JAC

Journal of Antimicrobial Chemotherapy (2000) 45, 183189

Antimicrobial resistance amongst Klebsiella spp. collected from

intensive care units in Southern and Western Europe in 19971998
Gioia S. Babini and David M. Livermore*
Antibiotic Resistance Monitoring and Reference Laboratory, Central Public Health Laboratory,
61 Colindale Avenue, London NW9 5HT, UK
A 1994 survey of 35 intensive care units (ICUs) in Western and Southern Europe found
extended-spectrum -lactamases (ESBLs) in 220/966 (23%) klebsiellae. A follow-up survey
from May 1997 to October 1998 collected klebsiellae from 24 ICUs, including 23 that participated in 1994. Twenty-one ICUs sent 433 eligible isolates, of which 110 (25%) had ESBLs.
The prevalence of ESBLs had not changed significantly from 1994 but the proportion of ESBLproducers resistant to piperacillin/tazobactam had risen from 31% to 63% (P < 0.001), and most
of this resistance was high level (MICs 128 4 mg/L). The proportion of Klebsiella oxytoca
isolates hyperproducing K1 -lactamase rose from 8% in 1994 to 21% in 19971998 (P < 0.001).
Most klebsiellae (99%) were very susceptible to meropenem (mode MIC 0.03 mg/L) but three
had decreased susceptibility (MICs 24 mg/L). These could not hydrolyse carbapenems.
Aminoglycoside resistance was not significantly changed in prevalence from 1994; ciprofloxacin resistance occurred in 31% of ESBL-producers in both years, but had increased
among non-producers (2% in 1994 versus 7% in 19971998, P < 0.001).

Introduction

Materials and methods

Klebsiellae are opportunistic pathogens which frequently

cause infections in immunocompromised patients.1 Since
the 1980s, they have become the major hosts for extendedspectrum -lactamases (ESBLs), most of which are
mutants of TEM- and SHV-type -lactamases.2 Klebsiellae
are also occasional hosts for plasmid-borne AmpC
-lactamases, and some Klebsiella oxytoca isolates owe
cephalosporin and aztreonam resistance to hyperproduction of their chromosomal K1 (KOXY) -lactamase.3
A survey of 35 intensive care units (ICUs) in Western
and Southern Europe in 19944 found ESBLs in 23% of 966
klebsiellae, AmpC enzymes in 1% and hyperproduction of
K1 enzyme in 8% of K. oxytoca. ESBL producers were
recovered at 23/35 intensive care units (ICUs), including
20/27 that sent more than ten isolates. A new survey was
conducted from May 1997 until October 1998, aiming to
monitor changes since 1994. We enrolled 24 centres,
including 23 that participated in 1994 (Table I).

Bacteria
Participating centres were asked to submit up to 30 consecutive non-replicate klebsiellae, irrespective of antibiogram, from clinically significant ICU infections. In
addition, the centres submitted 20 other Enterobacteriaceae with ESBLs from the same ICUs. They were
provided with Etest ESBL detection strips (AB Biodisk,
Solna, Sweden) to identify the latter organisms. Collection
was from May 1997 to October 1998. On receipt by the
Antibiotic Resistance Monitoring and Reference Laboratory (ARMRL), isolates were sub-cultured on MacConkey
agar (Oxoid, Basingstoke, UK) and were identified with
API 20E strips (bioMrieux, Lyons, France).

Antimicrobial agents and susceptibility testing

Ciprofloxacin was from Bayer, Newbury, UK; gentamicin,
amikacin, cloxacillin and cefuroxime from Sigma, Poole,
UK; ceftazidime from Glaxo Wellcome, Uxbridge, UK;

*Corresponding author. Tel: 44-181-200-4400; Fax:

183
2000 The British Society for Antimicrobial Chemotherapy

44-181-200-7449; E-mail: DLivermore@phls.nhs.uk

G. S. Babini and D. M. Livermore

Table I. Participating centres and isolates submitted in the 1994 and 19971998 surveys
Klebsiellae sent in
19971998 survey (n)

ESBL prevalence in
19971998 survey (%)

ESBL prevalence in
1994 survey (%)

Country

Centre

France

France1
France2
France3
Total France

30
0
29
59

3
0
62
32

5
2
54
19

UK

UK1
Total UK

23
23

9
9

0
0

Italy

Italy1
Italy2
Total Italy

29
29
58

52
24
38

11
18
15

Spain

Spain1
Spain2
Spain3
Spain4
Spain5
Total Spain

8
19
6
3
0
36

0
10.5
0
0
0
5.5

0
0
25
0
0
1

Turkey

Turkey1
Turkey2
Total Turkey

26
18
44

42
83
61

50
68
59

Germany

Germany1
Germany2
Germany3
Total Germany

19
21
29
69

5
24
34.5
23

Belgium

Belgium1
Belgium2
Belgium3
Total Belgium

29
0
30
59

17
0
37
32

49
0
22.5
31

Netherlands

Netherlands1
Netherlands2
Netherlands3
Total Netherlands

30
25
30
85

13
0
10
8

9.5
3
37
16

433

25

22

Total survey

0
NA
35
17

NA, not applicable (centre did not participate in 1994).

ceftriaxone from Roche, Welwyn Garden City, UK; Ro

48-1256 (an inhibitor of AmpC enzymes),5 from Roche,
Basel, Switzerland; piperacillin and tazobactam from
Wyeth, Taplow, UK; aztreonam, cefepime, cefotetan and
meropenem from Zeneca, Macclesfield, UK; imipenem
and cefoxitin from Merck, Hoddesdon, UK; and clavulanate from SmithKline Beecham, Harlow, UK. MICs were
determined, as previously,4 on IsoSensitest agar (Oxoid)
with inocula of 104 cfu/spot. -Lactamase inhibitors were

routinely used at 4 mg/L, but cloxacillin 100 mg/L was used

to inhibit AmpC enzymes.

Isolates with decreased susceptibility to meropenem

Crude extracts of these isolates were prepared as described
by Livermore & Williams6 and assayed against 0.1 mM
imipenem and meropenem by spectrophotometry at
297 nm.6

184

Antimicrobial resistance among klebsiellae

for ceftriaxone and aztreonam were also well stratified,
with 8894% of ESBL-producers resistant at 1 mg/L. Only
Data from the 1994 and 19971998 surveys were compared
five isolates inferred to have neither ESBLs nor hyperproby 2 tests.7
duction of K1 enzyme were resistant to aztreonam 1 mg/L,
and six were resistant to ceftriaxone 1 mg/L. Greater MIC
overlaps for ESBL-producers and non-producers were
seen for cefuroxime and cefoxitin; nevertheless, the modal
Results
MIC of cefuroxime for ESBL-producers (64 mg/L) greatly
Five hundred and six isolates from 21 hospitals were exceeded that for non-producers (2 mg/L). The modal MIC
received as klebsiellae and 484 were confirmed as Kleb - of cefoxitin was 4 mg/L for both ESBL-producers and
siella spp.: 342 as Klebsiella pneumoniae (including one non-producers; that of cefotetan was 0.51 mg/L for ESBLK. pneumonia ozaenae), 129 as K. oxytoca and three as producers, but 0.060.12 mg/L for non-producers.
Klebsiella ornithinolytica. Two centres collected only cefRegardless of inferred -lactamase production, virtually
tazidime-resistant klebsiellae; once these were excluded, all the klebsiellae were susceptible to meropenem 0.25
the number of klebsiellae analysable for ESBL prevalence mg/L; none the less, MICs 24 mg/L were recorded for
fell to 433, comprising 304 K. pneumoniae (including one three isolates: one from Italy and two from a hospital in
K. pneumoniae ozaenae), 126 K. oxytoca and three K. The Netherlands. These isolates were resistant to all the
ornithinolytica. Only six hospitals sent more than five other -lactams tested. Clavulanate 4 mg/L reduced their
non-Klebsiella spp. isolates inferred to have ESBLs and, ceftazidime MIC from 1024 mg/L to 864 mg/L, implying
for some of these, the inference was on methods other ESBL production. Synergy was not seen between cefthan the Etests provided. Because of these problems, tazidime and Ro 48-1256 4 mg/L nor cloxacillin 100 mg/L,
data for the non-Klebsiella spp. isolates will be analysed and was not increased when cloxacillin 100 mg/L was added
separately.
to ceftazidimeclavulanate. Cell-free extracts of these
isolates did not hydrolyse carbapenems.

Statistical analyses

Categorization of isolates by antibiogram

ESBL-positive klebsiellae were putatively identified based
on 16-fold synergy between ceftazidime and ceftazidime
clavulanate,4 and made up 110 of the 433 isolates (25%).
Of these, 94 were K. pneumoniae and 16 were K. oxytoca.
Table I gives their source details: ESBL-producers were
recovered from 15/19 centres, including 15/16 that sent ten
or more klebsiellae. The proportions of ESBL-producers
from individual hospitals ranged from 0 to 83%. The 314
isolates with ceftazidime:ceftazidime
clavulanate MIC
ratios of four or less were considered to lack ESBLs. Nine
isolates with ceftazidime:ceftazidime
clavulanate MIC
ratios of eight were viewed as a borderline group and were
excluded from further analysis (see Discussion). Twentyseven ESBL-negative K. oxytoca were identified as putative hyperproducers of K1 enzyme, based on susceptibility
to ceftazidime 1 mg/L but resistance to at least two of
cefuroxime 8 mg/L, ceftriaxone 0.5 mg/L and aztreonam
0.5 mg/L. Most were highly resistant to cefuroxime (MICs
256 mg/L) and aztreonam (MICs
16 mg/L). They
represented 21% of the 127 K. oxytoca collected, and were
from ten hospitals.

Susceptibility of ESBL-positive and -negative

isolates
Cephalosporins, aztreonam and meropenem
All the putative ESBL-producers were resistant to ceftazidime 2 mg/L, whereas all except two non-producers
(one K. pneumoniae and one K1 -lactamase-hyperproducing K. oxytoca) were susceptible. MIC distributions

Piperacillin and piperacillintazobactam

All the putative ESBL-producers and hyperproducers of
K1 enzyme were resistant to piperacillin 16 mg/L, as were
37/288 isolates with neither mode of resistance. Tazobactam
4 mg/L reduced the modal piperacillin MIC for ESBL nonproducers from 8 to 4 mg/L, and 41/110 were susceptible to
piperacillin/tazobactam at 16
4 mg/L; however, 69/110
ESBL-producers (63%) were resistant to piperacillin/
tazobactam at this concentration and 40% were resistant
at 512
4 mg/L. Every centre except one, that sent
ESBL producers included isolates that were resistant to
piperacillin/tazobactam 16
4 mg/L. All the hyperproducers of K1 enzyme were highly resistant to piperacillin
and piperacillin/tazobactam (MICs 128 mg/L). Of the 37
piperacillin-resistant klebsiellae (MIC
16 mg/L) with
neither ESBLs nor hyperproduction of K1 enzyme, 16
were resistant to piperacillin/tazobactam 16 4 mg/L.
Non -lactams
Aminoglycoside resistance was more frequent (P 0.001)
amongst ESBL-producers than non-producers: thus 61%
and 72% of ESBL-producers were resistant to amikacin
4 mg/L and gentamicin 1 mg/L, respectively, compared
with 4% and 9.5% of non-producers. Ciprofloxacin resistance (MIC
1 mg/L) was also more frequent among
ESBL-producers than non-producers (31% versus 7%,
P 0.001); nevertheless, resistance was scattered in both
groups; thus ciprofloxacin-resistant ESBL-producers were
found at seven hospitals and resistant non-producers were
found at ten.

185

G. S. Babini and D. M. Livermore

Isolates excluded from analysis

Nine isolates for which the ceftazidime:ceftazidime
clavulanate MIC ratio was eight were excluded from the
MIC comparisons (above). One was a K. oxytoca inferred
to hyperproduce K1 enzyme. Ceftazidime MICs for three
of the other eight isolates were 4 mg/L. These were also
resistant to aztreonam and ceftriaxone, with MICs
4
mg/L, and probably had ESBLs. Ceftazidime MICs for
the other five isolates were 0.122 mg/L and those of ceftriaxone and aztreonam were 0.25 mg/L. ESBL production seems unlikely in such organisms.

Laboratory reporting of resistance to cephalosporins

for putative ESBL-producers
Each isolate was sent to ARMRL with a case record form
indicating, inter alia, the source hospitals susceptibility
data. Between 10 and 38% of the putative ESBL-producers
had been reported as susceptible to a cephalosporin or to
aztreonam, and between 3 and 23.5% as intermediate
(Table II). Most of those reported susceptible (74%) had
low-level resistance to the compound in question (MICs
416 mg/L). In contrast, between 0.3% and 2% of susceptible isolates lacking ESBLs or hyperproduction of K1
enzyme had been reported resistant to one or more
cephalosporins. Five K. oxytoca isolates that hyperproduced K1 enzyme had been reported susceptible to
ceftriaxone, despite MICs of 832 mg/L.

Discussion: comparison of 1994 and 19971998

surveys
This study followed a similar survey in 1994, and aimed to
assess whether ESBLs had increased in prevalence among
ICU klebsiellae, and whether other Enterobacteriaceae
had acquired identical ESBLs to klebsiellae in the same
units. In the event, we received few non-klebsiellae, and,
for some of these, ESBL production had been inferred outside the protocol. Because of these problems the present
paper considers only the klebsiellae.
As in 1994, putative ESBL-producers were identified on
the basis of ceftazidime:ceftazidime
clavulanate MIC
ratios 16. The proportion of ESBL-producers was not
significantly changed (P
0.1), regardless of whether all
participating hospitals were considered or only those that
contributed in both years. Not all the centres that sent
ESBL-producers in the 1994 study sent ESBL-producers in
the 19971998 survey (Table I), and vice versa, but this
difference may have reflected the occurrence of outbreaks
and/or the fact that some centres sent few isolates. More
critically, and underscoring their widening distribution,
ESBL-producers were found at 15/16 hospitals that sent
ten or more klebsiellae in 19971998, compared with 20/27
in 1994 (P 0.05).
186

Antimicrobial resistance among klebsiellae

The resistance of ESBL-producers to aminothiazolyl
cephalosporins needs no discussion, but the status of
piperacillin/tazobactam deserves comment. The proportion of isolates resistant to this combination at 16 4 mg/L
rose from 30% in 1994 to 63% in 19971998 (P
0.001,
Table III), and most of this increase reflected isolates with
piperacillintazobactam MICs
512
4 mg/L (Figure).
Piperacillin/tazobactam-resistant ESBL-producers (MIC
16 4 mg/L) were found in 17/23 ICUs that sent ESBLproducers in 1994, but only three of the nine centres
that sent more than ten ESBL-producers had 50% of

piperacillin/tazobactam resistance (Table III). In 1997

1998, ESBL-producers resistant to piperacillin/tazobactam
were found in 14 of 15 hospitals that sent ESBL-producers
and, of the six centres that sent more than ten ESBLproducers, all but one had 50% piperacillin/tazobactam
resistance. Such resistance can reflect hyperproduction of
ESBLs, production of multiple ESBLs, or combinations of
-lactamase and impermeability,8 but the relative importance of these mechanisms is unclear.
Another change from 1994 was a significantly increased
proportion of K. oxytoca inferred to hyperproduce the

Table III. Summary: comparison of the findings of the 1994 and 19971998 surveys
Findings

1994

Prevalence of ESBL production

Prevalence of ESBL production in ICUs included in both studies
Prevalence of piptaz Ra (MICs 16 4 mg/L) in ESBL ve
isolates
Proportion of ICUs that sent 10 ESBL ve isolates with 50%
piptaz R
Prevalence of hyperproduction of K1 enzyme in K. oxytoca
Prevalence of gentamicin R (MICs 1 mg/L) in ESBL ve isolates
Prevalence of gentamicin R (MICs 1 mg/L) in ESBL ve isolates
Prevalence of amikacin R (MICs 4 mg/L) in ESBL ve isolates
Prevalence of amikacin R (MICs 4 mg/L) in ESBL ve isolates
Prevalence of ciprofloxacin R (MICs 1 mg/L) in ESBL ves
Prevalence of ciprofloxacin R (MICs 1 mg/L) in ESBL ves

19971998

220/966 (23%) 110/433 (25%)

133/640 (22%) 110/433 (25%)

P
0.1
0.1

66/220 (30%)

69/110 (63%)

0.001

3/9 (33%)
20/248 (8%)
168/220 (76%)
56/736 (8%)
114/220 (52%)
19/736 (3%)
69/220 (31%)
19/736 (2%)

5/6 (83%)
27/130 (21%)
79/110 (72%)
30/314 (9.5%)
67/110 (61%)
12/314 (4%)
34/110 (31%)
21/314 (7%)

0.05
0.001
0.1
0.1
0.05 P
0.1
0.1
0.001

R, resistance.

Figure. MIC distribution of piperacillin/tazobactam for klebsiellae with ESBLs in the 1994 ( ) and 19971998 ( ) surveys.

187

0.1

G. S. Babini and D. M. Livermore

K1 chromosomal -lactamase, up from 8% to 21% (P
0.001). Almost all the hyperproducers collected in 1994
were unique isolates,9 and it remains to be determined
whether the present increase was due to single-isolate
epidemics or to multiple separate isolates.
Cephamycins deserve mention, too. The modal MIC of
cefoxitin for ESBL-producers exceeded that for nonproducers in 1994 (16 mg/L versus 2 mg/L) whereas the
modal MIC for both groups was 4 mg/L in 19971998. Nevertheless, 17% of ESBL-producers collected in 19971998
were resistant to cefoxitin 16 mg/L compared with 4% of
non-producers (P 0.001). The modal MICs of cefotetan
in 19971998 were 0.51 mg/L for ESBL-producers compared with 0.060.12 mg/L for non-producers. The reduced
susceptibility of the ESBL-producers to cephamycins
conflicts with the fact that ESBLs do not protect transconjugants against these compounds.10
Aminoglycoside resistance was more common among
putative ESBL-producers than non-producers in both years
(Table III) and its prevalence had not significantly changed
between the two studies; likewise, there was no significant
change in the prevalence of ciprofloxacin resistance among
ESBL-producers (31% in both years, P 0.1, Table III).
On the other hand, ciprofloxacin resistance had increased
amongst the ESBL non-producers (2% in 1994 versus 7%
in 19971998, P 0.001, Table III). Moreover, most of the
ciprofloxacin resistance among ESBL-producers in 1994
reflected multiple inclusion of a serotype K25 strain, and it
may be that resistance is now more widespread among
ESBL-different strains.
Three ESBL producers (one from Italy and two from
The Netherlands) showed decreased meropenem susceptibility, with MICs of 24 mg/L. They were resistant to all
other -lactams, but lacked carbapenemase activity.
Decreased susceptibility to carbapenems in K. pneumoniae
has been linked to the simultaneous presence of an
acquired AmpC enzyme with the loss of a 42 kDa outer
membrane porin11 or to porin loss in the presence of a
hyperproduced SHV-type enzyme. 12 No synergy was seen
between ceftazidime and Ro 48-1256 or cloxacillin in the
present cases, contra-indicating AmpC production.
As in 1994, up to 40% of the ESBL-producers had been
reported as susceptible to cefotaxime and/or ceftriaxone,
and a further 2223% as intermediate. The continued
frequency of reporting ESBL-producers as susceptible to
these aminothiazolyl cephalosporins is disturbing when
we allow for the wide publicity13,14 given to the clinical
inactivity of the compounds against ESBL-producers.
Misreporting was rarer for ceftazidime and aztreonam,
doubtless reflecting the fact that resistance to these drugs
is generally more obvious than that to cefotaxime and
ceftriaxone.
In summary, the prevalence of ESBL production
amongst klebsiellae from European ICUs was unchanged
from 1994, but resistance to piperacillin/tazobactam had
increased among ESBL-producers and K. oxytoca isolates

hyperproducing K1 enzyme were more frequent. The

increased resistance to piperacillin/tazobactam casts a
doubt on its role in infections caused by ESBL-producers.

Acknowledgements
We are very grateful to the following for collecting isolates:
C. Bebear, Hopital Pellegrin, Universit de Bordeaux; G.
Bonfiglio, Istituto di Microbiologia, Universita di Catania;
F. Baquero, Hospital Ramon y Cajal, Madrid; R. Cisterna,
Hospial de Basurto, Bilbao; F. Crokaert, Institut Jules
Bordet, Brussels; T. Fosse, Hpital Saint-Roch, Nice; J. A.
Garcia-Rodriguez, Hospital Universitario de Salamanca; J.
J. A. Hoogkamp-Korstanje, St Radboud, Academisch
Ziekenhuis, Nijmegen; Jacobs, Universittsklinikum Carl
Gustav Carus, Dresden; J. G. M. Koelman, Academisch
Ziekenhuis, Vrije Universiteit, Amsterdam; V. Korten,
Marmara University Hospital, Istanbul; S. Lauwers,
Academisch Ziekenhuis, Vrije Universiteit Brussels; W. L.
Manson, Academisch Ziekenhuis, Groningen; Nunes da
Costa, Hospital General de Santo Antonio, Oporto; W.
Opferkuch, Medizinische Klinik, St Josef Hospital,
Bochum; B. Panzig, Ernst-Moritz-Ardnt-Universitt,
Greifswald; J. D. Perry, Freeman Hospital, Newcastle upon
Tyne; M. J. Salgado, Hospital de St Maria, Lisbon; M.
Salvado, Laboratoria de Referencia di Catalunya,
Barcelona; M. Segovia, Hospital Universitario de Murcia;
G. Schito, Istituto di Microbiologia, Universita di Genoa;
S. Unal, Hacettepe University School of Medicine,
Ankara; G. Verschraegen, Universiteit Gent. We are also
very grateful to Zeneca Pharmaceuticals, Alderley Park,
UK, for supporting this study financially, and to their
affiliates in Belgium, France, Germany, Italy, The Netherlands, Portugal, Spain and Turkey for liaising with the
survey participants.

References
1. Podschun, R. & Ullmann, U. (1998). Klebsiella spp. as nosocomial pathogens: epidemiology, taxonomy, typing methods, and
pathogenicity factors. Clinical Microbiology Reviews 11, 589603.
2. Bush, K., Jacoby, G. A. & Medeiros, A. A. (1995). A functional
classification scheme for -lactamases and its correlation with
molecular structure. Antimicrobial Agents and Chemotherapy 39,
121133.
3. Fournier, B., Arlet, G., Lagrange, P. H. & Philippon, A. (1994).
Klebsiella oxytoca: resistance to aztreonam by overproduction of the
chromosomally encoded -lactamase. FEMS Microbiology Letters
116, 316.
4. Livermore, D. M. & Yuan, M. (1996). Antibiotic resistance and
production of extended-spectrum -lactamases amongst Klebsiella
spp. from intensive care units in Europe. Journal of Antimicrobial
Chemotherapy 38, 40924.
5. Heinze-Krauss, I., Angehrn, P., Charnas, R. L., Gubernator, K.,
Gutknecht, E. M., Hubschwerlen, C. et al. (1998). Structure-based

188

Antimicrobial resistance among klebsiellae

design of -lactamase inhibitors. 1. Synthesis and evaluation of
bridged monobactams. Journal of Medicinal Chemistry 41, 396171.
6. Livermore, D. M. & Williams, J. D. (1996). -Lactams: mode of
action and mechanisms of antibacterial resistance. In Antibiotics in
Laboratory Medicine, 4th edn (Lorian, V., Ed.), pp. 50277. Williams
& Wilkins, Baltimore, MD.
7. Swinscow, T. D. (1976). Statistics at square one. XVThe chisquared tests (continued). British Medical Journal 2, 5134.
8. Livermore, D. M. (1993). Determinants of the activity of -lactamase inhibitor combinations. Journal of Antimicrobial Chemotherapy
31, Suppl. A, 921.
9. Gheorghiu, R., Yuan, M., Hall, L. M. & Livermore, D. M. (1997).
Bases of variation in resistance to -lactams in Klebsiella oxytoca
isolates hyperproducing K1 -lactamase. Journal of Antimicrobial
Chemotherapy 40, 53341.
10. Jacoby, G. A. & Carreras, I. (1990). Activities of -lactam antibiotics against Escherichia coli strains producing extended-spectrum
-lactamases. Antimicrobial Agents and Chemotherapy 34, 85862.

11. Bradford, P. A., Urban, C., Mariano, N., Projan, S. J., Rahal, J.
J. & Bush, K. (1997). Imipenem resistance in Klebsiella pneumoniae
is associated with the combination of ACT-1, a plasmid-mediated
AmpC -lactamase, and the loss of an outer membrane protein.
Antimicrobial Agents and Chemotherapy 41, 5639.
12. MacKenzie, F. M., Forbes, K. J., Dorai-John, T., Amyes, S. G. &
Gould, I. M. (1997). Emergence of a carbapenem-resistant Klebsiella pneumoniae. Lancet 350, 783.
13. Livermore, D. M. (1995). -Lactamases in laboratory and
clinical resistance. Clinical Microbiology Reviews, 8, 55784.
14. Bush, K. (1996). Is it important to identify extended-spectrum
-lactamase-producing isolates? European Journal of Clinical
Microbiology and Infectious Diseases 15, 3614.

Received 21 June 1999; returned 23 August 1999; revised 22

September 1999; accepted 11 October 1999

189