Antimicrobial Susceptibility Testing: A Review of General Practices PDF

INVITED ARTICLE MEDICAL MICROBIOLOGY L. Barth Reller and Melvin P. Weinstein, Section Editors Antimicrobial Susceptibility Tes...

INVITED ARTICLE MEDICAL MICROBIOLOGY L. Barth Reller and Melvin P. Weinstein, Section Editors Antimicrobial Susceptibility Testing: A Review of General Principles and Contemporary Practices James H. Jorgensen1 and Mary Jane Ferraro2,3 1 Department of Pathology, The University of Texas Health Science Center, San Antonio; and Departments of 2Pathology and 3Medicine, Massachusetts General Hospital and Harvard Medical School, Boston An important task of the clinical microbiology laboratory is the performance of antimicrobial susceptibility testing of significant bacterial isolates. The goals of testing are to detect possible drug resistance in common pathogens and to assure susceptibility to drugs of choice for particular infections. The most widely used testing methods include broth microdilution or rapid automated instrument methods that use commercially marketed materials and devices. Manual methods that provide flexibility and possible cost savings include the disk diffusion and gradient diffusion methods. Each method has strengths and weaknesses, including organisms that may be accurately tested by the method. Some methods provide quantitative results (eg, minimum inhibitory concentration), and all provide qualitative assessments using the categories susceptible, intermediate, or resistant. In general, current testing methods provide accurate detection of common antimicrobial resistance mechanisms. However, newer or emerging mechanisms of resistance require constant vigilance regarding the ability of each test method to accurately detect resistance. EMERGENCE OF ANTIMICROBIAL RESISTANCE OVERVIEW OF COMMONLY USED AND THE RATIONALE FOR PERFORMING SUSCEPTIBILITY TESTING METHODS SUSCEPTIBILITY TESTING Broth dilution tests. One of the earliest antimicrobial sus- The performance of antimicrobial susceptibility testing by the ceptibility testing methods was the macrobroth or tube-dilution clinical microbiology laboratory is important to confirm sus- method. This procedure involved preparing two-fold di- ceptibility to chosen empirical antimicrobial agents, or to detect lutions of antibiotics (eg, 1, 2, 4, 8, and 16 mg/mL) in a liquid resistance in individual bacterial isolates. Empirical therapy growth medium dispensed in test tubes [1, 2]. The antibiotic- continues to be effective for some bacterial pathogens because containing tubes were inoculated with a standardized bacterial resistance mechanisms have not been observed e.g., continued suspension of 1–5 ⫻ 10 5 CFU/mL. Following overnight incu- penicillin susceptibility of Streptococcus pyogenes. Susceptibility bation at 35C, the tubes were examined for visible bacterial testing of individual isolates is important with species that may growth as evidenced by turbidity. The lowest concentration of possess acquired resistance mechanisms (eg, members of the antibiotic that prevented growth represented the minimal in- Enterobacteriaceae, Pseudomonas species, Staphylococcus spe- hibitory concentration (MIC). The precision of this method cies, Enterococcus species, and Streptococcus pneumoniae). was considered to be plus or minus 1 two-fold concentration, due in large part to the practice of manually preparing serial dilutions of the antibiotics. The advantage of this technique was the generation of a quantitative result (ie, the MIC). The principal disadvantages of the macrodilution method were the Received 8 May 2009; accepted 27 July 2009; electronically published 26 October 2009. tedious, manual task of preparing the antibiotic solutions for Reprints or correspondence: Dr James H. Jorgensen, Dept of Pathology, University of Texas Health Science Center, 7703 Floyd Curl Dr, San Antonio, TX 78229-7750 (jorgensen@ each test, the possibility of errors in preparation of the antibiotic uthscsa.edu). solutions, and the relatively large amount of reagents and space Clinical Infectious Diseases 2009; 49:1749–55 required for each test.  2009 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2009/4911-0022$15.00 The miniaturization and mechanization of the test by use of DOI: 10.1086/647952 small, disposable, plastic “microdilution” trays (Figure 1) has MEDICAL MICROBIOLOGY CID 2009:49 (1 December) 1749 Downloaded from https://academic.oup.com/cid/article-abstract/49/11/1749/344384/Antimicrobial-Susceptibility-Testing-A-Review-of by guest on 20 September 2017  Figure 1. A broth microdilution susceptibility panel containing 98 reagent wells and a disposable tray inoculator made broth dilution testing practical and popular. Standard ent diffusion method uses the principle of establishment of an trays contain 96 wells, each containing a volume of 0.1 mL that antimicrobial concentration gradient in an agar medium as a allows approximately 12 antibiotics to be tested in a range of means of determining susceptibility. The Etest (bioMérieux AB 8 two-fold dilutions in a single tray [2, 4]. Microdilution panels BIODISK) (Figure 2) is a commercial version available in the are typically prepared using dispensing instruments that aliquot United States. It employs thin plastic test strips that are im- precise volumes of preweighed and diluted antibiotics in broth pregnated on the underside with a dried antibiotic concentra- into the individual wells of trays from large volume vessels. tion gradient and are marked on the upper surface with a Hundreds of identical trays can be prepared from a single mas- concentration scale. As many as 5 or 6 strips may be placed in ter set of dilutions in a relatively brief period. Few clinical a radial fashion on the surface of an appropriate 150-mm agar microbiology laboratories prepare their own panels; instead fro- plate that has been inoculated with a standardized organism zen or dried microdilution panels are purchased from one of suspension like that used for a disk diffusion test. After over- several commercial suppliers. The cost of the preprepared pan- night incubation, the tests are read by viewing the strips from els range from approximately $10 to $22 each. Inoculation of the top of the plate. The MIC is determined by the intersection panels with the standard 5 ⫻ 10 5 CFU/mL is accomplished us- of the lower part of the ellipse shaped growth inhibition area ing a disposable device that transfers 0.01 to 0.05 mL of stan- with the test strip. dardized bacterial suspension into each well of the microdi- The gradient diffusion method has intrinsic flexibility by lution tray or by use of a mechanized dispenser. Following being able to test the drugs the laboratory chooses. Etest strips incubation, MICs are determined using a manual or automated cost approximately $2–$3 each and can represent an expensive viewing device for inspection of each of the panel wells for approach if more than a few drugs are tested. This method is growth. best suited to situations in which an MIC for only 1 or 2 drugs The advantages of the microdilution procedure include the is needed or when a fastidious organism requiring enriched generation of MICs, the reproducibility and convenience of medium or special incubation atmosphere is to be tested (eg, having preprepared panels, and the economy of reagents and penicillin and ceftriaxone with pneumococci) [5–7]. Generally, space that occurs due to the miniaturization of the test. There Etest results have correlated well with MICs generated by broth is also assistance in generating computerized reports if an au- or agar dilution methods [5–9]. However, there are some sys- tomated panel reader is used. The main disadvantage of the tematic biases toward higher or lower MICs determined by the microdilution method is some inflexibility of drug selections Etest when testing certain organism-antimicrobial agent com- available in standard commercial panels. binations [6, 10]. This can represent a potential shortcoming Antimicrobial gradient method. The antimicrobial gradi- when standard MIC interpretive criteria derived from broth 1750 CID 2009:49 (1 December) MEDICAL MICROBIOLOGY Downloaded from https://academic.oup.com/cid/article-abstract/49/11/1749/344384/Antimicrobial-Susceptibility-Testing-A-Review-of by guest on 20 September 2017  Figure 2. A Staphylococcus aureus isolate tested by the Etest gradient diffusion method with vancomycin (VA), daptomycin (DM), and linezolid (LZ) on Mueller-Hinton agar. The minimum inhibitory concentration of each agent is determined by the intersection of the organism growth with the strip as measured using the scale inscribed on the strip. dilution testing are applied to Etest MICs that may not The advantages of the disk method are the test simplicity be identical. that does not require any special equipment, the provision of Disk diffusion test. The disk diffusion susceptibility categorical results easily interpreted by all clinicians, and flex- method [2, 11, 12] is simple and practical and has been well- ibility in selection of disks for testing. It is the least costly of standardized. The test is performed by applying a bacterial all susceptibility methods (approximately $2.50–$5 per test for inoculum of approximately 1–2 ⫻ 10 8 CFU/mL to the surface materials). The disadvantages of the disk test are the lack of of a large (150 mm diameter) Mueller-Hinton agar plate. Up mechanization or automation of the test. Although not all fas- to 12 commercially-prepared, fixed concentration, paper an- tidious or slow growing bacteria can be accurately tested by tibiotic disks are placed on the inoculated agar surface (Figure this method, the disk test has been standardized for testing 3). Plates are incubated for 16–24 h at 35C prior to deter- streptococci, Haemophilus influenzae, and N. meningitidis mination of results. The zones of growth inhibition around through use of specialized media, incubation conditions, and each of the antibiotic disks are measured to the nearest mil- specific zone size interpretive criteria. limeter. The diameter of the zone is related to the susceptibility Automated instrument systems. Use of instrumentation of the isolate and to the diffusion rate of the drug through the can standardize the reading of end points and often produce agar medium. The zone diameters of each drug are interpreted susceptibility test results in a shorter period than manual read- using the criteria published by the Clinical and Laboratory ings because sensitive optical detection systems allow detection Standards Institute (CLSI, formerly the National Committee of subtle changes in bacterial growth. There are 4 automated for Clinical Laboratory Standards or NCCLS) or those instruments presently cleared by the FDA for use in the United included in the US Food and Drug Administration (FDA)– States. Three of these can generate rapid (3.5–16 h) suscepti- approved product inserts for the disks. The results of the disk bility test results, while the fourth is an overnight system. diffusion test are “qualitative,” in that a category of suscepti- The MicroScan WalkAway (Siemens Healthcare Diagnostics) is bility (ie, susceptible, intermediate, or resistant) is derived from a large self-contained incubator/reader device that can incubate the test rather than an MIC. However, some commercially- and analyze 40–96 microdilution trays. The WalkAway utilizes available zone reader systems claim to calculate an approximate standard size microdilution trays that are hydrated and inoc- MIC with some organisms and antibiotics by comparing zone ulated manually and then placed in one of the incubator slots sizes with standard curves of that species and drug stored in in the instrument. The instrument incubates the trays for the an algorithm [14, 15]. appropriate period, examining them periodically with either a MEDICAL MICROBIOLOGY CID 2009:49 (1 December) 1751 Downloaded from https://academic.oup.com/cid/article-abstract/49/11/1749/344384/Antimicrobial-Susceptibility-Testing-A-Review-of by guest on 20 September 2017  Figure 3. A disk diffusion test with an isolate of Escherichia coli from a urine culture. The diameters of all zones of inhibition are measured and those values translated to categories of susceptible, intermediate, or resistant using the latest tables published by the CLSI. photometer or fluorometer to determine growth development. lution plates that can be inoculated with a Sensititre Gram-negative susceptibility test panels containing fluorogenic Autoinculator. Growth is determined by fluorescence mea- substrates can be read in 3.5–7 h. Separate gram-positive and surement after 18–24 h of incubation. Test panels are available gram-negative panels read using turbidimetric end points are for gram-positive and gram-negative bacteria, S. pneumoniae, ready in 4.5–18 hours. Haemophilus species, and nonfermentative gram-negative The BD Phoenix Automated Microbiology System (BD Di- bacilli. agnostics) has a large incubator reader with a capacity to process The Phoenix, Sensititre ARIS 2X, Vitek 1 and 2, and 99 test panels that contain 84 wells devoted to antibiotic dou- WalkAway instruments have enhanced computer software used bling dilutions and are inoculated manually. The Phoenix mon- to interpret susceptibility results including “expert systems” for itors each panel every 20 min using both turbidometric and analyzing test results for atypical patterns and unusual resis- colorimetric (oxidation-reduction indicator) growth detection. tance phenotypes. Two studies [17, 18] have shown that Test panels for gram-negative, gram-positive, S. pneumoniae, providing rapid susceptibility test results can lead to more b-hemolytic, and viridans group streptococci are available. MIC timely changes to appropriate antimicrobial therapy, substantial results are generated in 6–16 h. direct cost savings attributable to ordering of fewer additional The Vitek 2 System (bioMérieux) is highly automated and laboratory tests, performance of fewer invasive procedures, and uses very compact plastic reagent cards (credit card size) that a shortened length of stay. These benefits are best realized when contain microliter quantities of antibiotics and test media in a coupled with extended laboratory staffing schedules, and real- 64-well format. The Vitek 2 employs repetitive turbidimetric time, electronic transmission of verified results. One of the early monitoring of bacterial growth during an abbreviated incu- bation period. The instrument can be configured to accom- shortcomings of rapid susceptibility testing methods was a less- modate 30–240 simultaneous tests. The susceptibility cards al- ened ability to detect some types of antimicrobial resistance low testing of common, rapidly growing gram-positive, and including inducible b-lactamases and vancomycin resistance. gram-negative aerobic bacteria, and S. pneumoniae in a period However, the recently FDA-cleared instruments have made sig- of 4–10 h. An older, less automated, Vitek 1 System is still used nificant improvements in large part through modifications of in some laboratories. The system is more limited with a 45- the instruments’ computer software to either provide extended well card and does not include S. pneumoniae. incubation for problematic organism-drug combinations, or by The Sensititre ARIS 2X (Trek Diagnostic Systems) is an au- editing of susceptibility results using expert software to prevent tomated, overnight, incubation and reading system with a 64- unlikely results from being reported. In some cases these mod- panel capacity. The test panels are standard 96-well microdi- ifications result in prolonged incubation (ie, 110 h) of test 1752 CID 2009:49 (1 December) MEDICAL MICROBIOLOGY Downloaded from https://academic.oup.com/cid/article-abstract/49/11/1749/344384/Antimicrobial-Susceptibility-Testing-A-Review-of by guest on 20 September 2017  panels to assure accurate results, thus rendering them less ies results (including comparisons of MIC and zone diameter “rapid.” with microbiological eradication and clinical efficacy) obtained during studies prior to FDA approval and marketing of an SELECTION OF DRUGS FOR ROUTINE TESTING antibiotic. The laboratory must test and report the antimicrobial agents A “susceptible” result indicates that the patient’s organism that are most appropriate for the organism isolated, for the site should respond to therapy with that antibiotic using the dosage of the infection, and the institution’s formulary [13, 19]. The recommended normally for that type of infection and species CLSI provides tables that list the antimicrobial agents appro- [13, 20]. Conversely, an organism with a MIC or zone size priate for testing members of the Enterobacteriaceae, Pseudo- interpreted as “resistant” should not be inhibited by the con- monas, and other gram-negative glucose nonfermenters, staph- centrations of the antibiotic achieved with the dosages normally ylococci, enterococci, streptococci, Haemophilus species, etc. used with that drug [13, 20]. An “intermediate” result indicates. The listings include recommendations for agents that are that a microorganism falls into a range of susceptibility in which important to test routinely, and those that may be tested or the MIC approaches or exceeds the level of antibiotic that can reported selectively based on the institution’s formulary. ordinarily be achieved and for which clinical response is likely The availability of antimicrobial agents for testing by the to be less than with a susceptible strain. Exceptions can occur laboratory’s routine testing methodology must next be deter- if the antibiotic is highly concentrated in a body fluid such as mined. The disk diffusion and gradient diffusion procedures urine, or if higher than normal dosages of the antibiotic can offer the greatest flexibility including testing of newly available be safely administered (eg, some penicillins and cephalospo- drugs. Most broth microdilution or automated test panels con- rins). At times, the “intermediate” result can also mean that tain ⭐96 wells, effectively limiting the number of agents tested certain variables in the susceptibility test may not have been or the range of dilutions of each drug that can be included. properly controlled, and that the values have fallen into a Manufacturers of commercially prepared panels have attempted “buffer zone” separating susceptible from resistant strains [13, to deal with this problem by offering a number of different 20]. Generally, reporting of a category result of susceptible, standard panel configurations, or by including fewer dilutions intermediate, or resistant provides the clinician with the in- of each drug in a single panel. Another solution to this formation necessary to select appropriate therapy. Reporting problem is testing antimicrobial agents that have activities that of MICs could aid a physician is selecting from among a group are essentially the same as the desired formulary drugs. The of similar drugs for therapy of infective endocarditis or oste- CLSI susceptibility testing document lists groups of some omyelitis, in which therapy is likely to be protracted. antimicrobial agents with nearly identical activities that can It is important that the tables used for susceptibility test provide practical alternatives for testing. interpretations represent the most current criteria. Indeed, the CLSI documents are reviewed and updated frequently, usually INTERPRETATION OF SUSCEPTIBILITY TEST once per year. Use of old or outdated information from the RESULTS original editions of FDA-approved drug labels or older CLSI tables could represent a serious shortcoming in the reporting The results of a susceptibility test must be interpreted by the of patients’ results. laboratory prior to communicating a report to a patient’s phy- sician. Optimal interpretation of MICs requires knowledge of WHAT IS THE ACCEPTABLE ACCURACY OF A the pharmacokinetics of the drug in humans, and information SUSCEPTIBILITY TEST METHOD? on the likely success of a particular drug in eradicating bacteria at various body sites. This is best accomplished by referring When assessing the accuracy of various susceptibility testing to an expert source such as the CLSI, which publishes inter- methods as compared to standard reference methods, the terms pretive criteria for MICs of all relevant antibiotics for most very major and major errors have been used to describe false- bacterial genera. Indeed, both MIC values and disk dif- susceptible or false-resistant results, respectively. In evaluations fusion zone diameters must be interpreted using a table of of new susceptibility testing methods it is important to examine values that relate to proven clinical efficacy of each antibiotic a representative number of strains that are resistant to various and for various bacterial species. The CLSI zone size and drugs to verify the ability of the new test to detect resistance MIC interpretive criteria are established by analysis of 3 kinds and to test a number of susceptible strains to determine the of data: (1.) microbiologic data, including a comparison of rate of major errors that might be expected in a typical clinical MICs and zone sizes on a large number of bacterial strains, laboratory setting [16, 21]. To be cleared for marketing in the including those with known mechanisms of resistance that have United States, the FDA requires that very major errors attrib- been defined either phenotypically or genotypically; (2) phar- utable to a test device should be !1.5% for individual species/ macokinetic and pharmacodynamic data; and (3) clinical stud- drug comparisons, major errors should not exceed 3%, and an MEDICAL MICROBIOLOGY CID 2009:49 (1 December) 1753 Downloaded from https://academic.oup.com/cid/article-abstract/49/11/1749/344384/Antimicrobial-Susceptibility-Testing-A-Review-of by guest on 20 September 2017  overall essential MIC agreement of 190% of device MICs within and disk diffusion methods. In: Murray PR, Baron EJ, Jorgensen JH, Landry ML, Pfaller MA, eds. Manual of clinical microbiology. 9th ed. one doubling dilution of a CLSI reference MIC. A recent, Washington, DC: American Society for Microbiology, 2007:1152–72. international standard on susceptibility test device evaluation 3. Balows A. Current techniques for antibiotic susceptibility testing. proposes similar but not identical criteria for acceptable ac- Springfield, IL: Charles C. Thomas, 1972. 4. Clinical and Laboratory Standards Institute. Methods for dilution an- curacy. The emergence of new antimicrobial resistance timicrobial susceptibility testing for bacteria that grew aerobically. Ap- mechanisms, including some that may be difficult to detect (eg, proved Standard M7–A10. Wayne, PA: Clinical and Laboratory Stan- vancomycin intermediate susceptibility in S. aureus and car- dards Institute, 2009. bapenemase production in some gram-negative organisms) re- 5. Huang MB, Baker CN, Banerjee S, Tenover FC. Accuracy of the E test for determining antimicrobial susceptibilities of staphylococci, enter- quires that the performance of susceptibility devices be con- ococci, Campylobacter jejuni, and gram-negative bacteria resistant to stantly reassessed and updated when needed. In some cases, it antimicrobial agents. J Clin Microbiol 1992; 30:3243–8. has been necessary to employ special ancillary testing methods 6. Jorgensen JH, Ferraro MJ, McElmeel ML, Spargo J, Swenson JM, Ten- over FC. Detection of penicillin and extended-spectrum cephalosporin (eg, single concentration screening agars, modified Hodge test resistance among Streptococcus pneumoniae clinical isolates by use of for carbapenemase production) to supplement routine the E test. J Clin Microbiol 1994; 32:159–63. testing by a commercial instrument system. 7. Citron, DM, Ostovari MI, Karlsson A, Goldstein EJC. Evaluation of the E test for susceptibility testing of anaerobic bacteria. J Clin Mi- crobiol 1991; 29:2197–203. CURRENT TEST METHODS AND FUTURE 8. Baker CN, Stocker SA, Culver DM, Thornsberry C. Comparison of DIRECTIONS the E-test to agar dilution, broth microdilution, and agar diffusion susceptibility testing techniques by using a special challenge set of The antimicrobial susceptibility testing methods described in bacteria. J Clin Microbiol 1991; 29:533–8. this article provide reliable results when used according to the 9. Rennie R, Turnbull L, Brosnikoff C. Comparison of Oxoid M.I.C. Evaluator device with broth microdilution and E test device from AB procedures defined by the CLSI or by the manufacturers of the Biodisk for antimicrobial susceptibility testing of Enterobacteriaceae commercial products. However, there is considerable oppor- [abstract P859]. In: Program and abstracts of the 18th Annual Meeting tunity for improvement in the area of rapid and accurate rec- of the European Congress on Clinical Microbiololgy and Infectious ognition of bacterial resistance to antibiotics. There is a need Diseases (Barcelona). European Congress on Clinical Microbiololgy and Infectious Diseases, 2008. for development of new automated instruments that could pro- 10. Prakash V, Lewis JS II, Jorgensen JH. Vancomycin MICs with methi- vide faster results and also save money by virtue of lower re- cillin-resistant Staphylococcus aureus (MRSA) isolates differ based upon agent costs and reduced labor requirements. To accomplish this, the susceptibility test method used. Antimicrob Agents Chemother 2008; 52:4528. it will likely be necessary to explore different methodologic 11. Bauer AW, Kirby WMM, Sherris JC, Turk M. Antibiotic susceptibility approaches for detection of bacterial growth. The direct de- testing by a standardized single disk method. Am J Clin Pathol 1966;45: tection of resistance genes by polymerase chain reaction or 493–6. 12. Clinical and Laboratory Standards Institute. Performance standards for similar techniques has limited utility, because only a few resis- antimicrobial disk susceptibility tests. Approved standard M2–A10. tance genes are firmly associated with phenotypic resistance Wayne, PA: Clinical and Laboratory Standards Institute, 2009. (eg, mecA, vanA, and vanB). There are hundreds of b- 13. Clinical and Laboratory Standards Institute. Performance standards for lactamases, and numerous mutations, acquisitions, and ex- antimicrobial susceptibility testing. Nineteenth informational supple- ment M100–S19. Wayne, PA: Clinical and Laboratory Standards In- pression mechanisms that result in fluoroquinolone, amino- stitute, 2009. glycoside, and macrolide resistance ; too many to be easily 14. Korgenski EK, Daly JA. Evaluation of the BIOMIC video reader for detected by current molecular techniques. Thus, it seems likely determining interpretive categories of isolates on the basis of disk dif- fusion susceptibility results. J Clin Microbiol 1998; 36:302–4. that phenotypic measures of the level of susceptibility of bac- 15. Nijs A, Cartuyvels R, Mewis A, Peeters V, Rummens JL, Magerman K. terial isolates to antimicrobial agents will continue to be clin- Comparison and evaluation of Osiris and Sirscan 2000 antimicrobial ically relevant for years to come. susceptibility systems in the clinical microbiology laboratory. J Clin Microbiol 2003; 41:3627–30. Acknowledgments 16. Richter SS, Ferraro MJ. Susceptibility testing instrumentation and com- puterized expert systems for data analysis and interpretation. In: Mur- Potential conflicts of interest. J.H.J. and M.J.F. disclose their mem- ray PR, Baron EJ, Jorgensen JH, Landry ML, Pfaller MA, eds. Manual bership on microbiology advisory committees for BD Diagnostics and of clinical microbiology. 9th ed. Washington, DC: American Society bioMérieux. J.H.J. has advised Accelr8 Technology and has received research for Microbiology, 2007:245–56. support from BD Diagnostics, bioMérieux, Merck, Pfizer, and Siemens 17. Doern GV, Vautour R, Gaudet M, Levy B. Clinical impact of rapid in Healthcare. vitro susceptibility testing and bacterial identification. J Clin Microbiol 1994; 32:1757–62. References 18. Barenfanger J, Drake C, Kacich G. Clinical and financial benefits of rapid bacterial identification and antimicrobial susceptibility testing. J 1. Ericsson JM, Sherris JC. Antibiotic sensitivity testing: report of an Clin Microbiol 1999; 37:1415–8. international collaborative study. Acta Pathol Microbiol Scand 1971; 19. Jorgensen JH. Selection of antimicrobial agents for routine testing in 217 (Suppl):1–90. a clinical microbiology laboratory. Diagn Microbiol Infect Dis 1993; 2. Jorgensen JH, Turnidge JD. Antibacterial susceptibility tests: dilution 16:245–9. 1754 CID 2009:49 (1 December) MEDICAL MICROBIOLOGY Downloaded from https://academic.oup.com/cid/article-abstract/49/11/1749/344384/Antimicrobial-Susceptibility-Testing-A-Review-of by guest on 20 September 2017  20. Clinical and Laboratory Standards Institute. Development of in vitro testing and in vitro diagnostic test systems—susceptibility testing of susceptibility testing criteria and quality control parameters; approved infectious agents and evaluation of performance of antimicrobial sus- guideline. Third ed. CLSI document M23–A3. Wayne, PA: Clinical and ceptibility test devices. Part 2: evaluation of performance of antimi- Laboratory Standards Institute, 2008. crobial susceptibility test devices. ISO 20776–2. Geneva: International 21. Jorgensen JH. Selection criteria for an antimicrobial susceptibility test- Organization for Standardization, 2007. ing system. J Clin Microbiol 1993; 31:2841–4. 24. Tenover FC. DNA probes for antimicrobial susceptibility testing. Clin 22. US Food and Drug and Administration. Class II special controls guid- Lab Med 1989; 9:341–7. ance document: antimicrobial susceptibility test (AST) systems; guid- 25. Rice LB, Bonomo RA. Mechanisms of resistance to antibacterial agents. ance for industry and FDA. Rockville, MD: US Food and Drug Ad- In: Murray PR, Baron EJ, Jorgensen JH, Landry ML, Pfaller MA, eds. ministration, 2003. Manual of clinical microbiology. 9th ed. Washington, DC: American 23. International Organization for Standardization. Clinical laboratory Society for Microbiology, 2007:1114–45. MEDICAL MICROBIOLOGY CID 2009:49 (1 December) 1755 Downloaded from https://academic.oup.com/cid/article-abstract/49/11/1749/344384/Antimicrobial-Susceptibility-Testing-A-Review-of by guest on 20 September 2017