From laboratory sample to test portion: preparation according to ISO 6887, initial suspension and pooling for pathogens

The step nobody sees in the report

A microbiological result is reported as "absent in 25 g" or "3,200 CFU/g", but between the bag that arrives at the laboratory and the plate or broth there is a stage that almost never appears in the report: sample preparation. How much is weighed, in which diluent, how long it is homogenized, how long the suspension waits and whether portions are combined: each decision changes what the number means. Sampling in the plant is covered in surface sampling and the transfer to the laboratory in sample transport and storage; this article starts where those end, with the sample already received.

The international reference for this stage is the ISO 6887 series, from committee ISO/TC 34/SC 9. It has six parts: 6887-1:2017 sets the general rules and the other five apply them to product groups: meat (6887-2:2017), fish and fishery products (6887-3:2017), miscellaneous products such as eggs, vegetables, cocoa, herbs and spices (6887-4:2017), milk and milk products (6887-5:2020) and primary production samples, from farm to slaughterhouse (6887-6:2013). If the specific method for a microorganism has its own instructions, those take precedence.

The 1:10 initial suspension, step by step

Weighing and ratio

ISO 6887-1 requires weighing or measuring the test portion with a tolerance of ±5%, with a minimum of 10 g or 10 mL unless the method states otherwise, and adding 9 times that quantity of diluent, preferably measured by mass with a tolerance of ±2%. This gives the 1:10 dilution. The standard allows other ratios when needed: 1 in 20, 1 in 50 or 1 in 100 for products that give suspensions that are too thick, and 1 in 2 or 1 in 5 when low counts are sought, although it warns that the latter can unbalance the ratio between inoculum and medium and must be verified case by case. Any non-standard ratio carries through to the final calculation.

Diluents

For general use the standard defines two diluents: peptone salt solution (1 g of enzymatic digest of casein and 8.5 g of sodium chloride per liter) and buffered peptone water, both with a final pH of 7.0 ± 0.2 at 25 °C. There is also a double-strength version of buffered peptone water for highly acidic products. The final volume may not deviate by more than ±2% and diluents must pass a performance test: after 45 minutes to 1 hour at room temperature, the count of the control strain must remain within ±30% of the initial one. The performance requirements for diluents and media are detailed in ISO 11133.

Homogenization

The mixture is homogenized according to ISO 7218 (fourth edition, 2024), with a rotary homogenizer (laboratory blender) or a peristaltic one, the paddle device that beats a sterile bag and that laboratories usually call a stomacher. If necessary, large particles are allowed to settle for up to 15 minutes; bags with an integrated filter are an equivalent alternative. For hard or dry products, the standard limits the rotary homogenizer to 2.5 minutes at a time and prior chopping or grinding to 1 minute, to prevent the sample from heating up. The FDA manual for food (BAM, chapter 1) uses for counts 50 g of sample in 450 mL of Butterfield's phosphate diluent, blended for 2 minutes.

Maximum times

ISO 6887-1 establishes that no more than 45 minutes may elapse between the end of preparation of the initial suspension and the contact of the inoculum with the final culture medium, and no more than 30 minutes between the initial suspension and the start of the dilutions. If the laboratory is above 27 °C, both times should be shortened. The BAM is stricter: no more than 15 minutes from homogenization until all dilutions are in their media. In a nutrient-rich suspension microorganisms can start to multiply and the result comes out higher than the real sample.

Decimal dilutions

For each decimal step, 1 mL ± 0.02 mL is transferred to a tube with 9 mL ± 0.2 mL of diluent, without inserting the pipette more than 1 cm, mixing for 5 to 10 seconds and with a new tip at each step.

Rules by food type

Frozen foods

For small packages (generally under 2 kg), ISO 6887-1 allows bringing them to a consistency that permits taking the portion, at 18-27 °C for a maximum of 3 hours or at 5 ± 3 °C for a maximum of 24 hours, and advises against thawing in a water bath or under running water, because of the risk of contamination if the package is not watertight. The BAM prefers not to thaw and, if necessary, does so at 2-5 °C within a maximum of 18 hours, or below 45 °C for no more than 15 minutes with agitation in a thermostatically controlled bath.

Meat

ISO 6887-2 covers fresh, chilled, frozen, cured, fermented and minced meat, meat preparations, mechanically separated meat, cooked, dried and smoked meats, concentrated extracts and carcass samples taken by excision or swabbing. It distinguishes, among other cases, blocks and cuts, sausages with casing and surface or deep samples. Carcass sampling is the subject of ISO 17604. The industry context is in microbiology in the meat industry.

Dairy

ISO 6887-5 has its own procedures for liquid milk, milk and whey powder, cheese, caseins and caseinates (with a special case for rennet casein), butter, ice cream, dairy desserts with pH above 5, fermented milks and yogurts with pH below 5 and powdered infant formula. The BAM illustrates why not everything can be treated the same: for Salmonella in cheese it blends 25 g with 225 mL of lactose broth for 2 minutes, lets it stand for 60 minutes and adjusts the pH to 6.8 ± 0.2; instant nonfat dry milk, by contrast, is poured gently onto the surface of the broth and incubated without mixing.

Fish and shellfish

ISO 6887-3 covers whole or filleted fish, crustaceans, cephalopods, bivalve molluscs, gastropods, tunicates and echinoderms, as well as smoked and cooked products and frozen blocks, and includes special procedures for raw molluscs, tunicates and echinoderms from primary production areas.

Dry, acidic and fatty foods

• Low moisture: weigh the diluent first and then add the portion, to reduce osmotic shock, and allow to soak for up to 1 hour if the product is hard. The BAM, for powders, first adds about 15 mL of broth, forms a lump-free suspension and completes the volume in portions.
• Acidic: the suspension must return to pH 7.0 ± 0.5. Buffered peptone water is sufficient for products with pH 4.5 or higher; between pH 3.5 and 4.5 the double-strength version can be used, checking the pH the first time.
• Fatty (more than 20%): polysorbate 80 can be added to the diluent, between 1 and 10 g/L; in general 1 g/L for every 10% fat is enough.
• Inhibitory spices: the BAM adds 0.5% potassium sulfite for dehydrated onion and garlic, and dilutes allspice, cinnamon and oregano to 1:100 and cloves to 1:1000, because no way of neutralizing their effect is known.

Composite and pooled: not the same thing

Combining units from the same batch saves work and reagents. Amendment 1 of 2024 to ISO 6887-1 clarifies two concepts that are confused in practice:

• Composite sample: several units are mixed and a portion of the size indicated by the method, for example 25 g, is taken from the mixture. The result is still expressed per 25 g.
• Pooled sample (pooling): several units or portions are combined and the whole mixture is analyzed. The result is expressed per total mass, for example 100 g or 375 g.

Pooling can be done outside the laboratory, combining individual samples into a larger one, or inside it, combining portions into a single large pre-enrichment or combining already pre-enriched broths. There are firm rules: only for qualitative tests, only with units of the same origin (same batch or lot) and without combining two modes at once, for example pooling portions and then also pooling their broths.

25 g or 375 g: what changes when the portion is enlarged

The classic scheme comes from the BAM: a 25 g portion is taken from each sample unit and up to 15 portions can be combined into a 375 g composite, which is analyzed in full while keeping the 1:9 ratio (375 g in 3,375 mL of broth). The number of units depends on the food category: 60 for category I (no lethal step and intended for older adults, the sick or infants), 30 for II and 15 for III, that is, 4, 2 or 1 composite of 375 g. A single positive rejects the batch. The BAM excludes from compositing, among others, non-instant nonfat dry milk and whole milk powder.

The 2024 amendment sets technical conditions so that the large volume does not hide a positive:

• The pre-enrichment broth must be pre-warmed to the incubation temperature.
• The temperature profile of the large volume must be known. Taking the example of ISO 6579-1, whose pre-enrichment lasts between 16 and 20 hours, the mixture must be kept between 34 °C and 38 °C for at least 16 hours, and the time to reach that temperature must be minimized.
• The sample/diluent ratio of the validated method is not changed, except to increase it in inhibitory matrices such as onion powder, garlic, oregano, peppers, some teas and coffees, vitamin premixes or very salty products.
• A laboratory sample larger than the maximum is divided: 750 g are processed as two 375 g portions.

In addition, the larger size has to be validated. If it was validated in a study according to ISO 16140-2 or ISO 16140-5, each laboratory only needs to verify it according to ISO 16140-3 within the same food category; if not, it must validate it itself with the protocol in Annex H of Amendment 1:2024 to ISO 16140-4, demonstrating with the relative level of detection (RLOD) that the LOD50 does not worsen. Once 375 g have been validated, smaller portions can be used at the same ratio. These concepts are explained in validation and verification of methods.

The evidence supports the approach when it is done well. Tomás Fornés and colleagues validated pooling of portions up to 375 g for Salmonella with five methods (culture, ELISA and real-time PCR) in 23 foods (confectionery, infant formula, pet food, powdered beverages, among others). In 22 of the 23 the RLOD50 compared with 25 g was acceptable (below 2.5), on condition that the broth was pre-warmed and the maximum incubation time respected. Pooling also appears in regulation: Regulation (EC) 2073/2005 requires combining the neck skin of at least three poultry carcasses from the same flock of origin into a 26 g sample, and pooling the swabs from different sites of a carcass before analysis. How n and c are read in those plans is in attribute sampling plans, and why the broth stage is decisive, in microbiological enrichment.

Which TAAG kits and media are related

The kit defines which portion it is validated for. In the TAAG catalog, Elevia 1.1 Salmonella spp. (V-PAT04, RT-qPCR with AiGOR technology) declares a sensitivity of 1 CFU/375 g in food and 1 CFU/sample on surfaces; its AOAC certification covers 375 g portions of 62% cocoa chocolate, cocoa liquor, skimmed milk powder and peanuts, and TAAG also declares validated matrices such as filled chocolates, ground beef (90% lean), animal feed pellets, pasteurized liquid eggs and stainless steel surfaces. Its data sheet lists the supplement Potentia 1 Salmonella spp. (V-PET06), a growth enhancer for buffered peptone water. Ampliora 1.1 Salmonella spp. (V-SF97, real-time PCR) declares 1 CFU/25 g, with validated matrices of raw chicken thigh, frozen beef burger, longaniza-type sausage, mortadella and surfaces, and is associated with Augmentis 14 Universal Gram negative (V-FL46, 225 mL bottles, the broth volume for 25 g at 1:10). Using a kit with a portion, matrix or pooling scheme different from those declared requires the verification or validation called for by the ISO 6887-1 amendment. To choose by matrix, see how to choose a TAAG kit.