J Clin Pathol 1985;38:1132-1138

Four hour identification of Enterobacteriaceae withthe API Rapid 20E and Micro-ID systems

B HOLMES, PS HUMPHRY

From the National Collection of Type Cultures, Central Public Health Laboratory, London

SUMMARY One hundred strains of Enterobacteriaceae were examined in parallel with the APIRapid 20E and Micro-ID commercial four hour identification systems. With the API Rapid 20Esystem 78% of the strains were correctly identified, 15% were not id,entified, and 7% were

misidentified. The respective figures with the Micro-ID system were 74%, 11%, and 15%.

The Micro-ID system and its predecessor Pathotecwere among the earliest commercial systemsdesigned to yield identification of Enterobac-teriaceae after four hours of incubation. Later, APIUSA (Analytab Products, Plainview, New York),developed a data base to permit interpretation ofthe results obtained after incubating the standardAPI 20E system for only five hours (this data base isavailable only in America). More recently, APIFrance (API System SA, La Balme-les-Grottes,Montalieu Vercieu, France, a company independentof API USA and which supplies the United King-dom with API products) introduced a productspecifically designed for four hour identification ofEnterobacteriaceae, API Rapid 20E (this productwas available in America as the DMS Rapid E Sys-tem from DMS Laboratories, Flemington, NewJersey, but since October 1984 the product has beenmarketed by API USA as the API Rapid E system).Only a few authors have performed an evaluation

solely of the Micro-ID system.'-4 More often theMicro-ID system has been evaluated in parallel withanother system, generally API 20E as an 18-24hour identification system.5"'4 In one of thesestudies the authors included time and motion andcost comparisons in their evaluation as well as suc-cess in identification.6 In addition, in two of thesestudies other commercial identification systems wereincluded in the evaluation (AutoMicrobic System,Enterotube, and Micro-Media System,'2 API 1OS,Enterotube, Enterotube II, and Minitek'3). SomeAmerican authors have compared the Micro-ID sys-tem with API 20E, using the latter both as an 18-24hour and as a five hour identification system.'5-'7 Intwo of these studies other commercial identificationsystems were also included in the evaluation(Abbott MS-2,"5 Entero-Set 20'7). Gooch and HillAccepted for publication 16 May 1985

compared Micro-ID with API 20E, using the lattersolely as a five hour identification system.'8 Micro-ID has also been evaluated in parallel with theMinitek system'9 20 and with the r/b system.2' 22Some workers have used the Micro-ID system forthe direct identification of bacteria from blood23 -2(in two of these studies other commercial systemswere included in the evaluation: Abbott MS-2, API20E, and AutoMicrobic System,26 Minitek27) andurine,2829 whereas others have examined the effectof refrigeration of the inoculated system to permitdelayed reading of the reactions.3032 Cox et al mod-ified the procedure for inoculating the Micro-ID kit,using a single colony to minimise the possibility ofinoculating the kit with a mixed culture." TheMicro-ID system has also been used for the rapidbiochemical characterisation of Haemophilusspecies.34The API Rapid 20E system has been evaluated

alone35 but more commonly in parallel with thestandard API 20E system.36-"0 In two of the latterstudies other commercial identification systems werealso included in the evaluation (Enterotube II andMast ID-System,36 Abbott Quantum IP9). Therehave been two previous reports comparing the APIRapid 20E and Micro-ID systems,'412 although inone of these studies42 the systems were used todirectly identify bacteria from blood.We report here our experiences obtained from an

examination of 100 strains of Enterobacteriaceaeexamined in parallel with the API Rapid 20E andMicro-ID systems.

Material and methods

ORGANISMSAs only 100 units of each commercial system hadbeen made available to us by the respective manu-

1132

on May 1, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.10.1132 on 1 October 1985. D

ownloaded from

Four hour identification of Enterobacteriaceae with the API Rapid 20E and Micro-ID systems

facturers, only 100 strains could be used to assessthe success of the API Rapid 20E and Micro-IDsystems in identification (Table 1). The strains com-prised a mixture of reference strains (maintained inthe National Collection of Type Cultures) and fieldstrains referred to the National Collection foridentification. All the strains had been previouslyexamined for 50 conventional characteristics andtheir identity confirmed or determined by the com-puter identification method of Lapage et al43 andprobability matrix of Bascomb et al.44 Althoughsome field strains are referred to us because they areatypical in certain characteristics, others received foridentification are typical but have been submittedfor other reasons (such as test errors). Only fairlytypical field strains were used in this evaluation.

TESTSAll 100 strains were subcultured several times in anattempt to increase their enzyme activity. They wereinitially recovered on nutrient agar in Petri dishes.They were then allowed to grow for up to severaldays. This procedure was then repeated a furthertwo times except that on the final subculture onlyovernight incubation was permitted (before inocula-tion of the commercial systems). All were thenexamined with the API Rapid 20E system, whichconsists of a series of plastic cupules (containingdehydrated media) moulded to a strip of plastic. Thefollowing 21 tests were performed in the cupules:/3-galactosidase production (ONPG test), lysinedecarboxylase, ornithine decarboxylase, urease pro-duction, citrate utilisation, deamination of

Table 1 Identification of100 strains by the API Rapid 20E and Micro-ID systems

Taxon Total API Rapid 20E Micro-IDNo ofstrains Correctly Not Incorrectly Correctly Not Incorrectly

identified identified identified identified identified identifiedCitrQbacter freundii 3 0 1 2 0 2 1Citrobacter koseri 3 3 0 0 3 0 0Edwardsiella tarda 3 3 0 0 3 0 0Enterobacter aerogenes 3 2 0 1 3 0 0Enterobacter cloacae 3 2 1 0 3 0 0Escherichia adecarboxylata 3 0 3 0 3 0 0Escherichia coli 3 3 0 0 2 (1) 1 0Hafntia alvei 3 3 0 0 3 0 0KlebsieUa oxytoca 1 1 0 0 1 0 0Klebsiella pneunoniae

subspecies aerogenes 3 3 0 0 2 0 1KlebsieUa pneunoniae

subspecies ozaenae 3 2 (1) 0 1 2 0 1Klebsiella pneumoniae

subspeciespneumoniae 3 3 (1) 0 0 2 0 1Klebsiella pneumoniae

subspecies rhinoscleromatis 3 2 1 0 1 1 1MorganeUa morganii 3 3 (1) 0 0 3 0 0Proteus mirabilis 3 3 0 0 3 0 0Proteus vulgaris 3 3 (1) 0 0 3 0 0Providencia alcalifaciens 3 3 0 0 3 0 0Providencia rettgeri 3 3 (2) 0 0 2 0 1Providencia stuartii 3 1 2 0 1 0 2Salmonella choleraesuis 2 2 0 0 2 0 0Salmonella ferlac 2 2 (2) 0 0 2 0 0Salmonella gaUinarum 3 1 1 1 3 0 0Salmonella pullorum 2 0 1 1 2 0 0Salmonella subgenus I 3 3 0 0 3 0 0SalmoneUla subgenus II 3 0 3 0 0 3 0SalmoneUla subgenus III 3 3 0 0 3 0 0Salmonella subgenus IV 2 2 0 0 2 0 0Serratia liquefaciens 3 0 2 1 1 2 0Serrada marcescens 3 3 (1) 0 0 2 0 1Serratia marinorubra 3 3 (1) 0 0 1 0 2Serratia plymuthica 3 3 (2) 0 0 0 1 2Shigella boydii 3 3 0 0 3 0 0Shigellaflexneri 1 1 0 0 1 0 0Shigella sonnei 3 3 0 0 3 0 0Yersinia enterocolitica 3 3 0 0 0 1 2Yersinia pseudotuberculosis 3 3 0 0 3 0 0

Total 100 78 (12) 15 7 74(1) 11 15Numbers in parentheses are strains identified by reference to the full data base, either held on a floppy disc for use in a microcomputer(API Rapid 20E), or via the manufacturer (Micro-ID).Escherichia adecarboxylata is not specifically included in the data base of either product. Escherichia adecarboxylata strains have, however,been considered to be biochemically atypical strains ofErwinia herbicola (synonym Enterobacter agglomerans). Strains ofE adecarboxylatashould therefore either fail to identify on the system or identify to E herbicola, the E herbicola data base of the API Rapid 20E system doesnot allow for strains with the characteristics ofE adecarboxylata, unlike that of the Micro-ID. Serratia plymuthica is not included in the database of the Micro-ID system.

1133

on May 1, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.10.1132 on 1 October 1985. D

ownloaded from

1134

phenylalanine, malonate utilisation, aesculin hyd-rolysis, fermentation of arabinose, xylose, adonitol,rhamnose, cellobiose, melibiose, sucrose, trehalose,raffinose and glucose, production of indole, produc-tion of acetoin, and production of cytochrome oxid-ase.

In accordance with the manufacturer's instruc-tions, the Rapid 20E kits were each inoculated withone well isolated colony (or several colonies of iden-tical morphology where necessary) taken fromovernight cultures grown on nutrient agar. The bac-terial growth was removed from the plate using thetip of a polythene pipette (supplied with the strip).The bacterial growth was then homogenised in1-25 ml of sterile physiological saline at 085% (andcontaining no additives) to achieve a turbidity equi-valent to 05 on the McFarland scale (equivalent to1 or 2 medium sized colonies). The tests were theninoculated with the suspension; in the citrate testtwo drops were added so as to fill both tube andlower part of the cupule, and in the other testsjust sufficient inoculum was added to fill the tubesection only. The lysine decarboxylase, ornithinedecarboxylase, and urease tests were then overlayedwith mineral oil before the tests were incubated forfour hours at 37°C.The reactions obtained were recorded after refer-

ence to a written description, provided by the man-ufacturer, of the appearance of positive and negativereactions. (A colour chart was not available at thetime of this evaluation but. one has since been pro-duced; however, owing to limitations in printing inkssuch charts often offer no advantage over detailedwritten descriptions.) Reagents had to be added tothe cupules testing for acetoin and indole produc-tion. The oxidase reaction of the strains had beendetermined previously by conventional means, but itmay also be determined in the strip by the additionof the recommended reagent into the aesculin orphenylalanine cupules (if negative).

All the strains were also examined in parallel withthe Micro-ID system, which comprises a mouldedstyrene tray containing 15 reaction chambers and ahinged cover. The first five reaction chambers con-tain a substrate disc and a detection disc. Theremaining 10 reaction chambers each contain asingle combined substrate/detection disc. Thesefilter paper discs contain all substrate and detectionreagents required to perform the indicated biochem-ical test (except that for acetoin production). Thesurface of the tray is covered with transparentpolypropylene tape to contain the organism suspen-sion during incubation and also provides completevisibility. The hinged cover is opened to provideaccess to the inoculation ports. The inside surface ofthe cover contains a strip of filter paper to absorb

Holmes, Humphryany spillage resulting from mishandling. The follow-ing 15 tests were performed in the reaction cham-bers: acetoin production, nitrate reduction, deami-nation of phenylalanine, hydrogen sulphide produc-tion, indole production, ornithine decarboxylase,lysine decarboxylase, malonate utilisation, ureaseproduction, aesculin hydrolysis, /3-galactosidaseproduction (ONPG test), and fermentation ofarabinose, adonitol, inositol, and sorbitol.

In accordance with the manufacturer's instruc-tions, the Micro-ID kits were each inoculated withseveral well isolated morphologically identical col-onies taken from overnight cultures grown on nut-rient agar. The bacterial growth was removed fromthe plate using a wire loop and was then homogen-ised in 3-5 ml of physiological saline to achieve aturbidity equivalent to 2-0 on the McFarland scale(this is the inoculum density recommended for stockcultures; for fresh clinical isolates a turbidity equi-valent to 0-5 on the McFarland scale is recom-mended). With the Micro-ID tray flat on the benchand the cover open the tests were inoculated withthe suspension, about 0- 2 ml per test. With the coverclosed the Micro-ID tray was then incubated in anupright position for four hours at 37°C, after ensur-ing that the suspension was in contact with all subs-trate discs.

After incubation the Micro-ID tray was againplaced flat on the bench and two drops of 20% KOHadded to the acetoin test only. The tray was thenheld upright again to allow the KOH to flow downinto the acetoin production test solution. Finally, thetray was rotated clockwise through 900 so that theupper (detection) discs in the first five wells becamewet. The reactions obtained were recorded afterreference to a written description, provided by themanufacturer, of the appearance of positive andnegative reactions (a colour chart is available, how-ever, for those who prefer it).

IDENTIFICATION OF STRAINSThe reactions of the 21 API Rapid 20E tests arerecorded as a seven digit profile number based onoctal numbers.45 This seven digit number is lookedup in an index entitled API Rapid 20E AnalyticalProfie Index (product reference No 2079), and ourswas dated March 1982. If a profile number was notlisted in the profile index then the full data base,held on a floppy disc for microcomputer use, wasconsulted (the same information can be obtained bya telephone service offered by the manufacturer). Inwhichever way the data base is consulted a commentis given as to the level of acceptability of the sug-gested identification.The reactions of the 15 Micro-ID tests are

recorded in a similar way to yield a five digit profile

on May 1, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.10.1132 on 1 October 1985. D

ownloaded from

Four hour identification of Enterobacteriaceae with the API Rapid 20E and Micro-ID systemsnumber, which is looked up in an index entitledMicro-ID Identification Manual, and ours was datedJanuary 1981. If a profile number was not listed inthe index then the full data base was consultedthrough the manufacturer. For entries in the index a

comment is given, as with API Rapid 20E, on thelevel of acceptability of the suggested identification.

For some identifications in both the API Rapid20E and Micro-ID systems additional conventionaltests, as well as serological investigations, may berequired to further or confirm the identification. Theadditional tests were not carried out in this study; ifthe system indicated the correct taxon as one of thepossible identities this was counted as a correctidentification. If serological investigation wouldhave prevented a misidentification then that patternof results was counted as not identified.

Results

As can be seen from Table 1 the number of strainsnot identified was almost identical in both systems.API Rapid 20E had a higher identification rate(78%) than Micro-ID (74%). Without reference tothe full data base, however, the identification ratewith API Rapid 20E was only 66% compared withthe corresponding Micro-ID figure of 73%. The mis-identification rate with Micro-ID (15%) was twicethat with API Rapid 20E (7%). Both systemsencountered difficulty in identifying the strains ofCitrobacter freundii, Providencia stuarti, Salmonellasubgenus II, and Serratia liquefaciens. API Rapid20E proved much more effective in identifyingstrains of Serratia plymuthica and Yersiniaenterocolitica, but this is not surprising for theformer taxon as S plymuthica is not included in thedata base of the Micro-ID system. Conversely,Micro-ID proved more effective in identifyingEscherichia adecarboxylata. This is not surprisingeither, however, for reasons given in the footnote toTable 1. For the remaining taxa the identificationperformance of the two systems was about the same.

Table 2 details the misidentifications.

Discusson

In the few published evaluations of API Rapid 20Ethe identification success rate has generally rangedfrom about 90%36 to 96%.3' The lowest ratesreported have been around 85%,"35 although inone of these studies the identification rate rose to92-4% when use was made of the full data base heldby the manufacturer.4' Misidentification ratesreported have been low, around 1-5%.4' Theidentification rate we obtained for the API Rapid20E system (78%, which includes referrals to the

full data base), is considerably lower than thatreported by other authors. Similarly, the mis-identification rate we observed (7%) was higherthan that previously reported.

In the 30 or so published evaluations of theMicro-ID system the identification rate has beengenerally 90% or more, even when the kit wasdirectly inoculated with blood2327 or urine,2829 andhas sometimes exceeded 99%.20 33 A lower rate hasbeen reported (87%) in a study in which the authorswere inoculating the kit directly from blood cul-tures,42 and also in another study4' (74.2%) inwhich the author was working with atypical strainsthat could not be identified by conventional means.In the latter study,41 however, the author alsoexamined the strains with the API Rapid 20E sys-tem and obtained a good identification rate (92.4%)despite the use of atypical strains. Misidentificationrates reported with the Micro-ID system have variedbetween 1.5% 10 25 for typical strains and 16 7%4'for atypical strains. The identification rate weobtained with the Micro-ID system (74%, whichincludes referrals to the full data base), is consider-ably lower than that reported previously. Otherauthors, however, had noted a misidentification rate(16-7%) in excess of that which we observed (15%),although using atypical strains.4'

There are two probable reasons why our

Table 2 Misidentification ofstrains by theAPI Rapid 20Eand Micro-ID systems

Organism Incorrecdy identified as:

API Rapid 20ECitrobacter freundii Escherichia coliCitrobacter freundii Escherichia coliEnterobacter aerogenes Serratia fonticolaKiebsiella pneumoniae Yersinia pseudotuberculosis

ssp ozaenaeSalmonella gallinarum Serratia odoriferaSalmoneUa pullorum Hafnia alveiSerratia liquefaciens Serratia odorifera

Micro-IDCitrobacter freundii Yersinia enterocoliticaKlebsiella pnewnoniae Serratia rubidaea (synonym

ssp aerogenes Serratia marinorubra)KlebsieUa pneumoniae Yersinia pestis

ssp ozaenaeKlebsiella pneumoniae Klebsiella pneumoniae

ssp pneunoniae ssp ozaenaeKIebsiella pnewnoniae Klebsiella pneumoniae

ssp rhinoscleromans ssp ozaenaeProvidencia reugeri Providencia stuartiiProvidencia stuarti Providencia akcalifaciensProvidencia stuartii Providencia alcalifaciensSerratia marcescens Hafnia alveiSerratia marinorubra KlebsieUa pneumoniae

ssp ozaenaeSerratia marinorubra Klebsiella pneumoniae

ssp ozaenaeSerratia plymuthica* Enterobacter agglomeransSerratia ptyymuthica* Enterobacter agglomeransYersinia enterocolitica Yersinia pseudotuberculosisYersinia enterocolitica Yersinia pseudotuberculosis

*Not included in the data base of the Micro-ID system.

1135

on May 1, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.10.1132 on 1 October 1985. D

ownloaded from

1136

identification rate, for both systems, is lower and ourmisidentification rate higher than that reported bymost other authors. Firstly, we used only stock cul-tures, and other authors58 noted lower identificationrates for such strains (for example, 93-2% withMicro-ID8) than with fresh clinical isolates (forexample, 97-8% with Micro-ID8). This is becausestock cultures may show less enzyme activity thanfresh clinical isolates. Indeed, the manufacturer ofthe Micro-ID system recommends a more denseinoculum (approximately equal to a McFarland No2 turbidity standard) for use with stock cultures. Wesubcultured our strains several times in an effort toincrease their enzyme activity and increased theinoculum density as recommended, but it is not poss-ible to judge how successful these procedures were.The only way to test properly the effectiveness ofrepeated subculturing would be to examine eachstrain twice with the commercial system, once withminimum subculturing and once with repeated sub-culturing. The second possible reason for our loweridentification rate is that we chose a wider range oforganisms, including several rarer taxa, than wouldnormally be available to a routine clinical laboratoryperforming such an evaluation. For example,Escherichia coli represented only 3% of the strainsin our study, but in other studies this organismaccounted for 38-3% (143/373)5 and 59%14 of thestrains examined.Whatever the factors that caused our

identification rate to be lower than that in mostother published evaluations of these products, thosefactors apply equally to the two products. Therefore,a comparison of the two, of which there are at pres-ent only two such reports,4142 is of much greaterinterest. From the procedures used it can be seenthat the API Rapid 20E product had several appar-ent advantages over Micro-ID. Firstly, there were21 tests as opposed to 15. With more tests, particu-larly if wisely chosen, there can be greater discrimi-nation between taxa so that in theory, at least, thereshould be fewer instances of identification solely togenus level and fewer instances in which supplemen-tary tests need to be performed in order to furtherthe identification. Secondly, although both productsused the same density of inoculum (05 on theMcFarland scale) the volume of inoculum was con-siderably greater (3.5 ml) with Micro-ID than it waswith API Rapid 20E (1*25 ml). To achieve sufficientinoculum the Micro-ID kit will always require sev-eral colonies, but with API Rapid 20E a single col-ony will often suffice. This point will vary in impor-tance with different workers, but most workers willaccept that with certain specimens, in particularfaeces, there is a greater danger that morphologi-cally identical colonies may represent different

Holmes, Humphry

organisms. Others who have inoculated the Micro-ID system from primary isolation plates have foundthat a sufficient inoculum density could, in any case,be obtained only with a proportion of the cultures(74%,5 87%,8 92%,'4 83%18).

Since it may not always be convenient to readMicro-ID tests after four hours several authors havefound that identification at the species level is essen-tially unchanged when inoculated strips are refriger-ated overnight and then incubated, or incubated andthen refrigerated overnight, before reading.30-32 It isnot yet known if it will be possible to do this withAPI Rapid 20E. Finally, since there are fewer testswith the Micro-ID system than with API Rapid 20E,fewer combinations of test results are possible withMicro-ID so the data base for that system is smallerthan that for API Rapid 20E. The Micro-ID profileindex is therefore more comprehensive than that ofAPI Rapid 20E. Consequently, fewer referrals needto be made to the full Micro-ID data base (twostrains in this study) than to the full API Rapid 20Edata base (25 strains in this study). Such referralsare, however, certainly worthwhile as the use of thefull data base increased the identification rate forAPI Rapid 20E by 12% in this study and 7-6% inanother study4' (although the identification rate forMicro-ID was increased by only 1% in this study).The identification rate of API Rapid 20E at 78%

compared with that of Micro-ID at 74% suggeststhe former to be the more successful product, asdoes the overall misidentification rate for API Rapid20E (7%), which is only half that for Micro-ID(15%). The latter figures agree reasonably well withthose of Altwegg4' (1.5% and 16-7%, respectively).The proportion of strains not identified was aboutthe same with both systems.The misidentifications with the API Rapid 20E

system (Table 2) should cause few serious consequ-ences. The Salmonella pullorum strain that was mis-identified as a Hafnia alvei might cause problems.Given the biochemical similarity of these two organ-isms this problem could perhaps be avoided if, whenH alvei was the suggested identification, the man-ufacturer recommended testing with a polyvalentSalmonella antiserum to exclude the possibility of amisidentified Salmonella. Other authors have notedas we have, that occasional strains of Citrobacterfreundii may be misidentified as Escherichia coli39and occasional strains of Serrada liquefaciens may bemisidentified as Serratia odorifera37 with the APIRapid 20E system.The misidentifications with the Micro-ID system

(Table 2) were confined almost exclusively to thegenera Klebsiella, Providencia, Serratia, and Yer-sinia. With Providencia and Yersinia only was thespecies identification incorrect, and a similar obser-

on May 1, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.10.1132 on 1 October 1985. D

ownloaded from

Four hour identification of Enterobacteriaceae with the API Rapid 20E and Micro-ID systems

vation applies to half the misidentified Klebsiellastrains. The misidentifications with Micro-ID shouldcause few serious consequences, although theidentification of a strain of Klebsiella pneumoniaessp ozaenae as one of Yersinia pestis would causesome concern until the clinical condition of thepatient was reappraised. Our observations about theMicro-ID system have also been confirmed by otherauthors; the following misidentifications may occa-sionally be possible: Citrobacter as Yersiniaenterocolitica,7 Klebsiella pneumoniae as Serratiamarinorubra (synonym Serratia rubidaea),232728Providencia stuartii as Providencia alcalifaciens,27Serratia marcescens as Hafnia alvei,23 and Yersiniaenterocolitica as Yersinia pseudotuberculosis.2 Stuverand Matsen identified only nine of 16 strains of Pro-videncia stuariii using the Micro-ID system,4 andothers have considered the product inaccurate foridentifying Providencia as well as Serratia.I6 Othershave noted the misidentification of Klebsiella strainsto Serratia23 27 28 and vice versa,8 and this is in accordwith our own findings.Some misidentifications are inevitable with com-

mercial identification systems because a lowerthreshold identification level is used than in thereference laboratory situation.43 The overall mis-identification figures suggest that the data base forMicro-ID is more lenient than that for API Rapid20E. Potentially weak areas in the data bases ofboth products might warrant the attention of therespective manufacturers.

Prices can change relatively quickly and the accu-racy and ease of use of a commercial identificationsystem are likely to be given greater weight thancost, but for those interested, the current price forAPI Rapid 20E is £1-38 compared with £128 forAPI 20E. Since the two prices are so similar, amajor factor influencing which system to use will bewhether one wishes to identify a strain in four hoursor 18 to twenty four hours. The unit cost for Micro-ID is £2*70, which is almost twice that of API Rapid20E. Both companies offer discounts on bulk pur-chases, which will reduce the above prices by up to10-15%.

We thank API Laboratory Products Limited andGeneral Diagnostics for their respective gifts of APIRapid 20E and Micro-ID kits.

References

Buesching WJ, Rhoden DL, Esaias AO, Smith PB. Evaluation ofthe modified Micro-ID system for the identification ofEnterobacteriaceae. Abstracts of the Annual Meeting of theAmerican Society for Microbiology 1979:333.

2 Buesching WJ, Rhoden DL, Esaias AO, Smith PB, WashingtonII JA. Evaluation of the modified Micro-ID system foridentification of Enterobacteriaceae. J Clin Microbiol1979; 10:454-8.

Cox NA, Mercuri AJ. Accuracy of Micro-ID for identification ofSalmonella and other Enterobacteriaceae from clinical andfood sources. Journal of Food Protection 1979;42:942-5.

4Stuver JC, Matsen JM. Evaluation of revised Micro-ID kits forEnterobacteriaceae identification. Abstracts of the AnnualMeeting of the American Society for Microbiology 1978:304.

Alridge KE, Gardener BB, Clark SJ, Matsen JM. Comparison ofMicro-ID, API 20E, and conventional media systems inidentification of Enterobacteriaceae. J Clin Microbiol1978;7:507-13.

6 Bale MJ, Matsen JM. Time-motion and cost comparison study ofMicro-ID, API 20E, and conventional biochemical testing inidentification of Enterobacteriaceae. J Clin Microbiol1981; 14:665-70.

Barry AL, Badal RE. Rapid identification of Enterobacteriaceaewith the Micro-ID system versus API 20E and conventionalmedia. J Clin Microbiol 1979; 10:293-8.

Blazevic DJ, Mackay DL, Warwood NM. Comparison of Micro-ID and API 20E systems for identification of Enterobac-teriaceae. J Clin Microbiol 1979;9:605-8.

9 Edberg SC, Atkinson B, Chambers C, Moore H, Zorzon C,Singer JM. Clinical evaluation of the Micro-ID, API 20E, andconventional media systems for the identification ofEnterobacteriaceae. Abstracts of the Annual Meeting of theAmerican Society for Microbiology 1978:305.

'° Edberg SC, Atkinson B, Chambers C, Moore MH, Palumbo L,Zorzon CF, Singer JM. Clinical evaluation of the Micro-ID,API 20E, and conventional media systems for identification ofEnterobacteriaceae. J Clin Microbiol 1979; 10:161-7.

"Izard D, Gavini F, Ochin D, Enayeh E, Leclerc H. Evaluation ofMicro-ID, for identification Enterobacteriaceae. Pathol Biol(Paris) 1980;28:681-4.

12 Kelly MT, Latimer JM. Comparison of the AutoMicrobic Systemwith API, Enterotube, Micro-ID, Micro-Media Systems, andconventional methods for identification of Enterobacteriaceae.J Clin Microbiol 1980; 12:659-62.

3 Mastroeni P, Carbone M, Fera MT, Teri G, Burdash NM. Com-parison of six commercial systems for the identification ofEnterobacteriaceae. The Public Health Laboratory (Journal ofthe Conference of Public Health Laboratory Directors)1984;42: 50-63.

4 Turek P, Kasper JT. A comparative evaluation of Micro-ID andAPI 20E for identification of Enterobacteriaceae. Abstracts ofthe Annual Meeting of the American Society for Microbiology1979:332.

5 Brandt JA, Eastvold DJ, Lewis JE. Clinical evaluation of threerapid methods for Enterobacteriaceae identification. Abstractsof the Annual Meeting of the American Society for Microbiol-ogy 1981:302.

16 Depase CA, Stone LL. A comparison study of the 4 hr Micro-IDand the 5 hr API 20E for the identification of the Enterobac-teriaceae. Abstracts of the Annual Meeting of the AmericanSociety for Microbiology 1981:302.

1 Shahidi A, Flejzor B. A comparison of API 20E (4 and 18 hoursincubation) Micro-ID and Entero-Set 20 for identification ofGram negative bacilli. Abstracts of the Annual Meeting of theAmerican Society for Microbiology 1979:333.

18 Gooch III WM, Hill GA. Comparison of Micro-ID and API 20Ein rapid identification of Enterobacteriaceae. J Clin Microbiol1982; 15: 885-90.

'9Cox NA, Mercuri AJ, Carson MO, Tanner DA. Comparativestudy of Micro-ID, Minitek and conventional methods withEnterobacteriaceae freshly isolated from foods. Journal ofFood Protection 1979;42:735-8.

20 Cox NA, Bailey JS, Thomson JE. Comparison of Micro-ID andMinitek-serology systems for rapid identification of Sal-

1137

on May 1, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.10.1132 on 1 October 1985. D

ownloaded from

1138

monella. Journal ofFood Protection 1981; 44:442-4.21 Bannister ER, Burdash NM, West ME. Evaluation of the four

hour Micro-ID method with the r/b system for theidentification of the Enterobacteriaceae. Abstracts of theAnnual Meeting of the American Society for Microbiology1979:333.

22 Bannister ER, West ME, Burdash NM. Comparison of Micro-IDand r/b for identification of Enterobacteriaceae. The PublicHealth Laboratory (Journal ofthe Conference ofPublic HealthLaboratory Directors) 1981;39:67-73.

23 Appelbaum PC, Shick SF, Kellogg JA. Evaluation of the four-hour Micro-ID technique for direct identification of oxidase-negative, Gram-negative rods from blood cultures. J Clin Mic-robiol 1980; 12:533-7.

24 Edberg SC, Clare D, Slater MH, Singer JM. Use of the Micro-I-D system for the rapid identification of Enterobacteriaceaefrom blood culture. Abstracts of the Annual Meeting of theAmerican Society for Microbiology 1977:67.

25 Edberg SC, Clare D, Moore MH, Singer JM. Rapid identificationof Enterobacteriaceae from blood cultures with the Micro-IDsystem. J Clin Microbiol 1979; 10: 693-7.

26 Malloy PJ, Ducate MJ, Schreckenberger PC. Comparison of fourrapid methods for identification of Enterobacteriaceae fromblood cultures. J Clin Microbiol 1983; 17:493-9.

27 Wellstood S. Comparison of Micro-ID, a modified Minitek sys-tem, and conventional methods for four-hour identification ofEnterobacteriaceae directly from blood cultures. Am J MedTechnol 1982;48:769-73.

28 Applebaum PC, Olmstead CC. Evaluation of Gram-stain screenand Micro-ID methods for direct identification of Enterobac-teriaceae from urines. Med Microbiol Immunol (Berl)1982; 170:173-84.

29 Kelly MT, Hale DC, Matsen JM. Rapid identification by theMicro-ID system of Enterobacteriaceae detected by urinescreening. J Clin Microbiol 1981; 14:295-7.

30 Burdash NM, West ME. Effects of refrigeration on inoculatedMicro-ID strips. Abstracts ofthe Annual Meeting ofthe Ameri-can Society for Microbiology 1981:303.

31 Burdash NM, West ME. Effects of refrigeration on inoculatedMicro-ID strips. J Clin Microbiol 1981; 14:350-1.

32 Hale DC, Lee KA. Delayed reading of Micro-ID. Abstracts ofthe Annual Meeting of the American Society for Microbiology1981:302.

33 Cox NA, Bailey JS, Thomson JE, Carson MO, Posey DA, RiveraE. Modified procedure to inoculate Micro-ID. Journal ofFoodProtection 1980;43:774-7.

Holmes, Humphry

Edberg SC, Melton E, Singer JM. Rapid biochemical character-ization of Haemophilus species by using the Micro-ID. J ClinMicrobiol 1980; 11: 22-6.

3 Rhoden DL, Esaias AO, Smith PB. Evaluation of the DMSRapid E system for the identification of Enterobacteriaceaestrains. Abstracts ofthe Annual Meeting ofthe American Soci-ety for Microbiology 1984:241.

36 Dickgiesser N, Groth-Tittmann I. Identification of Enterobac-teriaceae by using Rapid 20E, ID-System, Enterotube II andAPI 20E. Zentralbl Bakteriol Mikrobiol Hyg[A]1984;256:412.

37 Izard D, Husson MO, Vincent P, Leclerc H, Monget D, Boeuf-gras JM. Evaluation of the four-hour Rapid 20E system foridentification of members of the family Enterobacteriaceae. JClin Microbiol 1984; 2051-4.

38 Keville MW, Doern GV. Evaluation of the DMS Rapid E systemfor identification of clinical isolates of the family Enterobac-teriaceae. J Clin Microbiol 1984; 20: 1010-1.

39 Murray PR, Gauthier A, Niles A. Evaluation of the Quantum IIand Rapid E identification systems. J Clin Microbiol1984;20:509-14.

40 Thomas B, Gayral JP, Monget D. A new 4-hour identificationsystem for Enterobacteriaceae: Rapid 20E. Progr Abstr XIIIInt Congr Microbiol, Boston, Mass 1982:67.

41 Altwegg M. Performance of two four-hour identification systemswith atypical strains of Enterobacteriaceae. Eur J Clin Mic-robiol 1983;2:529-33.

42 Mounier M, Denis F. Four-hour direct identification ofEnterobacteriaceae in blood cultures. Eur J Clin Microbiol1983;2:593-5.

43 Lapage SP, Bascomb S, Willcox WR, Curtis MA. Identificationof bacteria by computer: general aspects and perspectives. JGen Microbiol 1973;77:273-90.

4Bascomb S, Lapage SP, Curtis MA, Willcox WR. Identificationof bacteria by computer: identification of reference strains. JGen Microbiol 1973;77:291-315.

4 D'Amato RF, Holmes B, Bottone EJ. The systems approach todiagnostic microbiology. CRC Crit Rev Microbiol 1981;9: 1-44.

Requests for reprints to: Mr B Holmes, NCTC, CentralPublic Health Laboratory, 61 Colindale Avenue, LondonNW9 5HT, England.

on May 1, 2020 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.38.10.1132 on 1 October 1985. D

ownloaded from