The status of antimicrobial resistance in Taiwan among gram-negative pathogens: the Taiwan surveillance of antimicrobial resistance (TSAR) program, 2000

Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

www.elsevier.com/locate/diagmicrobio

Surveillance

The status of antimicrobial resistance in Taiwan among gram-negative pathogens: the Taiwan surveillance of antimicrobial resistance (TSAR) program, 2000 Tsai-Ling Lauderdale*, L. Clifford McDonald, Yih-Ru Shiau, Pei-Chen Chen, Hui-Ying Wang, Jui-Fen Lai, Monto Ho, TSAR Participating Hospitals Division of Clinical Research, National Health Research Institutes, Taipei, Taiwan, R.O.C. Received 15 July 2003; accepted 6 October 2003

Abstract In a nationwide surveillance of antimicrobial resistance (Taiwan Surveillance of Antimicrobial Resistance, TSAR), isolates were collected from 21 medical centers and regional hospitals throughout Taiwan over a three-month period in 2000 (TSAR II). This report summarizes susceptibility data of 7 common Gram-negative bacilli (Escherichia coli, Klebsiella pneumoniae, Serratia marcescens, Proteus mirabilis, Enterobacter cloacae, Pseudomonas aeruginosa, and Acinetobacter baumannii) in the TSAR II collection and compared selected key forms of resistance by epidemiologic factors and with isolates collected in 1998 (TSAR I) as well as with data from international surveillance studies. Resistance of the 5 Enterobacteriaceae species to most of the commonly prescribed “first-line” antimicrobials in Taiwan, such as ampicillin (78% in E. coli, 68% in P. mirabilis), gentamicin (19% in K. pneumonia to 66% in S. marcescens), and trimethoprim/sulfamethoxazole (29% in K. pneumoniae to 70% in P. mirabilis), was high, several of which are higher than other countries. Resistance to certain broad-spectrum antimicrobials is also more acute in Taiwan than most Western countries, such as ceftazidime resistant A. baumannii (73%) and ciprofloxacin resistant E. coli (12%). Differences in geographic regions and specimen types were associated with certain forms of resistance in TSAR II; however, the resistance problem is prevalent among both inpatients and outpatients of not only medical centers but also regional hospitals throughout Taiwan. © 2004 Elsevier Inc. All rights reserved. Keywords: Gram-negative bacteria; Antimicrobial; Antibiotic resistance in Taiwan

1. Introduction Antimicrobial resistance is an increasing global problem in developing countries as well as in developed countries (Kunin 1993; O’Brien 1997). Increasing antimicrobial resistance contributes to morbidity, mortality, and increased health care costs resulting from treatment failures and longer hospital stays (McGowan 2001). One of the first steps in the containment of antimicrobial resistance is surveillance. Several regional and global surveillance programs provide valuable resistance data that help focus international control efforts (Bax et al., 2001; Fridkin 1999; Jones and Masterton, 2001; Lewis, 2002; Livermore, 1998; Masterton, 2000). National surveillance is important for * Corresponding author. Tel.: ⫹886-2-26534401; fax: ⫹886-227890254 E-mail address: [email protected] (T.L. Lauderdale). 0732-8893/04/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.diagmicrobio.2003.10.005

monitoring current resistance rates and aids in pinpointing specific resistance problems that may exist locally. It also allows for comparison with other countries in the world and for assessing the impact of interventions. Understanding of local resistance data are also important in the global public health management of emerging resistance since resistant organisms may be acquired through international travel (Okeke and Edelman, 2001; Smith et al., 1999). Taiwan is a small mountainous island with a population of around 22 million. The majority of people live along the western coastal area with the North and the East being the most and least densely populated regions, respectively. In March 1995 Taiwan instituted a national health insurance system that covers over 95% of the population. The Bureau of National Health Insurance regulates the use of antimicrobials and first-line antimicrobials must be used unless there is documented resistance or failure of treatment or severe infections (Chang et al., 2001). The first-line antimi-

212

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

crobials include penicillinase-labile and –stable penicillins, first generation cephalosporins, gentamicin, erythromycin, tetracycline, clindamycin, and chloramphenicol. In 1997 the National Health Research Institutes (NHRI) of Taiwan embarked on a national program to survey antimicrobial resistance in Taiwan and to promote national action in controlling antimicrobial resistance. The Taiwan Surveillance of Antimicrobial Resistance (TSAR) was initiated following the pattern of successful national and international surveillance programs, such as SENTRY (Jones et al., 1997; Pfaller et al., 1998). The first round of TSAR was conducted in the fall of 1998 (TSAR I) with 44 hospitals including medical centers and regional hospitals from 4 geographic regions [North, Middle (West), South, and East] of Taiwan participating. The results of isolates from half of the 44 TSAR I hospitals were published and revealed that the major resistance problems in Taiwan were associated with the commonly used “first-line” antimicrobials (Ho et al., 1999; Ho 2001). The resistance data on the Gramnegative bacilli isolates from all 44 hospitals were later analyzed and found to be similar to those already published. The second round of national surveillance (TSAR II) was conducted in the spring of 2000. This report summarizes antimicrobial activities against the seven most common Gram-negative bacilli (GNB) in the TSAR II collection, namely: Escherichia coli, Klebsiella pneumoniae, Serratia marcescens, Proteus mirabilis, Enterobacter cloacae, Pseudomonas aeruginosa, and Acinetobacter baumannii. We also identified epidemiologic factors associated with several clinically important forms of resistance that were selected on the basis of these being either internationally recognized forms of emerging resistance, unusually high levels of a more common form of resistance, or resistance to a drug for which there are few therapeutic options to treat resistant phenotypes. Key forms of resistance rates between TSAR I and TSAR II isolates were also compared. Finally, we compared the resistance rates of TSAR II GNB isolates with the rates reported by the SENTRY program from their international and national surveillance studies.

2. Materials and methods 2.1. Participating hospitals A total of 21 hospitals including 8 medical centers and 13 regional hospitals representing 35% (8/23) and 20% (13/64) of the hospitals accredited in these two categories in Taiwan in 2000 participated in TSAR II; 18 of these also participated in TSAR I. The distribution of TSAR hospitals is shown in Fig. 1. The status of medical centers, regional, and local hospitals is accredited and updated every three years by the Department of Health in Taiwan based on the number of hospital beds, space, intensive care units, services, patient/physician ratio, specialties offered by the hospital, and medical school affiliation. The average number of

Fig. 1. Distribution of TSAR I and TSAR II hospitals in Taiwan.

beds for medical centers and regional hospitals was 1195, and 535, respectively. 2.2. Organisms For TSAR II, isolates were collected by the participating hospitals from March to May 2000. Non-duplicate sequential isolates were collected in each of several different categories including 25 isolates each from out-patient/emergency room patients (OPD), pediatric patients, patients with bacteremia (positive blood cultures), and adult intensive care unit (ICU) patients. In addition, 50 adult consecutive, non-blood, isolates were collected. All isolates were stored frozen in bead-containing Microbank cryovials (PRO-LAB Diagnostics, Austin, TX, USA) by the hospitals and sent frozen to our laboratory where they were stored at –70°C. At our laboratory, each isolate was subcultured onto appropriate agar plates (BBL, Becton Dickinson Microbiology System, Cockeysville, MD,USA) to check for purity and identification. Species identification was made using a combination of standard conventional biochemical tests (BBL) (Murray et al., 1999), Vitek Gram-negative Identification Plus cards (bioMe´ rieux Vitek, Inc. Hazelwood, MO, USA), and API 20E or 32GN (bioMe´ rieux, Marcy l’Etoile, France).

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

213

Table 1 Specimen source, patient location, hospital type of the 7 most common gram-negative bacilli collected during TSAR II Number of isolates, (%) Specimen source Blood Urine Respiratory-tract Wound* Other Patient location ICU Non-ICU OPD/ER Hospital type Medical Centers Regional Hosp.

E. coli (N ⫽ 591)

K. pneumoniae (N ⫽ 233)

S. marcescens (N ⫽ 105)

P. mirabilis (N ⫽ 92)

E. cloacae (N ⫽ 69)

P. aeruginosa (N ⫽ 328)

A. baumannii (N ⫽ 111)

Total (N ⫽1529)

161 (27) 318 (54) 22 (4) 57 (10) 33 (5)

68 (29) 38 (16) 87 (37) 20 (9) 20 (9)

17 (16) 28 (26) 38 (36) 13 (12) 9 (9)

9 (10) 36 (39) 23 (25) 19 (21) 5 (5)

15 (22) 10 (14) 20 (29) 15 (22) 9 (9)

29 (9) 36 (11) 176 (54) 66 (20) 21 (6)

8 (7) 8 (7) 65 (59) 19 (17) 11 (10)

307 (20) 474 (31) 431 (28) 209 (14) 108 (7)

78 (13) 297 (50) 215 (37)

67 (29) 115 (49) 51 (22)

42 (40) 49 (47) 14 (13)

25 (28) 39 (42) 28 (30)

26 (37) 33 (48) 10 (15)

114 (35) 165 (50) 49 (15)

66 (59) 37 (34) 8 (7)

418 (27) 735 (48) 375 (25)

228 (37) 363 (63)

95 (41) 138 (59)

54 (51) 51 (49)

29 (32) 63 (68)

25 (37) 44 (63)

157 (48) 171 (52)

57 (51) 54 (49)

645 (42) 884 (58)

* Including pus.

2.3. Antimicrobial susceptibility testing Minimum inhibitory concentrations (MIC) were determined using the broth micro-dilution method following the guidelines of the National Committee for Clinical Laboratory Standards (NCCLS) (NCCLS, 2000; NCCLS, 2001). Custom-designed 96 well panels (Sensititre, Trek Diagnostics, East Essex, England) were used with the final inoculum of 5 ⫻ 105 CFU/mL at 100 uL per well. Screening criteria

defined by NCCLS were used to identify E. coli and K. pneumoniae that were possible extended spectrum ␤-lactamase (ESBL) producers. Confirmatory testing for ESBLs was performed using Etest (AB Biodisk, Solna, Sweden) and disks (BBL) containing cefotaxime and ceftazidime with and without clavulanic acid following NCCLS guidelines (NCCLS, 2001) and manufacturer’s instructions. Antimicrobials (concentrations in ␮g/mL) tested for Enterobacteriaceae included ampicillin (1-16), amikacin (2-32),

Table 2 In vitro activities of 17 antimicrobials against E. coli and K. pneumoniae, S. marcescens, P. mirabilis, and E. cloacae from TSAR II Antimicrobial agent

E. coli (n ⫽ 591) %R†

Aminoglycosides: Amikacin 2 Gentamicin 33 Cephems: Cephalothin 58 Cefazolin 15 Cefuroxime 11 Cefoxitin 9 Ceftazidime 3 Cefotaxime 4 Cefepime 2 Penicillins: Ampicillin 78 Piperacillin 76 Quinolone and fluoroquinolone Nalidixic acid 34 Ciprofloxacin 12 Others Amoxicillin/CA* 18 Aztreonam 4 Imipenem 0 SXT** 62

MIC50/90

K. pneumoniae (n ⫽ 233)

S. marcescens (n ⫽ 105)

P. mirabilis (n ⫽ 92)

E. cloacae (n ⫽ 69)

%R

%R

%R

%R

MIC50/90

MIC50/90

MIC50/90

MIC50/90

ⱕ2/8 1/⬎8

9 19

ⱕ2/16 ⱕ0.5/⬎8

22 66

ⱕ8/⬎32 8/⬎8

5 35

16/⬎16 ⱕ2/⬎16 4/16 4/8 ⱕ0.5/ⱕ0.5 ⱕ0.5/ⱕ0.5 ⱕ0.25/ⱕ0.25

24 19 20 8 7 13 8

4/⬎16 ⱕ2/⬎16 4/⬎16 4/8 ⱕ0.5/4 ⱕ0.5/32 ⱕ0.25/8

100 100 100 94 5 48 24

⬎16/⬎16 ⬎16/⬎16 ⬎16/⬎16 ⬎16/⬎16 ⱕ0.5/8 8/⬎32 0.5/⬎16

30 27 6 0 0 2 0

8/⬎16 8/⬎16 2/4 4/8 ⱕ0.5/ⱕ0.5 ⱕ0.5/ⱕ0.5 ⱕ0.25/0.5

100 100 78 100 43 48 10

⬎16/⬎16 ⬎16/⬎16 ⬎16/⬎16 ⬎16/⬎16 2/⬎16 8/⬎32 ⱕ0.25/16

⬎16/⬎16 ⬎64/⬎64

98 26

⬎16/⬎16 8/⬎64

98 70

⬎16/⬎16 64/⬎64

68 22

⬎16/⬎16 4/⬎64

94 54

⬎16/⬎16 ⬎64/⬎64

ⱕ8/⬎16 ⱕ0.06/⬎2

12 6

ⱕ8/⬎16 ⱕ0.06/0.25

67 51

⬎16/⬎16 2/⬎2

25 20

ⱕ8/⬎16 ⱕ0.06/2

24 10

ⱕ8/⬎16 ⱕ0.06/1

8/16 ⱕ0.5/ⱕ0.5 ⱕ0.25/ⱕ0.25 ⬎2/⬎2

11 11 0 29

2/16 ⱕ0.5/16 ⱕ0.25/0.5 ⱕ0.5/⬎2

99 24 0 62

⬎16/⬎16 2/⬎16 1/2 ⬎2/⬎2

11 1 0 70

2/16 ⱕ0.5/ⱕ0.5 2/4 ⬎2/⬎2

100 46 0 62

⬎16/⬎16 2/⬎16 0.5/1 ⬎2/⬎2

* CA, Clavulanic acid. ** SXT, Trimethoprim/Sulfamethoxazole. † %R include resistance and intermediates.

4/16 2/⬎8

8 48

ⱕ2/16 1/⬎8

214

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

Table 3 In vitro activities of 16 antimicrobials against P. aeruginosa and A. baumannii Antimicrobial agent

P. aeruginosa (n ⫽ 328) %R*

Aminoglycosides: Amikacin Gentamicin Tobramycin Cephems: Cefepime Cefoperazone Ceftazidime Ceftriaxone B-lactam/B-lactamase inhibitor: Piperacillin/Tazobactam Ticarcillin/Clavulanic acid Penicillins: Carbenicillin Piperacillin Others: Aztreonam Chloramphenicol Ciprofloxacin Imipenem SXT

A. baumannii (n ⫽ 111) MIC50/90

%R

10 30 25

ⱕ4/16 2/⬎8 1/⬎8

63 82 69

⬎32/⬎32 ⬎8/⬎8 ⬎8/⬎8

16 23 17 94

4/16 8/⬎32 2/32 ⬎32/⬎32

60 97 73 92

16/⬎16 ⬎32/⬎32 32/⬎32 ⬎32/⬎32

MIC50/90

12 22

8/⬎64 32/⬎128

74 72

⬎64/⬎64 ⬎128/⬎128

23 14

64/⬎256 8/⬎64

94 86

⬎256/⬎256 ⬎64/⬎64

98 97 69 2 78

⬎16/⬎16 ⬎16/⬎16 ⬎2/⬎2 1/2 ⬎2/⬎2

24 —** 14 8 94

8/⬎16 — 0.12/⬎2 1/4 ⬎2/⬎2

* %R include resistance and intermediates. ** Not tested.

aztreonam (0.5-16), amoxicillin/clavulanic acid (0.5/0.2516/8), cefazolin (2-16), cefepime (0.25-16), cefotaxime (0.5-32), cefoxitin (2-16), ceftazidime (0.5-16), cefuroxime (1-16), cephalothin (2-16), ciprofloxacin (0.06-2), gentamicin (0.5-8), imipenem (0.25-8), nalidixic acid (8-16), piperacillin (4-64), trimethoprim/sulfamethoxazole (SXT) (0.5/ 9.5-2/38). For Pseudomonas spp. and Acinetobacter spp., amikacin (1-32), aztreonam (1-16), carbenicillin (16-256), cefepime (0.25-16), cefoperazone (4-32), ceftazidime (132), ceftriaxone (2-32), ciprofloxacin (0.06-2), chloramphenicol (2-16), gentamicin (0.25-8), imipenem (0.25-8), piperacillin (4-64), piperacillin/tazobactam (2/4-64/4), ticarcillin/clavulanic acid (4/2-128/2), tobramycin (0.5-8), SXT (0.5/9.5-2/38) were tested. Quality control was performed on each day of testing using E. coli ATCC 25922, S. aureus ATCC 29213, E faecalis ATCC 29212, P. aeruginosa ATCC 27853, E. coli ATCC 35218 and K. pneumoniae ATCC 700603. 2.4. Data analysis Duplicate isolates were removed from the database for analysis. For resistance comparison, intermediate results were included. Univariate analysis was performed using Epi Info 6.04 (CDC, Atlanta, GA), a database and statistical program for public health. Significance of differences in frequencies and proportions was tested by the ␹2 test with Yates’ correction. To account for differences in the number of isolates from the geographic regions, hospital types,

specimen sources, and patient locations between TSAR I and II, multivariate analysis was performed by binary stepwise logistic regression using SPSS for Windows 10 (SPSS, Chicago, IL, USA).

3. Results A total of 1769 non-duplicate aerobic Gram-negative (GN) isolates were collected in TSAR II. Table 1 lists the specimen sources (blood, urine, respiratory tract, wound, and others), patient location (ICU, non-ICU, and Out-patient), hospital types (medical centers vs. regional hospitals) of the 7 most common GN isolates in the collection, E. coli, K. pneumoniae, S. marcescens, P. mirabilis, E. cloacae, P. aeruginosa and A. baumannii, which accounted for 86% (1529/1769) of the GN in the collection. Forty-two percent of these isolates came from medical centers and 58% came from regional hospitals. Among these 7 species, 39% were from the most densely populated northern part of the island, 20% from the middle, 30% from the south, and 11% from the least populated eastern region of the island (data not shown). Table 2 lists the in vitro activities (% resistant and MIC50/90) of 17 antimicrobial agents tested against E. coli and K. pneumoniae, S. marcescens, P. mirabilis, and E. cloacae. In E. coli, there was an overall high rate of resistance to the commonly used “first-line” antimicrobials in Taiwan, such as ampicillin (78%), cephalothin (58%), gen-

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

215

Table 4 Comparison of resistance rates (%)* between intensive care unit (ICU), non-ICU, and outpatient (OPD) isolates Antimicrobial agent

E. coli ICU (78)**

Aminoglycosides: Amikacin Gentamicin Cephems: Cephalothin Cefazolin Cefuroxime Cefotaxime Ceftazidime Cefepime Penicillins Ampicillin Piperacillin Quinolone and fluoroquinolone: Nalidixic acid Ciprofloxacin Others: Amoxicillin/Clavulanic acid Aztreonam Imipenem Piperacillin/Tazobactam SXT

K. pneumoniae non-ICU (297)

OPD (215)

ICU (67)

non-ICU (115)

6 44

2 33

0 28

18 36

5 15

60 23 17 6 1 5

58 17 11 5 5 3

63 8 7 1 1 1

42 34 36 28 16 18

83 80

76 76

78 75

42 19

37 12

27 5 0 —‡ 71

20 5 0 — 66

A. baumannii†

P. aeruginosa OPD (51)

ICU (114)

non-ICU (165)

OPD (49)

ICU (66)

non-ICU (37)

4 6

6 30

12 29

6 29

67 86

62 76

18 14 12 7 4 4

14 8 14 6 2 4

— — — — 18 18

— — — — 18 17

— — — — 8 12

— — — — 73 65

— — — — 70 57

100 40

97 24

98 14

— 13

— 16

— 8

85

87

28 11

19 11

10 4

8 2

— 15

— 15

— 14

— 71

— 65

12 1 0 — 54

18 25 2 — 42

8 5 0 — 26

8 4 0 — 16

— 22 11 11 97

— 26 7 13 95

— 20 2 6 90

— 99 3 77 83

— 97 0 70 68

* Resistant and intermediate combined. ** Number of isolates. † A. baumannii OPD not presented due to small number of isolates. ‡ Not tested.

tamicin (33%), and SXT (62%). The one exception among commonly used drugs was cefazolin, to which 15% of E. coli were resistant. In addition, 12% of the E. coli isolates were resistant to the fluoroquinolone (FQ) ciprofloxacin and 34% were resistant to the non-fluorinated quinolone, naladixic acid. Nalidixic acid resistance with ciprofloxacin susceptibility provides an estimate of the rate of reducedsusceptibility to the FQs (22%). ESBL-production confirmatory testing was performed on 58 E. coli and 48 K. pneumoniae isolates having an MIC of ⬎ 1 ␮g/mL for aztreonam, ceftazidime, or cefotaxime. A total of 26 E. coli and 30 K. pneumoniae were confirmed as ESBL producers. Thus the ESBL producer rates were 4% (26/591) for E. coli and 13% (30/233) for K. pneumoniae in the TSAR II collection. Eight of the E. coli and 1 of the K. pneumoniae isolates were positive for ESBL based upon confirmatory testing by cefotaxime only. All other isolates were positive for ESBLs by both cefotaxime and ceftazidime. The 30 ESBL producing K. pneumoniae made up 27% (18/67) of ICU isolates and 10% (11/115) of non-ICU isolates (data not shown). Over half (52%) of S. marcescens were resistant to ciprofloxacin and 48%, 24%, and 23% were resistant to cefotaxime, aztreonam, and cefepime, respectively. In E. cloacae, resistance to aztreonam, ceftazidime, and cefotaxime

ranged from 44 to 48%. Of the 32 (48%) gentamicinresistant E. cloacae, 27 (84%) were also resistant to ceftazidime. In P. mirabilis, highest rates of resistance were seen in ampicillin (68%), SXT (70%), and gentamicin (36%). The in vitro activities of 16 antimicrobials tested against P. aeruginosa, and A. baumannii are presented in Table 3. A large proportion of A. baumannii had overall high rates of resistance ranging from 50-90% to extended spectrum ␤-lactams, ␤-lactams/ ␤-lactamase inhibitors, aminoglycosides, and cirpofloxacin. In contrast, imipenem appeared to remain quite active against both P. aeruginosa (92% susceptible) and A. baumannii (98% susceptible). Comparison of resistance among isolates from ICU, nonICU, and outpatients on E. coli, K. pneumoniae, P. aeruginosa, and A. baumannii, the top four species in the TSAR II collection, is presented in Table 5. Overall, isolates from ICU had the highest resistance, followed by non-ICU, and outpatients, with the most notable differences seen in K. pneumoniae and E. coli. Factors independently associated with key forms of resistance among these 7 species are presented in Table 5. These multivariate analysis results were obtained by stratifying resistance rates by patient location, specimen source, hospital type, and geographic region. For example, in the case of E. coli, the overall ciprofloxacin-resistance was

216

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

Table 5 Factors independently associated with key forms of gram-negative resistance in TSAR II isolates Resistance phenotype

Total isolates in analysis

Proportion (%) with resistant phenotype

Factors

Odds ratio

95% confidence interval

p

Ciprofloxacin-resistant E. coli

591

12

Ciprofloxacin-resistant P. mirabilis Ceftazidime resistant E. cloacae Gentamicin-resistant E. cloacae Ceftazidime-resistant P. aeruginosa Ceftazidime-resistant A. baumannii

92

20*

Middle region** East region** Respiratory specimen† Respiratory specimen†

2.61 0.26 2.82 4.09

1.52–4.48 0.61–1.09 1.06–7.48 1.38–12.06

0.001 0.065 0.037 0.011

69

43

Urine specimen†

4.00

0.93–17.14

0.062

69

48

None

328

17

Respiratory specimen†

2.94

1.23–4.27

0.009

111

73

South region** Medical center‡

3.78 0.18

1.12–12.79 0.067–0.49

0.033 0.001

* The majority (15%) of ciprofloxacin non-susceptible P. mirabilis was in the intermediate category. ** vs. North geographic region. † vs. Blood specimen. ‡ vs. Regional hospital.

12%. However, E. coli from the Middle (West) geographic region of Taiwan were two to three times more likely (Odds ratio [OR] 2.61, 95% confidence interval [CI95] 1.52-4.48; p ⬍ 0.001), and isolates from the Eastern region were four times less likely (OR 0.26, CI95 0.61-1.09, p ⫽ 0.065), than referent Northern isolates to be resistant regardless of specimen source. Meanwhile, respiratory isolates were nearly 3 times more likely than referent blood isolates to be resistant (OR 2.82, CI95 1.06-7.48, p ⫽ 0.037) regardless of geographic region (Table 5). Comparison of resistance rates was made on selected key forms of resistance between TSAR I and TSAR II isolates from the 18 hospitals that participated in both surveys (Table 6). Rates of ampicillin, cefazolin, and ciprofloxacin resistance in E. coli from these 18 hospitals did not change from TSAR I to TSAR II. A significant increase in resistance was seen in ceftazidime resistant P. aeruginosa (7% in TSAR I to 18% in TSAR II, p ⬍ 0.001). The majority of increase in ciprofloxacin non-susceptible P. mirabilis (4% in TSAR I to 22% in TSAR II, p ⬍ 0.001) in TSAR II was in the intermediate category (15%) (Data not shown). The statistically significant increase of ciprofloxacin-resistance in P. mirabilis (OR 6.70, CI95 2.08-21.30; p ⫽ 0.001) and ceftazidime resistance in P. aeruginosa (OR 2.68, CI95 1.53-4.71, p ⫽ 0.001) was confirmed using multivariate analysis to control for other factors (e.g., patient location or specimen source). Finally, Table 7 compares selected key forms of resistance among TSAR II isolates with data previously published by the SENTRY group from different studies. Several of the resistance phenotypes in TSAR II were higher than those from North America and Europe.

4. Discussion Several studies on antimicrobial usage have found high rates of antimicrobial use among outpatients and inpatients in Taiwan, and overuse and inappropriate use of the “firstTable 6 Comparison of selected and key forms of resistance between TSAR I (1998) and TSAR II (2000) isolates from 18 hospitals that participated in both surveys Resistant organism

TSAR I (N ⫽ 63–518)

TSAR II (N ⫽ 62–524)

p

Ampicillin resistant E. coli Cefazolin resistant E. coli Ciprofloxacin resistant E. coli Cefazolin resistant K. pneumoniae Ciprofloxacin resistant P mirabilis Ceftazidime resistant E. cloacae Gentamicin resistant E. cloacae Ceftazidime resistant P. aeruginosa Imipenem resistant P. aeruginosa Ceftazidime resistant A. baumannii Imipenem resistant A. baumannii

80

78

NS

20

16

NS

12

13

NS

13

21

NS

4

22*

⬍0.001

35

47

NS

37

50

NS

7

18

⬍0.001

6

8

NS

65

73

NS

3

2

NS

* The majority (15%) of ciprofloxacin non-susceptible P. mirabilis was in the intermediate category.

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

217

Table 7 Comparison of selected key forms of resistance between ISAR II and published SENTRY data Resistance phenotype

Specimen type

TSAR II % resistant

Ampicillin resistant E. coli Gentamicin resistant E. coli SXTa resistant E. coli Ampicillin resistant P. mirabilis Gentamicin resistant P. mirabilis SXT resistant P. mirabilis ESBL producing E. coli

Blood Blood Blood Urine Urine Urine All

74% 26% 64% 58% 22% 72% 4%†

ESBL producing K. pneumoniae

All

13%†

Ceftazidime resistant in A. baumannii Imipenem resistant in A. baumannii Ceftazidime resistant in P. aeruginosa

All All All

73% 2% 17%

Imipenem resistant in P. aeruginosa

All

8%

SENTRY data % resistant

Country of isolates

Reference

46–49% 4–6% 19–26% 11% 8% 9% 0.2–25%† 3–9%c 6–42%† 5–45%‡ 20–64%** 3–10%** 16%** 15–33%** 4%** 8–28%**

United States & Canada United States & Canada United States & Canada United States & Canada United States & Canada United States & Canada Asia-Pacific region Europe and the Americas Asia-Pacific region Europe and the Americas North and Latin Americas North and Latin Americas Asia-Pacific Americas & Europe Asia-Pacific Americas & Europe

Pfaller et al., 1998 Pfaller et al., 1998 Pfaller et al., 1998 Mathai et al., 2001 Mathai et al., 2001 Mathai et al., 2001 Bell et al., 2002 Winokur et al., 2001 Bell et al., 2002 Winokur et al., 2001 Gales et al., 2001a Gales et al., 2001a Gales et al., 2001b Gales et al., 2001b Gales et al., 2001b Gales et al., 2001b

* SXT, Trimethoprim/Sulfamethoxazole. † Confirmed extended spectrum ␤-lactamase (ESBL) producer. ‡ % of isolates expressing ESBL phenotypes. ** 1999 data.

line” antimicrobials; such as aminopenicillins, first generation cephalosporins and aminoglycosides, are common. (Chang et al., 2001; Ho et al., 2002; Liu et al., 1999; McDonald et al., 2001a). Among ambulatory patients in public health clinics between 1996 and 1999, common cold was the most frequent diagnosis for which antimicrobials were prescribed and 31% received antimicrobials (Chang et al., 2001). Patients treated for community-acquired infections in Taiwan hospitals received a mean of 2.25 antibiotics for 8.2 days; of these, 79.2% were first-line antibiotics (Ho et al., 2002). In another study among adult inpatients in 14 acute-care hospitals in Taiwan, 67% of patients received antibiotics, and many received multiple antibiotics, for an overall rate of 813 antibiotic days/1000 patient days; first generation cephalosporins (39%) and aminoglycoside (24%) use accounted for the majority of all antibiotic-days (McDonald et al., 2001a). However, TSAR II data showed that resistance to ampicillin was 78% in E. coli and 68% in P. mirabilis, while resistance to trimethoprim/sulfamethoxazole (SXT) ranged from 29% in K. pneumoniae to 70% in P. mirabilis and resistance to gentamicin ranged from 19% in K. pneumoniae to 66% in S. marcescens. These results highlight the fact that the majority of commonly prescribed “first-line” antimicrobials in Taiwan are largely ineffective against these 5 Enterobacteriaceae species. One exception is cefazolin, which has moderate activity against E. coli (85% susceptibility) and K. pneumoniae (81% susceptibility), the two most common species of Enterobacteriaceae. We also found that several of the ESBL producers in E. coli and K. pneumoniae were detected only by cefotaxime

and not by ceftazidime. Thus it is important to use more than one antimicrobial to screen and confirm ESBL producers as substrate specificity can vary in different geographic areas (Bell et al., 2002; NCCLS, 2001; Winokur et al., 2001). K. pneumoniae isolates from ICU, non-ICU and outpatient (OPD) settings had the greatest stepwise decrease in most resistance rates, in part attributable to the fact that 27% of the ICU K. pneumoniae isolates were ESBL producers. While resistance rates among E. coli isolates also demonstrated a stepwise decrease in isolates from ICU, non-ICU, and OPD settings, resistance in OPD isolates to ampicillin (78%), gentamicin (28%), and SXT (54%) remains a serious problem. Resistance among A. baumannii to extended spectrum ␤-lactams, ␤-lactam/␤-lactamase inhibitors and other last-line antimicrobials is a concern as the rate of resistance to most of these antimicrobials was at least 50% with the exception of imipenem (2%). It is unknown whether regional differences in antimicrobial use can explain the independent associations between geographic regions and specimen types with fluoroquinolone and ceftazidime resistance (Table 5). Prior fluoroquinolone use and chronic underlying diseases including cancer and urinary tract infections have been found to be risk factors for increased fluoroquinolone resistance in E. coli (McDonald et al., 2001b; Pena et al., 1995). Cross-transmission of resistance within high-risk patient populations could also serve as an alternate explanation for regional differences in the rates of these resistant phenotypes. Resistance to commonly used first-line antimicrobials, such as ampicillin and cefazolin in E. coli and cefazolin in K. pneumoniae, did not change significantly between TSAR

218

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219

I and II among the 18 hospitals that participated in both surveys. However, resistance and reduced susceptibility of Gram-negative bacilli to fluoroquinolones are emerging antimicrobial resistance problems in Taiwan and other countries (Fridkin et al., 2002; Livermore et al., 2002; McDonald et al., 2001; Sheng et al., 2002). While there was no significant increase in the rate of ciprofloxacin resistance in E. coli between TSAR I and II (12 to 13%) among the 18 hospitals compared, there was a significant increase in P. mirabilis (4 to 22%); albeit the majority were in the intermediate category (15%). These rates are much higher than the ⬍4% reported by researchers from the United States and United Kingdom (Fridkin et al., 2002; Livermore et al., 2002). Several of the resistance rates found in TSAR II GNB are higher than other countries compared to data reported by SENTRY (Bell et al., 2002;Gales et al., 2001a & 2001b; Mathai et al., 2001; Pfaller et al., 1998). This is especially apparent in resistance to ampicillin in E. coli blood isolates (74%) and in P. mirabilis urine isolates (58%) compared to 49% and 11% on isolates from North America, respectively. In contrast, ESBL producing E. coli and K. pneumoniae, and resistance of A. baumannii and P. aeruginosa to extended spectrum antimicrobials were similar to the rates of some countries in Asia-Pacific, Europe, North and Latin Americas, except ceftazidime resistance in A. baumannii (73%), which was higher than the rates in North and Latin Americas. These results may reflect differences in prescribing behavior between these geographic regions. For example, hospital-wide rates of cefazolin use appeared higher in Taiwan hospitals whereas third-generation cephalosporin and glycopeptide use appeared lower when compared to the use of these drugs in U.S. hospitals (McDonald et al., 2001a). Resistance of several pathogens from clinical (inpatient and outpatient) isolates of medical centers and resistance of ICU isolates from one major medical center in Taiwan has been previously reported (Chang et al., 2000; Hsueh et al., 2002a & 2002b). Data from TSAR demonstrates that the prevalence of resistance is not only in ICU and Non-ICU inpatient but also in outpatient isolates of medical centers as well as regional hospitals in Taiwan. Comparison of national resistance with data from international surveillance studies is important in understanding the severity of resistance in Taiwan, which provides the basis for advocating appropriate antimicrobial use to help in the containment of antimicrobial resistance.

Acknowledgments We express our sincere appreciation to the TSAR participating hospitals: Northern Region: Armed Forces Sung Shan Hospital, Cathay General Hospital, Chang Gung Memorial Hospital at Keelung, Chang Gung Memorial Hospital at Linkou, Cheng Hsin Rehabilitation Center, Hsin-Chu

Hospital, DOH, the Executive Yuan, Koo Foundation Sun Yat-sen cancer center, Lo-Hsu Foundation Inc. Lo-Tung Poh Ai Hospital, Miin Sheng General Hospital, National Taiwan University Hospital, St. Mary Hospital, Taipei Hospital, DOH, the Executive Yuan, Taipei Medical College Hospital, Taipei Medical College-affiliated Taipei Municipal Wan-fang Hospital, Taipei Municipal Chung Hsiao Hospital, Taipei Municipal Chung-Hsin Hospital, Taipei Municipal Yang-Ming Hospital, Taiwan Adventist Hospital, Tao-Yuan General Hospital, DOH, the Executive Yuan, Tri Service General Hospital; Middle Region: Chang-Hwa Christian Hospital, Cheng Ching Hospital, China Medical College Hospital, Chung Shan Medical Dental College Hospital, Kuan-Tien General Hospital, Shalu Tung’s Memorial Hospital, Show Chwan Memorial Hospital, Veterans General Hospital—Taichung, Zen Ai General Hospital; Southern Region: Chang Gung Memorial Hospital at Kaohsiung, Chiayi Christian Hospital, Chi-Mei Foundation Hospital, God’s Help Hospital, Jen Ai General Hospital, Kaohsiung Military Hospital, Kaohsiung Medical College Chung-Ho Memorial Hospital, St. Martin De Porres Hospital, Tainan Municipal Hospital, Veterans General Hospital—Kaohsiung; Yuan’s General Hospital, Eastern Region: Buddhist Tzu-Chi General Hospital, Hua-Lien Hospital, DOH, the Executive Yuan, Hua-Lien Mennonite Church Hospital, Mackay Memorial Hospital Taitung Branch. We would also like to thank Ms. Hsiao Chun Yin for drawing the map of TSAR hospitals. References Bax, R., Bywater, R., Cornaglia, G., Goossens, H., Hunter, P., Isham, V., Jones, R., Phyllips, I., Sahm, D., Senn, S., Struelens, M., Taylor, D., & White, A. (2001). Surveillance of antimicrobial resistance-what, how and whither? Clin Microbiol and Infect Dis 7, 316 –325. Bell, J. M., Turnidge, J. D., Gales, A. C., Pfaller, M. A., Jones, R. N., & the SENTRY APAC Study Group (2002). Prevalence of extended spectrum B-lactamase (ESBL)-producing clinical isolates in the AsiaPacific region and South Africa: regional results from SENTRY Antimicrobial Surveillance Program (1998-99). Diag Microbiol Infect Dis 42, 193–198. Chang, S. C., Hsieh, W. C., Liu, C. Y., & The Antibiotic Resistance Study Group of the Infectious Disease Society of the Republic of China. (2000). High prevalence of antibiotic resistance of common pathogenic bacteria in Taiwan. Diag Microbiol Infect Dis 36, 107–112. Chang, S. C., Shiu, M. N., & Chen, T. J. (2001). Antibiotic usage in primary care units in Taiwan after the institution of national health insurance. Diag Microbiol Infect Dis 40, 137–143. Fridkin, S. K. (1999). Intensive Care Antimicrobial Resistance Epidemiology (ICARE) surveillance report, data summary from January 1996 through December 1997. Am J Infect Control 27, 279 –284. Fridkin, S. K., Hill, H. A., Volkova, N. V., Edwards, J. R., Lawton, R. M., Gaynes, R. P., McGowan, J. E., & the Intensive Care Antimicrobial Resistance Epidemiology (ICARE) project hospitals (2002). Temporal changes in prevalence of antimicrobial resistance in 23 US hospitals. Emerg Infect Dis 8, 697–701. Gales, A. C., Jones, R. N., Forward, K. R., Linares, J., Sader, H. S., & Verhoef, J. (2001a). Emerging importance of multidrug-resistant Acinetobacter species and Stenotrophomonas maltophilia as pathogens in seriously ill patients: geographic patterns, epidemiological features,

T.L. Lauderdale et al. / Diagnostic Microbiology and Infectious Disease 48 (2004) 211–219 and trends in the SENTRY Antimicrobial Surveillance Program (19971999). Clin Infect Dis 32, S104 –S113. Gales, A. C., Jones, R. N., Turnidge, J., Rennie, R., & Ramphal, R. (2001b). Characterization of Pseudomonas aeruginosa isolates: occurrence rates, antimicrobial susceptibility patterns, and molecular typing in the global SENTRY Antimicrobial Surveillance Program, 19971999. Clin Infect Dis 32, S146 –55. Ho, M., McDonald, L. C., Lauderdale, T. L., Yeh, L. L. L., Chen, P. C., & Shiau, Y. R. (1999). Surveillance of antibiotic resistance in Taiwan, 1998. J Microbiol Immuno Infect 32, 239 – 49. Ho, M. (2001). Taiwan seeks to solve its resistance problems. Science 291, 2550b. Ho, M., Chang, F. Y., Yin, H. C., Ben, R. J., Chang, L. Y., Chen, P. Y., Cheng, S. H., Chen, S. T., Huang, F. L., Lin, H. C., Lu, D., Want, N. C., & Wang, J. T. (2002). Antibiotic usage in community-acquired infections in hospitals in Taiwan. J Formos Med Assoc 101, 34 – 42. Hsueh, P. R., Chen, M. L., Sun, C. C., Chen, W. H., Pan, H. J., Yang, L. S., Chang, S. C., Ho, S. W., Lee, C. Y., Hsieh, W. C., & Luh, K. T. (2002a). Antimicrobial drug resistance in pathogens causing nosocomial infections at a university hospital in Taiwan, 1981-1999. Emerg Infect Dis 8, 63– 68. Hsueh, P. R., Liu, C. Y., & Luh, K. T. (2002b). Current status of antimicrobial resistance in Taiwan. Emerg Infect Dis 8, 132–136. Jones, R. N., Marshall, S. A., Pfaller, M. A., Wilke, W. W., Hollis, R. J., & Erwin, M. E. (1997). Nosocomial enterococcal blood stream infections in the SCOPE Program: antimicrobial resistance, species occurrence, molecular testing results, and laboratory testing accuracy. SCOPE Hospital Study Group. Diag Microbiol Infect Dis 29, 95–102. Jones, R. N., & Masterton, R. (2001). Determining the value of antimicrobial surveillance programs. Diag Microbiol Infect Dis 41, 171–175. Kunin, C. M. (1993). Resistance to antimicrobial drugs-a worldwide calamity. Annals Intern Med 118, 557–561. Lewis, D. (2002). Antimicrobial resistance surveillance: methods will depend on objectives. J Antimicrob Chemother 49, 3–5. Liu, Y. C., Huang, W. K., Huang, T. S., & Kunin, C. M. (1999). Extent of antibiotic use in Taiwan shown by antimicrobial activity in urine. Lancet 354, 1360. Livermore, D. M., Macgowan, A. P., & Wale, M. C. J. (1998). Surveillance of antimicrobial resistance. Br Med J 317, 614 – 615. Livermore, D. M., James, D., Reacher, M., Graham, C., Nichols, T., Stephens, P., et al. (2002). Trends in fluoroquinolone (ciprofloxacin) resistance in Enterobacteriaceae from bacteremias, England and Wales, 1990-1999. Emerg Infect Dis 8, 473– 478. Masterton, R. G. (2000). Surveillance studies: how can they help the management of infection? J Antimicrob Chemother 46, 53–58. Mathai, D., Jones, R. N., Pfaller, M. A., & SENTRY participant group North America. (2001). Epidemiology and frequency of resistance among pathogens causing urinary tract infections in 1510 hospitalized patients: A report from the SENTRY Antimicrobial Surveillance Program (North America). Diag Microbiol Infect Dis 40, 129 –136.

219

McDonald, L. C., Yu, H. T., Yin, H. C., Hsiung, C. A., Hung, C. C., Ho, M., & the Antibiotic Use Working Group. (2001a). Correlates of antibiotic use in Taiwan hospitals. Infect Control Hosp Epidemiol 22, 565–571. McDonald, L. C., Chen, F. J., Lo, H. J., Yin, H. C., Lu, P. L., Huang, C. H., Chen, P., Lauderdale, T. L., & Ho, M. (2001b). Emergence of reduced susceptibility and resistance to fluoroquinolones in Escherichia coli in Taiwan and contributions of distinct selective pressures. Antimicrob Agents Chemother 45, 3084 –3091. McGowan, J. E. (2001). Economic impact of antimicrobial resistance. Emerg Infect Dis 7, 286 –292. Murray, P. R., Baron, E. J., Pfaller, M. A., Tenover, F. C., & Yolken, R. C. (ed). (1999). Manual of clinical microbiology (7th ed.). American Society for Microbiology, Washington, DC. National Committee for Clinical Laboratory Standards (2000). Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically (5th ed.). Approved standard. M7-A5. Wayne, PA: National Committee for Clinical Laboratory Standards. National Committee for Clinical Laboratory Standards (2001). Performance standards for antimicrobial susceptibility testing; eleventh information supplement M100-S11; 2001. Wayne, PA: National Committee for Clinical Laboratory Standards. O’Brien, T. F. (1997). The global epidemic nature of antimicrobial resistance and the need to monitor and manage it locally. Clin Infect Dis 24, S2–S8. Okeke, I. N., & Edelman, R. (2001). Dissemination of antibiotic-resistant bacteria across geographic borders. Clin Infect Dis 33, 364 –369. Pena, C., Albareda, J. M., Pallares, R., Pujol, M., Tubau, F., & Ariza, J. (1995). Relationship between quinolone use and emergence of ciprofloxacin-resistant Escherichia coli in bloodstream infections. Antimicrob Agents Chemother 39, 520 –524. Pfaller, M. A., Jones, R. N., Doern, G. V., & Kugler, K. (1998). Bacterial pathogens isolated from patients with bloodstream infection: Frequencies of occurrence and antimicrobial susceptibility patterns from the SENTRY Antimicrobial Surveillance Program (United States and Canada, 1997). Antimicrob Agents Chemother 42, 1762–1770. Sheng, W. H., Chen, Y. C., Wang, J. T., Chang, S. C., Luh, K. T., & Hsieh, W. C. (2002). Emerging fluoroquinolone-resistance for common clinically important gram-negative bacteria in Taiwan. Diag Microbiol Infect Dis 43, 141–147. Smith, K. E., Besser, J. M., Hedberg, C. W., Leano, F. T., Bender, J. B., Wicklund, J. H., Johnson, B. P., Moore, K. A., Osterholm, M. T., & the Investigation Team (1999). Quinolone-resistant Campylobacter jejuni infections in Minnesota, 1992-1998. Investigation Team. NEJM 340, 1525–1532. Winokur, P. L., Canton, R., Casellas, J. M., & Legakis, N. (2001). Variations in the prevalence of strains expressing an extended-spectrum B-lactamase phenotype and characterization of isolates from Europe, the Americas, and the Western Pacific Region. Clin Infect Dis 32, S94 –S103.