Yeasts and molds can cause various degrees of food decomposition.

Transcription

1 296 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, 2003 FOOD BIOLOGICAL CONTAMINANTS Enumertion of Totl Yests nd Molds in Foods by the SimPlte Yest nd Mold Color Indictor Method nd Conventionl Culture Methods: Collbortive Study PHILIP T. FELDSINE, ANDREW H. LIENAU, STEPHANIE C. LEUNG, nd LINDA A. MUI BioControl Systems, Inc., SE 32nd St, Bellevue, WA Collbortors: S. Al-Hsni; G. Anderson; A. Ashmore; P. Audet; D. Brhm; C. Beulieu; F. Cook; M. Cornec; R. Dvis; A. Dens; S. DeLncey; L. DeWinter; M. Divine; S. Drocco; M. Feist; D. Hickson; F. Humbert; L. Humes; J. Jckson; R. Klinowski; J. Knickerbocker; M. Lisby; C. McDonld; J. Terry; K. Thmmsouk; E. Tuncn; A. Vini; D. Vrn; W. Wrren; S. White The reltive effectiveness of the SimPlte Yest nd Mold Color Indictor method (Y&M CI) ws compred to the U.S. Food nd Drug Administrtion s (FDA) Bcteriologicl Anlyticl Mnul (BAM) method nd the proposed Interntionl Orgniztion for Stndrdiztion (ISO) method, ISO/CD 21527, for enumerting yests nd molds in foods. Test portions were prepred nd incubted ccording to the conditions stted in both the BAM nd ISO methods. Six food types were nlyzed: frozen corn dogs, nut mets, frozen fruits, cke mix, cerel, nd fresh cheese. Nut mets, frozen fruits, nd fresh cheese were nturlly contminted. All other foods were rtificilly contminted with either yest or mold. Seventeen lbortories throughout North Americ nd Europe prticipted in the study. Three method comprisons were conducted. In generl, there ws <0.3 men log count difference in recovery between the SimPlte method nd the 2 corresponding reference methods. Moreover, men log counts between the 2 reference methods were lso very similr. The repetbility (s r ) nd reproducibility (s R ) stndrd devitions were comprble between the 3 method comprisons. These results indicte tht the BAM method nd the SimPlte method re equivlent for enumerting yest nd mold popultions in foods. Similrly, the SimPlte method is comprble to the proposed ISO method when test portions re prepred nd incubted s defined in the proposed ISO method. Submitted for publiction December The recommendtion ws pproved by the Methods Committee on Microbiology nd Extrneous Mterils s First Action. See Officil Methods Progrm Actions, (2003) Inside Lbortory Mngement, Mrch/April issue. Corresponding uthor s e-mil: ptf@biocontrolsys.com. Yests nd molds cn cuse vrious degrees of food decomposition. Their bility to proliferte in wide rnge of ph nd tempertures, nd in foods of vrious wter contents, mke it necessry to detect fungl spoilge t the erly stges. Current methods for enumertion of yests nd molds include the U.S. Food nd Drug Administrtion s (FDA) Bcteriologicl Anlyticl Mnul (BAM) culture method (1) nd the Interntionl Orgniztion for Stndrdiztion (ISO) method (2). Both methods use gr pltes tht re incubted for 5 7 dys ( h) before obtining results. The SimPlte Yest nd Mold Color Indictor (Y&M CI) method uses Binry Detection Technology (BDT) to enumerte yests nd molds in foods fter minimum of 56 h of incubtion, thereby offering significnt timesving dvntge over the current culture methods. The SimPlte device lso elimintes the uncertinty in enumertion of plte counts becuse of fctors such s overcrowding, spreding colony types, nd food prticulte interference on conventionl plte methods. In the SimPlte multiple test formt, prepred food test portions re plced onto the center of SimPlte device, nd Y&M CI liquid medium is then dded. For the single test formt, premixed test portion/medium homogente is dispensed into the test device. The test portion/medium homogente is distributed into fixed number of individul incubting wells: 84 for the norml counting rnge SimPlte test nd 198 for the high counting rnge. With the SimPlte method, foodborne microorgnisms re suspended in nutritionlly defined growth medium. To enhnce performnce, medium supplements re provided for use with certin food mtrixes. These supplements blnce certin inherent ttributes of some mtrixes such s incresed cidity found in foods contining high levels of vitmin C. Discrete liquots re seprtely comprtmentlized nd isolted from ech other in the incubting wells where biochemicl ctivities of vible microorgnisms re monitored in liquid environment. Detection by this biochemicl process requires fewer microorgnisms to produce detectble signl in SimPlte well thn the number required to form clerly visible colony on n gr plte. Enumertion is mesured by

2 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, Tble 1. SimPlte conversion tble No. of positive wells = popultion per plte 1=2 31=76 61=216 2=4 32=80 62=224 3=6 33=84 63=232 4=8 34=86 64=240 5=10 35=90 65=248 6=12 36=94 66=256 7=14 37=96 67=266 8=16 38=100 68=276 9=18 39=104 69= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = > = = 178 If there re no positive 26=62 56=184 wells nd the sponge is positive, then the 27=64 57=190 popultion is 1 28 = = 196 If there re no positive 29=70 59=202 wells nd the sponge is negtive, then the 30=74 60=208 popultion is <1 The popultion reflects the number of microorgnisms per plte. To determine the number of microorgnisms per grm (ml) food product, see I (Reding nd Interprettion of Results). simple binry rection; ech well is either positive or negtive. Any color chnge from the originl bckground color in ech well or in the sponge is interpreted s positive rection. Yests nd molds re enumerted by counting the numbers of wells in ech plte tht exhibit color chnge (positive wells) fter incubtion. The finl count per plte is derived from Tble 1 tht is bsed upon the principle of the Poisson distribution. The lrge number of wells llows n ccurte mesure of the true microbil popultion in test smple. A recent study compred the SimPlte Y&M CI method nd BAM method for recovery nd quntifiction of yests nd molds in 20 different foods (publiction pending). Overll, the results indicted tht the SimPlte Y&M CI method is equivlent to the BAM reference method. This report describes seprte interntionl multilbortory collbortive study compring the reltive effectiveness of the SimPlte Y&M CI method to the BAM plte method for recovery of yests nd molds. The SimPlte method nd the proposed ISO culture method were evluted in this sme study. Collbortive Study Design of Study Six food types, representtive of wide rnge of food ctegories, were evluted: frozen corn dogs (frnkfurter encsed in cornbred), frozen fruits, cerel, fresh cheese, cke mix, nd nut mets. Frozen fruits, fresh cheese, nd nut mets were nturlly contminted. For these foods, 2 test portions from 3 different lots were nlyzed for ech food type (totl of 6 test portions per food type). Cerel, cke mix, nd frozen corn dogs were rtificilly contminted with either yest or mold. Inocultion levels in foods rtificilly contminted were s follows: high, pproximtely ; medium, 1000; low 100 fungi/g of food. Foods tht required Dichlorn rose bengl gr (DRBC) spred pltes s the reference culture method were rtificilly contminted t these inocultion levels: high, pproximtely ; medium, ; low 1000 (minimum detection level by the BAM method) fungi/g of food. Two test portions from 3 different inocultion levels nd uninoculted controls were nlyzed for ech rtificilly inoculted food type (totl of 8 test portions per food type). Collbortors were sent 2 sets of rndomized test portions; one set ws used for nlysis by the BAM culture method, using 0.1% peptone diluent, nd the corresponding SimPlte method (SIM B). The other set ws used for nlysis by the ISO culture method, using peptone slt solution, nd the corresponding SimPlte method (SIM I). The cerel, cke mix, nd nut mets were stored t room temperture until the dy of nlysis. Fresh cheeses were refrigerted; frozen fruits nd corn dogs were kept frozen ( 20 C) until the dy of nlysis. For one set of test portions, collbortors were instructed to prepre the initil suspension nd further deciml dilutions s recommended by BAM (1). The pproprite dilutions were nlyzed by the BAM plte method nd the SimPlte method (SIM B), both incubted t 25 C. Two types of gr pltes were used for the reference method. Dichlorn 18 glycerol (DG18) gr ws used when the wter ctivity of the food ws <0.95; otherwise, DRBC gr ws used for ll other food types. For the second set of test portions, collbortors prepred the initil suspension nd further deciml dilutions ccording to ISO 6887 (3). The pproprite dilutions were nlyzed ccording to ISO/CD (2) culture method (DG18 gr) nd the SimPlte method (SIM I), both incubted t 25 C. The food types with higher wter ctivity ( w > 0.95) were lso tested with DRBC gr pltes. An overview of the interlbortory study design is shown in Figure 1. Results obtined were submitted on summry forms with the pproprite

3 298 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, g Test portion 450 ml 0.1% peptone wter g Test portion 450 ml peptone slt diluent 2. Homogenize test portion 2. Homogenize test portion 3. Perform pproprite dilutions 3. Perform pproprite dilutions SimPlte test FDA s BAM SimPlte test ISO 1.0 ml Diluted smple onto device * 1.0 ml Diluted smple onto triplicte 1.0 ml Diluted smple onto device 1.0 ml Diluted smple onto duplicte + 9 ml SimPlte Y&M CI medium Petri pltes ml DG18 gr + 9 ml SimPlte Y&M CI medium Petri pltes + 15 ml DG18 gr OR spred 0.1 ml diluted smple AND, only for foods with w > 0.95, onto triplicte DRBC gr pltes spred 0.1 ml diluted smple onto duplicte DRBC gr pltes C for 56 h minimum in drk 25 C for 5 7 dys in drk C for 56 h minimum in drk C for 5 dys 1. Count number of positive wells 1. Count colonies on gr pltes 1. Count number of positive wells 1. Count colonies on gr pltes 2. Use SimPlte conversion tble to 2. Fctor in dilution for finl CFU/g 2. Use SimPlte conversion tble to 2. Fctor in dilution for finl CFU/g determine totl number of fungi per plte determine totl number of fungi per plte 3. Fctor in dilution for the totl 3. Fctor in dilution for the totl number of fungi per grm number of fungi per grm Figure 1. SimPlte Y&M CI, BAM, nd ISO methods for enumertion of yests nd molds. * = Spred or pour plte ws selected bsed on the wter ctivity of the food nlyzed. rw dt. A minimum of 8 lbortories submitted vlid dt for ech food type. Preprtion of Inocul nd Test Portions Frozen corn dogs, cke mix, nd cerel test portions were inoculted with Cndid lbicns, Torulspor delbrueckii, nd Aspergillus clvtus, respectively. These fungi were chosen becuse they re commonly isolted from food products. Test cultures were grown on potto dextrose gr (PDA) pltes t 25 C for 3 7 dys. Growth from these pltes ws resuspended into Butterfield s phosphte buffered dilution wter (BPBD). This suspension ws diluted nd used to rtificilly contminte foods. For cke mix, C. lbicns ws grown on PDA pltes t 25 C for minimum of 3 dys. The vible colonies were collected, wshed, nd resuspended into 10% nonft dry milk for freeze-drying. Lyophilized cultures were stored t 20 C until used. Inoculted test portions were prepred t lest 1 week before nlysis to llow popultions to stbilize. The nturlly contminted foods were purchsed t the retil level in Settle, WA. The bulk foods were thoroughly mixed nd then pckged into 50 g test portions for shipment to collbortors. Test Portion Distribution Ech collbortor received 2 sets of 6 8 test portions for ech week the study ws conducted. Test products were distributed by n overnight delivery from the United Sttes to lbortories throughout North Americ nd Europe. Frozen foods were shipped on dry ice; cerel, cke mix, nd nut met test portions were shipped t room temperture. Fresh cheese ws shipped on ice. Test Portion Anlysis The BAM nd ISO culture methods use different diluents for the initil suspension nd deciml dilutions of test portions. This required collbortors to set up seprte pired smples with one test portion prepred with the BAM diluent nd the other with the ISO diluent. Anlysts were instructed to follow both BAM nd ISO reference procedures without devition. Designted food test portions were prepred s specified by BAM (8th Ed., Rev. A, Ch. 18; 1). For ll food types including nut mets, 50 g test portions were dded to 450 ml 0.1% peptone wter, nd homogenized by stomching for 2 min. Frozen test portions were dded directly to the diluent (unthwed) nd then stomched to obtin liquid homogente. After homogeniztion, suspensions were serilly diluted into 0.1% peptone wter. The number of dilutions necessry to perform the test depended on the food type being nlyzed. Subsequent 10-fold dilutions were prepred by dding 10 ml of the previous dilution to 90 ml sterile diluent nd shken 25 times in 30 cm rc. BioControl provided the suggested dilution series for ech food type the week before nlysis ws initited. For low moisture food types (cke mix, cerel, nd nut mets), 1 ml from ech pproprite dilution of the test portion ws trnsferred to sterile Petri dish in triplicte. A 20 ml mount of DG18 gr, tempered to C, ws dded nd swirled to evenly distribute the test dilution nd gr. Plte counts of ech dilution level were performed in triplicte. For high moisture food types (frozen corn dog, frozen fruits, nd cheese), 0.1 ml from ech pproprite dilution of the test portion ws trnsferred to sterile pre-poured DRBC dish in triplicte nd spred cross the surfce of the plte. Plte counts of ech dilution level were performed in triplicte. Frozen fruits were spred plted on DRBC gr pltes contining chlortetrcycline. All pltes were incubted in the drk t 25 C for 5 7 dys. The number of fungi present on ech plte ws enumerted fter the required incubtion. Dt were recorded onto the pproprite worksheets. For the BAM method, pltes contining colonies were counted. Designted food test portions were lso prepred s specified by ISO method For ll food types, 50 g test portions were dded to 450 ml peptone slt solution nd homogenized by mcerting for 1 min. After homogeniztion, liquots were serilly diluted into peptone slt solution. The

4 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, number of dilutions necessry to perform the test depended on the food type being nlyzed. Subsequent 10-fold dilutions were prepred by dding 10 ml of the previous dilution to 90 ml sterile diluent nd mixed well. BioControl provided the suggested dilution series for ech food type the week before nlysis ws initited. For low moisture food types (cke mix, cerel, nd nut mets), 1 ml from ech pproprite dilution of the test smple ws trnsferred to sterile Petri dish in duplicte. Approximtely 15 ml DG18 gr, tempered to C, ws dded nd swirled to evenly distribute test liquot nd gr. Plte counts of ech dilution level were performed in duplicte. For high moisture food types (frozen corn dog, frozen fruits, nd cheese), 0.1 ml from ech pproprite dilution of test smple ws trnsferred to sterile pre-poured DRBC plte in duplicte nd spred cross the surfce of the plte. Frozen fruits were spred plted on DRBC gr pltes contining chlortetrcycline. Plte counts of ech dilution level were performed in duplicte. All pltes were incubted t 25 C for 5 7 dys. The number of fungi per plte were enumerted fter the required incubtion. Dt were recorded onto the pproprite worksheets. Diluted test portions prepred for the BAM nd ISO reference methods were lso used for SimPlte nlyses. One set of SimPlte devices (SIM B) ws set up with dilutions from the BAM method. A seprte set of SimPlte devices (SIM I) ws set up with dilutions from the ISO method. A1mL mount of diluted test portion ws trnsferred to the center of the SimPlte device. The device ws overlyed with 9 ml rehydrted Y&M CI medium, for finl volume of 10 ± 0.2 ml. After the plte ws covered with the lid, the test smple/medium homogente ws mixed nd swirled into the wells. Excess liquid ws decnted into the device collection sponge. All SimPlte devices were incubted in the drk t 25 C for minimum of 56 h. After incubtion, the number of wells with color chnge from the originl bckground color were counted. Tble 1 ws used to clculte the yest nd mold popultion per SimPlte device. The number of positive wells counted ws mtched to the corresponding popultion given in Tble 1. If dilution ws mde to the originl food test portion before inocultion of the SimPlte device, then the fungi per grm of food ws clculted by multiplying the popultion from Tble 1 by the dilution fctor. Sttisticl Anlysis For ech lot of food, triplicte (duplicte for ISO) pltes were verged nd reported s colony forming units (CFU)/g for the BAM or ISO reference methods nd the SimPlte methods. For the BAM method, if none of the pltes for test portion hd t lest 10 colonies, the exct counts from the lest dilute test smple were verged. The bse 10 logrithms (log 10 ) of SimPlte counts nd CFU/g for ech of the reference culture methods were used for sttisticl nlysis of the test portions. In few instnces, indefinite vlues were reported for dt vlues for the test portions. To perform sttisticl nlysis on these prticulr dt vlues, the indefinite sign ws truncted nd method performnce ws determined (4). Repetbility (s r ) nd reproducibility (s R ) stndrd devitions, reltive stndrd devitions of repetbility (RSD r ) nd reproducibility (RSD R ), nd repetbility (r) nd reproducibility (R) vlues were clculted ccording to the methods of Youden nd Steiner (5) fter the Cochrn test ws pplied to eliminte outliers. The Cochrn test ws used to remove lbortories showing significntly greter vribility mong replicte nlyses thn other lbortories for prticulr set of test portions. An F sttistic tht computes the rtio of the 2 vrinces ws used to compre repetbility nd reproducibility vrinces (6, 7). Men responses between the 2 methods were compred by using 2-smple (pired) t-test (6, 7). AOAC Officil Method Detection nd Quntifiction of Yests nd Molds in Foods SimPlte Yest nd Mold Color Indictor (Y&M CI) Method First Action 2002 (Applicble to detection nd quntifiction of yests nd molds in chocolte, cke mix, spices, nut mets, diry foods, red mets, poultry mets, sefoods, fermented mets, frozen corn dogs, cerel, pst, egg products, flour, prepckged fresh sld, frnkfurters, vegetbles, fruits, nd fruit juice.) Cution: Test portion dilutions nd incubted SimPlte devices from food products could contin pthogenic fungi if the prticulr test portion ws so contminted. Use stndrd septic microbiologicl lbortory technique, including decontmintion of ny spills with disinfectnt. See Tbles A C for the results of the interlbortory study supporting cceptnce of the method. A. Principle SimPlte Yest nd Mold Color Indictor (Y&M CI) method is used to detect nd quntify yest nd mold popultions. The Y&M CI medium nd test smple mixture is dispensed into SimPlte device nd incubted for minimum of 56 h. The medium chnges color in the presence of yests nd molds. The yest nd mold count is determined by counting the wells with chnged color nd referring to Tble 1. B. Medi nd Regents () Dehydrted Y&M CI medium. Individully pckged single or multiple test formt with supplement(s), vilble from BioControl Systems, Inc. (12822 SE 32nd St, Bellevue, WA 98005, (b) SimPlte devices. In pcks of 20 (BioControl). (c) Peptone wter diluent, 0.1% (used in BAM method). Dissolve 1.0 g peptone in 1 L deionized wter. Autoclve t 121 C for 15 min. Finl ph is 7.0 ± 0.2. (d) Peptone slt solution (used in ISO method). Use 1.0 g enzymtic digest of csein nd 8.5 g sodium chloride. Suspend ingredients in 1 L deionized wter. Autoclve t 121 C for 15 min. Finl ph is 7.0 ± 0.2.

5 300 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, 2003 C. Apprtus () Incubtor. Mintining 25 C in drk environment. (b) Micropipetor. Accurtely dispensing 0.1 nd 1.0 ml. (c) Pipets. Glss or plstic sterile pipets, 1 ml with 0.01 ml grdutions; nd 10 ml with 0.1 ml grdutions. (d) Stomcher/mstictor. IUL Instruments (Cincinnti, OH), or equivlent, for mcerting test portions. D. Generl Instructions Do not use expired medium. Store reconstituted medium between 15 nd 25 C in the drk nd use within 12 h. Dispose of medium in decontmintion continer, nd sterilize before discrding. E. Test Suspension Preprtion () Weigh 50 g test portion into 450 ml sterile diluent [e.g., 0.1% peptone wter (BAM method) or peptone slt solution (ISO method)]. This is 1:10 dilution. Mcerte mixture. (b) If lterntive test portion size is specified in testing procedure, prepre 10% (w/v) suspension. (c) If necessry, prepre 10-fold seril dilutions pproprite for nticipted popultion of test portion. F. Y&M CI Test Procedure, Single Test Medium For 1.0 ml test suspension. () Suspend powdered medium with 9.0 ml sterile deionized wter contining 1 ml Supplement A per 100 ml wter. Add 1.0 ml prepred test portion nd mix well. Do not count this reconstitution s dilution. For 0.1 ml test suspension. (b) Suspend powdered medium with 9.9 ml sterile deionized wter contining 1 ml Supplement A per 100 ml wter. Add 0.1 ml prepred test portion nd mix well. This is n dditionl 1:10 dilution from E. The finl volume of test portion/medium mixture in the continer should be ml. Note: For rw mets nd spices, dd 0.1 ml Supplement M to hydrted medium. For undiluted fruit juice contining vitmin C or dry pet food, dd 0.1 ml Supplement V. For foods contining spreding mold popultions, dd 0.1 ml Supplement D. Supplements re vilble from BioControl Systems, Inc. The finl volume of the test portion/medium mixture in the continer should be 10 ± 0.2 ml. (c) Remove lid from the SimPlte device nd trnsfer test portion/medium mixture onto center of plte. Immeditely replce lid. Continue with H. G. Y&M CI Test Procedure, Multiple Test Medium () Empty contents of one continer, B(), into 100 ml sterile deionized wter contining 1 ml Supplement A per 100 ml wter. Shke to completely dissolve. Note: For rw mets nd spices, dd 1.0 ml Supplement M to hydrted medium. For undiluted fruit juice contining vitmin C or dry pet food, dd 1.0 ml Supplement V. For foods contining spreding mold popultions, dd 1.0 ml Supplement D. Supplements re vilble from BioControl Systems, Inc. (b) Remove lid from SimPlte device. Pipet prepred test suspension onto center of plte. If prepred test suspension size is 1.0 ml, overly test suspension with 9.0 ml medium. Do not count this medi ddition s dilution. (c) For 0.1 ml of prepred test portion, overly with 9.9 ml medium; this is n dditionl 1:10 dilution of test suspension from E. The finl volume of test portion/medium mixture on the plte should be 10 ± 0.2 ml. Immeditely replce lid. Continue with H. H. Test Procedure for Single nd Multiple Tests () Gently swirl to distribute test portion/medium mixture into ll wells. Hold plte with both hnds tilted slightly to help distribute medium into wells. (b) Pour off excess medium by holding lid ginst plte on either side of sponge cvity. Tip plte towrd you to llow liquid to drin into sponge. Observe bckground color of wells. Bckground is defined s color of test portion/medium mixture inside wells. (c) Do not invert the SimPlte device. Incubte t room temperture (22 25 C) for h in the drk. If test portion being nlyzed is known to contin slow-growing fungl popultions, such s certin mold gener, incubtion time my be extended to totl of 72 h. If Supplement V is used, incubte SimPlte devices for 72 h in the drk. I. Reding nd Interprettion of Results () If reding is tken between 56 nd 63 h fter inocultion, plce SimPlte device on illuminted bckground to help visulize positive results. After 63 h this is not necessry. After incubtion, observe color chnge of liquid in wells. Disregrd prticulte mtter if present. Count number of wells showing color chnge from bckground color. The common color chnges produced by microorgnisms re ornge, pech, red, nd white. Note: For mixtures contining Supplement V, count only the number of wells tht fluoresce blue by holding UV light (365 nm wvelength) pproximtely 5 cm (2 in.) bove the SimPlte device. Do not count nonfluorescent wells tht only exhibit color chnge. (b) To determine the popultion, clculte s follows: (1) Count the number of positive wells on the plte; (2) use Tble 1 to determine the totl popultion per plte. (c) To clculte the number of fungi/g, multiply the number in I(b)(2) by the pproprite dilution fctor (see E nd F for single test or E nd G for multiple test). Ref.: J. AOAC Int. 86, (2003) Results The SimPlte Y&M CI method ws compred to the BAM method nd the proposed ISO method (ISO/CD 21527) for enumertion of yests nd molds in foods. Test portions were prepred nd incubted ccording to the conditions stted in both the BAM nd ISO methods. Seventeen lborto-

6 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, Tble A. Sttisticl nlysis of interlbortory results for totl fungi recovery by the BAM culture method nd the SimPlte Y&M CI method Men b S r e RSD r, % f r g S R h RSD R, % i R j Food group Lot/level N BAM c SIM d BAM c SIM d BAM c SIM d BAM c SIM d BAM c SIM d BAM c SIM d BAM c SIM d Frozen corn dog Low Medium High Frozen fruit Rspberries Blckberries Strwberries Cerel Low k Medium High Cheese Blue k Mozzrell k Cheddr k l Cke mix Low Medium High m Nut mets Hzelnut Pecn n l Wlnut n 0.27 k o m Number of lbortories with vlid dt. Men log count of totl fungi/g. U.S. Food nd Drug Administrtion s Bcteriologicl Anlyticl Mnul culture method. SimPlte Yest nd Mold Color Indictor method. Repetbility stndrd devition. Repetbility reltive stndrd devition. Repetbility vlues, 2.8 S r. Reproducibility stndrd devition. Reproducibility reltive stndrd devition. Reproducibility vlues, 2.8 S R. Significntly different repetbility p < 0.05; sttisticlly not different p < Significntly different reproducibility p < Significntly different men log counts p < 0.05; sttisticlly not different p < Significntly different men log counts p < Significntly different reproducibility p < 0.05; sttisticlly not different p < b c d e f g h i j k l n o

7 302 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, 2003 Tble B. Sttisticl nlysis of interlbortory results for totl fungi recovery by the ISO culture method nd the SimPlte Y&M CI method Men b S r e RSD r, % f r g S R h RSD R, % i R j Food group Lot/level N ISO c SIM d ISO c SIM d ISO c SIM d ISO c SIM d ISO c SIM d ISO c SIM d ISO c SIM d Frozen corn dog Low k Medium High Frozen fruit Rspberries Blckberries Strwberries Cerel Low l Medium High m Cheese Blue n o Mozzrell n 0.07 m o Cheddr p Cke mix Low o Medium k High k Nut mets Hzelnut Pecn k p Wlnut n m Number of lbortories with vlid dt. Men log count of totl fungi/g. Interntionl Orgniztion for Stndrdiztion culture method. SimPlte Yest nd Mold Color Indictor method. Repetbility stndrd devition. Repetbility reltive stndrd devition. Repetbility vlues, 2.8 S r. Reproducibility stndrd devition. Reproducibility reltive stndrd devition. Reproducibility vlues, 2.8 S R. Significntly different men log counts p < 0.05; sttisticlly not different p < Significntly different repetbility p < 0.05; sttisticlly not different p < Significntly different repetbility p < Significntly different men log counts p < Significntly different reproducibility p < 0.05; sttisticlly not different p < Significntly different reproducibility p < b c d e f g h i j k l n o p

8 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, Tble C. Sttisticl nlysis of interlbortory results for totl fungi recovery by the ISO method nd the BAM method Men b S r e RSD r, % f r g S R h RSD R, % i R j Food group Lot/level N BAM c ISO d BAM c ISO d BAM c ISO d BAM c ISO d BAM c ISO d BAM c ISO d BAM c ISO d Frozen corn dog Low Medium High Frozen fruit Rspberries Blckberries Strwberries Cerel Low Medium k High l Cheese Blue Mozzrell Cheddr Cke mix Low Medium High Nut mets Hzelnut Pecn l Wlnut Number of lbortories with vlid dt. Men log count of totl fungi/g. U.S. Food nd Drug Administrtion s Bcteriologicl Anlyticl Mnul culture method. Interntionl Orgniztion for Stndrdiztion culture method. Repetbility stndrd devition. Repetbility reltive stndrd devition. Repetbility vlues, 2.8 S r. Reproducibility stndrd devition. Reproducibility reltive stndrd devition. Reproducibility vlues, 2.8 S R. Significntly different reproducibility p < 0.05; sttisticlly not different p < Significntly different repetbility p < 0.05; sttisticlly not different p < b c d e f g h i j k l

9 304 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, 2003 ries throughout North Americ nd Europe prticipted in the study (Tble 2). Eight lbortories nlyzed ll 6 food types, 5 lbortories nlyzed 5 food types, one nlyzed 3 food types, nd 3 nlyzed one food type (Tble 2). The numbers of fungi per grm of food (reported in log 10 vlues) recovered from individul test portions re presented in Tbles 3 8. Repetbility nd reproducibility nlyses re presented in Tbles A C. Frozen Corn Dogs Frozen corn dog test portions inoculted t low, medium, nd high levels were nlyzed (Tble 3). Uninoculted controls were included on the dy of nlysis. Thirteen lbortories prticipted in the nlysis of frozen corn dogs. Lbortories 6, 7, 10, 11, nd 17 did not follow study instructions. The dt from these lbortories were excluded from sttisticl nlysis. Dt generted by the BAM method nd the corresponding SimPlte method (SIM B) were compred sttisticlly. Eight lbortories submitted vlid dt for ll 3 levels of frozen corn dogs. Men log counts recovered from low, medium, nd high level test portions were not sttisticlly different between the 2 methods. The repetbility (s r ) nd reproducibility (s R ) stndrd devitions (SDs) of the 2 methods were sttisticlly nlyzed (Tble A). The RSD r nd RSD R vlues were not sttisticlly different for ll 3 levels between the 2 methods (Tble A). Dt generted by the ISO method nd the corresponding SimPlte method (SIM I) were compred sttisticlly. Eight lbortories submitted vlid dt for ll 3 levels of frozen corn dogs. Men log counts recovered from medium nd high levels test portions were not sttisticlly different between the 2 methods. The SimPlte method recovered higher men log counts thn did the ISO method for the low level (p < 0.05). However, the men counts were very similr, with <0.3 log units difference between the 2 methods. The RSD r nd RSD R vlues were not sttisticlly different between the 2 methods for ll 3 levels (Tble B). Dt generted by the BAM nd ISO methods were compred sttisticlly. Eight lbortories submitted vlid dt for ll 3 levels. Men log counts recovered from ll 3 levels were not sttisticlly different between the 2 methods. The RSD r nd RSD R vlues were not sttisticlly different for ll 3 levels between the 2 methods (Tble C). Frozen Fruits Three types of frozen fruit (strwberries, blckberries, nd rspberries) were nlyzed (Tble 4). These 3 fruits were chosen becuse they re known to contin significnt levels of fungi contmintion. Lbortory 12 did not complete nlysis of test portions. The dt from this lbortory were excluded from sttisticl nlysis. Dt generted by the BAM method nd the SimPlte method (SIM B) were compred sttisticlly. Twelve lbortories submitted vlid dt for frozen fruit test portions. The men log counts from the BAM nd the SIM B methods were not sttisticlly different for the 3 lots nlyzed. The RSD r nd Tble 2. Collbortor prticiption for SimPlte Y&M Cl interlbortory study by food type Lb Frozen corn dog Frozen fruit Cerel Cheese Cke mix Nut mets 1 Y Y Y Y Y Y 2 Y Y Y Y Y Y 3 Y Y Y Y Y Y 4 Y Y Y Y Y Y 5 N Y Y Y Y Y 6 Y b Y Y Y Y Y 7 Y b Y Y Y Y Y 8 N Y Y Y Y Y 9 N Y Y b Y Y c Y 10 Y b Y Y Y Y N 11 Y b Y Y Y Y Y 12 Y Y d Y d Y N Y 13 Y Y Y Y Y Y 14 N N N Y Y Y 15 Y N N N N N 16 Y N N N N N 17 Y b N N N N N Totl e b c d e Y = Collbortor nlyzed this food type; N = collbortor did not nlyze this food type. Lbortory did not follow study instructions. Results were not included in the sttisticl nlysis for the designted food types. Lbortory did not follow the SimPlte directions for use. Results were not included in the sttisticl nlysis for the designted food types. Lbortory did not complete/strt nlysis. Results were not included in the sttisticl nlysis for the designted food types. Totl number of lbortories prticipting in the nlysis of this food type. RSD R vlues for the frozen fruits test portions were not sttisticlly different between the 2 methods (Tble A). Dt generted by the ISO method nd the SimPlte method (SIM I) were compred sttisticlly. Lbortory 3 ws determined to be n outlier by the Cochrn test for the blckberry test portions. This pired dt set ws excluded from sttisticl nlysis. Twelve lbortories submitted vlid dt for the strwberry nd rspberry test portions, nd 11 lbortories submitted vlid dt for blckberry test portions. The men log counts from the BAM nd the SIM I methods were not sttisticlly different for the 3 lots nlyzed. The RSD r nd RSD R vlues for the frozen fruits test portions were not sttisticlly different between the 2 methods (Tble B). Dt generted by the BAM nd ISO methods were compred sttisticlly. Twelve lbortories submitted vlid dt for ll 3 lots of frozen fruits test portions. Men log counts recovered from these 3 lots were not sttisticlly different between the 2 methods. The RSD r nd RSD R vlues were not

10 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, sttisticlly different for ll 3 lots between the 2 methods (Tble C). Cerel Cerel test portions inoculted t low, medium, nd high level were nlyzed (Tble 5). Uninoculted controls were included on the dy of nlysis. Lbortory 9 did not follow the SimPlte directions for use. Lbortory 12 did not complete nlysis of the test portions. The dt from these lbortories were excluded from sttisticl nlysis. Throughout the nlysis of cerel test portions, some lbortories reported plte counts for certin test portions tht were below the suitble counting rnge for the BAM method. These dt were estimted nd used for sttisticl nlysis. Dt generted by the BAM nd SimPlte methods were compred sttisticlly. Eleven lbortories submitted vlid dt for ll 3 levels. The men log counts from the BAM nd SIM B methods were not sttisticlly different for ll 3 levels nlyzed. The RSD r vlues from the BAM nd SIM B methods were not significntly different for the medium nd high test portions. The BAM method s dt reported lower RSD r for the low test portions (p < 0.05). The RSD R vlues from the dt for the BAM nd SIM B methods were not sttisticlly different for the ll 3 levels. Dt generted by the ISO nd SimPlte methods were compred sttisticlly. Lbortories 11 nd 4 were determined to be outliers by the Cochrn test for the low nd high test portions, respectively. These pired dt were excluded from sttisticl nlysis. Eleven lbortories submitted vlid dt for the medium level nd 10 lbortories submitted vlid dt for the low nd high levels. The men log counts from the ISO nd SIM I methods were not sttisticlly different for ll 3 levels nlyzed. The RSD r vlues from the ISO nd SIM I methods were not significntly different for the medium level test portions. The ISO method s dt showed lower RSD r for the low nd high test portions. The RSD R vlues from the ISO nd SIM I methods were not sttisticlly different for the ll 3 levels (Tble B). Dt generted by the BAM nd the ISO methods were compred sttisticlly. Lbortory 11 ws determined to be n outlier by the Cochrn test for the low level test portions. These pired dt were excluded from sttisticl nlysis. Ten lbortories submitted vlid dt for the low level nd 11 lbortories submitted vlid dt for the medium nd high levels. Tble 3. Yest nd mold counts for frozen corn dog test portions (log 10 CFU/g) by SimPlte Y&M CI (SIM B nd I), BAM culture method, nd ISO culture method Uninoculted Low Test portion 1 Test portion 2 Test portion 3 Test portion 4 Lb SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO 1 <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 < <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 < <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 < <2.00 < <2.00 <2.00 < < < < < < <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 < <2.00 <2.00 <2.00 < < < <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 <2.00 < Medium High Test portion 5 Test portion 6 Test portion 7 Test portion 8 SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B = Test portions prepred ccording to the FDA s BAM method; SIM I = test portions prepred ccording to the proposed ISO method.

11 306 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, 2003 Tble 4. Yest nd mold counts for frozen fruits test portions (log 10 CFU/g) by SimPlte Y&M CI (SIM B nd I), BAM culture method, nd ISO culture method Rspberries Test portion 1 Test portion 2 Lb SIM B BAM SIM I ISO SIM B BAM SIM I ISO Blckberries Test portion 3 Test portion 4 SIM B BAM SIM I ISO SIM B BAM SIM I ISO b 5.72 b b 3.72 b Strwberries Test portion 5 Test portion 6 SIM B BAM SIM I ISO SIM B BAM SIM I ISO b SIM B = Test portions prepred ccording to the FDA s BAM method; SIM I = test portions prepred ccording to the proposed ISO method. Outlier; dt not used in nlysis for method comprison of SIM I nd ISO.

12 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, The men log counts from the BAM nd ISO methods were not sttisticlly different for ll 3 levels nlyzed. The RSD r vlues from the ISO nd BAM methods were not significntly different for the low nd medium level test portions. The ISO method s dt reported lower RSD r for the high test portions (p < 0.05). The RSD R vlues from the ISO nd BAM methods were not sttisticlly different for the low nd high test portions. The ISO method s dt showed lower RSD R for the medium test portions (p < 0.05; Tble C). Fresh Cheese Three types of fresh cheese (blue, mozzrell, nd cheddr) were nlyzed (Tble 6). Dt generted by the BAM nd SimPlte methods were compred sttisticlly. Lbortory 6 ws determined to be n outlier by the Cochrn test for blue nd mozzrell cheese test portions; therefore, the pired dt were excluded from sttisticl nlysis. Thirteen lbortories submitted vlid dt for blue cheese nd mozzrell cheese test portions nd 14 lbortories submitted vlid dt for cheddr cheese test portions. The men log counts from the BAM nd SIM B methods were not sttisticlly different for test portions of ll 3 cheeses nlyzed. The BAM method s dt showed lower RSD r mozzrell nd cheddr cheese test portions (p < 0.05). The SimPlte method s dt, however, reported lower RSD r (p < 0.05) for the blue cheese test portions. The RSD R vlues from the BAM nd SIM B methods were Tble 5. Yest nd mold counts for cerel test portions (log 10 CFU/g) by SimPlte Y&M CI (SIM B nd I), BAM culture method, nd ISO culture method Uninoculted Low Test portion 1 Test portion 2 Test portion 3 Test portion 4 Lb SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO 1 <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 < <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 < < <1.00 <1.00 < < <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 < < < < < <1.00 <1.00 <1.00 <1.00 <1.00 < < <1.00 <1.00 < <1.00 <1.00 < <1.00 <1.00 <1.00 < <1.00 <1.00 < <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 < b < b b b <1.00 <1.00 <1.00 <1.00 <1.00 <1.00 < Medium High Test portion 5 Test portion 6 Test portion 7 Test portion 8 SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO SIM B BAM SIM I ISO c 4.04 c c 3.78 c b c SIM B = Test portions prepred ccording to the FDA s BAM method; SIM I = test portions prepred ccording to the proposed ISO method. Outlier; dt not used in nlysis for method comprison of BAM nd ISO. Outlier; dt not used in nlysis for method comprison of SIM I nd ISO.

13 308 FELDSINE ET AL.: JOURNAL OF AOAC INTERNATIONAL VOL. 86, NO. 2, 2003 Tble 6. Yest nd mold counts for cheese test portions (log 10 CFU/g) by SimPlte Y&M CI (SIM B nd I), BAM culture method, nd ISO culture method Blue Test portion 1 Test portion 2 Lb SIM B BAM SIM I ISO SIM B BAM SIM I ISO b <6.00 b,c c 8.48 b 7.32 b,c c c 7.83 d 8.73 c,d c <6.00 c 7.76 c,d Mozzrell Test portion 3 Test portion 4 SIM B BAM SIM I ISO SIM B BAM SIM I ISO b 7.45 b,c c 7.84 b 6.87 b,c c Cheddr Test portion 5 Test portion 6 SIM B BAM SIM I ISO SIM B BAM SIM I ISO b c d SIM B = Test portions prepred ccording to the BAM method; SIM I = test portions prepred ccording to the proposed ISO method. Outlier; dt not used in nlysis for method comprison of SIM B nd BAM. Outlier; dt not used in nlysis for method comprison of ISO nd BAM. Outlier; dt not used in nlysis for method comprison of SIM I nd ISO.