l PCR mixtures without further purification. REP-PCR, which uses consensus primers for the REP sequences found in many bacterial chromosomes, was used in the genotyping of A. baumannii clones (23). This highly conserved REP sequence is approximately 35 nucleotides long, includes an inverted repeat, and can occur in the genome singly or as multiple adjacent copies. The primer pair REP 1 (5 -IIIGCGCCGICATCAGGC-3 ) and REP2(5 -ACGTCTTATCAGGCCTAC-3 ) was used to amplify putative REPlike elements in the bacterial DNA. Amplification PCRs were performed as described previously (23). A negative control to detect reagent contamination was included in each PCR, containing all components except the DNA extract, which was replaced by 5 l of sterile distilled H 2 O. Aliquots (12 l) of each sample were subjected to electrophoresis in a 1.5% agarose gel. Amplified products were detected by being stained with ethidium bromide and photographed with Polaroid type 665 film. In order to group isolates for photographic documentation, the REP-PCR fingerprints of strains were exposed to UV light, photographed, and compared by visual inspection. The molecular sizes of fragments generated by electrophoresis were judged by comparison with concurrently run molecular weight standards. Fingerprints were considered to be highly similar when all visible bands from two isolates had the same apparent migration distance. Variations in the intensity or shape of bands were not taken into account. The absence of up to two bands from a fingerprint was allowed, when all other visible bands in the fingerprints matched, before isolates were considered different by visual inspection (17, 23). Each isolate was run in duplicate. Fingerprint profiles were interpreted without knowledge of the clinical data. A source of bacteremia was considered definite when the strain from the probable source of bacteremia had the same REP-PCR genotype as the bacteremic strain. Statistical analyses. Results were expressed as mean standard deviation, median, and range or as a proportion of the total number of patients. The Fisher exact test was used. All statistical analyses were carried out by using Statistical Package for the Social Sciences software (version 10.0; Statistical Package for the Social Sciences Inc., Chicago, Ill.). TABLE 1. Antimicrobial susceptibility of A. baumannii strains Antimicrobial agent Breakpoint(s) ( g/ml) b RESULTS % of strains found to be: Susceptible Intermediate Resistant Ampicillin Ampicillin-sulbactam a 8/ Aztreonam Imipenem Cefotaxime Ceftazidime Gentamicin Amikacin Tetracycline Ciprofloxacin Trimethoprimsulfamethoxazole 2/ Colistin a The values for ampicillin-sulbactam apply to only 80 strains. b Where two breakpoints are listed, they are those of the individual drugs in the combination. Clinical characteristics of patients. From April 1997 to March 1999, 67 patients (41 males and 26 females) with true A. baumannii bacteremia were included, which represented an incidence of 0.54 episodes/1,000 admissions. Patients had a mean age of 52.2 years (range, 12 to 85 years). Fifty-one (76%) were in the ICU (44 in the general ICU, 5 in the burn unit, and 2 in the transplantation unit), 11 (16.5%) were in medical wards, and 5 (7.5%) were in surgical wards. Sixty-six (98.5%) patients had intravascular catheters, 63 (94%) had urinary catheters, 61 (91%) had received antimicrobial treatment for a median of 9 days (range, 0 to 16 days), 45 (67%) had mechanical ventilation, and 27 (40%) had undergone surgical procedures. The median interval between hospital admission and the onset of bacteremia was 13 days (range, 0 to 228). The mean Acute Physiology and Chronic Health Evaluation II score was Bacteremia was polymicrobial in 16 (24%) of the 67 patients. Staphylococcus epidermidis (37.5%) and Enterococcus faecalis (31%) were the most frequent concomitant isolates. There were severe sepsis in 26 cases (39%) and septic shock in 13 (19%). Twenty-six (38.8%) of the 67 cases died in the 30 days after the diagnosis of bacteremia. In surviving patients, the median hospital stay was 37 days (range, 3 to 144 days). Susceptibility testing. Table 1 details the percentage of resistant isolates among 109 A. baumannii isolates (67 from blood cultures and 42 from probable sources of bacteremia). Twenty-two isolates (20%) were resistant to all classes of antimicrobial agents tested except for colistin. Conversely, only four strains (4%) were susceptible to all classes of drugs. Genotyping. Isolates from blood cultures in the 67 patients were grouped according to the REP-PCR fingerprints in six different genotypes, named clones I to VI, which included 21 (31.5%), 27 (40%), 3 (4.5%), 9 (13.5%), 3 (4.5%), and 4 (6%) strains, respectively. Most of these clones were distributed throughout the study period with the exception of clone V, which appeared only during the first 6 months of the study (Fig. 1). Source of bacteremia. In the 67 patients, according to the standard clinical and/or microbiological criteria, the source of

3 VOL. 40, 2002 REP-PCR AND SOURCES OF A. BAUMANNII BACTEREMIA 4573 TABLE 3. Confirmation by REP-PCR of source of bacteremia established by clinical and/or microbiological criteria No. of cases for which source was: Source Probable, according to clinical and/or microbiological criteria Definite, according to REP-PCR (%) Refuted by REP-PCR (%) FIG. 1. Distribution of clones (I to VI) throughout the study period grouped by quarters. Respiratory tract (68) 6 (32) Surgical sites 12 8 (67) 4 (33) Intravascular catheters 5 4 (80) 1 (20) Burns 3 3 (100) 0 Urinary tract 3 2 (67) 1 (33) Total (71) 12 (29) bacteremia had a known origin in 42 patients (63%) (Table 2). Strains isolated from the probable source of bacteremia were available in these 42 patients. The distribution of sources was the following: respiratory tract (n 19), surgical sites (n 12), intravascular catheters (n 5), burns (n 3), and urinary tract (n 3). According to REP-PCR genotypes, the definite source of bacteremia could be established in 30 out of the 42 patients (71%), whereas, in the remaining 12 patients (29%) the strains from blood cultures and other sites showed different genotypes, the probable source being refuted by this method. Figure 2 shows fingerprints of representative strains of the six genotypes, both from blood and definite source strains. The definite sources of bacteremia according to genotype were as follows: respiratory tract in 13 patients (31%), surgical sites in 8 (19%), intravascular catheters in 4 (9%), burns in 3 (7%), and urinary tract in 2 patients (5%) (Table 3). The distribution of clones by the definite sources of bacteremia is shown in Table 4. A comparison of strains from blood cultures and other sites with regard to their REP-PCR and antimicrobial resistance profiles was also made to assess the usefulness of the antibiotype for epidemiological purposes. The strains were considered to have different antibiotypes when differences in susceptibility to two or more antimicrobial agents were observed (18, 22). In the 30 cases with the same genotype in blood and the definite source of bacteremia, 23 (77%) had the same antibiotype and 7 (23%) had different antibiotypes. In the 12 cases with different genotypes in blood and the probable source of bacteremia, 7 (58%) had the same antibiotype and 5 (42%) had different ones. Therefore, when REP-PCR was taken as the gold standard for the comparison of samples, the positive predictive value of the antibiotype was 77% and the negative predictive value was 42%. The crude mortality rate of A. baumannii bacteremia at 30 days was 39%. With respect to genotypes, the mortality rate was as follows: for clone I, 11 of 21 (52%); for clone II, 7 of 27 (26%); for clone III, 1 of 3 (33%); for clone IV, 3 of 9 (33%); for clone V, 2 of 3 (67%); and for clone VI, 2 of 4 (50%); these differences were not significant. DISCUSSION The clinical characteristics of nosocomial A. baumannii bacteremia in this study were similar to those reported from our institution in 1993 (5), although the incidence has decreased from 1.8 episodes per 1,000 admissions in 1993 to 0.54 episodes during the period of this study. The A. baumannii bacteremia predominantly occurs in patients admitted to the ICU, with high frequency of underlying chronic disease, extended previous hospital stay, and multiple nosocomial procedures. In addition, A. baumannii bacteremia is frequently associated with severe sepsis and/or septic shock and the mortality rate at 30 days is near 40%, similar to that reported in other data (5, 21, 26). A. baumannii resistance to antimicrobials was greater than TABLE 2. Probable sources of 67 episodes of nosocomial A. baumannii bacteremia Probable source of bacteremia a No. of cases attributed % Of cases attributed Unknown Respiratory tract Surgical sites Intravascular catheters 5 7 Urinary tract 3 5 Burns 3 5 a In accordance with clinical and/or microbiological criteria. FIG. 2. Representative REP-PCR fingerprints of A. baumannii isolates from blood and definite bacteremia sources, corresponding to the different genotypes. Lane 1, lambda EcoRI-HindIII molecular size marker (Master Diagnostica, Granada, Spain); lanes 2 and 3, genotype I (strains from blood and respiratory tract, respectively); lanes 4 and 5, genotype II (strains from blood and intravascular catheter, respectively); lanes 6, genotype III (strain from blood); lanes 7 and 8, genotype IV (strains from blood and surgical site, respectively); lane 9, genotype V (strain from blood); lanes 10 and 11, genotype VI (strains from blood and burn, respectively); and lane 12, negative control. Note: this figure is a composite of lanes from different gels.

4 4574 MARTÍN-LOZANO ET AL. J. CLIN. MICROBIOL. TABLE 4. Distribution of the different clones by the definite sources of bacteremia Source I II No. of examples of: III IV V VI Respiratory tract Surgical sites Intravascular catheter Burn Urinary tract that reported in 1993 (5). So, the resistance rate to ampicillinsulbactam and imipenem has risen from 7 to 33% and from 19 to 34%, respectively. The nosocomial A. baumannii bacteremias in this study were polyclonal, six different genotypes having been identified, as occurs in situations where disease is endemic (4). On the contrary, epidemic outbreaks are mainly due to a single clone of A. baumannii (1, 4). The distribution of clones throughout the study period shows that in our institution there is an situation where disease is endemic, with additional, different, small outbreaks. The polyclonal character of A. baumannii explains why 29% of 42 patients with two A. baumannii isolates in different clinical samples had genotypically different A. baumannii strains. The present study determined the probable source of A. baumannii bacteremia in 63% of cases by using clinical and/or microbiological criteria. The respiratory tract was the most common source, followed by surgical sites, intravascular catheters, urinary tract, and burns. In other studies, the respiratory tract was also the most common source of A. baumannii bacteremia, ranging from 31 to 41% of the cases with a known source (5, 21), and surgical sites and intravascular catheters were the second and third most common sources of A. baumannii bacteremia, in 10 and 6% of the instances (5). In another study, intravascular catheters were the most frequent source of A. baumannii bacteremia, being the cause of 22% of the cases (26). This discrepancy in results probably reflects the different number of patients admitted to the ICU: 84% in the present study versus 69% in the work of Wisplinghoff et al. (26). Finally, the urinary tract and burns are rarely reported as the source of bacteremia. In this study, using a molecular method with REP-PCR, we have proven that the diagnosis of source of bacteremia using conventional clinical and/or microbiological criteria was not always accurate. So, in 29% of the cases the conventional diagnosis of the probable source of bacteremia was incorrect, notably 32% in the respiratory tract, 33% from surgical sites, and 20% from intravascular catheters. These results illustrate the limitations of conventional methods in diagnosing the source of A. baumannii bacteremia. The usefulness of antibiotyping for the diagnosis of the source of A. baumannii bacteremia was limited, with a positive predictive value of 77% and negative predictive value of 42%. These results of antibiotyping should be interpreted with precaution because of the discrepancies of automated methods, such as MicroScan, with the standard broth dilution. However, in a study using 200 strains of gram-negative bacilli, when MicroScan was compared to broth dilution, in the 1,600 organism-antimicrobial combinations agreement occurred in 92%; moreover, the five discrepancies found with MicroScan were with Pseudomonas spp. and Providencia stuartii and in no case occurred when Acinetobacter calcoaceticus was being tested (16). These results confirm the low discrimination power of antibiotyping in comparison of different isolates, as has been reported in other epidemiological studies (9, 10, 13, 18, 22). In conclusion, the diagnosis of the source of nosocomial A. baumannii bacteremia by use of conventional clinical and/or microbiological criteria, including antibiotyping, is of limited utility, as demonstrated by REP-PCR. Additionally, the most common definite source of nosocomial A. baumannii bacteremia is the respiratory tract, followed by surgical sites and intravascular catheters. ACKNOWLEDGMENT This work was supported in part by a research grant from Consejería de Salud, Andalucía, Spain. REFERENCES 1. Beck-Sagué, C. M., W. R. Jarvis, J. H. Brook, D. H. Culver, A. Potts, E. Gay, B. W. Shotts, B. Hill, R. L. Anderson, and M. P. Weinstein Epidemic bacteremia due to Acinetobacter baumannii in five intensive care units. Am. J. Epidemiol. 132: Bergogne-Bérézin, E., and K. J. Towner Acinetobacter spp. as nosocomial pathogens: microbiological, clinical, and epidemiological features. Clin. Microbiol. Rev. 9: Bou, G., G. Cerveró, M. A. Domínguez, C. Quereda, and J. Martínez-Beltrán PCR-based DNA fingerprinting (REP-PCR, AP-PCR) and pulsed-field gel electrophoresis characterization of a nosocomial outbreak caused by imipenem- and meropenem-resistant Acinetobacter baumannii. Clin. Microbiol. Infect. 6: Bou, G., G. Cerveró, M. A. Domínguez, C. Quereda, and J. Martínez-Beltrán Characterization of a nosocomial outbreak caused by a multiresistant Acinetobacter baumannii strain with a carbapenem-hydrolyzing enzyme: high-level carbapenem resistance in A. baumannii is not due solely to the presence of -lactamases. J. Clin. Microbiol. 38: Cisneros, J. M., M. J. Reyes, J. Pachón, B. Becerril, F. J. Caballero, J. L. García-Garmendia, C. Ortiz, and A. R. Cobacho Bacteremia due to Acinetobacter baumannii: epidemiology, clinical findings, and prognostic features. Clin. Infect. Dis. 22: Dijkshoorn, L., H. M. Aucken, P. Gerner-Smidt, M. E. Kaufmann, J. Ursing, and T. L. Pitt Correlation of typing methods for Acinetobacter isolates from hospital outbreaks. J. Clin. Microbiol. 31: Garner, J. S., W. R. Jarvis, T. G. Emori, T. C. Horan, and J. M. Hughes CDC definitions for nosocomial infections. Am. J. Infect. Control 16: Gerner-Smidt, P., and W. Frederiksen Acinetobacter in Denmark: I. Taxonomy, antibiotic susceptibility, and pathogenicity of 112 clinical strains. Acta Pathol. Microbiol. Immunol. Scand. 101: Gerner-Smidt, P., and I. Tjernberg Acinetobacter in Denmark: II. Molecular studies of the Acinetobacter calcoaceticus-acinetobacter baumannii complex. Acta Pathol. Microbiol. Immunol. Scand. 101: Gerner-Smidt, P., I. Tjernberg, and J. Ursing Reliability of phenotypic tests for identification of Acinetobacter species. J. Clin. Microbiol. 29: Grundmann, H. J., K. J. Towner, L. Dijkshoorn, P. Gerner-Smidt, M. Maher, H. Seifert, and M. Vaneechoutte Multicenter study using standardized protocols and reagents for evaluation of reproducibility of PCRbased fingerprinting of Acinetobacter spp. J. Clin. Microbiol. 35: Horan, T. C., R. P. Gaynes, W. J. Martone, W. R. Jarvis, and T. G. Emori CDC definitions of nosocomial surgical site infections, 1992: a modification of CDC definitions of surgical wound infections. Infect. Control Hosp. Epidemiol. 13: Kämpfer, P., I. Tjernberg, and J. Ursing Numerical classification and identification of Acinetobacter genomic species. J. Appl. Bacteriol. 75: Knaus, W. A., E. A. Draper, D. P. Wagner, and J. E. Zimmerman APACHE-II: a severity of disease classification system. Crit. Care Med. 13: National Committee for Clinical Laboratory Standards Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 4th ed. Approved standard. NCCLS document M7-A4. National Committee for Clinical Laboratory Standards, Wayne, Pa.

5 VOL. 40, 2002 REP-PCR AND SOURCES OF A. BAUMANNII BACTEREMIA Peterson, E. M., K. D. Evans, J. T. Shigei, M. T. Pezzlo, and L. M. de la Maza Evaluation of four anti-microbic susceptibility testing systems for gram-negative bacilli. Am. J. Clin. Pathol. 86: Reboli, A. C., E. D. Houston, J. S. Monteforte, C. A. Wood, and R. J. Hamill Discrimination of epidemic and sporadic isolates of Acinetobacter baumannii by repetitive element PCR-mediated DNA fingerprinting. J. Clin. Microbiol. 32: Scerpella, E. G., A. R. Wanger, L. Armitage, P. Anderlini, and C. D. Ericsson Nosocomial outbreak caused by a multiresistant clone of Acinetobacter baumannii: results of the case-control and molecular epidemiologic investigations. Infect. Control Hosp. Epidemiol. 16: Seifert, H., and P. Gerner-Smidt Comparison of ribotyping and pulsed-field gel electrophoresis for molecular typing of Acinetobacter isolates. J. Clin. Microbiol. 33: Seifert, H., A. Schulze, R. Baginski, and G. Pulverer Comparison of four different methods for epidemiologic typing of Acinetobacter baumannii. J. Clin. Microbiol. 32: Seifert, H., A. Strate, and G. Pulverer Nosocomial bacteremia due to Acinetobacter baumannii. Clinical features, epidemiology, and predictors of mortality. Medicine 74: Siau, H., K. Y. Yuen, P. L. Ho, S. S. Wong, and P. C. Woo Acinetobacter bacteremia in Hong Kong: prospective study and review. Clin. Infect. Dis. 28: Snelling, A. M., P. Gerner-Smidt, P. M. Hawkey, J. Heritage, P. Parnell, C. Porter, A. R. Bodenham, and T. Inglis Validation of use of whole-cell repetitive extragenic palindromic sequence-based PCR (REP-PCR) for typing strains belonging to the Acinetobacter calcoaceticus-acinetobacter baumannii complex and application of the method to the investigation of a hospital outbreak. J. Clin. Microbiol. 34: Traub, W. H., and M. Spohr Antimicrobial drug susceptibility of clinical isolates of Acinetobacter species (A. baumannii. A. haemolyticus, genospecies 3, and genospecies 6). Antimicrob. Agents Chemother. 33: Vila, J., M. A. Marcos, and M. T. Jiménez de Anta A comparative study of different PCR-based DNA fingerprinting techniques for typing of the Acinetobacter calcoaceticus-a. baumannii complex. J. Med. Microbiol. 44: Wisplinghoff, H., M. B. Edmond, M. A. Pfaller, R. N. Jones, R. P. Wenzel, and H. Seifert Nosocomial bloodstream infections caused by Acinetobacter species in United States hospitals: clinical features, molecular epidemiology, and antimicrobial susceptibility. Clin. Infect. Dis. 31: Downloaded from on November 18, 2018 by guest