Comparison of drug sensitivity testing with microdilution quantitative minimum inhibitory concentration and the Autobac I system

0095-1137/80/05-0471/05$02.00/0 Vol.11, No.5

Comparison of Drug Sensitivity

Testing with Microdilution

Quantitative Minimum

Inhibitory Concentration

and the

Autobac I

System

C. V.JACOB* ANDJ. KLEINEMAN

Department of Pathology,Saint Luke's Hospital, Cleveland, Ohio44104

Antibiotic susceptibility studieswerecompared bythe quantitative minimum inhibitory concentration and Autobac I. Four hundred common isolates from clinical materials including both gram-negative and -positive organisms were evaluated. Results demonstrate that the AutobacI methodcanprovide reliable semiquantitative estimation of bacterial susceptibilitythat correlates well with

an acceptable standard procedure. The specific advantages of the Autobac I

system in addition to the reliability demonstrated are speed, objectivity of

interpretation, and abilitytoprovideaprintedresult.These advantagescanresult

in increased speed of reporting, which may reduce hospital stay and patient

morbidity as well asproviding needed information morerapidly in critically ill patients with bacterial infections.

Drug sensitivity

tests are

routinely

performed in

the clinical

laboratory by

a number of meth-ods such asthe disk

diffusion and drug

dilution

techniques

(2, 3, 8,

11). In disk

diffusion

methods the tests are

reported

in

qualitative

terms,

whereas the

results

of drug dilution

tests are

reported

quantitatively.

Recently a number of

semiautomated

modifications of

these methods

have

come

into

use

which give

a rapid test result for

specific antibiotics (5,

9).

These

are based on

the

original method of Bauer (2)

as modified by U.S. Food and

Drug Administration (4) and the

National

Committee for Clinical Laboratory

Standards

(6).

This

investigation

will

report the results of

comparison studies of antibiotic sensitivity

stud-ies with

the semiautomated Autobac

I system

and the

microdilution

quantitative minimum

in-hibitory

concentration

(MIC)

system.

MATERIALS AND METHODS

Testand controlorganisms. Organisms used in

thisstudywerefresh isolates from theclinical

micro-biology laboratory at Saint Luke's Hospital,

Cleve-land, Ohio. The cultures were selected at random

within each bacterialclassification andwereidentified

by conventional biochemicalprocedures.Control

cul-tures were Escherichia coli ATCC 29194, E. coli

ATCC 25922, Staphylococcus aureus ATCC 25923,

and Pseudomonas aeruginosa ATCC 27853

pur-chased from theAmerican Type Culture Collection,

Rockville,Md., inlyophilizedstateand subculturedat

leastoncebeforeuse.

Media and reagents.For themicrodilution

quan-titativeMIC, the inoculationplastic trays containing

Muller-Hinton broth with andwithout antibiotic

di-lutionsweremanufactured andsuppliedinfrozenstate

by the Micro-mediaSystems Inc.,SanJose, Calif. The

concentrationsof the antimicrobial agent used in

Mi-crodilutionQuantitativeMIC systemwereinmultiples

oftwostartingfrom the lowest concentration(in

mi-crograms permilliliter)for each antibioticasfollows:

ampicillin (gram-positive organism),0.12 to8.0;

am-picillin (gram-negative organism), erythromycin, and

gentamicin,0.25 to 16.0;carbenicillin and

nitrofuran-toin,8.0 to512.0; cephalothinandkanamycin, 1.0 to

64.0;chloramphenicol,0.5 to32.0;penicillinG,0.06 to

4.0; and trimethoprim-sulfamethoxazole from 0.5:4.5

to 32:608ng/ml.

For the AutobacI, the eugonic broth and the

anti-microbialdrugcontaining disks withampicillin (0.22

,ugfor gram-positive organisms and4.5

jig

for

gram-negative organisms), Carbenicillin (120.0

ILg),

cepha-lothin and nitrofurantoin (15.0,tg), chloramphenicol

(4.0 ,ug), erythromycin (2.5 ug), gentamicin (9.0 ,ug), kanamycin(22.0 ,ug),penicillinG(0.2

pg), tetracycline

(0.5 ,ug for gram-positive organisms and 1.2 ,ug for

gram-negative organisms), and

trimethoprim-sulfa-methoxazole (18.0 ug) were purchased from Pfizer

Diagnostics and stored according to the

manufac-turer'ssuggestion.

Procedures. For the microdilution

quantitative

MIC, thestationary-phase inoculation standardization

as described by Barry et al. (1) was used with an

inoculum size of105to106colony-formingunits

(CFU)

perml. This inoculum sizewasrecommendedbythe

collaborative studyreported byEricssonandSherris

(3).Inbrief,from theprimary agarplates,fourtofive

identicalcolonieswereinoculatedinto 0.5ml ofbrain

heartinfusion broth andincubatedat35°Cfor4 to 6

h. This will give approximately 109 CFU/ml. From

this, a standardized inoculum in distilled water was

obtainedbypipetting0.05mlinto

screw-capped

tube

containing25 ml ofsterile distilled water

yielding

a

bacterialinoculumof

approximately

2x106

CFU/ml.

The frozen inoculationtrayswereallowedtothaw

just

465

on February 7, 2020 by guest

http://jcm.asm.org/

before use, the transfer lid was removed, and the

standardized inoculumwaspouredintothe seedtray.

The transfer lidwasreappliedwithsufflcientpressure

sothat allprongstouched the inoculum.

Thedesign of thetrayis suchthat eachprongpicks

up5[L of the inoculum anddeposits itinto 100,ulof

broth medium. This results in thefinal concentration ofapproximately 105 CFU/ml. The inoculated trays

werestacked and coveredtopreventevaporation. The stackswereincubated15 to 18 h at35°Cinaregular

incubator. The interpretationof thedrug susceptibil-itycriteria used for this studyis outlined in Table 1.

For the Autobac I, the inoculum standardization

andsensitivitytestsystem setupwas carried outas

described byThornsberryetal.(10).Inbrief, threeto four morphologically similar bacterial colonies from bloodagarplatesareselected,andanevensuspension

is prepared in sterile normalsaline. The turbidity of thebacterial suspensionisadjustedontheAutobac I

photometer to yield approximately 107 to 2 x 107 CFU/ml. The standardizedinoculum is diluted further 1:10ineugonic broth. Pseudomonas strainsarediluted

1:50 instead of 1:10 in eugonic broth to discourage bacterialflaking. Antimicrobialelution disksarethen addedtoeach cuvette,and the standardized inoculum isintroduced. Cuvettesareincubatedat35'C for 3to 5h inanincubator shaker. Thelight-scattering

read-ing for eachdrug concentration is comparedwith that ofthegrowth control and isrecordedas aratio. The

susceptibility interpretations usingthelight scattering indexareshown in Table 1.

RESULTS

The susceptibility testresults obtained with

MIC and Autobac I and their comparisons are

tabulated in Tables 2 and 3. Out ofthe 2,250

drug sensitivity tests performed with various

gram-negative organisms, theresults agreed in 96% by both methods. In the four groups of gram-negative organisms tested (Table 2), the highest incidence of discrepancywasfound with the tetracyclines(10.4%). Lesssignificant

differ-enceswereevident with ampicillin (6.89%),

ceph-alothin (5.2%), and chloramphenicol (4.0%). An overallagreement ofgreaterthan 90% between thetwomethodswasobserved with

nitrofuran-toin, kanamycin, gentamicin, chloramphenicol, cephalothin, carbenicillin, and

trimethoprim-sulfamethoxazole.

Among the gram-positive organisms, 93.8% showed identicalresultswithbothtestsystems.

The differences between thetwoprocedures

var-iedfrom 1.7to5.2%for gram-positive organisms

and were referred to as major and minor

dis-crepancies (Table 3). A minor discrepancy is

defined as onein which the organism's

suscep-tibility by one method is placed one category

higherorlower than that observedbythe alter-native method. Amajor discrepancy isdefined as one in which one group reports assensitive

and other reports as resistant or vice versa.

Minor discrepancies with gram-positive

orga-nisms varied from 1.7 to 2.8% with a mean of

2.3%,andmajordiscrepanciesvaried from 3.2to 5.2% withameanof3.9% foralldrugs.

Theoverallcorrelation between the two

meth-ods for both gram-positive and gram-negative

organisms sensitivity test averaged 95%. With

approximately 37% ofgram-negative organisms (Table 2) and 35% ofgram-positive organisms,

thediscrepanciesseen wereminor (Table3).

TABLE 1.

Interpretive

criteriausedin drug susceptibility testing with Autobac I and microdilution

quantitativeMIC

Microdilution quantitative MIC Autobac ILSI

Antimicrobial agents (_g/ml)

Autobac_I _LSI"

Rb I S R I S

Ampicillin (gram-negativeorga- C16.0 8.0 -4.0 C0.5 0.51-0.59 -0.60

nisms)

Ampicihin (gram-positiveorga- C0.5 -0.25 >0.50 0.51-0.59 -0.60

nisms)

Carbeniciflin <64.0 32.0 -16.0 C0.50 0.51-0.59 .0.60

CarbeniciUin (Pseudomonas) C512.0 256.0 -128.0 C0.50 0.51-0.59 -0.60

Cephalothin C32.0 16.0 -8.0 C0.50 0.51-0.59 -0.60

Chloramphenicol C32.0 16.0 -8.0 O0.50 0.51-0.59 -0.60

Erythromycin C8.0 4.0 -2.0 C0.50 0.51-0.59 -0.60

Gentamicin '16.0 8.0 -4.0 C0.50 0.51-0.59 -0.60

Kanamycin C32.0 16.0 -8.0 C0.50 0.51-0.59 -0.60

Nitrofurantoin c 128.0 64.0 -32.0 O0.50 0.51-0.59 -0.60

Penicillin

Gram-negativeorganisms C4.0 2.0 -1.0 0.0-0.91 0.91-1.0

Gram-positiveorganisms C0.25 -0.12 0.0-0.91 0.91-1.0

Trimethoprim-sulfamethoxazole <8:152 4:76

.2:38

C0.50 0.51-0.59 -0.60

LSI, Light-scatteringindex.

R, Resistant; I, intermediate; S,sensitive.

CLIN. MICROBIOL.

on February 7, 2020 by guest

http://jcm.asm.org/

TABLE 2. Antimicrobialsusceptibility ofgram-negative organisms asdeternined by Autobac I and microdilution quantitative MIC

No. of strains tested

Antimicrobial agent Coelation E. coli Pseudomo- mirabili K.K pneu- A organism%

(1O0)a nas (50) monia

(50) (50) Ampicillin Carbenicillin Cephalothin Chloramphenicol Gentamicin Kanamycin Nitrofurantoin Tetracycline Trimethoprim-sulfamethoxazole

Alldrugs(%)

Same

Minor'

Major' Same Minor Major Same Minor Major Same Minor Major Same Minor Major Same Minor Major Same Minor Major Same Minor Major Same Minor Major Same Minor Major 93 1 6 98 0 2 90 4 6 98 0 2 98 1 1 99 0 1 100 0 0 90 1 9 100 0 0 866(96.2) 8(0.8) 27(3.0) 50 0 0 47 0 3 50 0 0 47 0 3 50 0 0 46 4 0 50 0 0 40 4 6 43 0 7 40 0 10 50 0 0 47 3 0 45 4 1 50 0 0 50 0 0 47 3 0 49 0 1 49 0 1 423(94.0) 8(1.8) 19(4.2) 50 0 0 50 0 0 50 0 0 50 0 0 47 3 0 50 0 0 93.2 0.4 6.4 98.0 0.0 2.0 94.8 2.8 2.4 96.0 1.6 2.4 98.0 1.6 0.4 98.0 1.6 0.4 45 96.8 5 3.2 0 0.0 47 0 3 50 0 0 427(94.9) 10(2.2) 13(2.9) 89.6 2.0 8.4 96.0 0.0 3.2 439(97.5) 8(1.8) 3(0.7) 95.6(95.7 1.4(1.6) 2.1(2.7)

aTotal number of strains tested.

b Minor: byonemethod the

organism

wasplacedonecategory higherorlower than thatdescribedbythe

other method.

eMajor: total disagreement, onegroupreported assensitive and the otherreported asresistant and vice

versa.

DISCUSSION

The results with the MIC system and the

Autobac I willprovide asemiquantitative

esti-mateofbacterial susceptibilityto antibacterial

drugs in vitro. Earlier MIC agar diffusion and dilutionstudies (2, 3, 8)werefoundtocorrelate

wellwith results obtainedbytheprocedurethat

has been proposed as a reference method for testing susceptibility (1, 7, 10). The

microdilu-tion quantitative

MIC values corresponding

to

the reference

Kirby-Bauer inhibition

zones

of

intermediate,

sensitive, and resistant as

pro-posed by

the

Manual

of Clinical Microbiology

and other studies (7, 10) are

outlined

in

Table

1. Atotal of3,450 antimicrobial sensitivity tests were

compared

in this

study

with

microdilution

quantitative MIC

and the

Autobac

I system. The

overall

agreement

between the

twomethods for all

organisms

was 95%. In

only

5.4% of all 11,

on February 7, 2020 by guest

http://jcm.asm.org/

TABLE 3. Antimicrobialsusceptibility ofgram-positive organismsasdeterminedbyAutobac I and microdilution quantitative MIC systems

No. of strains tested

Antimicrobicalagents Correlation

All

organisnms

S. aureus S.

epidermis

Enterococci (%)

(50)b (50) (50)

Ampicillin Same 50 50 50 100.0

Minorc 0 0 0 0.0

Majord 0 0 0 0.0

Cephalothin

Erythromycin

Gentamicin

Nitrofurantoina

Penicillin

Tetracycline

Trimethoprim-sulfamethoxazole

Alldrugs (%)

Same Minor Major

Same Minor

Major

Same Minor

Major

Same Minor

Major

Same Minor

Major

Same Minor Major

Same Minor Major

Same Minor Major

48 1 1 46 2 2

50

0 0 48 2 0 44

2

4 49 0 1

46 0 4 46 0 4 50 0 0 46 4 0 42 4 4 46

3

1 44

0 6

379(94.8)

7(1.7) 14(3.5)

50 0 0 40

0

10 37 9 4 43 0 7 50 0 0 50 0 0

50

0 0

376(94.0) 11

(2.8)

13

(3.2)

50

0 0

370

(92.5)

9

(2.3)

21

(5.2)

96.6

0.0

3.3

88.0

1.3

10.6

91.3 6.0

3.6

91.3 4.0 4.6

90.6

4.0

5.3

96.0 2.0

1.3

96.0

0.0

4.0

93.8(93.8) 2.1(2.3)

3.9(3.9)

a Used forurinarytractinfection.

'Totalnumber of strains tested.

'By onemethod, the organism was placed one category higher or lower than that described by the other

method.

d Totaldisagreement,onegroupreported as sensitive and the other reported as resistant or vice versa.

tests, the

susceptibility

as

determined

by

the disk

elution

method

(Autobac

I)

differed

from that

obtained

by

the

microdilution

broth MIC procedure.Among the

gram-negative organisms

the minor

discrepancy

varied

from

0.8to 1.8% of tests, an average of

1.6%; major discrepancy

variedfrom0.7 to4.2%, withanaverage of 2.7%.

Thus,

withgram-negative organisms, 37% of the

discrepancies observed

wereminor (Table 2).

Among

the 1,200

sensitivity

tests

performed

on gram-positive organisms by both methods, the

minor and

major

discrepancies

varied

from

1.7 to 2.8% and 3.2 to 5.2% (Table 3),

respec-tively. Further analysis showed

that the minor

variation alone accounted for 35% of the ob-serveddifferences.

Theoverallcorrelation between thetwo

meth-odsaveraged 95%. This is excellent agreement,

considering

the many differences in

technique

and reagents usedin the twomethods. For

ex-ample,

theMIC and Autobac Isystem use dif-ferent basalmedia; in the MIC method,a

single

antibiotic concentration is mixed with the basal

medium,

whereas in the Autobac I method, the antibiotic is released from the disk during the cultureinoculation andincubation.

These results demonstrate that the Autobac Imethod, when appropriately used,provides a

J. CLIN. MICROBIOL.

on February 7, 2020 by guest

http://jcm.asm.org/

I

reliable semiquantitative estimate of bacterial

susceptibility that correlates

well

with

an

ac-ceptable

standard procedure.

The

specific advantages of

the Autobac I sys-tem other

than its

reliability

areits speed,

ob-jectivity for

interpretation, and ability

toprovide a

printed

result. This will

reduce technical

vari-ation and coping

errors.

These

advantages can

be

used

to

improve the effectiveness of

the

mi-crobiological section of

the

clinical laboratory

in

providing patient

care.

ACKNOWLEDGMENT

Wethank M. Alousi forhelpful suggestionsand B.Semple fortechnicalassistance.

LITERATURE CITED

1. Barry, A.L, F. Garcia, and L D. Thrupp. 1970. An improvedsingle disc method for testing the antibiotic susceptibilityofrapidly growingpathogens. Am. J.Clin. Pathol.53:149-158.

2. Bauer, A. W., W. M. M. Kirby, J. C. Sherris, and M. Turck. 1966. Antibioticsusceptibilitytesting by a stan-dardizedsingle disk method. Am. J. Clin. Pathol.45: 493-497.

3. Ericsson,H. M., and J. C. Sherris. 1971. Antibiotic sensitivity testing: report of aninternational collabora-tive study. Acta Pathol. Microbiol. Scand. Sect. B. (Suppl.) 21:1-90.

4. Food andDrug Administration. 1972. Rules and regu-lations:antibiotic susceptibility discs. Fed.Regist. 37: 20525.

5. Mekie, J., R. Borovoy, J. Dooley, G. Evanega, G. Mendoza, F. Meyer, M. Moody, D. Packer, J. Prag-lin,and H. Smith. 1974. Autobac I: a 3 hr automated antimicrobialsusceptibilitysystem. II.Microbiological studies, p. 209-242. In C.G. Heden and T.Deni(ed.), Automation in microbiology and immunology. John Wiley andSons, Inc., New York.

6. National Committee forClinicalLaboratory Stand-ards. 1973. Performance standards for antimicrobial discsusceptibility tests, as used in clinical laboratories: revised tentative standard. National Committee for ClinicalLaboratory Standards, LosAngeles. 7. Paik, G. 1970. Reagents, stains and test procedures, p.

678.In J. E. Blair, E. H. Lennette, and J. P. Truant. (ed.),Manual of clinical microbiology. American Soci-ety forMicrobiology, Bethesda, Md.

8. Ryan, K. J., G. M.Needham, C. L Dunsmoor, and J. C.Sherris. 1970. Stability of antibiotics and chemo-therapeutics in agar plated. Microbiology 20:447. 9. Steers,E.,E. L.Foltz,B. S.Graves,and J. Riden.

1959. Aninoculareplicatingapparatus for routine test-ing ofbacterialsusceptibility toantibiotics. Antibiot. Chemother.9:307-311.

10.Thornsberry,C.,T. LGavan,J. C.Sherris,A.

Bal-ows, J. M.Matsen,L. D.Sabath,F.Schoenknecht,

L. D.Thrupp,and J. A.WashingtonH. 1975. Lab-oratoryevaluation ofarapid,automaticsusceptibility

testingsystem: report ofacollaborativestudy. Antimi-crob.Agents Chemother. 7:466-480.

11.Washington,J.A.,II.1971.Theagar-dilution method, p. 127-141. In T. L.Gavan, H. W. McFadden, Jr., and E.L. Cheatte(ed.), Antimicrobialsusceptibility testing. AmericanSociety ofClinicalPathologists,Commission onContinuing Education, Chicago.

VOL. 11,

on February 7, 2020 by guest

http://jcm.asm.org/