- Email: [email protected]
Coagulase-negative Staphylococci in Danish blood cultures: species distribution and antibiotic susceptibility
Journal
of Hospital
Infection
(1996)
32,
217-227
Coagulase-negative Staphylococci cultures: species distribution susceptibility
in Danish blood and antibiotic
J. 0. Jarlsv*+, T. Hsjbjergz, C. Busch-Serensen’, J. Scheibel§, J. K. Melled, H. J. J. Kolmos** and D. A. Wandal1f-f *Division of Microbiology, Statens Seruminstitut; f-Department of Clinical Microbiology, National University Hospital (Rigshopitalet), Copenhagen; $Department of Clinical Microbiology, Aalborg Hospital; $Department of Clinical Microbiology, Herlev Hospital, Copenhagen; TDepartment of Clinical Microbiology, Aarhus University Hospital; **Department of Clinical Microbiology, Hvidovre Hospital, Copenhagen; and f+Department of Clinical Microbiology, Bispebjerg Hospital, Copenhagen, Denmark Received
21 September
1995; revised manuscript
accepted 23 November
1995
Summary:
The distribution and antibiotic susceptibility of coagulase-negative staphylococci (CONS) isolated from blood cultures was examined in samples from hospitals covering most of Denmark. A total of 499 CONS isolates were detected in 477 blood cultures from 340 patients and speciated as Staphylococcus epidermidis, 285; Staphylococcus hominis, 61; Staphylococcus haemolyticus, 43; Staphylococcus warneri, 12; Staphylococcus cohnii, 7; Staphylococcus saprophyticus, 4; Staphylococcus capitis, 2 and Staphylococcus Zugdunensis, 1. Seventy-eight isolates could not be identified to species level and six were Micrococcus spp. In 108 (22.6%) blood culture sets, more than one CONS strain were found, as detected by species identification, antibiogram and biotyping. Significantly more blood cultures from patients in university hospitals were drawn from central venous catheters. Comparing university and non-university hospitals, the overall antibiotic susceptibility among CONS was only slightly different, except for methicillin and amikacin. The prevalence of methicillin-resistant strains was 35.1% in the university hospital strains vs. 25.3% in the non-university hospital strains. The overall prevalence of methicillin resistance was 32%. Great geographic variation in both species distribution and antibiotic resistance was observed. The high prevalence of S. epidermidis makes subtyping of this species important. Keywords: antibiograms.
Coagulase-negative
staphylococci;
blood
cultures;
identification;
Introduction
Isolation of coagulase-negative staphylococci (CONS) from blood cultures always raises the question of whether the bacteria represents a true inCorrespondence Juliane Maria
to: Jens Otto Jarlov, Department Vej 22, DK 2100 Copenhagen
0,
of Clinical Denmark.
Microbiology,
Afsnit
9301,
Rigshospitalet,
218
J. 0. Jarlrav et aZ.
fection.lm7 CONS are normal inhabitants of the skin and mucous membranes, but may cause infection in the immunocompromised host, neonates,4.7 intensive-care patients and others especially those with indwelling i.v. devices.lm9 Antibiotic resistance among CONS in Denmark, especially methicillin resistance, seems to follow the same pattern as in other countries”‘” in contrast to methicillin-resistant Staphylococcus auwus, which has become less common, with more than half of cases being imported.” In a previous study dealing with staphylococci in blood cultures from non-university hospitals in two smaller Danish counties,13 the prevalence of methicillin resistance was 40% for Staphylococcus epidermidis and 21% for other CONS species. Furthermore, we found a high number of specimens that grew more than one CONS strain.13 In order to investigate the prevalence of methicillin resistance, resistance to other antibiotics and the extent of growth of multiple CONS strains in Danish blood cultures, we conducted the present study, which included hospitals from five of the 14 Danish counties and Copenhagen Municipality. Materials
and
methods
The following five Danish counties: Northern Jutland, Aarhus, Funen, Roskilde, Copenhagen, representing different parts of Denmark, were included in this study together with Copenhagen Municipality. The investigated regions represent 2.9 million inhabitants out of the total Danish population of 5.2 million. All blood cultures that were CONS positive in a two-month period in the autumn 1992 or winter 1992-i 993 were included. From Roskilde county the collection time was expanded to three months. The blood culture systems used are listed, together with the number of CONS-positive specimens in Table I. It was noted whether the patients had one or more central venous catheters (CVC) and if the blood culture was drawn from a CVC. Species identi$cation and susceptibility testing All positive blood cultures were subcultured locally and either the agar plate or a heavy inoculum from the agar plate was referred to the Staphylococcus Laboratory, Statens Seruminstitut, Copenhagen. In order to detect the possible presence of more than one CONS species or type, a heavy inoculum (to obtain confluent growth) was spread on a blood agar plate and susceptibility testing performed. In addition, a stepwise cultivation was done, and if either different colony morphologies or susceptibility-patterns were observed, each different colony was further tested: (1) species identification was performed with a modified Kloos and Schleifer scheme14 supplemented by an ornithine-decarboxylase test, (2) biotyping was performed by means of Minibact-S,‘” (3) a secondary susceptibility test with semiconfluent growth was performed with penicillin (breakpoint 0.25 pg), methicillin
CONS Table
Roskilde Copenhagen Couyty Copenhagen Mumcqxlity: Bispebjerg Hvidovre Funen County Aarhus County Northern Jutland County Total
in
I. Distribution
Danish
blood
of patients
cultures
219
and blood
cultures
No. of patients
No. of patients with CVC Yes/No/ Unknolvn
No. of blood cultures
36 79
512813 19/40/20
40 158
5:
21 30 67 48 113 477
I 8 11 1
19 2.5 60 44 77 340*
4/l s/o 11/10/4 25/l 3122 2117125 9/52/l 6 75/175/90
* 175 females and 165 males. ‘rhe following blood culture systems u-ere used: ‘Colorbact’: Northern Jutland County. ‘Bactec’: Plarhus County. Roche: system (Statens Seruminstitut): TTunen County. CVC, central venous catheter.
No. of cvcdrawn blood cultures
Roskilde, Bispebjerg.
No. of blood cultures with >l strain
ki
Copenhagen Conventional
No. of strains
7 25
45 128
10 6 21 13 26 108
31 32 88 59 116 116
County, blood
Hvidovre, culture
(8 pg), cephalothin (4 pg), cefuroxime (4 pg), cefotaxime (2 pg), streptomycin (8 pg), gentamicin (4 pg), netilmicin (4 pg), amikacin (4 pg), tetracycline (2 pg), chloramphenicol(6 pg), ciprofloxacin (1 pg), erythromycin (1 pg), fusidic acid (2 pg), rifampicin (0.5 pg), teicoplanin (2 pg) and vancomycin (2 pg) using ‘Neosensitabs’ (Rosco, Taastrup, Denmark).‘” The plates were incubated at 3 5°C and read after 18 h. Methicillin susceptibility was tested on an agar plate with 7.5% NaCl. The latter plates were read after 18 and 48 h and zone-diameter less than 21 mm, or the presence of colonies within the inhibition zone, was used as indicative of methicillin resistance. Identical strains found in different blood cultures from the same patient was included only once in the analysis. Statistical analysis Chi-square test with Yates correction or Fisher’s exact test were used to evaluate differences. PqO.05 was considered significant. Results
A total of 477 blood cultures from 340 patients grew CONS. More than one blood culture (range 2-1.5) were received from 71 patients. Four hundred and ninety-nine different CONS strains were identified and speciated as S. epidermidis, 285; Staphylococcus hominis, 61; Staphylococcus haemolyticus, 43; Staphylococcus warneri, 12; Staphylococcus cohnii, 7; Staphylococcus saprophyticus, 4; Staphylococcus capitis, 2; and one Staphylococcus lugdunensis. Seventy-eight isolates could not be identified to species level and six strains were identified as Micrococcus spp.
220
J. 0. Jarlcav et GE. Table
Percentage resistant strains according to subspecies
II.
Percentage Species
Staphylococcus epidermidis N=285 Staphylococcus hominis N=61 Staphylococcus haemolyticus N=43 Staphylococcus warneri N=l2 Staphylococcus cohnii N=7 Staphylococcus saprophyticus N=4 Staphylococcus capitis N=2 Staphylococcus lugdunensis N=l Micrococcus spp.
resistance
pen
met
str
gen
net
ami
77
41
5
33
5
28
54
18
12
10
0
61
33
23
16
42
0
8
57
71
71
100
0
00 S
S
chl
cry
fus
rif
tip
17
10
28
30
4
5
7
26
8
28
10
0
7
0
20
26
5
33
5
0
9
8
8
0
17
8
0
0
0
0
29
14
57
57
57
57
0
57
43
tet
0
0
0
25
0
0
0
0
0
0
0
0
0
0
0
100
0
0
0
0
R
S
s
s
s
s
s
S
s
s
N=6 Unidentified N=78
67
50
50
17
0
33
17
17
17
17
17
51
15
6
13
1
10
18
3
9
17
0
0
Total
67.5
32.1
8.8
24.2
3.2
21-8
19.4
10.2
24.6
22.4
3-4
5.6
All strains were sensitive to vancornycin and teicoplanin. Pen, penicillin G; met, methicillin; str, streptomycin; gen, gentamicin; net, netilmicin; tetracyclin; chl, chloramphenicol; ery, erythromycin; fus, fusidic acid; rif, rifampicin;
17
ami, amikacin; tet, tip, ciprofloxacin.
More than one CONS strain were found in 108 (22.6%) (two strains in 102 and three in six) blood culture sets from 97 patients. In 78 of these blood cultures, the strains belonged to different CONS-species. In 25 blood cultures, the strains differed in both susceptibility patterns and biotype, but belonged to the same species, mainly S. epidemzidis. Major differences in susceptibility patterns of the same species were found in four blood cultures, and major differences in biotype were found in one blood culture. The antibiotic susceptibility-patterns of the CONS-species are noted in Table II, and the number of strains resistant to one or more antibiotics are listed in Table III. The most resistant species seemed to be S. cohnii followed by S. epidermidis and S. haemolyticus. Significant geographic variation in the species distribution and in antibiotic susceptibility was observed (Tables IV and V). Methicillin resistance ranged from 40% (51/
CONS in Danish Table
III.
The
distribution
Suscew tible. to all
Species Staphylococcus epidermidis N=285 Staphylococcus hominis N=61 Staphylococcus haemolyticus N=43 Staphylococcus %*arneri N=12 Staphylococcus cohnii N=7 Staphylococcus saprophyticus N=4 Staphylococcus capitis N=2 Staphylococcus lugdunensis N=l Micrococcus spp. N=6 Unidentified N=78 * Shown
in Table
of CONS-strains
Number
blood
resistant antibiotics*
of strains
221
cultures to various
resistant
numbers
to various
of the investigated
numbers
of antibiotics
1
2
3
4
5
6
7
8
9
10
11
46
69
37
25
35
32
21
14
6
0
0
0
17
16
15
4
3
4
0
2
0
0
0
9
17
2
2
6
0
2
3
1
0
0
5
4
2
1
0
0
0
0
0
0
0
0
1
0
2
0
0
0
1
1
1
1
0
3
1
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
2
0
0
1
1
1
0
0
0
0
0
25
30
6
6
4
2
3
2
0
0
0
II
128) in Copenhagen County to 21% (24/l 16) in Northern Jutland County (Table V). The p revalence of S. epidermidis ranged from 73% (46/63) in Copenhagen Municipality to 46% (53/l 16) in Northern Jutland County. Those CONS strains, which could not be identified, were more prevalent in Aarhus and Northern Jutland counties than in the Copenhagen area (Table IV). The most pronounced differences in antibiotic susceptibility between the different areas are listed in Table V. In general, the highest frequency of antibiotic resistance was found in the Copenhagen area and in Aarhus county (Table V). No pronounced differences were found in antibiotic susceptibility between university and non-university hospitals, except for methicillin where 35.1% (121/345) were resistant in the university hospitals (Table hospitals and 25.3% (39/l 54) resistant in the non-university V) and amikacin 26% (89/345) vs. 13% (20/154). Eighty-eight blood cultures from 75 patients were drawn through a CVC (Table I). No information about the presence of a CVC could be obtained
J. 0. Jarlev
222 Table
Species
Roskilde County N=45
Staphylococcus epidermis N=285 58 Staphylococcus hominis N=61 13 Staphylococcus haemolyticus N=43 2 Staphylococcus warneri N=12 4 Staphylococcus cohnii N=7 2 Staphylococcus saprophyticus N=4 0 Staphylococcus capitis N=2 0 Staphylococcus lugdunensis N=l 0 Micrococcus spp. N=6 0 Unidentified N=78 20 * Significantly
different
IV.
Percentage
Copenhagen County N=128
distribution Copenhagen Municipality N=63
et al.
of CONS-specied
Funen County N=88
Aarhus County N=59
in Danish
regions
Northern Jutland County N=116
University hospitals N=345
Nonuniversity hospitals N=154
57
73*
66
49
46*
57.7
55.8
15
10
6
9
17
11.0
14.9
12
6
8
14
7
9.8
5.8
2
2
5
2
2
2.3
2.6
2
2
0
2
1
1.7
0.6
0
0
5
0
0
1.1
0
1
0
0
0
1
0.6
0
1
0
0
0
0
0.3
0
1
0
1
2
3
1.4
0.6
8-t
10-t
24~
24$
14.8
17.5
lot (BO.05);
+ vs.$
Significantly
different
(P
concerning the blood cultures from 90 patients. These blood cultures, which were similar to the blood cultures from patients without CVC with regard to number of CONS strains in the blood culture and susceptibilitypatterns, are not included in the calculations concerning CVC. Significantly more blood cultures from patients in university hospitals were CVC-drawn [82/252 (32.5%) vs. 6/90 (6*7%)] and more CONS isolates from patients with CVC were methicillin resistant [54/120 (45%) vs. 67/ 255 (26*3%)] (Table VI). Furthermore, significantly more strains from patients with CVC were S. epidermidis [84/120 (70%) vs. 135/255 (53%), P
CONS in Danish Table
\r.
Percentage
blood
resistant
cultures
strains
according
Percentage Region Roskilde Count) ,v=45 Copenhagen County *V=128 Copenhagen Municipality X=67 Funen County N=X8 Aarhus County N=59 Northern Jutland Count) X=116 University hospitals TV= 345 Non-universit) hospitals N=l54 Total
pen
met
str
f-w
net
223 to regions
resistance
ami
tet
chl
cry
fus
rif
tip
67
29
21
27
2
13
20
4
27
27
2
4
69
40*
14-t
23
2
27t
17
9
32*
27t
5
4
81
39
6
27
8
25
16
1X
31
24
12*
IO
66
31
6
17
1
26
9
7
IS*
19
1
3
63
34
IO
32t
7
25
12
12
20
2.i
2
IO
63
21*
9
16-t
2
13t
20
11
21
161
1*
4
68.7
35.11
9-5
25.6
3.7
25.9*
19.X
II.5
24.4
24.1
4.3
5..i
65.6
25.3-t
7.1
214
I.9
13*
18.2
7.1
24.7
18.2
I.9
5.8
67.7
32.1
8.8
24.3
3.2
21.9
19.3
IO.2
24.5
22.3
3.6
5.6
* Significantly different (PO.01) from -t Significantly different (P
other other and
result in the result in the teicoplanin.
same same
column column
with with
the the
same same
mark. mark.
patients without CVC, but significantly fewer CI/C drawn blood cultures grew more than one CONS-strain; comparing only the blood cultures from the patients known to have a CVC, this difference was the same [lo/88 (11.4%) vs. lo/28 (35.70/o), Table VI] and no significant difference in methicillin resistance among the CONS strains could be found [28/63 (44.4%) vs. 16/33 (48..5%)].
Discussion
Many papers dealing with antibiotic resistance, especially methicillin resistance, in CONS have focused on tertiary care/university hospitals.‘-” A high degree of antibiotic resistance has been found. The skin flora of nontreated individuals has been found to include a vast number of multiply resistant CONS.” Methicillin-resistant CONS were found in a low number on the skin. In a previous paper,13 concerning two small Danish counties, it was noted
present
VI.
The
of central
2 25 228 108
141
number of blood cultures (N=477)
Total
influence
catheters
on CONS
18* 6* 2 2 31
1:: 77
(BO.05).
more
Blood cultures from nonuniversity hospitals (N=133)
(CVC)
123* 82* 26
cultures from university hospitals (N= 344)
Blood
venous
The blood cultures with no information about CVC are not included in the statistical calculations. *Significantly different (RO~OOOOl) from other result in the same row with the same mark. T Significantly different (P
Blood
cultures CVC-drawn Blood cultures not CVC-drawn Unknown CVC not present No information about CVC
CVC
Table
than
cultures
one
isolate.
This
54/120 (45)t 28/63 (44.4)§() 16/33 (48.5)ij lo/24 (41.7) 67/255 (26.3)t 39/124 (31.5)
Methicillin resistant/total isolates (%) 160/499 (32.1)
in blood
difference
is significant
25/141 (17,7)9 10/88 (11.4)$1 10/28 (35.7)s S/25 (20) 58/228 (25.4)s 25/108 (23.1)
Number of blood cultures with >I CONS isolate/total Number of cultures (%) 1081477 (22.6)
t? g c it p -
9
4
w P
CONS in Danish
blood
cultures
225
that growth of multiple CONS strains in blood cultures was a frequent event. The present study was conducted to investigate this phenomenon and to survey antibiotic susceptibility among CONS isolates from different regions in Denmark. Remarkable geographic variation in the prevalence of different CoNSspecies and antibiotic susceptibility was observed. This could not only be explained by differences in hospital categories because all the regions studied, except for Roskilde County, included both university hospitals and smaller hospitals. In a previous Danish study some geographic variability in the prevalence of CONS species and phage-typability was found.18 These findings connected with our results might indicate the occurrence in hospitals of not only a local hospital flora, but also a local community flora from which the local hospital flora is derived. The antibiotic policy in the local area is probably important in selecting resistant CONS-strains in the hospitals.“~” Resistance to methicillin and amikacin were more prevalent among CONS strains from university hospitals compared with the smaller hospitals. Because amikacin is rarely used in Denmark, the rather high incidence of amikacin resistance was surprising. A marked co-variation between amikacin, gentamicin and methicillin was found (data not shown), pointing towards co-selection and this could in part explain the difference between the types of hospitals. However, it should be noted that a large number of the amikacin resistant strains had zone diameters near the breakpoint zone and these strains may not be truly resistant, in vitro. The influence of the presence of a CVC is difficult to evaluate. Not surprisingly, more patients from university hospitals had a CVC. Blood cultures from patients with CVC usually had growth on one single CONS strain, and a high number were methicillin resistant S. epidermidis. This might indicate that these strains were selected in the hospital, had colonized the CVC and, thereby, caused bacteraemia. Two or more CONS strains were frequently found in one blood culture. The significance of a blood culture with growth of multiple CONS strains is a matter for debate. In some instances, it seems reasonable to regard the patient to be truly infected with one or more of these multiple isolates, but the majority of cases should be considered insignificant.13’20 It is important to detect the CONS strain that has the highest degree of antibiotic resistance in order to treat the patient with the correct antibiotics. Several colonies of CONS from the blood culture should, therefore, be included when primary antibiotic susceptibility testing is performed. If only one or two colonies are included the more resistant isolates may be overlooked. Several papers have discussed the method that should be used for the detection of methicillin resistance.21-24 Isolates, which were found to be resistant in our agar diffusion method with 7.5% NaCl, had minimum
J. 0. Jarlev
226
et al.
inhibitory concentrations (MI&) corresponding to 38 mg/L, in a platedilution method using Columbia agar with 4.5% NaCl and an inoculum of approximately 1O6cfu (unpublished data). However, our plates were difficult to read. A NaCl concentration of 795% is probably not optimal with CONS, which grow with some variation on this medium. MIC breakpoints recommended by others vary from 4 mg/L to 16 mg/L methicillin,22-24 depending on the methods used. And even the breakpoint for oxacillin (i.e., 62 mg/L), recommended by the National Committee for Clinical Laboratory Standards, has been questioned recently because this breakpoint did not correlate well with the presence of met-A gene.25 Further investigation is needed before a definite method for the detection of methicillin resistance can be recommended. We thank Dr Bente Gahm-Hansen, for supplying supported in part by Eli Lilly Denmark A/S.
the specimens
from
Funen
County. This work was
References 1. Dominguez-de Villota E, Algora-Weber A, Millan I, Galdos P, Mosguera JM. Early evaluation of coagulase-negative staphylococcus in blood samples of intensive care unit patients. A clinically uncertain judgement. Intensive Care Med 1987; 13: 390-394. 2. Kirchoff LV, Sheagren JN. Epidemiology and clinical significance of blood cultures positive for coagulase-negative staphylococcus. Infect Control 1985; 6: 479-486. 3. Leibovici L, Gransden WR, Eykyn SJ, Konsiberger H, Deucker M, Pitlik SD, Phillips I. Clinical index to predict bacteraemia caused by staphylococci. J Intern Med 1993; 234: 83-89. Marshall RJ, Davies AJ, Kirk R, Reeves DS. The laboratory interpretation of coagulasenegative staphylococcal bacteraemia in neonates. J Hosp Infect 1988; 13: 29.5-299. Martin MA, Pfaller MA, Wenzel RP. Coagulase-negative staphylococcal bacteremia. Ann Int Med 1989; 110: 9-16. Ringberg H, Thoren A, Bredberg A. Evaluation of coagulase-negative staphylococci in blood cultures. ScandJ Infect Dis 1991; 23: 315-323. Sidebottom DG, Freeman J, Platt R, Epstein MF, Goldmann DA. Fifteen-year experience with bloodstream isolates of coagulase-negative staphylococci in neonatal intensive care. J Clin Microbial 1988; 26: 713-718. 8. Ziert CH. Evidence of transient Staphylococcus epidermidis bacteremia in patients and in healthy humans. J Clin Microbial 1983; 17: 628-630. 9. Eykyn SJ. Staphylococcal sepsis. The changing pattern of disease and therapy. Lancet 1988; 331: 100-103. 10. Meller JK. Observations on multiple drug resistance in coagulase-negative staphylococci isolated in hospitals from 1975 to 1985. r Hosp Infect 1988; 11: 26-35. 11. Moller JK. Antimicrobial usage and microbial resistance in a university hospital during a seven-year period. J Antimicrob Chemother 1989; 24: 983-992. VT, Knudsen AM. The decline of methicillin resistance among Danish 12. Rosdahl Staphylococcus aureus strains. Infect Control Hosp Epidemiol 1991; 12: 83-88. 13. Jarlov JO, Prag J, Rosdahl VT, Espersen F. Evaluation of staphylococci from a blood culture system (‘Colorbact’). APMIS 1995; 103: 383-387. 14. Gahrn-Hansen B, Heltberg 0, Rosdahl VT, Sogaard P. Evaluation of a conventional routine method for identification of clinical isolates of coagulase-negative Staphylococcus and Micrococcus species. Acta Path01 Microbial Immunol Stand Sect B 1987; 95: 283-292. B, Rosdahl VT, Espersen F. Identification of coagulase-negative 1.5. Jarlov JO, Nissen staphylococci and typing of Staphylococcus epidermidis by a 4 h micromethod. APMIS
1994; 102: 272-278. 16. Casals JB, Pedersen Acta
Path01
Microbial
OG. Tablet Immunol
sensitivity Stand Sect
testing: B 1972;
a comparison 80: 806-816.
of different
methods.
CONS in Danish 17. 18. 19.
20. 21. 22. 23. 24.
25.
blood
cultures
227
Cove JH, Eady EA, Cunliffe WJ. Skin carriage of antibiotic-resistant coagulase-negative staphylococ; in untreated subjects. J Antimicrobial Chemother 1990; 25: 459-469. Rosdahl VT, Gahrn-Hansen B, Moller JK, Kjzldgaard P. Phage-typing of coagulasenegative staphylococci. Factors influencing typability. APSIS 1990; 98: 299-304. Mouton RP, Hermans J, Simoons-Smit AM, Hoogkamp-Korstanje JAA, Degener JE, van Klingeren B. Correlations between consumption of antibiotics and methicillin resistance in coagulase-negative staphylococci. J Antimicrobial Chemother 1990; 26: 573-583. Ryding J, Espersen F, Jarlov JO, Rosdahl VT. Coagulase-negati\re staphylococci in a newborn with a cardiac vascular prosthesis. APSIS 1992; 100: 246-248. Baker CS, Huang MB, Tenover FC. Optimizing testing of methicillin-resistant Stnphylococcus species. Diagn Microbial Infect Dis 1994; 19: 167-170. Hedin G, Hambraeus A. Screening tests for the detection of methicillin resistance in Staphylococcus epidermidis. J AntiGicvob Chemother 1991; 28: 681-694. Hedin G. Liifdahl S. Detecting methicillin-resistant Staahvlococcus ebidermidis-disc diffusion,‘broth breakpoint or F;blymerase chain reaction? LA?MZS 1995; 101: 3 11-318. Law D, lVIegson GM, Keaney MGL, Ganguli LA. The influence of salt concentration on the detection of methicillin resistance in coagulase-negative staphylococci. J Antinzicrob Chemother 1992; 30: 603-614. McDonald CL, Maher \$‘E, Fass RJ. Revised interpretation of oxacillin nIICs for Staphylococczcs epidermidis based on met-A-detection. Antimicroh L!!gents Chenzothe? 1995: 39: 982-984.
of Hospital
Infection
(1996)
32,
217-227
Coagulase-negative Staphylococci cultures: species distribution susceptibility
in Danish blood and antibiotic
J. 0. Jarlsv*+, T. Hsjbjergz, C. Busch-Serensen’, J. Scheibel§, J. K. Melled, H. J. J. Kolmos** and D. A. Wandal1f-f *Division of Microbiology, Statens Seruminstitut; f-Department of Clinical Microbiology, National University Hospital (Rigshopitalet), Copenhagen; $Department of Clinical Microbiology, Aalborg Hospital; $Department of Clinical Microbiology, Herlev Hospital, Copenhagen; TDepartment of Clinical Microbiology, Aarhus University Hospital; **Department of Clinical Microbiology, Hvidovre Hospital, Copenhagen; and f+Department of Clinical Microbiology, Bispebjerg Hospital, Copenhagen, Denmark Received
21 September
1995; revised manuscript
accepted 23 November
1995
Summary:
The distribution and antibiotic susceptibility of coagulase-negative staphylococci (CONS) isolated from blood cultures was examined in samples from hospitals covering most of Denmark. A total of 499 CONS isolates were detected in 477 blood cultures from 340 patients and speciated as Staphylococcus epidermidis, 285; Staphylococcus hominis, 61; Staphylococcus haemolyticus, 43; Staphylococcus warneri, 12; Staphylococcus cohnii, 7; Staphylococcus saprophyticus, 4; Staphylococcus capitis, 2 and Staphylococcus Zugdunensis, 1. Seventy-eight isolates could not be identified to species level and six were Micrococcus spp. In 108 (22.6%) blood culture sets, more than one CONS strain were found, as detected by species identification, antibiogram and biotyping. Significantly more blood cultures from patients in university hospitals were drawn from central venous catheters. Comparing university and non-university hospitals, the overall antibiotic susceptibility among CONS was only slightly different, except for methicillin and amikacin. The prevalence of methicillin-resistant strains was 35.1% in the university hospital strains vs. 25.3% in the non-university hospital strains. The overall prevalence of methicillin resistance was 32%. Great geographic variation in both species distribution and antibiotic resistance was observed. The high prevalence of S. epidermidis makes subtyping of this species important. Keywords: antibiograms.
Coagulase-negative
staphylococci;
blood
cultures;
identification;
Introduction
Isolation of coagulase-negative staphylococci (CONS) from blood cultures always raises the question of whether the bacteria represents a true inCorrespondence Juliane Maria
to: Jens Otto Jarlov, Department Vej 22, DK 2100 Copenhagen
0,
of Clinical Denmark.
Microbiology,
Afsnit
9301,
Rigshospitalet,
218
J. 0. Jarlrav et aZ.
fection.lm7 CONS are normal inhabitants of the skin and mucous membranes, but may cause infection in the immunocompromised host, neonates,4.7 intensive-care patients and others especially those with indwelling i.v. devices.lm9 Antibiotic resistance among CONS in Denmark, especially methicillin resistance, seems to follow the same pattern as in other countries”‘” in contrast to methicillin-resistant Staphylococcus auwus, which has become less common, with more than half of cases being imported.” In a previous study dealing with staphylococci in blood cultures from non-university hospitals in two smaller Danish counties,13 the prevalence of methicillin resistance was 40% for Staphylococcus epidermidis and 21% for other CONS species. Furthermore, we found a high number of specimens that grew more than one CONS strain.13 In order to investigate the prevalence of methicillin resistance, resistance to other antibiotics and the extent of growth of multiple CONS strains in Danish blood cultures, we conducted the present study, which included hospitals from five of the 14 Danish counties and Copenhagen Municipality. Materials
and
methods
The following five Danish counties: Northern Jutland, Aarhus, Funen, Roskilde, Copenhagen, representing different parts of Denmark, were included in this study together with Copenhagen Municipality. The investigated regions represent 2.9 million inhabitants out of the total Danish population of 5.2 million. All blood cultures that were CONS positive in a two-month period in the autumn 1992 or winter 1992-i 993 were included. From Roskilde county the collection time was expanded to three months. The blood culture systems used are listed, together with the number of CONS-positive specimens in Table I. It was noted whether the patients had one or more central venous catheters (CVC) and if the blood culture was drawn from a CVC. Species identi$cation and susceptibility testing All positive blood cultures were subcultured locally and either the agar plate or a heavy inoculum from the agar plate was referred to the Staphylococcus Laboratory, Statens Seruminstitut, Copenhagen. In order to detect the possible presence of more than one CONS species or type, a heavy inoculum (to obtain confluent growth) was spread on a blood agar plate and susceptibility testing performed. In addition, a stepwise cultivation was done, and if either different colony morphologies or susceptibility-patterns were observed, each different colony was further tested: (1) species identification was performed with a modified Kloos and Schleifer scheme14 supplemented by an ornithine-decarboxylase test, (2) biotyping was performed by means of Minibact-S,‘” (3) a secondary susceptibility test with semiconfluent growth was performed with penicillin (breakpoint 0.25 pg), methicillin
CONS Table
Roskilde Copenhagen Couyty Copenhagen Mumcqxlity: Bispebjerg Hvidovre Funen County Aarhus County Northern Jutland County Total
in
I. Distribution
Danish
blood
of patients
cultures
219
and blood
cultures
No. of patients
No. of patients with CVC Yes/No/ Unknolvn
No. of blood cultures
36 79
512813 19/40/20
40 158
5:
21 30 67 48 113 477
I 8 11 1
19 2.5 60 44 77 340*
4/l s/o 11/10/4 25/l 3122 2117125 9/52/l 6 75/175/90
* 175 females and 165 males. ‘rhe following blood culture systems u-ere used: ‘Colorbact’: Northern Jutland County. ‘Bactec’: Plarhus County. Roche: system (Statens Seruminstitut): TTunen County. CVC, central venous catheter.
No. of cvcdrawn blood cultures
Roskilde, Bispebjerg.
No. of blood cultures with >l strain
ki
Copenhagen Conventional
No. of strains
7 25
45 128
10 6 21 13 26 108
31 32 88 59 116 116
County, blood
Hvidovre, culture
(8 pg), cephalothin (4 pg), cefuroxime (4 pg), cefotaxime (2 pg), streptomycin (8 pg), gentamicin (4 pg), netilmicin (4 pg), amikacin (4 pg), tetracycline (2 pg), chloramphenicol(6 pg), ciprofloxacin (1 pg), erythromycin (1 pg), fusidic acid (2 pg), rifampicin (0.5 pg), teicoplanin (2 pg) and vancomycin (2 pg) using ‘Neosensitabs’ (Rosco, Taastrup, Denmark).‘” The plates were incubated at 3 5°C and read after 18 h. Methicillin susceptibility was tested on an agar plate with 7.5% NaCl. The latter plates were read after 18 and 48 h and zone-diameter less than 21 mm, or the presence of colonies within the inhibition zone, was used as indicative of methicillin resistance. Identical strains found in different blood cultures from the same patient was included only once in the analysis. Statistical analysis Chi-square test with Yates correction or Fisher’s exact test were used to evaluate differences. PqO.05 was considered significant. Results
A total of 477 blood cultures from 340 patients grew CONS. More than one blood culture (range 2-1.5) were received from 71 patients. Four hundred and ninety-nine different CONS strains were identified and speciated as S. epidermidis, 285; Staphylococcus hominis, 61; Staphylococcus haemolyticus, 43; Staphylococcus warneri, 12; Staphylococcus cohnii, 7; Staphylococcus saprophyticus, 4; Staphylococcus capitis, 2; and one Staphylococcus lugdunensis. Seventy-eight isolates could not be identified to species level and six strains were identified as Micrococcus spp.
220
J. 0. Jarlcav et GE. Table
Percentage resistant strains according to subspecies
II.
Percentage Species
Staphylococcus epidermidis N=285 Staphylococcus hominis N=61 Staphylococcus haemolyticus N=43 Staphylococcus warneri N=l2 Staphylococcus cohnii N=7 Staphylococcus saprophyticus N=4 Staphylococcus capitis N=2 Staphylococcus lugdunensis N=l Micrococcus spp.
resistance
pen
met
str
gen
net
ami
77
41
5
33
5
28
54
18
12
10
0
61
33
23
16
42
0
8
57
71
71
100
0
00 S
S
chl
cry
fus
rif
tip
17
10
28
30
4
5
7
26
8
28
10
0
7
0
20
26
5
33
5
0
9
8
8
0
17
8
0
0
0
0
29
14
57
57
57
57
0
57
43
tet
0
0
0
25
0
0
0
0
0
0
0
0
0
0
0
100
0
0
0
0
R
S
s
s
s
s
s
S
s
s
N=6 Unidentified N=78
67
50
50
17
0
33
17
17
17
17
17
51
15
6
13
1
10
18
3
9
17
0
0
Total
67.5
32.1
8.8
24.2
3.2
21-8
19.4
10.2
24.6
22.4
3-4
5.6
All strains were sensitive to vancornycin and teicoplanin. Pen, penicillin G; met, methicillin; str, streptomycin; gen, gentamicin; net, netilmicin; tetracyclin; chl, chloramphenicol; ery, erythromycin; fus, fusidic acid; rif, rifampicin;
17
ami, amikacin; tet, tip, ciprofloxacin.
More than one CONS strain were found in 108 (22.6%) (two strains in 102 and three in six) blood culture sets from 97 patients. In 78 of these blood cultures, the strains belonged to different CONS-species. In 25 blood cultures, the strains differed in both susceptibility patterns and biotype, but belonged to the same species, mainly S. epidemzidis. Major differences in susceptibility patterns of the same species were found in four blood cultures, and major differences in biotype were found in one blood culture. The antibiotic susceptibility-patterns of the CONS-species are noted in Table II, and the number of strains resistant to one or more antibiotics are listed in Table III. The most resistant species seemed to be S. cohnii followed by S. epidermidis and S. haemolyticus. Significant geographic variation in the species distribution and in antibiotic susceptibility was observed (Tables IV and V). Methicillin resistance ranged from 40% (51/
CONS in Danish Table
III.
The
distribution
Suscew tible. to all
Species Staphylococcus epidermidis N=285 Staphylococcus hominis N=61 Staphylococcus haemolyticus N=43 Staphylococcus %*arneri N=12 Staphylococcus cohnii N=7 Staphylococcus saprophyticus N=4 Staphylococcus capitis N=2 Staphylococcus lugdunensis N=l Micrococcus spp. N=6 Unidentified N=78 * Shown
in Table
of CONS-strains
Number
blood
resistant antibiotics*
of strains
221
cultures to various
resistant
numbers
to various
of the investigated
numbers
of antibiotics
1
2
3
4
5
6
7
8
9
10
11
46
69
37
25
35
32
21
14
6
0
0
0
17
16
15
4
3
4
0
2
0
0
0
9
17
2
2
6
0
2
3
1
0
0
5
4
2
1
0
0
0
0
0
0
0
0
1
0
2
0
0
0
1
1
1
1
0
3
1
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
2
0
0
1
1
1
0
0
0
0
0
25
30
6
6
4
2
3
2
0
0
0
II
128) in Copenhagen County to 21% (24/l 16) in Northern Jutland County (Table V). The p revalence of S. epidermidis ranged from 73% (46/63) in Copenhagen Municipality to 46% (53/l 16) in Northern Jutland County. Those CONS strains, which could not be identified, were more prevalent in Aarhus and Northern Jutland counties than in the Copenhagen area (Table IV). The most pronounced differences in antibiotic susceptibility between the different areas are listed in Table V. In general, the highest frequency of antibiotic resistance was found in the Copenhagen area and in Aarhus county (Table V). No pronounced differences were found in antibiotic susceptibility between university and non-university hospitals, except for methicillin where 35.1% (121/345) were resistant in the university hospitals (Table hospitals and 25.3% (39/l 54) resistant in the non-university V) and amikacin 26% (89/345) vs. 13% (20/154). Eighty-eight blood cultures from 75 patients were drawn through a CVC (Table I). No information about the presence of a CVC could be obtained
J. 0. Jarlev
222 Table
Species
Roskilde County N=45
Staphylococcus epidermis N=285 58 Staphylococcus hominis N=61 13 Staphylococcus haemolyticus N=43 2 Staphylococcus warneri N=12 4 Staphylococcus cohnii N=7 2 Staphylococcus saprophyticus N=4 0 Staphylococcus capitis N=2 0 Staphylococcus lugdunensis N=l 0 Micrococcus spp. N=6 0 Unidentified N=78 20 * Significantly
different
IV.
Percentage
Copenhagen County N=128
distribution Copenhagen Municipality N=63
et al.
of CONS-specied
Funen County N=88
Aarhus County N=59
in Danish
regions
Northern Jutland County N=116
University hospitals N=345
Nonuniversity hospitals N=154
57
73*
66
49
46*
57.7
55.8
15
10
6
9
17
11.0
14.9
12
6
8
14
7
9.8
5.8
2
2
5
2
2
2.3
2.6
2
2
0
2
1
1.7
0.6
0
0
5
0
0
1.1
0
1
0
0
0
1
0.6
0
1
0
0
0
0
0.3
0
1
0
1
2
3
1.4
0.6
8-t
10-t
24~
24$
14.8
17.5
lot (BO.05);
+ vs.$
Significantly
different
(P
concerning the blood cultures from 90 patients. These blood cultures, which were similar to the blood cultures from patients without CVC with regard to number of CONS strains in the blood culture and susceptibilitypatterns, are not included in the calculations concerning CVC. Significantly more blood cultures from patients in university hospitals were CVC-drawn [82/252 (32.5%) vs. 6/90 (6*7%)] and more CONS isolates from patients with CVC were methicillin resistant [54/120 (45%) vs. 67/ 255 (26*3%)] (Table VI). Furthermore, significantly more strains from patients with CVC were S. epidermidis [84/120 (70%) vs. 135/255 (53%), P
CONS in Danish Table
\r.
Percentage
blood
resistant
cultures
strains
according
Percentage Region Roskilde Count) ,v=45 Copenhagen County *V=128 Copenhagen Municipality X=67 Funen County N=X8 Aarhus County N=59 Northern Jutland Count) X=116 University hospitals TV= 345 Non-universit) hospitals N=l54 Total
pen
met
str
f-w
net
223 to regions
resistance
ami
tet
chl
cry
fus
rif
tip
67
29
21
27
2
13
20
4
27
27
2
4
69
40*
14-t
23
2
27t
17
9
32*
27t
5
4
81
39
6
27
8
25
16
1X
31
24
12*
IO
66
31
6
17
1
26
9
7
IS*
19
1
3
63
34
IO
32t
7
25
12
12
20
2.i
2
IO
63
21*
9
16-t
2
13t
20
11
21
161
1*
4
68.7
35.11
9-5
25.6
3.7
25.9*
19.X
II.5
24.4
24.1
4.3
5..i
65.6
25.3-t
7.1
214
I.9
13*
18.2
7.1
24.7
18.2
I.9
5.8
67.7
32.1
8.8
24.3
3.2
21.9
19.3
IO.2
24.5
22.3
3.6
5.6
* Significantly different (PO.01) from -t Significantly different (P
other other and
result in the result in the teicoplanin.
same same
column column
with with
the the
same same
mark. mark.
patients without CVC, but significantly fewer CI/C drawn blood cultures grew more than one CONS-strain; comparing only the blood cultures from the patients known to have a CVC, this difference was the same [lo/88 (11.4%) vs. lo/28 (35.70/o), Table VI] and no significant difference in methicillin resistance among the CONS strains could be found [28/63 (44.4%) vs. 16/33 (48..5%)].
Discussion
Many papers dealing with antibiotic resistance, especially methicillin resistance, in CONS have focused on tertiary care/university hospitals.‘-” A high degree of antibiotic resistance has been found. The skin flora of nontreated individuals has been found to include a vast number of multiply resistant CONS.” Methicillin-resistant CONS were found in a low number on the skin. In a previous paper,13 concerning two small Danish counties, it was noted
present
VI.
The
of central
2 25 228 108
141
number of blood cultures (N=477)
Total
influence
catheters
on CONS
18* 6* 2 2 31
1:: 77
(BO.05).
more
Blood cultures from nonuniversity hospitals (N=133)
(CVC)
123* 82* 26
cultures from university hospitals (N= 344)
Blood
venous
The blood cultures with no information about CVC are not included in the statistical calculations. *Significantly different (RO~OOOOl) from other result in the same row with the same mark. T Significantly different (P
Blood
cultures CVC-drawn Blood cultures not CVC-drawn Unknown CVC not present No information about CVC
CVC
Table
than
cultures
one
isolate.
This
54/120 (45)t 28/63 (44.4)§() 16/33 (48.5)ij lo/24 (41.7) 67/255 (26.3)t 39/124 (31.5)
Methicillin resistant/total isolates (%) 160/499 (32.1)
in blood
difference
is significant
25/141 (17,7)9 10/88 (11.4)$1 10/28 (35.7)s S/25 (20) 58/228 (25.4)s 25/108 (23.1)
Number of blood cultures with >I CONS isolate/total Number of cultures (%) 1081477 (22.6)
t? g c it p -
9
4
w P
CONS in Danish
blood
cultures
225
that growth of multiple CONS strains in blood cultures was a frequent event. The present study was conducted to investigate this phenomenon and to survey antibiotic susceptibility among CONS isolates from different regions in Denmark. Remarkable geographic variation in the prevalence of different CoNSspecies and antibiotic susceptibility was observed. This could not only be explained by differences in hospital categories because all the regions studied, except for Roskilde County, included both university hospitals and smaller hospitals. In a previous Danish study some geographic variability in the prevalence of CONS species and phage-typability was found.18 These findings connected with our results might indicate the occurrence in hospitals of not only a local hospital flora, but also a local community flora from which the local hospital flora is derived. The antibiotic policy in the local area is probably important in selecting resistant CONS-strains in the hospitals.“~” Resistance to methicillin and amikacin were more prevalent among CONS strains from university hospitals compared with the smaller hospitals. Because amikacin is rarely used in Denmark, the rather high incidence of amikacin resistance was surprising. A marked co-variation between amikacin, gentamicin and methicillin was found (data not shown), pointing towards co-selection and this could in part explain the difference between the types of hospitals. However, it should be noted that a large number of the amikacin resistant strains had zone diameters near the breakpoint zone and these strains may not be truly resistant, in vitro. The influence of the presence of a CVC is difficult to evaluate. Not surprisingly, more patients from university hospitals had a CVC. Blood cultures from patients with CVC usually had growth on one single CONS strain, and a high number were methicillin resistant S. epidermidis. This might indicate that these strains were selected in the hospital, had colonized the CVC and, thereby, caused bacteraemia. Two or more CONS strains were frequently found in one blood culture. The significance of a blood culture with growth of multiple CONS strains is a matter for debate. In some instances, it seems reasonable to regard the patient to be truly infected with one or more of these multiple isolates, but the majority of cases should be considered insignificant.13’20 It is important to detect the CONS strain that has the highest degree of antibiotic resistance in order to treat the patient with the correct antibiotics. Several colonies of CONS from the blood culture should, therefore, be included when primary antibiotic susceptibility testing is performed. If only one or two colonies are included the more resistant isolates may be overlooked. Several papers have discussed the method that should be used for the detection of methicillin resistance.21-24 Isolates, which were found to be resistant in our agar diffusion method with 7.5% NaCl, had minimum
J. 0. Jarlev
226
et al.
inhibitory concentrations (MI&) corresponding to 38 mg/L, in a platedilution method using Columbia agar with 4.5% NaCl and an inoculum of approximately 1O6cfu (unpublished data). However, our plates were difficult to read. A NaCl concentration of 795% is probably not optimal with CONS, which grow with some variation on this medium. MIC breakpoints recommended by others vary from 4 mg/L to 16 mg/L methicillin,22-24 depending on the methods used. And even the breakpoint for oxacillin (i.e., 62 mg/L), recommended by the National Committee for Clinical Laboratory Standards, has been questioned recently because this breakpoint did not correlate well with the presence of met-A gene.25 Further investigation is needed before a definite method for the detection of methicillin resistance can be recommended. We thank Dr Bente Gahm-Hansen, for supplying supported in part by Eli Lilly Denmark A/S.
the specimens
from
Funen
County. This work was
References 1. Dominguez-de Villota E, Algora-Weber A, Millan I, Galdos P, Mosguera JM. Early evaluation of coagulase-negative staphylococcus in blood samples of intensive care unit patients. A clinically uncertain judgement. Intensive Care Med 1987; 13: 390-394. 2. Kirchoff LV, Sheagren JN. Epidemiology and clinical significance of blood cultures positive for coagulase-negative staphylococcus. Infect Control 1985; 6: 479-486. 3. Leibovici L, Gransden WR, Eykyn SJ, Konsiberger H, Deucker M, Pitlik SD, Phillips I. Clinical index to predict bacteraemia caused by staphylococci. J Intern Med 1993; 234: 83-89. Marshall RJ, Davies AJ, Kirk R, Reeves DS. The laboratory interpretation of coagulasenegative staphylococcal bacteraemia in neonates. J Hosp Infect 1988; 13: 29.5-299. Martin MA, Pfaller MA, Wenzel RP. Coagulase-negative staphylococcal bacteremia. Ann Int Med 1989; 110: 9-16. Ringberg H, Thoren A, Bredberg A. Evaluation of coagulase-negative staphylococci in blood cultures. ScandJ Infect Dis 1991; 23: 315-323. Sidebottom DG, Freeman J, Platt R, Epstein MF, Goldmann DA. Fifteen-year experience with bloodstream isolates of coagulase-negative staphylococci in neonatal intensive care. J Clin Microbial 1988; 26: 713-718. 8. Ziert CH. Evidence of transient Staphylococcus epidermidis bacteremia in patients and in healthy humans. J Clin Microbial 1983; 17: 628-630. 9. Eykyn SJ. Staphylococcal sepsis. The changing pattern of disease and therapy. Lancet 1988; 331: 100-103. 10. Meller JK. Observations on multiple drug resistance in coagulase-negative staphylococci isolated in hospitals from 1975 to 1985. r Hosp Infect 1988; 11: 26-35. 11. Moller JK. Antimicrobial usage and microbial resistance in a university hospital during a seven-year period. J Antimicrob Chemother 1989; 24: 983-992. VT, Knudsen AM. The decline of methicillin resistance among Danish 12. Rosdahl Staphylococcus aureus strains. Infect Control Hosp Epidemiol 1991; 12: 83-88. 13. Jarlov JO, Prag J, Rosdahl VT, Espersen F. Evaluation of staphylococci from a blood culture system (‘Colorbact’). APMIS 1995; 103: 383-387. 14. Gahrn-Hansen B, Heltberg 0, Rosdahl VT, Sogaard P. Evaluation of a conventional routine method for identification of clinical isolates of coagulase-negative Staphylococcus and Micrococcus species. Acta Path01 Microbial Immunol Stand Sect B 1987; 95: 283-292. B, Rosdahl VT, Espersen F. Identification of coagulase-negative 1.5. Jarlov JO, Nissen staphylococci and typing of Staphylococcus epidermidis by a 4 h micromethod. APMIS
1994; 102: 272-278. 16. Casals JB, Pedersen Acta
Path01
Microbial
OG. Tablet Immunol
sensitivity Stand Sect
testing: B 1972;
a comparison 80: 806-816.
of different
methods.
CONS in Danish 17. 18. 19.
20. 21. 22. 23. 24.
25.
blood
cultures
227
Cove JH, Eady EA, Cunliffe WJ. Skin carriage of antibiotic-resistant coagulase-negative staphylococ; in untreated subjects. J Antimicrobial Chemother 1990; 25: 459-469. Rosdahl VT, Gahrn-Hansen B, Moller JK, Kjzldgaard P. Phage-typing of coagulasenegative staphylococci. Factors influencing typability. APSIS 1990; 98: 299-304. Mouton RP, Hermans J, Simoons-Smit AM, Hoogkamp-Korstanje JAA, Degener JE, van Klingeren B. Correlations between consumption of antibiotics and methicillin resistance in coagulase-negative staphylococci. J Antimicrobial Chemother 1990; 26: 573-583. Ryding J, Espersen F, Jarlov JO, Rosdahl VT. Coagulase-negati\re staphylococci in a newborn with a cardiac vascular prosthesis. APSIS 1992; 100: 246-248. Baker CS, Huang MB, Tenover FC. Optimizing testing of methicillin-resistant Stnphylococcus species. Diagn Microbial Infect Dis 1994; 19: 167-170. Hedin G, Hambraeus A. Screening tests for the detection of methicillin resistance in Staphylococcus epidermidis. J AntiGicvob Chemother 1991; 28: 681-694. Hedin G. Liifdahl S. Detecting methicillin-resistant Staahvlococcus ebidermidis-disc diffusion,‘broth breakpoint or F;blymerase chain reaction? LA?MZS 1995; 101: 3 11-318. Law D, lVIegson GM, Keaney MGL, Ganguli LA. The influence of salt concentration on the detection of methicillin resistance in coagulase-negative staphylococci. J Antinzicrob Chemother 1992; 30: 603-614. McDonald CL, Maher \$‘E, Fass RJ. Revised interpretation of oxacillin nIICs for Staphylococczcs epidermidis based on met-A-detection. Antimicroh L!!gents Chenzothe? 1995: 39: 982-984.