Laboratory- and species-specific interpretive breakpoints for disk diffusion tests of chloramphenicol susceptibility of Haemophilus influenzae

Setting interpretive breakpoints for antimicrobial susceptibility testing using disk diffusion

Antimicrobial susceptibility testing plays a key role in clinical microbiology. The disk diffusion test dates back to the 1940s and became standardised from the 1950s, with the International Collaborative Study (ICS) and National Committee for Clinical Laboratory Standards (NCCLS) as the two major standards. Interlaboratory variation of disk test results was recognised early but has never been dealt with in a satisfactory manner. The error-rate bounded method was described in 1974 and its role is discussed. Species-specific susceptibility interpretation was coined in 1980 for Proteus mirabilis and chloramphenicol. In the late 1970s, more extensive use of species-specific breakpoints was introduced in Lund (Sweden). At the same time, P. Mouton constructed species-specific regression lines and pointed out the difficulties with narrow ranges of minimal inhibitory concentration (MIC) values. A more general use of species-specific regression lines was made possible with single-strain regression analysis, using one well-defined strain tested in disk diffusion with a range of disk contents. This method made it possible to calibrate the disk test in an individual laboratory. Other methods to achieve such calibration are also described. A recent method, 'MIC-coloured zone diameter histogram-technique', has proven useful for the validation of species-specific interpretive breakpoints. The microbiological breakpoint proposed by Williams in 1990 has experienced a renaissance with the European Committee on Antimicrobial Susceptibility Testing (EUCAST) epidemiological cut-off value (ECOFF). MIC and zone diameter distributions with accompanying ECOFFs for species-antimicrobial combinations are published on the EUCAST website. A method for the reconstruction of wild-type zone diameter populations, namely normalised resistance interpretation, is described. This method can produce resistance figures that are truly comparable between laboratories.

Normalized resistance interpretation as a tool for establishing epidemiological MIC susceptibility breakpoints

Normalized resistance interpretation (NRI) utilizes the fact that the wild-type population on the sensitive side is not affected by resistance development, and therefore a normalized reconstruction of the peak can be performed. The method was modified for MIC distributions by the introduction of helper variables, in-between values assigned the mean of the neighboring numbers of isolates. This method was used on Staphylo- coccus aureus and Escherichia coli MIC distributions for 27 antimicrobials each and obtained from the EUCAST (European Committee on Antimicrobial Susceptibility Testing) website (http://www.eucast.org/mic_distributions/). The number of isolates in each of the 54 distributions ranged from 40 to 124,472. NRI produced normalized distributions in all cases. Cutoff values were calculated for +2.0 and +2.5 standard deviations (SD) above the means and then rounded up to nearest regular MIC dilution step. EUCAST also show cutoff values, ECOFF values, which were used as the reference. The NRI generated +2.0 SD values showed the best agreement with 26 of 27 within ±1 dilution step and 17 exactly on the ECOFF values for Staphylococcus aureus, and 25 of 27 within ±1 dilution step and 14 right on the ECOFF values for Escherichia coli. NRI offers an objective method for the reconstruction of the wild-type population in an MIC distribution for a given bacterial species and an antimicrobial agent. This method offers a new tool in comparative susceptibility studies such as global surveillance of resistance, as well as in quality control in individual laboratories.

A new method for normalized interpretation of antimicrobial resistance from disk test results for comparative purposes

OBJECTIVE

To evaluate a calibration method for disk diffusion antibiotic susceptibility tests, using zone diameter values generated in the individual laboratory as the internal calibrator for combinations of antibiotic and bacterial species.

METHODS

The high-zone side of zone histogram distributions was first analyzed by moving averages to determine the peak position of the susceptible population. The accumulated percentages of isolates for the high zone diameter values were calculated and converted into probit values. The normal distribution of the ideal population of susceptible strains was then determined by using the least-squares method for probit values against zone diameters, and the ideal population was thereby defined, including mean and standard deviation. Zone diameter values were obtained from laboratories at the Karolinska Hospital (KS) and Växjö Hospital (VX), and from two laboratories (LabA, LabB) in Argentina. The method relies on well standardized disk tests, but is independent of differences in MIC limits and zone breakpoints, and does not require the use of reference strains. Resistance was tentatively set at below 3 SD from the calculated, ideal mean zone diameter of the susceptible population.

RESULTS

The method, called normalized interpretation of antimicrobial resistance, was tested on results from the KS and VX clinical microbiology laboratories, using the disk diffusion method for antimicrobial susceptibility tests, and for two bacterial species, Staphylococcus aureus and Escherichia coli. In total, 114 217 test results were included for the clinical isolates, and 3582 test results for control strains. The methodology at KS and VX followed the standard of the Swedish Reference Group for Antibiotics (SRGA). Zone diameter histograms for control strains were first analyzed to validate the procedure, and a comparison of actual means with the calculated means showed a correlation coefficient of r = 0.998. Results for clinical isolates at the two laboratories showed an excellent agreement for 54 of 57 combinations of antibiotic and bacterial species between normalized interpretations and the interpretations given by the laboratories. There were difficulties with E. coli and mecillinam, and S. aureus and tetracycline and rifampicin. The method was also tested on results from two laboratories using the NCCLS standard, and preliminary results showed very good agreement with quality-controlled laboratory interpretations.

CONCLUSIONS

The normalized resistance interpretation offers a new approach to comparative surveillance studies whereby the inhibition zone diameter results from disk tests in clinical laboratories can be used for calibration of the test.

Fluconazole and voriconazole multidisk testing of Candida species for disk test calibration and MIC estimation

Fluconazole and voriconazole MICs were determined for 114 clinical Candida isolates, including isolates of Candida albicans, Candida glabrata, Candida krusei, Candida lusitaniae, Candida parapsilosis, and Candida tropicalis. All strains were susceptible to voriconazole, and most strains were also susceptible to fluconazole, with the exception of C. glabrata and C. krusei, the latter being fully fluconazole resistant. Single-strain regression analysis (SRA) was applied to 54 strains, including American Type Culture Collection reference strains. The regression lines obtained were markedly different for the different Candida species. Using an MIC limit of susceptibility to fluconazole of < or =8 microg/ml, according to NCCLS standards, the zone breakpoint for susceptibility for the 25-microg fluconazole disk was calculated to be > or =18 mm for C. albicans and > or =22 mm for C. glabrata and C. krusei. SRA results for voriconazole were used to estimate an optimal disk content according to rational criteria. A 5-microg disk content of voriconazole gave measurable zones for a tentative resistance limit of 4 microg/ml, whereas a 2.5-microg disk gave zones at the same MIC level for only three of the species. A novel SRA modification, multidisk testing, was also applied to the two major species, C. albicans and C. glabrata, and the MIC estimates were compared with the true MICs for the isolates. There was a significant correlation between the two measurements. Our results show that disk diffusion methods might be useful for azole testing of Candida isolates. The method can be calibrated using SRA. Multidisk testing gives direct estimations of the MICs for the isolates.

Antibiotic medication and bacterial resistance to antibiotics: a survey of children in a Vietnamese community

OBJECTIVE

To investigate antibiotic use and antibiotic susceptibility of respiratory tract pathogens in children aged 1-5 years in Bavi, Vietnam.

METHOD

Nasopharynx and throat specimens were collected from 200 children from randomly selected households in a demographically defined population. Respiratory isolates were tested for antibiotic susceptibility according to the standard disk diffusion method. A questionnaire survey of carers elicited information on type of antibiotic used, duration of treatment, where the antibiotics had been purchased, type of treatment information retained by carers and episodes of illness preceding the study.

RESULTS

82% of the children had at least one symptom of acute respiratory tract infection (ARI) in the 4 weeks prior to the study, and of these 91% were treated with antibiotics. The most commonly used antibiotics were ampicillin (74%), penicillin (12%), amoxicillin (11%), erythromycin (5%), tetracycline (4%) and streptomycin (2%). Ampicillin was used for 3.3 days on average (SD:1.8) and penicillin for 2.6 days (SD:0.7). When deciding which antibiotic to use, 67% of the carers consulted the pharmacy seller, 11% decided themselves and 22% followed the doctor's prescription. The carrier rate of Streptococcus pneumoniae, Haemophilus influenzae and Moraxella catarrhalis was 50%, 39% and 17%, respectively. Isolates from 145 children were susceptibility tested, and 74% were found to carry resistant pathogens. Of the tested isolates, 90% of S. pneumoniae, 68% of H. influenzae and 74% of M. catarrhalis were resistant to at least one antibiotic. The mean number of antibiotics (susceptible strains excluded) to which resistance was found was 2.0 (SD:1.2), 2.5 (SD:1.8) and 2.1 (SD:0.9), respectively. S. pneumoniae and H. influenzae showed high resistance to tetracycline (88% and 32%, respectively), trimethoprim/sulphonamide (32% and 44%), and chloramphenicol (25% and 24%). 23% of S. pneumoniae were erythromycin-resistant and 18% of H. influenzae isolates were resistant to ampicillin. There was a significant difference in ampicillin and penicillin resistance between the group of children previously treated with beta lactam antibiotics and the group of children who did not receive antibiotics.

CONCLUSION

As reported by the carers, children in Bavi are treated with antibiotics frequently. Most antibiotics were obtained without consulting a doctor. High levels of antibiotic resistance and high prevalence of multidrug-resistant strains were found among respiratory pathogens. The existence of a large reservoir of resistance genes among children in low-income countries represents a threat to the success of antibiotic therapy throughout the world. Multi-faceted programmes to improve rational use of antibiotics in Vietnam are urgently needed.

MIC determination of fusidic acid and of ciprofloxacin using multidisk diffusion tests

OBJECTIVE

To investigate the possibility of estimating the MICs of fusidic acid and ciprofloxacin for bacterial isolates using series of antibiotic disk concentrations in diffusion tests, so-called M-tests.

METHODS

Thirty Staphylococcus aureus and S. epidermidis strains were tested for fusidic acid susceptibility. Sixty-one clinical isolates of eight bacterial species were tested for ciprofloxacin susceptibility. Disk diffusion was standardized according to the Swedish reference group for antibiotics (SRGA). For fusidic acid, a series of disks (1.5, 5.0, 15, 50 and 150 microg) was used. Ciprofloxacin was applied in four different diffusion sources (1, 3, 10 and 30 microg) on a single strip, the M-strip, and used. True MIC values were determined using the standardized agar dilution method according to the SRGA. Single-strain regression analysis (SRA) was employed to calculate critical concentration equivalents (Qzero).

RESULTS

Fusidic acid and ciprofloxacin critical concentrations were determined for the bacterial isolates. The mean conversion factors for Qzero to yield the true MIC were 2.06 (range 0.34-8.9) for fusidic acid and 2.05 (range 0.37-8.1) for ciprofloxacin. There was a correlation between true MIC values (all MICs expressed as 2 log + 9) and the calculated MIC values (Qzero x conversion factor) for both fusidic acid (R = 0.9822) and ciprofloxacin (R = 0.9696).

CONCLUSIONS

MIC values of clinical isolates can be estimated using SRA calculations on zone measurements in disk tests with several concentrations of the antibiotic in diffusion sources.

Calibration of the disk diffusion test for trovafloxacin susceptibility testing of four anaerobic species

OBJECTIVES

To study trovafloxacin susceptibility among clinical isolates of four anaerobic bacterial species using minimum inhibitory concentrations (MIC) determinations, E test assays and disk diffusion test results and to calibrate the disk diffusion method for these species using single strain regression analysis (SRA).

METHODS

One-hundred and eighty-seven clinical isolates of four anaerobic bacterial species were included. Trovafloxacin MIC determinations were performed using the agar dilution technique and MIC estimations using the E test. The disk diffusion test was performed according to Swedish Reference Group for Antibiotics standardization. NCCLS limits for susceptibility categories were applied. SRA was performed using 1, 3, 10, 30, and 100 microg trovafloxacin disk contents and ATCC control strains. The regression lines obtained permitted the calculation of zone equivalents to MIC limits as well as an evaluation of various disk potencies.

RESULTS

Trovafloxacin susceptibility (S + I) was noted in 98.9, 100, 100, and 97% of Bacteroides fragilis, Bacteroides thetaiotaomicron, Clostridium perfringens, and Peptostreptococcus magnus strains, respectively, as judged by MIC determinations. Agar dilution and E test estimations gave the same results, but E test values were consistently lower than MIC values by the reference method. Regression lines calculated for the four species using SRA showed different equation constants indicating species-related differences. Interpretive zone diameter breakpoints were calculated for the four species and used for the interpretation of susceptibility.

CONCLUSIONS

The disk diffusion test was successfully calibrated for trovafloxacin susceptibility testing of four anaerobic species using single strain regression analysis, SRA. There was a good agreement between the results of MIC-tests and disk testing. Interpretive errors of type I are prone to occur among Bacteroides isolates and might require species-related MIC limits. SRA calculations permitted the testing of the effect of different disk potencies on inhibition zones produced at the interpretive MIC limits. Criteria for the selection of a minimal disk content showed that 5 microg trovafloxacin is sufficient, but a 10 microg disk will safeguard against residual laboratory variation without producing too large inhibition zones for very susceptible strains.

Whole cell protein and partial 16S rRNA gene sequence analysis suggest the existence of a second Rothia species

OBJECTIVE: To subject ten clinical isolates grouped together based on their biochemical and microbiological profile to further investigations aimed at correct species identification. METHODS: The 16S rRNA gene was partially sequenced using nested amplification. Whole cell protein analysis (SDS-PAGE) and cluster analysis were performed on the 10 strains and also for comparison on 31 reference strains. The API Coryne biochemical kit as well as API 20 Strep were used for analysis of the phenotypic diversity of the strains by use of computerized numerical identification procedures. Antibiotic susceptibility testing was performed using a standardized disk diffusion test. RESULTS: The 265--556-bp-long 16S rRNA gene sequences of all 10 strains showed highest similarity to Rothia dentocariosa. Three strains showed complete identity between the sequences obtained and the sequence of the type strain of Rothia dentocariosa 16S rRNA gene (M59055), and the other seven ranged between 99.7% and 98.3% similarity. Detailed analysis of the sequences revealed a clustering of the strains into two groups. One group consisted of four isolates with the highest degrees of similarity with the reference strain (type I), while the members of another group (type II) showed differences in their nucleotide sequence at four distinct positions in the variable V7 region. T was replaced by C at position 597, C by T at position 608, T by C at position 610, and G by A at position 684 (position numbers according to reference sequence M59055, EMBL/GenBank). Whole cell protein analysis (SDS-PAGE) and cluster analysis also segregated the 10 Rothia dentocariosa strains into two different clusters, with one cluster containing all four strains belonging to 16S rRNA gene type I, and a second cluster containing all six strains belonging to 16S rRNA gene type II. CONCLUSIONS: Partial sequence data of the 16S rRNA gene as well as whole cell protein analysis showed a subdivision of the Rothia species into two groups, genomovar I (Rothia dentocariosa sensu stricto) and genomovar II, a possible new Rothia species.

Detecting erythromycin resistance in Streptococcus pyogenes: reliability of the disk diffusion method and the breakpoint susceptibility testing method. Finnish Study Group for Antimicrobial Resistance (FIRE)

Erythromycin susceptibility of clinical Streptococcus pyogenes isolates was determined at 19 Finnish clinical microbiology laboratories by their routine disk diffusion method and by a screening method adapted from the breakpoint susceptibility testing method. Results obtained at 12 laboratories using 4 major variants of the disk method were further evaluated. From these laboratories, 286 consecutive resistant and 349 consecutive susceptible isolates were sent to the Antimicrobial Research Laboratory, Turku where the MIC of erythromycin was determined. 96% and 97% of the disk results were correct, as compared with MIC results, when general and laboratory-specific breakpoints, respectively, were used. The results of the screening method were comparable to those of the disk method.

Disk diffusion susceptibility tests: need for laboratory-specific breakpoints

Significant changes were observed in the measured resistance levels of Escherichia coli and Staphylococcus aureus when the Biodisk disk susceptibility test method used in 1992 was replaced with another commercial method, Oxoid. For example, when non-species-specific breakpoints were used, the frequency of cephalotin-resistant E. coli was 12% of all isolates in 1992 but only 4% in 1993; the corresponding figures for the intermediately resistant isolates were 84% and 8%. The population distribution histograms were however, practically unchanged. Thus, the resistance percentages apparently did not reflect the real development of resistance. Similar findings were also made for several other antimicrobials. Susceptibility test breakpoints should therefore be examined separately for all bacterial species in each laboratory, and the application of adjusted laboratory-specific breakpoints should be considered. For this purpose, the WHONET computer program provides excellent assistance.

Antibiotic disk diffusion testing revisited. Single strain regression analysis. Review article

The standardized (NCCLS, ICS, DIN etc.) disk diffusion method is the most widespread technique for antibiotic susceptibility testing. Interpretive zone breakpoints are calculated from the regular regression line between minimum inhibitory concentrations (MIC) of bacterial isolates and the corresponding inhibition zone diameters around the disk containing the antibiotic. Studies of the regression line has revealed marked differences between different bacterial species. A newly described equation, the single strain regression analysis (SRA) equation, can be used to determine the regression line constants for individual strains. This method was applied to ciprofloxacin and S. aureus, E. faecalis, E. coli, P. mirabilis, P. aeruginosa, and P. maltophilia. The slope and intercept constants were determined for all 40 strains and showed a strong similarity within each species. A close similarity was also observed between the two Pseudomonas species and between S. aureus and E. faecalis. When the regression lines calculated by SRA for individual strains were extrapolated towards higher MIC values, the lines obtained for the more susceptible strains predicted the zones of more resistant strains within the species. The applications of SRA to several other antibiotics and bacterial species in earlier studies were reviewed. One exception to the predictive power of SRA has been detected earlier, H. influenzae and erythromycin. This led to the formulation of the standard curve regression analysis (SCA) equation which requires the use of two or more strains. Methodological aspects of SRA/SCA applications were presented. Three areas are particularly well suited for the use of SRA/SCA: 1. Calculation of interpretive zone breakpoints corresponding to recommended MIC limits in the individual laboratory. 2. Analysis of the effects of various disk contents of antibiotic on the resulting inhibition zones for various bacteria when new antibiotics are introduced. 3. Analytical tool as part of external quality control programmes.