Validation of antibiotic susceptibility testing guidelines in a routine clinical 1

microbiology laboratory exemplifies general key challenges in setting clinical 2

breakpoints 3

4

Michael Hombach1‡, Patrice Courvalin2, Erik C. Böttger1 5

6

1) Institut für Medizinische Mikrobiologie, Universität Zürich, 8006 Zürich, Schweiz 7

2) Unite des Agents Antibactériens, Institut Pasteur, 75724 Paris, France 8

9

Short title: Irregularities with EUCAST clinical breakpoints 10

11

Keywords: ECOFF, clinical breakpoints, diameter distributions, AST, disk diffusion, 12

CLSI, EUCAST 13

14

‡Corresponding author: 15

Michael Hombach, M.D. 16

Institut für Medizinische Mikrobiologie 17

Universität Zürich 18

Gloriastr. 30/32 19

8006 Zürich 20

Switzerland 21

Phone: 0041 44 634 27 00 22

Fax: 0041 44 634 49 06 23

Email: mhombach@imm.uzh.ch 24

25

26

27

AAC Accepts, published online ahead of print on 28 April 2014Antimicrob. Agents Chemother. doi:10.1128/AAC.02489-13Copyright © 2014, American Society for Microbiology. All Rights Reserved.

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

Abstract 28

29

This study critically evaluated the new European Committee for Antimicrobial 30

Susceptibility Testing (EUCAST) antibiotic susceptibility testing guidelines on the basis of 31

a large set of disk diffusion diameters determined for clinical isolates. We report several 32

paradigmatic problems illustrating key issues in the selection of clinical susceptibility 33

breakpoints, which are of general importance not only for EUCAST but for all guidelines 34

systems, i.e. i) the need for species-specific determination of clinical 35

breakpoints/epidemiological cut-offs (ECOFFs), ii) problems arising from pooling data 36

from various sources, and iii) the importance of the antibiotic disk content for separating 37

non-wild type and wild-type populations. 38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

56

Introduction 57

58

Since 2010 the European Committee for Antimicrobial Susceptibility Testing 59

(EUCAST) has issued regularly updated antibiotic susceptibility testing (AST) guidelines 60

including clinical susceptibility breakpoints (CBPs), which are determined mainly on the 61

basis of epidemiological cut-offs (ECOFFs), pharmacokinetic/pharmacodynamic 62

parameters (PK/PD), and, in part, clinical outcome data (1-3). EUCAST aims at full 63

transparency in the still complex, mostly consensus driven, process of CBP determination 64

by open source publishing of documents on diameter/MIC distributions, and ECOFF data 65

(4). Diameter and MIC distributions may readily be retrieved from the EUCAST webpage 66

under http://www.eucast.org/zone_diameter_distributions. The Clinical and Laboratory 67

Standards Institute (CLSI) uses a similar process for establishing clinical breakpoints (5). 68

In 2010, the Institute of Medical Microbiology (IMM), University of Zurich, changed 69

its guidelines from CLSI to EUCAST. During this shift we have established a database 70

containing non-duplicate disk diffusion diameters of > 30.000 clinical isolates with the 71

view to validate EUCAST AST guidelines as part of the quality control management in our 72

clinical microbiology laboratory (6, 7). Despite our data being largely consistent with 73

EUCAST, we encountered some problems and inconsistencies in EUCAST AST 74

guidelines exemplifying several important key issues in setting CBPs, that, incidentally, 75

also apply to other systems such as CLSI (8): i) the need for species-specific determination 76

of CBPs/ECOFFs, ii) problems arising from pooling data from various sources, and iii) the 77

importance of the antibiotic disk content for the separation of resistant populations from 78

the wild-type population. 79

80

81

82

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

Methods 83

84

Clinical isolates. 85

All non-duplicate clinical isolates included in this study were isolated over a 3-year-86

period from 2010 until 2012 in the clinical microbiology laboratory of the Institute for 87

Medical Microbiology, University of Zurich, mainly serving a 750-bed tertiary-care 88

hospital (University Hospital of Zurich). Isolates of the same species were considered 89

duplicate(s) if they i) originated from the same patient, and ii) showed one major AND two 90

minor differences in AST interpretation at maximum. All duplicates were excluded from 91

the analysis resulting in a reduction of included isolates by 33% in average (species 92

specific reduction: Staphylococcus aureus 39%, coagulase-negative staphylococci (CoNS) 93

32%, Enterococcus faecalis 23%, Enterobacter cloacae 31%, and Klebsiella pneumoniae 94

30%). Only isolates considered clinically relevant were included. The absolute numbers of 95

species/drug combinations for Staphylococcus aureus, CoNS, Enterococcus faecalis, 96

Enterobacter cloacae, and Klebsiella pneumoniae are available from Figures 1 and 2. 97

98

Susceptibility testing. 99

For susceptibility testing the disk (i2a, Montpellier, France) diffusion method was 100

carried out using Mueller-Hinton agar (Becton Dickinson, Franklin Lakes, NJ, USA) and a 101

McFarland 0.5 dilution from overnight cultures as an inoculum followed by incubation at 102

35°C for 16-18 h according to EUCAST recommendations (9). Inhibition zone diameters 103

were determined and recorded using the semi-automated Sirweb/Sirscan system (i2a). 104

105

ECOFF determination 106

ECOFFs were determined by visual inspection of diameter distributions (consensus of 107

five experienced persons, i.e. the ”eyeball method”) (8). EUCAST does not recommend a 108

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

method for ECOFF determination nor indicates the method used to determine EUCAST 109

ECOFFs. 110

111

Software. 112

Calculations were done using Microsoft Excel 2010 software (Microsoft Corporation, 113

Redmond, WA, USA). 114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

Results and Discussion 136

137

Reliable CBPs/ECOFFs require species-specific determination 138

Inhibition zone diameter and MIC distributions are species specific in terms of ranges, 139

means, and standard deviations (8). Thus, ECOFFs and CBPs, by definition, need to be 140

determined species-specific; a principle acknowledged by EUCAST and CLSI (10, 11). To 141

avoid increasing guideline complexity group-specific CBPs have been suggested, e.g. the 142

majority of staphylococcal CBPs are defined at the genus level, or Enterobacteriaceae 143

CBPs that apply at the group level (3, 5). However, it is agreed that CBPs must never split 144

the wild-type population due to biological variation within the wild-type isolates (11). A 145

practical solution to reduce the highly complex process of CBP setting for a large number 146

of drugs, is the use of drug class representatives with AST extrapolation to the other class 147

members (3, 5). 148

For sufficient statistical power, an analysis should include a minimum of 30 isolates to 149

form a Gaussian distribution suitable for ECOFF determination (central limit theorem) 150

(12). The 21,434 Enterobacteriaceae isolates tested for antibiotic susceptibility to 27 151

individual drugs in the IMM clinical laboratory from 2010 to 2012 comprised 11597 (54%) 152

E. coli, 3016 (14%) K. pneumoniae, 1685 (8%) Enterobacter cloacae, 1089 (5%) Proteus 153

mirabilis, 905 (4.2%) Klebsiella oxytoca, 735 (3.4%) Serratia marcescens, 535 (2.5%) 154

Citrobacter koseri, 481 (2.2%) Enterobacter aerogenes, 454 (2.1%) Citrobacter freundii, 155

404 (1.9%) Morganella morganii, 276 (1.3%) Proteus vulgaris, 121 (0.6%) Hafnia alvei, 156

50 (0.23%) Providencia rettgeri, 37 (0.17%) Serratia liquefaciens, and 33 (0.15%) 157

Pantoea agglomerans. For all other Enterobacteriaceae species the IMM database of 158

contained less than 30 isolates. The species listed above accounted for >99.5% of all 159

Enterobacteriaceae isolated in our clinical laboratory. Given these numbers that may serve 160

as an approximation for other laboratories, sufficient AST data would be available for 161

>99.5% of clinical Enterobacteriaceae isolates. E. coli, K. pneumoniae, and E. cloacae 162

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

alone, as the three most frequently isolated species, accounted for 76.1% of all isolates. 163

Automated diameter reading and documentation systems can facilitate the collection of 164

comparably high numbers of diameter values. 165

Evaluating our clinical isolate database we observed that genus and group-specific 166

CBPs result in several problems including potential interpretation errors. This situation is 167

best exemplified by fluoroquinolones and staphylococci, and ertapenem and E. cloacae 168

and K. pneumoniae: 169

170

The ciprofloxacin EUCAST ECOFFs for S. aureus (20 mm) and for CoNS (24 mm) are 171

consistent with both the EUCAST and IMM diameter distributions (Figure 1A and 1B) 172

(13). Likewise, the EUCAST levofloxacin ECOFF of 24 mm for CoNS is consistent with 173

the IMM diameter distribution (Figure 1D). EUCAST does not define an S. aureus 174

levofloxacin ECOFF, most likely due to the low number of available data points (n=107). 175

Despite different S. aureus and CoNS ciprofloxacin ECOFFs, and a missing levofloxacin 176

ECOFF for S. aureus, both EUCAST and CLSI CBP tables define uniform staphylococcal 177

CBPs for fluoroquinolones, which do not differentiate S. aureus from non-S. aureus 178

staphylococci, e.g. uniform staphylococcal ciprofloxacin CBPs for susceptibility of ≥ 20 179

mm and resistance of < 20 mm (EUCAST), or ≥ 21 mm for susceptibility and resistance of 180

≤ 15 mm (CLSI), and uniform staphylococcal levofloxacin CBPs for susceptibility of 181

≥ 22 mm, and resistance of < 19 mm (EUCAST), or ≥ 19 mm for susceptibility and 182

resistance of ≤ 15 mm (CLSI) (3, 5). The IMM S. aureus ciprofloxacin ECOFF (20 mm, n 183

= 5492 isolates, Figure 1A) correlates to EUCAST and CLSI susceptible breakpoints. 184

However, both EUCAST and CLSI susceptible CBPs applied to CoNS split the “out of 185

ECOFF” population (usually considered as “non-wild-type”) to different interpretative 186

categories (Figure 1B). Similarly, the uniform staphylococcal EUCAST and CLSI 187

levofloxacin CBPs split a homogeneously distributed “out of ECOFF” (non-wild type) 188

CoNS population (see IMM distribution, n = 3832 isolates tested with levofloxacin, Figure 189

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

1D) between all three interpretative categories (susceptible, intermediate, and resistant). In 190

contrast, the susceptible EUCAST levofloxacin CBP (≥ 22 mm) and the susceptible CLSI 191

CBP (≥ 19 mm) are largely adequate for S. aureus assigning only the wild-type population 192

to the susceptible category (Figure 1C). We assume that the uniform EUCAST 193

staphylococcal ciprofloxacin and levofloxacin CBPs were extrapolated from S. aureus 194

data, a procedure apparently inappropriate for CoNS. Consequently, fluoroquinolone CBPs 195

for staphylococci must be either defined in a species-specific fashion (e.g. S. aureus), or at 196

the group-level (coagulase-negative staphylococci) as done by EUCAST and CLSI for 197

cefoxitin S. aureus, and Staphylococcus lugdunensis or, most recently, by EUCAST for 198

Staphylococcus saprophyticus and S. pseudointermedius (3, 5). 199

200

EUCAST does not define an ertapenem ECOFF for E. cloacae. Most probably, this is 201

due to a shoulder in the population caused by a resistant sub-population overlapping the 202

wild type, which becomes clearly visible in the E. cloacae/ertapenem IMM diameter 203

distribution as compared to available EUCAST data (n = 1089 isolates IMM, see 204

Figure 1 E, versus n = 109 isolates in the EUCAST distribution) (13). Despite lack of an 205

ECOFF for ertapenem and E. cloacae, EUCAST recommends uniform susceptible (≥ 25 206

mm) /resistant (< 22 mm) ertapenem CBPs for all Enterobacteriaceae species as does 207

CLSI (susceptible and resistant CBPs of ≥22 mm, and ≤18 mm, respectively (3, 5). 208

However, testing ertapenem susceptibility in E. cloacae appears to be unreliable in terms 209

of clinical categorisation: Wild-type and resistant populations cannot safely be 210

distinguished by inhibition zone diameter determination alone (Figure 1E). Ertapenem 211

AST for E. cloacae may, consequently, be discouraged, or isolates should be tested for the 212

presence of resistance mechanisms. 213

For K. pneumoniae EUCAST and IMM ertapenem ECOFFs are identical (25 mm, 214

Figure 1F). The ECOFF matches the EUCAST susceptible Enterobacteriaceae CBP of 215

≥ 25 mm, appropriately assigning wild-type isolates to the susceptible category (13). IMM 216

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

distributions show a minor non-wild type population, which overlaps the wild-type, 217

although the situation is less pronounced when compared with E. cloacae (compare Figure 218

1F with 1E). The minor K. pneumoniae non-wild-type population is categorised 219

intermediate by EUCAST Enterobacteriaceae CBPs, which seems reasonable. According 220

to the susceptible ertapenem CLSI CBP (≥22 mm) this minor non-wild-type populations is 221

categorized susceptible. Clinical and PK/PD data, and molecular characterization for 222

resistance mechanisms are mandatory to decide if such isolates constitute a distinct 223

population and can be reliably be treated. 224

For both E. cloacae and K. pneumoniae the overlapping non-wild type populations are, 225

most likely, caused by increased beta-lactamase production combined with an outer 226

membrane porin deficiency causing reduced permeability. This combination of resistance 227

mechanisms is commonly found in E. cloacae and K. pneumoniae, but rarely in E. coli 228

(14-16). 229

The above examples indicate that CBPs are preferably set species-specific to prevent 230

erroneous treatment recommendations. Combinations of resistance mechanisms can result 231

in complex population diameter distributions lacking clear discrimination of wild-type and 232

non-wild-type populations. This phenotypical indifference may warrant determination of 233

resistance mechanisms (biochemically or genotypically) rather than mere assessment of 234

MIC/diameter CBPs. The reduced discriminative power of CBPs becomes even more 235

limiting when one considers measurement inaccuracy that further hampers correct 236

assignment of isolates to interpretative clinical categories (17). The unavoidable increase in 237

complexity of species-specific CBPs may prompt the use of software systems for AST 238

categorization, e.g. zone diameter classification (18). 239

240

241

242

243

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

Pooling data from various sources may lower population discrimination 244

The EUCAST E. coli ertapenem ECOFF was set at 29 mm with a minor population 245

between 27 and 29 mm that was obviously considered non-wild-type (13). This decision 246

was made on the basis of 579 data points from 2 sources. The IMM distribution comprises 247

6974 ertapenem diameter values of clinically relevant, non-duplicate E. coli isolates. Our 248

data suggest an ECOFF of 25 mm to appropriately assign wild-type and non-wild-type 249

isolates (Figure 2A) as an ECOFF of 29 mm would split the Gaussian distribution of wild-250

type isolates. The S. aureus amikacin EUCAST ECOFF is 18 mm (13). However, the 251

EUCAST distribution (841 data points from 4 sources) seems to contain a resistant S. 252

aureus sub-population overlapping the wild type, which is split by the amikacin CBPs 253

(S/R= ≥ 18mm, < 16 mm) (13). This overlapping resistant population is missing in the 254

IMM distribution (5702 data points, Figure 2B). 255

Diameter and MIC values are influenced by the method used for their determination 256

(19). EUCAST distribution data represent a compilation from numerous sources (13). This 257

raises some concerns about data consistency and bias due to the various methods used and 258

the differing measurement precision in individual laboratories. Technical variation between 259

individual laboratories may decrease the discriminative power of diameter distributions 260

due to higher spread of distribution curves, and, therefore, could hamper clear 261

differentiation of wild-type and non-wild-type populations. On the other hand, analysing 262

data from multiple sources is required to limit bias due to local epidemiology and to make 263

AST data more robust. The zone diameters presented in this study have been determined 264

by an automated reading system (Sirscan). Reading zone diameters by automats is 265

officially approved by EUCAST and CLSI (5, 9). A weekly quality control is performed in 266

our laboratory. Using EUCAST recommended quality control isolates, median diameter 267

values, standard deviations, and diameter ranges closely matched EUCAST criteria (see 268

Supplemental Figure 1) indicating that a comparison of IMM and EUCAST data is 269

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

appropriate. In addition, we have recently shown that diameter readings by the Sirscan 270

zone reader can reliably be compared to the EUCAST recommended manual method (20). 271

Possible reasons for the discrepancies observed include different epidemiology, lower 272

number of data points in the EUCAST distribution, pooling results from various sources, 273

bias due to low number of laboratories contributing data (single source in our data, 2 and 4 274

sources for the EUCAST data), and differences in ECOFF determination. In our study we 275

used the “eyeball”-method for ECOFF determination, i.e. a visual consensus value (8). 276

EUCAST does not officially recommend an ECOFF method nor provides information on 277

how EUCAST ECOFFs were determined. The eyeball method is, however, frequently used 278

and can at least provide a reasonable ECOFF estimation. Next to the visual inspection of 279

data distribution several statistical methods have been described to determine ECOFFs in 280

order to deal with statistical problems such as skewed and/or outlier distributions, but no 281

single method has been commonly accepted as the gold standard to date (21-25). 282

The data used in this study are limited as they originate from a single source (the 283

clinical laboratory of the IMM Zurich) and, thus, represent a local epidemiology. While the 284

problems described for our setting may not be encountered by all clinical laboratories, the 285

CBP inconsistencies we observed are of a more general character. 286

287

Antibiotic disk content is critical to delineate distinct non-wild-type and wild-type 288

populations 289

For determination of high level aminoglycoside resistance in enterococci antibiotic 290

disks with high drug loads have long been recommended by CLSI and the Comité de 291

l'Antibiogramme de la Societe Francaise de Microbiologie (CA-SFM), i.e. a load of 292

gentamicin 100 µg/disk by CLSI, and 500 µg/disk by the CA-SFM (5, 26). EUCAST 293

recommends a lower disk content of 30 µg/disk in order to enhance the sensitivity of high 294

level resistance detection (3). This approach can cause practical problems in separating 295

resistant and wild-type populations as illustrated by E. faecalis and gentamicin: The IMM 296

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

diameter distribution for the 30 µg disk (Figure 2C) shows a wild type population that is 297

left-shifted close to the resistant population. Reliable assignment of isolates to the resistant 298

or wild-type population is hampered due to measurement inaccuracy as per the low 299

distance between the wild-type and non-wild-type populations. As a consequence, 300

interpretation errors will inevitably occur due to technical measurement inaccuracy (17). It 301

may be advisable to either increase the disk content, thereby increasing the distance / 302

separation between wild-type and high level resistant populations, or to introduce a grey, 303

or “buffer” zone surrounding the ECOFF. Figure 2 D shows distribution data for 304

E. faecalis and gentamicin 500 µg/disk according to the 2013 CA-SFM CBPs. Clearly, the 305

wild-type population (deemed susceptible) can be separated from high-level resistant 306

isolates. Wild-type and high-level resistant populations are separated by a transitional 307

population of uncertain assignment (deemed intermediate). Not the least, to assign such 308

transitional populations to the susceptible or resistant category guideline issuing 309

organisations, such as CLSI and EUCAST, increasingly advocate to use determination of 310

resistance mechanisms to validate the phenotypic interpretive criteria (27). 311

312

In contrast to the ECOFF definition, CBP setting cannot be based on microbiological 313

data alone, but needs to consider clinical and PK/PD data as mandatory sources of 314

information. However, microbiological data should not be weighed less than PK/PD data. 315

Preliminary CBPs should preferably be consistent with both systems and should ultimately 316

be validated by clinical outcome studies. In conclusion, our data illustrate some issues of 317

general importance in determining CBPs: i) Reliable CBPs/ECOFFs require species-318

specific criteria, ii) pooling data from different sources may influence diameter 319

distributions, and iii) antibiotic disk content is critical for separating resistant and wild-type 320

populations. 321

322

323

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

324

Acknowledgments 325

We are grateful to the laboratory technicians of the Institute of Medical Microbiology, 326

University of Zurich for their dedicated help, and to Reinhard Zbinden, Guido Bloemberg, 327

Vanessa Deggim, Florian Maurer, Vera Rüegger, and Andrea Zbinden for valuable 328

discussions. 329

330

Funding 331

This work was supported by the University of Zurich 332

333

Transparency declarations 334

There are no conflicts of interest to declare. 335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

351

References 352

353

1. Courvalin P. 2008. Can pharmacokinetic-pharmacodynamic parameters provide 354

dosing regimens that are less vulnerable to resistance? Clin Microbiol Infect 355

14:989-994. 356

2. European Committee on Antimicrobial Susceptibility Testing. Breakpoint 357

tables for interpretation of MICs and zone diameters. Version 1.0. 2009. 358

http://www.eucast.org/clinical_breakpoints/previous_versions_of_tables/ (18th 359

January 2014, date last accessed). 360

3. European Committee on Antimicrobial Susceptibility Testing. Breakpoint 361

tables for interpretation of MICs and zone diameters. Version 4.0. 2014. 362

http://www.eucast.org/clinical_breakpoints/ (18th January 2014, date last 363

accessed). 364

4. European Committee on Antimicrobial Susceptibility Testing. Rationale 365

documents. http://www.eucast.org/documents/rd/ (18th January 2014, date last 366

accessed). 367

5. Clinical and Laboratory Standards Institute. Performance Standards for 368

Antimicrobial Susceptibility Testing; Twenty-fourth Informational Supplement. 369

CLSI document M 100-S 24. CLSI, Wayne, PA, USA, 2014. 370

6. Hombach M, Bloemberg GV, Bottger EC. 2012. Effects of clinical breakpoint 371

changes in CLSI guidelines 2010/2011 and EUCAST guidelines 2011 on antibiotic 372

susceptibility test reporting of Gram-negative bacilli. J. Antimicrob. Chemother. 373

67:622-632. 374

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

7. Hombach M, Wolfensberger A, Kuster SP, Bottger EC. 2013. Influence of 375

Clinical Breakpoint Changes from CLSI 2009 to EUCAST 2011 AST Guidelines 376

on Multi Drug Resistance Rates of Gram-negative Rods. J. Clin. Microbiol. 377

51:2385-2387. 378

8. Turnidge J, Paterson DL. 2007. Setting and revising antibacterial susceptibility 379

breakpoints. Clin. Microbiol. Rev. 20:391-408. 380

9. European Committee on Antimicrobial Susceptibility Testing. 381

Disk diffusion manual v. 3.0. http://www.eucast.org/eucast_susceptibility_testing/ 382

disk_diffusion_methodology (18th January 2014, date last accessed). 383

10. Clinical and Laboratory Standards Institute. Development of In Vitro 384

Susceptibility Testing Criteria and Quality Control Parameters; Approved 385

Guideline Third Edition. CLSI document M23-A3. CLSI, Wayne, PA, USA, 2008. 386

11. European Committee on Antimicrobial Susceptibility Testing. Setting 387

breakpoints for new agents, EUCAST SOP 1.0, 2010. 388

http://www.eucast.org/documents/sops/ (18th January 2014, date last accessed). 389

12. Horng-Jinh C, Kuo-Chang H, Chao-Hsien W. 2006. Determination of sample 390

size in using central limit theorem for Weibull distribution. Information and 391

Management Sciences 17:31-46. 392

13. European Committee on Antimicrobial Susceptibility Testing. Zone diameter 393

distributions. http://www.eucast.org/zone_diameter_distributions/ (18th January 394

2014, date last accessed). 395

14. Garcia-Sureda L, Domenech-Sanchez A, Barbier M, Juan C, Gasco J, Alberti 396

S. 2011. OmpK26, a novel porin associated with carbapenem resistance in 397

Klebsiella pneumoniae. Antimicrob. Agents Chemother. 55:4742-4747. 398

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

15. Wang XD, Cai JC, Zhou HW, Zhang R, Chen GX. 2009. Reduced susceptibility 399

to carbapenems in Klebsiella pneumoniae clinical isolates associated with plasmid-400

mediated beta-lactamase production and OmpK36 porin deficiency. J. Med. 401

Microbiol. 58:1196-1202. 402

16. Woodford N, Dallow JW, Hill RL, Palepou MF, Pike R, Ward ME, Warner 403

M, Livermore DM. 2007. Ertapenem resistance among Klebsiella and 404

Enterobacter submitted in the UK to a reference laboratory. Int. J. Antimicrob. 405

Agents 29:456-459. 406

17. Hombach M, Bottger EC, Roos M. 2013. The critical influence of the 407

intermediate category on interpretation errors in revised EUCAST and CLSI 408

antimicrobial susceptibility testing guidelines. Clin. Microbiol. Infect. 19:E59-71. 409

18. Winstanley T, Courvalin P. 2011. Expert systems in clinical microbiology. Clin. 410

Microbiol. Rev. 24:515-556. 411

19. Koeth LM, King A, Knight H, May J, Miller LA, Phillips I, Poupard JA. 2000. 412

Comparison of cation-adjusted Mueller-Hinton broth with Iso-Sensitest broth for 413

the NCCLS broth microdilution method. J. Antimicrob. Chemother. 46:369-376. 414

20. Hombach M, Zbinden R, Bottger EC. 2013. Standardisation of disk diffusion 415

results for antibiotic susceptibility testing using the sirscan automated zone reader. 416

BMC Microbiol. 13:225. 417

21. Canton E, Peman J, Hervas D, Iniguez C, Navarro D, Echeverria J, Martinez-418

Alarcon J, Fontanals D, Gomila-Sard B, Buendia B, Torroba L, Ayats J, 419

Bratos A, Sanchez-Reus F, Fernandez-Natal I, Group FS. 2012. Comparison of 420

three statistical methods for establishing tentative wild-type population and 421

epidemiological cutoff values for echinocandins, amphotericin B, flucytosine, and 422

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

six Candida species as determined by the colorimetric Sensititre YeastOne method. 423

J. Clin. Microbiol. 50:3921-3926. 424

22. Kronvall G, Kahlmeter G, Myhre E, Galas MF. 2003. A new method for 425

normalized interpretation of antimicrobial resistance from disk test results for 426

comparative purposes. Clin. Microbiol. Infect. 9:120-132. 427

23. Meletiadis J, Mavridou E, Melchers WJ, Mouton JW, Verweij PE. 2012. 428

Epidemiological cutoff values for azoles and Aspergillus fumigatus based on a 429

novel mathematical approach incorporating cyp51A sequence analysis. Antimicrob. 430

Agents Chemother. 56:2524-2529. 431

24. Pfaller MA, Boyken L, Hollis RJ, Kroeger J, Messer SA, Tendolkar S, 432

Diekema DJ. 2011. Wild-type MIC distributions and epidemiological cutoff values 433

for posaconazole and voriconazole and Candida spp. as determined by 24-hour 434

CLSI broth microdilution. J. Clin. Microbiol. 49:630-637. 435

25. Turnidge J, Kahlmeter G, Kronvall G. 2006. Statistical characterisation of 436

bacterial wild-type MIC value distributions and the determination of 437

epidemiological cut-off values. Clin. Microbiol. Infect. 12:418-425. 438

26. Comité de l'Antibiogramme de la Societe Francaise de Microbiologie. 439

Recommandations 2013. http://www.sfm-microbiologie.org/ (18th January 2014, 440

date last accessed). 441

27. European Committee on Antimicrobial Susceptibility Testing. EUCAST 442

guidelines for detection of resistance mechanisms 443

and specific resistances of clinical and/or epidemiological importance. 444

http://www.eucast.org/resistance_mechanisms/ (18th January 2014 last accessed). 445

446

447

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

Figures

Figure 1: Species specific breakpoints and overlapping populations

Inhibition zone diameter distributions of isolates tested in the clinical laboratory of the Institute of

Medical Microbiology, University of Zurich. Values were determined by the EUCAST recommended

disk diffusion procedure. Categorization according to EUCAST clinical breakpoints: White bars,

susceptible; grey bars, intermediate; black bars, resistant. Black arrow, EUCAST ECOFF as listed in

EUCAST diameter distributions (13); white arrow, IMM ECOFF (present study). Thin arrows, CLSI

clinical breakpoints. Dotted lines, predicted population distributions. CoNS, coagulase-negative

staphylococci.

A

330 899

C

D

E F

B

SR

S R

SR

SR

S RS R

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from

Figure 2: Diameter distributions, data pooling from various sources, and importance of antibiotic

disk content

Diameter distributions originate from isolates tested in the clinical laboratory of the Institute of

Medical Microbiology, University of Zurich. Values were determined by the EUCAST recommended

disk diffusion procedure. White bars, EUCAST susceptible category; grey bars, EUCAST intermediate

category; black bars, EUCAST resistant category. Black arrows, EUCAST ECOFFs as listed in

EUCAST diameter distributions (13); white arrows, IMM ECOFFs (present study). The 500 µg

gentamicin disk content is recommended by CA-SFM with according CBPs as shown. CoNS,

coagulase-negative staphylococci.

on April 25, 2018 by guest

http://aac.asm.org/

Dow

nloaded from