Nosocomial Klebsiella pneumoniae Bacteremia: Clinical Features and Antimicrobial Susceptibilities of Isolates

全文

(1)156. Nosocomial Klebsiella pneumoniae Bacteremia: Clinical Features and Antimicrobial Susceptibilities of Isolates 1. Wen-Liang Yu , Cheng-Wen Lin , Jen-Heisen Wang, Huei-Shan Cheng 1. 2. 2. Section of Infectious Diseases, Department of Medicine, Department of Clinical Laboratory, the Committee of 1. Nosocomial Infection Control, China Medical College Hospital; School of Medical Technology, China Medical College, Taiwan, R.O.C.. From July 1997 through June 1998, 24 consecutive adult patients with nosocomial Klebsiella pneumoniae bacteremia were diagnosed at the China Medical College Hospital. Herein, we present the clinical features and the prognostic factors of the outcome in these 24 patients. All these patients had one or more underlying diseases including liver cirrhosis, diabetes mellitus and malignancy. Localized infections were noted in nine (37%) patients, including six with pneumonia, and three with intra-abdominal infection. Of the 15 (63%) patients with primary bacteremia, three met the criteria for catheter-related infection. The portal of entry was unknown in the remaining 12 patients. Of the 16 patients with central venous catheters in place at the onset of bacteremia, four of the six removed catheters were sent for tip culture, and three grew K. pneumoniae. The survival rate was significantly higher in patients whose catheters were removed than in those whose catheters remained (83% vs 10%, p = 0.008). The overall mortality rate was 58% (14/24); 11 (46%) patients died of K. pneumoniae bacteremia. Mutivariate analysis demonstrated that patients with central venous catheters and high acute physiology and chronic health evaluation (APACHE) II scores had significantly higher risks for death. All isolates were susceptible to imipenem; 96% of the strains were susceptible to moxalactam, 71% to ceftazidime, 67% to amoxicillin/clavulanic acid, 58% to aztreonam, ceftriaxone and cefotaxime, 54% to cephalothin, gentamicin and amikacin; however, none was susceptible to ampicillin. In conclusion, to improve the outcome of patients with nosocomial K. pneumoniae bacteremia, early removal of the indwelling central venous catheter with tip culture is mandatory. Moxalactam and imipenem should be drugs of choice for the treatment of multi-drug resistant K. pneumoniae bacteremia. ( Mid Taiwan J Med 1999;4:156-63 ). Key words Klebsiella pneumoniae, nosocomial bacteremia. INTRODUCTION. noteworthy that several investigations have. Klebsiella pneumoniae, a member of the. shown an increasing trend in the nosocomial. family Enterobacteriaceae, is an important. bloodstream infection (BSI) rate [2,3]. Gram-. pathogen of nosocomial infections [1]. It is. negative bacilli have been recognized as the. Received December 28, 1998.. Revised March 3, 1999.. Accepted May 19, 1999.. most important pathogens in nosocomial BSIs [4]. K. pneumoniae has become the most. Wen-Liang Yu, Section of. common gram-negative pathogen of noso-. Infectious Diseases, Department of Medicine China Medical. comial BSI at the China Medical College. Address reprint requests to. College Hospital, No 2, Yuh-Der Road, Taichung 404, Taiwan, R.O.C.. Hospital [5]. The predisposing factors for.

(2) Wen-Liang Yu, et al.. 157. infection by this organism include malignancy. bacteremia without an obvious source.. or chronic debilitating diseases, indwelling. Patients with intravascular catheter-associated. intravenous catheter or urinary catheter, and. infection were included in this category [9].. prior antimicrobial therapy [4]. Recently, strains. Central venous catheter (CVC)-related infec-. of K. pneumoniae with multi-drug resistance. tion was defined as local inflammation at the. to newer cephalosporins have developed [6]. In. catheter site with concordant growth of the. Taiwan, several studies have also suggested. organism from blood and catheter tip cultures. that nosocomial K. pneumoniae isolates are. or exit site, in the absence of other obvious. usually resistant to some third-generation. localized infections [10]. Using a semiquan-. cephalosporins, including cefotaxime, cef-. titative procedure (roll plate method), a. triaxone, ceftazidime and aztreonam [7,8].. positive catheter tip culture was considered. Antimicrobial therapy is usually started. when 15 or more colonies were obtained from. empirically, based on the patient’s condition. a rolled out catheter [11]. Severity of illness was. including underlying disease, severity of. graded using the Acute Physiology and. illness, primary or secondary infection, and. Chronic Health Evaluation (APACHE) II. antimicrobial activities against the organisms.. scoring system at the onset of bacteremia [12].. We conducted this study to evaluate the. Empirical antibiotic prescription was. clinical features, predisposing factors, antimi-. started at the onset of bacteremia and was. crobial susceptibility profiles of isolates and. adjusted according to antimicrobial suscep-. outcomes in patients with nosocomial K.. tibility testing. The antibiotic regimen was. pneumoniae bacteremia.. considered to be appropriate only if the infecting organism was susceptible to at least one of the administered drugs for adequate. MATERIALS AND METHODS. Patient Selection and Definitions From July 1, 1997 through June 30, 1998, 24. duration (> 3 days) of therapy. If the infection caused by K. pneumoniae was under control at the time of death, the mortality was. consecutive adult patients with nosocomial K.. categorized as unrelated.. pneumoniae bacteremia diagnosed at the. Bacteriology. China Medical College Hospital were prospec-. Blood specimens were processed with the. tively enrolled into this study. Nosocomial K.. BACTEC NR-860 system (Becton Dickinson. pneumoniae bacteremia was defined as two. Microbiology System, Cockeysville, Mary-. or more separate blood cultures which. land). K. pneumoniae was identified on the. yielded K. pneumoniae or a single blood. basis of colonial morphology in the standard. culture which yielded K. pneumoniae and. biochemical tests [13]. This organism was also. associated systemic signs of infection in. confirmed using the VITEK system (Bio-. patients who has been admitted to the. Merieux Vitek, Inc, Hazelwood, Missouri, USA).. hospital for more than 48 hours. The date of. For the disk diffusion test, a 0.5 McFarland. specimen collection of the first positive. inoculum was placed on Mueller-Hinton agar. culture was regarded the onset time of. (BBL Sensi-Disc, Becton Dickinson Micro-. bacteremia. Polymicrobial bacteremia was. biology System, Cockeysville, Maryland, USA),. defined as a bacteremic episode in which. followed by various antibiotic discs [14]. The. more than one organism was simultaneously. antibiotics routinely tested were ampicillin,. isolated from blood culture. A localized. cephalothin, gentamicin, amikacin, amoxi-. infection was considered the portal of entry of. cillin/clavulanic acid, cefonicid, aztreonam,. K. pneumoniae bacteremia if it was microbi-. ceftazidime, cefotaxime, ceftriaxone, moxalac-. ologically and clinically documented.. tam, and imipenem (BBL Sensi-Disc, Becton. Primary bacteremia was defined as. Dickinson)..

(3) Nosocomial Klebsiella pneumoniae Bacteremia. 158. Table 1 . Clinical features of 24 patients with nosocomial K. pneumoniae bacteremia Patient No/age (yr) /sex. Underlying disease. 1/80/M 2/78/M. CVA CHF. 3/77/F. Cirrhosis. 4/74/F. Hepatoma Cirrhosis DM Cirrhosis ESRD CHF CVA. 5/74/M 6/73/F 7/70/F 8/69//F 9/67/F 10/66/F 11/66/F. DM CHF DM CHF CVA. Clinical diagnosis. Prior hospital day. Devices implanted. Pneumonia Pneumonia ARDS Primary bacteremia Primary bacteremia Primary bacteremia Pneumonia. 14 51. None Ventilator CVC CVC**. Primary bacteremia Primary bacteremia Primary bacteremia Subhepatic abscess Peritonitis. 44. None ND. S D (1d). 11,170. 21. ND. S. ND. S. None ND. D (25d) (unrelated) D (1d). 13. 6. None. 18,800. 19. Ventilator CVC CVC. 12,870. 26. 14,320. 28. Ceftazidime. ND. D (12d). 3. CVC. 22,770. 26. Inappropriate. ND. D (1d). 14. CVC. 3,170. 25. ND. D (1d). 2. None. 15,050. 15. Cephradine Gentamicin Ceftazidime. None. 40. CVC**. 13,200. 20. Imipenem. ( ). D (20d) (unrelated) D (15d). Pneumonia. 6. None. 24,310. 25. Cefotaxime. None. S. Primary bacteremia Pneumonia. 5. CVC. 4,760. 24. Cephradine. ND. D (1d). 7,870. 25. Inappropriate. ND. D (2d). 10. Ventilator CVC None. 12,950. 13. None. S. 78. CVC. 45,050. 22. Cefotaxime Gentamicin Inappropriate. ND. D (14d). 102. CVC**. 13,740. 20. Imipenem. (+). S. 16. None. 18,140. 17. Ceftazidime. None. S. Ventilator CVC None. 4,700. 12. Inappropriate. ND. D (3d). 5,160. 26. ND. D (2d). 25. CVC**. 5,320. 20. Cephradine Gentamicin Imipenem. (+). S. 20. CVC. 140. 16. ND. S. 12. CVC**. 8,460. 8. Ceftriaxone Amikacin Cefotaxime. (+). S. 3. None. 22,000. 24. Ceftriaxone. None. D (54d) (unrelated). 16/61/M. 19/59/F. Colon cancer DM Primary ESRD bacteremia* Hepatoma Primary Cirrhosis bacteremia Lung cancer Pneumonia. 20/55/M. Cirrhosis. 21/50/F. ESRD. 22/44/M. Leukemia. 23/41/M. Esophageal cancer ESRD. 24/23/F. Moxalactam Ceftazidime. 2,510. Cirrhosis. 18/60/F. 18 26. CVC**. 15/61/F. 17/60/F. 36,200 3,530. 14. 14/64/M. 13/65/F. CVC tip Outcome culture. Cephradine Gentamicin Cephradine Gentamicin Cephradine Gentamicin Inappropriate. DM CVA DM COPD Cirrhosis ESRD ARDS. 12/66/F. 3. WBC 3 APACHE Antibiotic (per mm ) II score treatment. Primary bacteremia Peritonitis. Primary bacteremia Primary bacteremia* Primary bacteremia Primary bacteremia* Primary bacteremia. 14. 10. 7 3. Note. CVA = cerebral vascular accident; CHF = congestive heart failure; DM = diabetes mellitus; COPD = chronic obstructive pulmonary disease; ESRD = end stage renal disease; ARDS = acute respiratory distress syndrome; CVC = central vascular catheter; WBC = white blood cell count; ND = not done; (+) = positive result of culture for K. pneumoniae; (. ) = negative result of cultures for K. pneumoniae; S = survived and discharged; D = died; d =. days after onset of bacteremia. * = catheter-related infection; ** = CVC removed. 2. Statistical analysis was done using or Fisher's exact tests for categorical variables and. Student’s t -test for continuous variables. Predictors of survivors were determined according to logistic regression by both.

(4) Wen-Liang Yu, et al.. 159. Table 2. Clinic al manif est at ions of 24 pat ient s wit h nosocomial K. pneumoniae bacteremia Manifestation. No. (%) of patients. Fever ( > 38 0C) Leukocytosis ( WBC count > 10,000/mm 3) Thrombocytopenia (platelet count < 100,000/mm 3) Jaundice (total bilirubin > 2.0 /dl) Leukopenia (WBC count < 4,000/mm 3) Hypoxemia (PaO2 < 60 mmHg) Septic shock Metabolic acidosis (pH < 7.20) Disseminated intravascular coagulation. 22 (92) 14 (64) 10 (42) 6 (25) 4 (17) 3 (14) 2 (8) 2 (8) 1 (5). univariate and multivariate analyses using the SAS 6.11 computer package. A p value of less than 0.05 was considered statistically significant.. RESULTS. Clinical Features Twenty-four adult patients (M : F = 8 : 16), mean age 63 years (range, 23 to 80 years) , were diagnosed with nosocomial K. pneumoniae bacteremia during the study period. Their demographic data and clinical characteristics are summarized in Table 1. All of them had clinical features of sepsis and at least one severe underlying disease, including liver cirrhosis (n = 7), diabetes mellitus (n = 6), malignancy (n = 6), end-stage renal disease (n = 5), cerebral vascular accident (n = 4), congestive heart failure (n = 4), chronic obstructive pulmonary disease (n = 1) and acute respiratory distress syndrome (n = 1). Nine patients contracted the infection in intensive care units. The mean APACHE II score was 19 (range, 8 to 28). The mean white 9 blood cell count was 13.6 10 (range, 1.4 to 45 9 10 ). The clinical manifestations of these patients are listed in Table 2. Localized infections were noted in nine (37%) patients, including six with pneumonia. and three with intra-abdominal infection. Although 16 patients had CVC in place at the onset of bacteremia, only four of the six catheters were removed and sent for tip culture; three grew K. pneumoniae. The duration of CVC insertion before the onset of bacteremia ranged from 3 to 17 days (median, 8 days). Six patients with localized infections also had indwelling CVC at the onset of bacteremia. Of the 15 (63%) patients with primary K. pneumoniae bacteremia, three met the criteria of catheter-related infection. The portal of entry was unknown in the remaining 12 (50%) patients. Four patients had concurrent bacteremia, including two with Enterococcus (patients 12 and 13) and one each with Burkholderia cepacia (patient 11) and Escherichia coli (patient 20). Two (patients 11 and 12) were associated with localized infections. Antimicrobial Therapy and Outcome The overall mortality rate was 58% (14/24) in which 11 (46%) died of K. pneumoniae bacteremia. Appropriate antibiotics were administered in 19 (79%) patients, including six with first-generation cephalosporin, 10 with third-generation cephalosporin and three with imipenem. However, nine of them died. Six died of K. pneumoniae bacteremia and the remaining three died of hepatic failure (patient 5), recurrent intracerebral hemorrhage (patient 10) and uremia (patient 24), respectively. Patient 10 developed intracerebral hemorrhage and subsequent nosocomial Pseudomonas aeuginosa pneumonia with acute respiratory distress syndrome. These three patients had been afebrile for at least one week after adequate treatment for the infections caused by K. pneumoniae, which were under control at the time of death. The efficacy of appropriate antimicrobial therapy was 68% (13/19). All of the five patients with inappropriate antimicrobial therapy died; the causes of death in these patients were all related to K. pneumoniae bacteremia. In the comparison of the 11 patients who died of K. pneumoniae related bacteremia with the 13 survivors (including three unrelated moralities), significant differences.

(5) Nosocomial Klebsiella pneumoniae Bacteremia. 160. Table 3. Univariate analysis of the predictors attributable to fatality in patients with nosocomial K. pneumoniae bacteremia Parameter. Age (mean SD, yr) Sex (M : F) Underlying disease DM Cirrhosis Malignancy CVA CHF ESRD Clinical syndrome Primary bacteremia Pneumonia Intra-abdominal infection Central venous catheter Prior hospital day (median) Polymicrobial bacteremia 9 WBC ( 10 /L, mean SD) APACHE II score (mean SD) Appropriate antimicrobial therapy. Surviving patients* (N = 13). p value. Deceased patients (N = 11). 0.2 1. 66 6 4:7. 59 6 4:9. Odds ratio. 95% CI. 1.051 0.778. 0.974 0.142. 1.134 4.265. 3 5 4 2 0 3. 3 2 2 2 4 2. 1 0.386 0.649 1 0.031 1. 1.250 0.356 0.5 1.222 16.2 0.741. 0.196 0.053 0.072 0.143 0.763 0.100. 7.956 2.368 3.454 10.480 343.80 5.490. 10 2 1. 5 4 2. 0.206 0.357 0.576. 0.25 3.143 2.667. 0.043 0.450 0.208. 1.443 21.958 34.197. 0.039 0.527 0.3 0.642 0.017. 11.67 1.011 4.5 0.982 1.358. 1.139 0.978 0.395 0.908 1.056. 119.542 1.045 51.298 1.061 1.747. 0.011. 0.044. 0.002. 0.917. 10 14 3 12.49 12.38 24 4 6. 6 12 1 14.54 9.76 18 5 13. *: including mortality unrelated to K. pneumoniae bacteremia. CI = confidence interval; SD = standard deviation; DM = diabetes mellitus; CVA = cerebral vascular accident; CHF = congestive heart failure; ESRD = end-stage renal disease. Table 4. Antibiotic susceptibilities of 24 strains of K. pneumoniae isolated from nosocomial bacteremia Antibiotics Imipenem Moxalactam Ceftazidime Amoxicillin/clavulanic acid Aztreonam Ceftriaxone Cefotaxime Cefonicid Cephalothin Amikacin Gentamicin Ampicillin. No. (%) of with susceptibility 24 23 17 16 14 14 14 13 13 13 13 0. (100) (96) (71) (67) (58) (58) (58) (54) (54) (54) (54) (0). were noted in four parameters which included congestive heart failure (p = 0.031), indwelling CVC (p = 0.039), APACHE II score (p = 0.017) and antibiotic treatment (p = 0.011) (Table 3). Multivariate analysis of these parameters using logistic regression revealed indwelling CVC (p = 0.039, odds ratio = 19.778, 95% CI = 1.169 to 334.519) and APACHE II score (p = 0.029, odds. ratio = 1.410, 95% CI = 1.036 to 1.918) were independent risk factors. Of the 16 patients with indwelling CVC at the onset of bacteremia, the survival rate was significantly higher in patients whose catheters were removed (5/6, 83%) than in those whose catheters remained in place (1/10, 10%; p = 0.008). Antimicrobial Susceptibility The results of antimicrobial susceptibility test were shown in Table 4. All the isolates were susceptible to imipenem but resistant to ampicillin. All but one were susceptible to moxalactam. DISCUSSION. K. pneumoniae has been recognized as a major cause of nosocomial infection [1,2]. Colonization in the bowel, oropharynx, or tracheostomy in hospitalized patients has been considered the main reservoir of Klebsiella [1,15]. Yinnon et al reported that the most frequent illnesses associated with nosocomial.

(6) Wen-Liang Yu, et al.. K. pneumoniae bacteremia were urinary tract infection (28%) and pneumonia (25%), while only seven of 40 (17%) had primary bacteremia [16]. In Taiwan, however, the first report of K. pneumoniae bacteremia did not demonstrate the common portals of entry in nosocomial cases [7]. In a later review, the most frequent portals of entry in 88 patients with nosocomial K. pneumoniae bacteremia were pneumonia (30%), unknown origin (17%) and intravascular catheter-related infection (10%) [8]. In contrast to previous reports, 15 (63%) of our patients had primary bacteremia. Only six patients had pneumonia and none had urinary tract infection. Sixteen patients had central venous catheters in place at the onset of bacteremia, however, only four of the six removed catheters were sent for tip culture and three proved to have CVC-related infections. Ten patients had continuous CVC insertion leading to a high mortality rate. Since most catheter tips were not sent for culture, this may be partially due to the clinician’s lack of concern regarding the existence of CVCrelated infection, which resulted in a failure to properly remove the involved catheter. We believe CVC-related infection may not have been limited to the three cases if routine catheter tip culture had been performed. In our study, the most frequent clinical manifestations included fever (92%), leukocytosis (64%) and thrombocytopenia (42%). This is similar to the report by Hu et al, which demonstrated patients with fever (84%), thrombocytopenia (56%) and leukocytosis (56%) [8]. However, Hu et al reported 28 (32%) patients had septic shock, while only two of our 24 patients had septic shock at onset. Four (17%) of our patients were associated with polymicrobial bacteremia as in agreement with previous study [8]. Although three of them died, there was no statistical significance to predict the outcome. The overall patient fatality rate was 58% (14/24) with 46% (11/24) died of K. pneumoniae bacteremia. The overall mortality rate of nosocomial K. pneumoniae bacteremia in previous reports varied, from 33% to 55% [1,7,8,16]. It has been suggested that the high. 161. mortality rate reflected the severity of the patients’ underlying disease [1,7]. APACHE II, which is a classification system for disease severity, is useful in predicting the subsequent risk of hospital death for acutely ill patients [12]. In our study, the patients who died had a significantly higher APACHE II score than those who survived. Another significant prognostic factor by multivariate analysis was indwelling CVC at the onset of bacteremia. Early removal of CVC should be considered in order to reduce the mortality rate of patients with nosocomial K. pneumoniae bacteremia, with an indwelling central venous catheter. Treatment failure has been attributed to the emerging antibiotic resistance of K. pneumoniae [1,16]. Recently, extended-spectrum -lactamase (ESBL)-producing strain of K. pneumoniae resistant to broad-spectrum antibiotics has caused serious infections [17,18]. ESBL has varied resistance to third-generation cephalosporins, but it is uniquely susceptible to cephamycins (cefoxitin, cefmetazole and moxalactam) and carbapenem (imipenem and meropenem) [19]. Treatment with an apparently susceptible third-generation cephalosporin may lead to clinical failure due to ESBL production [20]. In our study, all 24 K. pneumoniae isolates were resistant to ampicillin. Nine strains exhibited resistance to cephalothin, cefonicid, a third-generation cephalosporin (cefotaxime, ceftriaxone, ceftazidime, or aztreonam), but they were susceptible to moxalactam and imipenem, which could be considered to have an ESBLrelated resistance phenotype [19]. The true prevalence of ESBL-producing strain of nosocomial Klebsiella isolates needs further investigation. Of the nine strains with ESBLrelated resistance phenotype, three were ceftazidime-sensitive. This phenomenon may be related to clinical failure with ceftazidime therapy. Our result is similar to a recent report [20], in which up to 33% of putative ESBLproducing strains were apparently susceptible to third-generation cephalosporins or aztreonam according to disc diffusion method. In our study, as reported by Hu et al [8], susceptibilities of K. pneumoniae isolates to.

(7) Nosocomial Klebsiella pneumoniae Bacteremia. 162. moxalactam and imipenem were excellent. In conclusion, nosocomial K. pneumoniae bacteremia often develops in patients with severe underlying diseases, particularly in those with indwelling central venous catheters. Patients with indwelling central venous catheters at the onset of bacteremia and high APACHE II scores had significantly higher mortality rate. CVC-related bacteremia is highly suggested in patients with nosocomial K. pneumoniae bacteremia if there is no other infectious focus caused by this organism. Removal of the CVC and semiquantitative culture for the catheter tip may improve the diagnosis and survival. For the treatment of infections caused by multi-drug resistant K. pneumoniae, appropriate antimicrobial agents (such as moxalactam and imipenem) should be considered the drugs of choice.. Formos Med Assoc 1990;89:756-63. 8.. Hu BS, Lau YJ, Shi-ZY, et al. Nosocomial Klebsiella pneumoniae bacteremia. Nosocom Infect Control J 1997;7:286-92.. 9.. Banerjee SN, Emori TG, Culvr DH, et al. Secular trends in nosocomial primary bloodstream infections in the United States, 1980-1989. National Nosocomial Infections Surveillance System. Am J Med 1991 ;91(Suppl 3B):86-9.. 10. Dickinson GM, Bisno AL. Infection associated with indwelling devices: concepts of pathogenesis; infections associated with intravascular devices. [Review] Antimicrob Agents Chemother 1989;33:597601. 11. Maki DG, Weise CE, Sarafin HW: A semiquantitative culture method for identifying intravenous-catheterrelated infections. N Engl J Med 1977;296:1305-9. 12. Knaus WA, Draper EA, Wanger DP, et al. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818-29. 13. Farmer III JJ. Enterobacteriaceae: Introduction and identification. In: Murray PR, Pfaller MA, Tenover FC, Yolken RH, eds. Manual of Clinical Microbiology.. ACKNOWLEDGMENTS. This study was supported by a grant from the China Medical College (CMC86-TH-03).. American Society for Microbiology: Washington D.C., 1995:438-49. 14. National Committee for Clinical Laboratory Standards. 1990. Approved Standard M2-A4 Performance standards for antimicrobial disc susceptibility tests, 4th ed. NCCLS, Villanova, Pa. 15. Montgomerie JZ. Epidemiology of Klebsiella and. REFERENCES 1.. Terman JW, Aflord RH, Bryant RE. Hospital-acquired Klebsiella bacteremia. Am J Med Sci 1972;264:191-6.. 2. 3.. 4.. bacteremia: community versus nosocomial infection.. overview. Am J Med 1981;70:719-32.. QJM 1996;89:933-41.. Morrison AJ Jr, Freer CV, Searcy MA, et al.. 17. Sirot J, Chanal C, Petit A, et al. Klebsiella pneu-. Nosocomial bloodstream infections: secular trends in. moniae and other Enterobacteriaceae producing. a statewide surveillance program in Virginia. Infect. novel plasmid-mediated beta-lactamases markedly. Control 1986;7:550-3.. active against third-generation cephalosporins:. McGowan JE Jr, Barnes MW, Finland M. Bacteremia. epidemiologic studies. [Review] Rev Infect Dis. during 12 selected years (1935-1972), with special. -lactamases conferring transfer-. broad-spectrum. 1975;132:316-35.. able resistance to newer beta-lactam agents in. Yu WL, Lea KC, Huang CC. Major trends in the. Enterobacteriaceae: hospital prevalence and. teaching hospital in Taichung, 1991 to 1995. Med J. susceptibility patterns. Rev Infect Dis 1988;10:867-78. 19. Jan IS, Hsueh PR, Teng LJ, et al. Antimicrobial. CMCH 1997;2:70-5.. susceptibility testing for Klebsiella pneumoniae. Meyer KS, Urban C, Eagan JA, et al. Nosocomial. isolates resistant to extended-spectrum beta-lactam. outbreak of Klebsiella infection resistant to lategeneration cephalosporins. Ann Intern Med 1993;119:353-8. 7.. 1988;10:850-9. 18. Jarlier V, Nicolas MH, Fournier G, et al. Extended. reference to hospital-acquired cases. J Infect Dis. etiology of nosocomial bloodstream infections at a. 6.. 1979;1:736-53. 16. Yinnon AM, Butnaru A, Raveh D, et al. Klebsiella. Maki D. Nosocomial bacteremia: An epidemiologic. at Boston City Hospital: Occurrence and mortality. 5.. hospital-associated infections. Rev Infect Dis. antibiotics. J Formos Med Assoc 1998;97:661-6. 20. Livermore DM, Yuan M. Antibiotic resistance and production of extended-spectrum beta-lactamaes. Wang LS, Lee FY, Cheng DL, et al. Klebsiella. amongst Klebsiella spp. from intensive care units in. pneumoniae bacteremia: analysis of 100 episodes. J. Europe. J Antimicrob Chemother 1996;38:409-24..

(8) 163. Klebsiella pneumoniae 1. 2 2 1. 1997. 7. K. pneumoniae (29%). 1998. 6. 24. (25%). (25%). (25%). (12%). (63%). K. pneumoniae. 16. 50%. 3 17%. 58%. 46%. K. pneumoniae APACHE II. ampicillin. imipenem. moxalactam (96%). ceftazidime (71%) ceftriaxone (58%). amoxicillin/clavulanic acid (67%) cefotaxime (58%). gentamicin (54%). cefonicid (54%). amikacin (54%) K. pneumoniae. imipenem 1999;4:156-63. 404 12/28/1998 5/19/1999. 91 3/3/1999. aztreonam (58%) cephalothin (54%) moxalactam.

(9)