Protocols for the isolation and detection of lactic acid bacteria with bacteriocinogenic potential

The objective of this study was to evaluate culture media and methodologies for isolation and detection of lactic acid bacteria (LAB) capable to produce bacteriocin-like substances. Samples of milk and cheese were pour plated on de Mann-Rogosa-Sharpe agar (MRS) and Kang-Fung-Sol agar (KFS) (both at 35 °C/48 h, under anaerobiosis), from which 389 and 256 LAB cultures were selected. The antagonistic activity of them was evaluated using the spot-on-the-lawn and two culture media: brain-heart infusion agar with catalase (BHI + C) and M17 (both at 35 °C/24 h). The proteinaceous nature of the antagonistic cultures was verified using: spot-on-the-lawn (MRS, 25 °C/24 h, under anaerobiosis) and well-diffusion (cultures amplified on modified MRS broth at 25 °C/24 h, and then neutralized using NaOH). Listeria monocytogenes ATCC 7644 was used as indicator. A larger number of antagonist cultures were isolated from MRS (83 by M17 and 65 by BHI + C) in comparison to KFS (24 by M17 and 15 by BHI + C). The spot-on-the-lawn identified a higher frequency of LAB capable of producing bacteriocin-like substances. MRS was considered to be the best culture media for the isolation of LAB capable to produce bacteriocin-like substances, activity that was better identified using the spot-on-the-lawn methodology.     

Incidences of problematic organisms on petrifilm aerobic count plates used to enumerate selected meat and dairy products

Petrifilm aerobic count plates are similar to or better than conventional pour plates. Petrifilm has its problems, however; some microorganisms can liquefy the Petrifilm gel and others do not produce the necessary color change with the indicator dye used. Petrifilm aerobic count plates were compared with the pour plates for determining the incidence and identification of problematic organisms in 329 meat and dairy products. Petrifilm plates produced higher mean counts with better repeatability than did pour plates. There was also close correlation between methods with coefficients of 0.97 to 1.0. Bacillus subtilis, Bacillus licheniformis, and a group D Streptococcus liquefied Petrifilm gels in 17.4% of the samples tested: dairy products accounted for 16.0%, and meats accounted for the remaining 1.4%. Liquefaction hindered enumeration in 3.2% of the Petrifilm plates used. Streptococcus viridans was not detectable on Petrifilm plates after the recommend incubation period, and this organism occurred in 0.3% of the Petrifilm plates used. These results indicate that Petrifilm plates would be unsuitable for samples with large numbers of these organisms. Knowledge of the contaminating flora may be an asset when utilizing Petrifilm aerobic count plates for the enumeration of microbes in food.     

Technical note: Enumeration of mesophilic aerobes in milk: Evaluation of standard official protocols and Petrifilm aerobic count plates

Enumeration of mesophilic aerobes (MA) is the main quality and hygiene parameter for raw and pasteurized milk. High levels of these microorganisms indicate poor conditions in production, storage, and processing of milk, and also the presence of pathogens. Fifteen raw and 15 pasteurized milk samples were submitted for MA enumeration by a conventional plating method (using plate count agar) and Petrifilm Aerobic Count plates (3M, St. Paul, MN), followed by incubation according to 3 official protocols: IDF/ISO (incubation at 30°C for 72 h), American Public Health Association (32°C for 48 h), and Brazilian Ministry of Agriculture (36°C for 48 h). The results were compared by linear regression and ANOVA. Considering the results from conventional methodology, good correlation indices and absence of significant differences between mean counts were observed, independent of type of milk sample (raw or pasteurized) and incubation conditions (IDF/ISO, American Public Health Association, or Ministry of Agriculture). Considering the results from Petrifilm Aerobic Count plates, good correlation indices and absence of significant differences were only observed for raw milk samples. The microbiota of pasteurized milk interfered negatively with the performance of Petrifilm Aerobic Count plates, probably because of the presence of microorganisms that poorly reduce the dye indicator of this system.     

Evaluation of various selective media for the detection of Pseudomonas species in pasteurized milk

Pseudomonas spp. are common gram-negative, post-pasteurization contaminants that contribute to spoilage of pasteurized dairy products. This study evaluated 5 common selective media for detecting Pseudomonas spp. in pasteurized milk. The performance of each selective medium for recovering 12 different Pseudomonas isolates (selected to represent a diversity of pasteurized milk isolates) was compared with that of standard plate count agar pour plates. Pseudomonas isolates showed varying abilities to produce colonies on different selective media. For 2 of 12 isolates, a 48-h incubation time was required for colony formation on any of the media tested. Violet red bile agar and coliform Petrifilm (3M, St. Paul, MN) were less effective than standard plate count agar pour plates at recovering Pseudomonas, regardless of incubation time, and MacConkey agar showed poor detection efficiency compared with SPCP after a 48-h incubation (R(2) = 0.26). Therefore, the use of violet red bile agar, MacConkey agar, or coliform Petrifilm may not be sufficient for detecting common Pseudomonas spp. in milk. The methods showing the highest detection efficiencies were crystal violet tetrazolium agar (CVTA) pour plates (R(2) = 0.95) and CVTA plates inoculated by spiral plating (R(2) = 0.89) incubated at 32 °C for 48 h. Overall, plating milk samples on CVTA followed by a 48-h incubation at 32 °C was the most effective selective method for recovering a diversity of Pseudomonas spp. from milk.     

Comparison of dry culture medium and conventional plating techniques for enumeration of bacteria in pasteurized fluid milk

Standard plate counts, psychrotrophic bacterial counts, and coliforms were determined by conventional plating techniques and by Petrifilm TM plates, a dry culture medium, for 48 commercially processed milk samples (24 whole milk and 24 skim milk). The Petrifilm SM plate counts were compared with counts on standard methods agar for the standard plate count, psychrotrophic bacterial count, and rapid psychrotrophic bacterial count. The Petrifilm violet red bile plate counts were compared with counts on violet red bile agar for coliform test with a solid medium and the preliminary incubation method for detection of coliforms. Standard plate counts were determined within 24 h of packaging and after 7, 10, and 14 d of storage at 6.1 degrees C. Psychrotrophic bacterial counts and coliform counts were determined with 24 h of packaging and after 7 d storage. There was a strong linear relationship between Petrifilm SM and standard methods agar plates (excluding counts on samples plated within 24 h of packaging) and for the psychrotrophic bacterial count method. Petrifilm SM had a weak linear relationship with Standard Methods Agar plates for the rapid psychrotrophic bacterial count. Coliform counts determined on Petrifilm violet red bile plates were generally within the same range as counts on violet red bile agar plates. The positive predictive values for the Petrifilm violet red bile plates and violet red bile agar plates were essentially the same for samples plated within 24 h of packaging.     

Evaluation of dry sheet medium culture plate (Compactdry TC) method for determining numbers of bacteria in food samples

The Compactdry, a ready-to-use and self-diffusible dry medium sheet culture system, has been developed by the Nissui Pharmaceutical Co. Ltd. for enumerating bacteria in food. The Compactdry consists of special spread sheet with culture medium that is the same as standard method nutrients, a cold water-soluble gelling agent, and a unique plastic dish. The procedure for bacterial examination in a sample solution (1 ml) is to just inoculate a test solution into the center of the self-diffusible medium and incubate at 35 degrees C for 48 h. The Compactdry TC (CTC) for the enumeration of total aerobic bacteria from 97 food samples was compared with the standard plate count (SPC) method and 3M Petrifilm aerobic count plates (PAC). The correlation coefficients between the CTC and SPC method, the CTC and PAC, and the PAC and SPC method were 0.97, 0.99, and 0.97, respectively. The Compactdry system is useful for the enumeration of total aerobic bacteria in food and may be a possible suitable alternative to the conventional pour-plate or the Petrifilm plate methods.     

Evaluation of petrifilm series 2000 as a possible rapid method to count coliforms in foods

This research note is a preliminary comparison between the Petrifilm 2000 method and a widely used traditional enumeration method (on violet red bile agar); six batches of different foods (egg, frozen green beans, fresh sausage, a bakery product, raw minced meat, and raw milk) were studied. The reliability of the presumptive counts taken at 10, 12, and 14 h of incubation using this method was also verified by comparing the counts with the total confirmed counts at 24 h. In all the batches studied, results obtained with Petrifilm 2000 presented a close correlation to those obtained using violet red bile agar (r = 0.860) and greater sensitivity (93.33% of the samples displayed higher counts on Petrifilm 2000), showing that this method is a reliable and efficient alternative. The count taken at 10-h incubation is of clear interest as an early indicator of results in microbiological food control, since it accounted for 90% of the final count in all the batches analyzed. Counts taken at 12 and 14 h bore a greater similarity to those taken at 24 h. The Petrifilm 2000 method provides results in less than 12 h of incubation, making it a possible rapid method that adapts perfectly to hazard analysis critical control point system by enabling the microbiological quality control of the processing.     

Enumeration of bifidobacteria using Petrifilm™ AC in pure cultures and in a fermented milk manufactured with a commercial culture of Streptococcus thermophilus

Bifidobacteria are probiotic microorganisms that are widely used in the food industry. With the aim of using of Petrifilm™ Aerobic Count (AC) plates associated with selective culture media, aliquots of sterile skim milk were inoculated separately with four commercial cultures of bifidobacteria. These cultures were plated by both the conventional method and Petrifilm™AC, using the culture media NNLP and ABC. The cultures were incubated under anaerobiosis at 37 °C for 24, 48 and 72 h. No significant differences (p > 0.05) were observed between the obtained counts at 48 and 72 h. Bifidobacteria counts in ABC were usually higher than in NNLP, independent of the plating method. Subsequently, fermented milk was prepared with a Streptococcus thermophilus strain, and aliquots were inoculated with the same bifidobacteria. Then, the fermented milks were submitted to microbiological analysis for bifidobacteria enumeration using the same culture media and methodologies previously described, incubated under anaerobiosis at 37 °C for 48 h. Again, bifidobacteria counts in ABC were higher than in NNLP, with significant differences for some cultures (p < 0.05). The counts obtained by both methodologies presented significant correlations (p < 0.05). The results indicate the viability of Petrifilm™AC as an alternative method for bifidobacteria enumeration when associated to specific culture media, specially the ABC.     

Comparison of the Baird-Parker agar and 3M Petrifilm Staph Express Count plate methods for enumeration of Staphylococcus aureus in naturally and artificially contaminated foods

The recently developed 3M Petrifilm Staph Express Count plate (PFSE) method was compared with the U.S. Food and Drug Administration Bacteriological Analytical Manual's Baird-Parker agar spread plate (B-P) method for enumeration of Staphylococcus aureus in naturally contaminated, mechanically separated poultry (MSP; n = 92) and raw milk (n = 12). In addition, mozzarella and Parmesan cheeses and hot-smoked rainbow trout and chub were surface inoculated with a three-strain mixture of S. aureus, stored at 5 degrees C, and periodically analyzed with both methods for numbers of S. aureus. For naturally contaminated raw milk and MSP samples, the PFSE method yielded counts that were not significantly different (P > 0.05) from counts obtained using the B-P method. From raw milk and MSP samples, 60% (21 of 35) and 55% (124 of 226), respectively, of confirmed (DNAse-positive) isolates from PFSE plates were identified by further testing as S. aureus. Corresponding S. aureus identification rates for isolates forming typical colonies on B-P plates were 53% (19 of 36) and 50% (125 of 248). For both methods, other staphylococci composed the vast majority of tested isolates that were not identified as S. aureus. For inoculated hot-smoked fish, S. aureus counts from the PFSE method were not significantly different from counts from the B-P method. Compared to the B-P method, significantly lower numbers of inoculated S. aureus were recovered using the PFSE method in analyses of mozzarella cheese stored 28 and 42 days at 4 degrees C. The PFSE and B-P methods were not significantly different for inoculated cheeses at all other sampling times. DNAse-positive isolates from PFSE analyses of inoculated cheeses and smoked fish were identified as S. aureus 98% (51 of 52) and 86% (36 of 42) of the time, respectively, as compared with 100% (58 of 58) and 95% (40 of 42) of the time for typical B-P isolates. Overall, the PFSE and B-P methods appeared to perform similarly in enumeration of S. aureus in animal-derived foods.     

Comparison of culture media, simplate, and petrifilm for enumeration of yeasts and molds in food

The efficacy of three culture media, dichloran rose bengal chloramphenicol (DRBC), dichloran 18% glycerol agar (DG18), and potato dextrose agar (PDA) supplemented with two antibiotics, were compared with the Simplate and Petrifilm techniques for mold and yeast enumeration. The following foods were analyzed: corn meal, wheat flour, cassava flour, bread crumbs, whole meal, sliced bread, ground peanuts, mozzarella cheese, grated parmesan cheese, cheese rolls, orange juice, pineapple pulp, pineapple cake, and mushroom in conserve. Correlation coefficients of DRBC versus PDA and DG18 for recovering total mold and yeast counts from the composite of 14 foods indicated that the three media were generally equivalent. Correlation coefficients for Petrifilm versus culture media were acceptable, although not as good as between culture media. Correlation coefficients of Simplate versus DRBC, DG18, PDA, and Petrifilm for recovering total yeasts and molds from a composite of 11 foods demonstrated that there was no equivalence between the counts obtained by Simplate and other culture media and Petrifilm, with significant differences observed for the most foods analyzed.     

Evaluation of the Petrifilm rapid coliform count plate method for coliform enumeration from surimi-based imitation crab slurry

The 3M Petrifilm rapid coliform count (RCC) plate method was compared with two conventional methods, namely violet red bile agar (VRBA) and desoxycholate lactose agar (DLA), for enumerating coliforms. The VRBA plating method is a reference method in the Bacteriological Analytical Manual and the DLA plating method is the method recommended by the Food Sanitation Law of Korea for enumeration of coliforms. Serratia sp., a coliform that was isolated from frozen surimi, was incubated in surimi-based imitation crab (SBIC) slurries and enumerated on the Petrifilm RCC, VRBA, and DLA plates. Results from the Petrifilm RCC plate were not significantly different from results from VRBA or DLA plates at P < 0.05 level. The correlation coefficient for Petrifilm RCC plates versus the VRBA method and for Petrifilm RCC plates versus the DLA method were 0.994 and 0.996, respectively. With the Petrifilm RCC plate method, we were able to estimate presumptive coliforms (except Serratia sp.) after 14 h and to enumerate confirmed coliforms (including Serratia sp.) after 24 h.     

Rapid and simple estimation of microbiological quality of raw milk using chromogenic Limulus amoebocyte lysate endpoint assay

A rapid chromogenic Limulus amoebocyte lysate (LAL) endpoint assay for the enumeration of total mesophilic microbial loads and coliforms was investigated as a means to assess the microbiological quality of raw milk. For experiment 1, raw milk samples (n = 25) were stored in a refrigerator (2 +/- 2 degrees C) and then analyzed at regular intervals (1, 5, 10, and 15 days). For experiment 2, fresh raw milk samples (n = 50) were tested to determine the utility of the LAL assay for fresh raw milk. The sample was diluted threefold in a 96-well microtiter plate with pyrogen-free water and assayed with a chromogenic LAL kit to find a final reaction point. The LAL results were compared with standard plate counts (SPC) and coliform counts determined by conventional plating methods. The results of the LAL assay were strongly correlated to conventional SPC (r2 = 0.93; n = 100) and were highly correlated to coliforms (r2 = 0.74; n = 100). A highly significant linear relationship (r2 = 0.82; n = 50) was also observed between the predicted SPC based on the LAL value and the actual SPC. The results of LAL testing were classified into one of seven contamination groups. The data set for SPC was effectively differentiated using the LAL technique (P < 0.01). The chromogenic LAL assay was found to be a rapid (within 16 min) and simple (not requiring specific instruments) method for monitoring microbial levels in raw milk. This method may be successfully implemented to rapidly determine highly microbial contaminated raw milk (> 3.0 log10 CFU/ml of SPC).     

Evaluation of dry medium (Sanita-Kun) for enumeration of coliforms and Escherichia coli in milk, ice cream, ham, and codfish fillet

This study was performed to compare the efficiency and accuracy of Sanita-Kun dry medium with those of conventional culture and Petrifilm dry medium methods for enumerating coliforms and Escherichia coli in milk, ice cream, ham, and codfish fillet. All 3 methods showed high correlations each other and no significant differences in enumerating coliforms and E. coli. In addition, the Sanita-Kun method was easier to count microbial load than the Petrifilm method. Therefore, we concluded that the Sanita-Kun method can be used as an alternative to conventional culture and Petrifilm methods for the analysis of coliforms and E. coli in milk, ice cream, ham, and codfish fillet products.     

Use of human lysozyme transgenic goat milk in cheese making: effects on lactic acid bacteria performance

Genetically engineered goats expressing elevated levels of the antimicrobial enzyme lysozyme in their milk were developed to improve udder health, product shelf life, and consumer well-being. The purpose of this study was to evaluate the effect of lysozyme on the development of lactic acid bacteria (LAB) throughout the cheese-making process. Raw and pasteurized milk from 7 lysozyme transgenic goats and 7 breed-, age-, and parity-matched nontransgenic controls was transformed into cheeses by using industry methods, and their microbiological load was evaluated. The numbers of colony-forming units of LAB were determined for raw and pasteurized goat milk, whey, and curd at d 2 and at d 6 or 7 of production. Selective plating media were used to enumerate lactococcal species separately from total LAB. Although differences in the mean number of colony-forming units between transgenic and control samples in raw milk, whey, and cheese curd were non-significant for both total LAB and lactococcal species from d 2 of production, a significant decrease was observed in both types of LAB among d 6 transgenic raw milk cheese samples. In pasteurized milk trials, a significant decrease in LAB was observed only in the raw milk of transgenic animals. These results indicate that lysozyme transgenic goat milk is not detrimental to LAB growth during the cheese-making process.     

A collaborative study of a method for the enumeration of probiotic bifidobacteria in animal feed

An enumeration method to be used as an official control method in the framework of Council Directive 70/524/EEC for probiotic bifidobacteria used as feed additives was validated. Seventeen laboratories in 11 European Countries carried out a collaborative study. A spread plate method following BS ISO 15214:1998 using four different agars, Man Rogosa Sharpe (MRS), acidified MRS, MRS with triphenyl tetrazolium chloride (TTC) and a selective bifidobacteria medium, was validated. Precision data in terms of repeatability (r) and reproducibility (R) of the method for each medium using different feeding stuffs with a high and a low inoculation level were determined. Bifidobacteria were present in the samples as a single component or in mixtures with other probiotics. The enumeration of bifidobacteria on all agars showed a relative standard deviation of repeatability (RSD(r)) between 1.2% and 6.3% and a relative standard deviation of reproducibility (RSD(R)) between 2.6% and 8.7%. MRS agar was preferred, followed by acidified MRS and MRS+TTC agar. The selective bifidobacteria medium gave similar counts as the MRS media. For routine analysis, the use of MRS agar with supplementation of cysteine hydrochloride (the selective bifidobacteria medium without antibiotics) is recommended. Depending on the presence and concentration of other probiotics such as enterococci, lactobacilli and pediococci, acidified MRS or MRS+TTC agar is recommended. The selective bifidobacteria medium was selective for bifidobacteria.An official control method for enumeration of probiotic bifidobacteria as a single component and in mixtures with other probiotic microorganisms in feeding stuffs was validated. The methodology is not applicable to mineral feed. The results are intended for consideration for adaptation as CEN and ISO standards.     

Evaluation of de-lipidated egg yolk and yeast autolysate as growth supplements for lactic acid bacteria culture

Four strains of lactic acid bacteria (LAB) were cultivated on deproteinated whey supplemented with yeast autolysate (YA) and/or de-lipidated egg yolk (DEY). Growth was compared to that recorded on the widely used MRS medium. The feasibility of the use of YA in place of yeast extract was confirmed. In the presence of YA in the medium, cell numeration gained nearly 1 log for only 0.4% DEY added in the culture medium, namely 0.1 g/L additional nitrogen, showing that DEY also supplies LAB growth in trace elements and growth factors like vitamins. If compared to the MRS medium, final cell numerations were about 1 log lower on deproteinated whey supplemented with YA and DEY, while the protein nitrogen ratio of both media was 2.7, showing the high potential of the tested supplements, and especially DEY. Subsequent works with higher and hydrolysed DEY may be therefore useful.     

Evaluation of culture media for selective enumeration of probiotic strains of lactobacilli and bifidobacteria in combination with yoghurt or cheese starters

A series of culture media were evaluated for selective enumeration of 10 commercial probiotic cultures. The media M17 and MRS 5.2 were most suitable to enumerate the yoghurt starters Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus, respectively. For the enumeration of probiotic Lb. acidophilus cultures, the MRS-clindamycin medium was found to be most optimal. Selective recovery results of the tested Bifidobacterium sp. were not uniform but indicated strain-to-strain variations depending on whether NPLN or MRS-LP were used. For the enumeration of probiotic cultures of Lb. rhamnosus and Lb. paracasei, LC medium (for yoghurt products) and MRS-AC medium (for cheese products) are recommended. In conclusion, our data indicate that the choice of culture medium and methodology for selective enumeration of commercial probiotic strains in combination with starters depends strongly on the product matrix, the target group and the taxonomic diversity of the bacterial background flora in the product.     

Laboratory evaluation of 3M Petrifilms and University of Minnesota Bi-plates as potential on-farm tests for clinical mastitis

The objective was to determine test characteristics and compare 2 potential on-farm culture systems for clinical mastitis, the Minnesota Easy Culture System II Bi-plate and Petrifilm. The tests were evaluated using clinically positive mastitic milk samples (n = 282) to determine their ability to differentiate appropriate treatment groups; all cases that had gram-positive growth were considered treatment candidates (n = 161), whereas cases that grew gram-negative organisms only or yielded no bacterial growth were classified as no treatment (n = 121). For Petrifilm, both undiluted and 1:10 diluted milk samples were used. To create treatment categories, 2 types of Petrifilms were used, Aerobic Count (AC) and Coliform Count (CC). Both Bi-plates and Petrifilms were read after 24 h of incubation. Analysis was conducted at various colony count thresholds for the Petrifilm test system. The combination of Petrifilms that had the highest sensitivity classified a case as gram-negative if there were ≥20 colonies present on the CC. If there were <20 colonies present on the CC and >5 colonies present on the AC, a case would be classified as gram-positive. The Bi-plate had a sensitivity of 97.9% and a specificity of 68.6%. The Petrifilm test system had a sensitivity of 93.8% and a specificity of 70.1%. There was no significant difference in the sensitivities between the tests. All Bi-plates and Petrifilms were read by a laboratory technician and a group of masked readers with limited microbiology training. Kappa values for the masked readers were 0.75 for Bi-plates and 0.84 and 0.86 for AC and CC Petrifilms, respectively. The Bi-plate and Petrifilm were able to successfully categorize clinical cases of mastitis into 2 treatments based on their ability to detect the presence of a gram-positive organism. Neither method had the ability to determine if a sample was contaminated. The results of this study indicate that both tests were able to appropriately categorize cases, which could potentially result in a reduction in the quantity of antibiotics used to treat clinical cases of mastitis.     

Enumeration of coagulase and thermonuclease-positive Staphylococcus spp. in raw milk and fresh soft cheese: An evaluation of Baird-Parker agar,Rabbit Plasma Fibrinogen agar and the Petrifilm™ Staph Express count system

Staphylococcus spp. are microorganisms that are naturally present in milk and dairy products and are often associated with food-borne diseases outbreaks due to the ability of some strains to produce thermostable enterotoxins. This ability is usually associated with coagulase and thermonuclease production, characteristics that are considered in the microbiological analyses for the control of such microorganisms. The objective of this study was to evaluate the culture media and the methodologies used for the enumeration of coagulase and thermonuclease-positive Staphylococcus spp. in raw milk and fresh soft cheese. Samples of artificially contaminated milk (with coagulase-positive Staphylococcus reference strains) and samples of naturally contaminated raw milk and cheese were submitted for enumeration in Baird-Parker agar (BP), Rabbit Plasma Fibrinogen agar (RPFA) and in the Petrifilm™ Staph Express count system (STX). No significant differences (P > 0.05) were observed between the mean counts obtained in all of the evaluated culture media. RPFA and STX had good correlation indices between the total and typical colony counts as well as with coagulase and the thermonuclease-positive colony counts. Thus, there is a better association between coagulase and thermonuclease production to typical colony morphology developed on these culture media, leading to more accurate and reliable results than with BP, which demonstrated lower correlation indices between these counts.     

Application of a bacteriological on-farm test to reduce antimicrobial usage in dairy cows with purulent vaginal discharge

The objective of this study was to assess the effect of a selective antibiotic treatment strategy based on a quick bacteriological on-farm test (Petrifilm, 3M Corp., St. Paul, MN) compared with the conventional antibiotic treatment of all cows having clinical endometritis (CE) defined by the presence of purulent vaginal discharge on both clinical cure rate and reproductive performance. The study was simultaneously conducted with dairy cows reared under a highly supplemented rotational grazing system in Argentina and in a freestall system in Slovakia. Cows having an abnormal vaginal discharge (VD, indicative of clinical endometritis) on 21 to 35 d in milk (DIM) were randomly allocated to 1 of 2 study groups: selective treatment (ST) or conventional treatment (CT). All cows in the CT group (n = 174) received a single intrauterine administration of 500 mg of cephapirin. In the ST group (n = 178), treatment decision was made according to the results of the bacteriological on-farm test. For this test, we collected intrauterine samples with the cytobrush technique and stroke the brushes onto 2 different Petrifilm plates, one for aerobic count and another for Enterobacteriaceae count, incubated the plates, and counted the number of colonies after 24 h. Positive cows (≥5 colonies in one or both plates) received a single intrauterine treatment with 500 mg of cephapirin, whereas negative cows (<5 colonies) remained untreated. Clinical cure rate was assessed by direct vaginal inspection at 14 d after treatment (VD-0). The odds for conception at first artificial insemination, artificial insemination by 80 DIM, pregnancy by 100 DIM, and for nonpregnancy by 200 DIM were estimated with mixed logistic regression models. The hazard of conception was also assessed with proportional hazard regression model. The selective antibiotic treatment strategy based on the outcome of Petrifilm test reduced the number of required treatments (57%) and maintained similar efficacy in terms of clinical cure and reproductive performance as the conventional antibiotic treatment of all endometritic cows.