10 Steps to Validating and Verifying Allergen Cleaning Procedures

That cleaning is a vital component of any allergen management program is beyond dispute. But how do food producers know whether their cleaning regimen is working? Paul Bagshaw of Holchem Laboratories guides us through the ins and outs of cleaning validation and verification.

Food manufacturers and processors rely on a variety of differing policies and procedures to enforce allergen controls. These include personnel controls, such as a hand-washing procedure and the use of protective clothing and equipment (PPE), process controls such as segregated storage and color-coded equipment, production controls such as dedicated equipment and time segregation, and – most importantly – cleaning.

Cleaning regimens are subject to rigorous validation. Global Food Safety Initiative (GFSI) retailer approval schemes such as the British Retail Consortium (BRC) state, “Where cleaning procedures are part of a defined prerequisite plan to control the risk of a specific hazard the cleaning and disinfection procedures and frequency shall be validated…” In simple terms, this means that the validation process should demonstrate that the cleaning procedure a site is using reduces the hazard – in this case, an allergen – to a level deemed to be acceptable.

The validation guidelines at Holchem Laboratories Ltd. have been developed in accordance with the principles of the European Hygienic Engineering and Design Group (EHEDG) Cleaning Validation subgroup1 and Campden BRI2, together with in-house best practices. The overarching principle can be summed up in this way: validation should be carried out under worst-case scenarios. Here, we take a look at the steps involved in setting up a validation program and then verifying that program.

1. Determine the objective of the cleaning

The process begins by determining the objective of the cleaning: with allergens, the goal is to ensure the absence of detectable allergens in food products that undergo processing following the cleaning. The site should first define the products and process lines that the validation will cover as well as the type of cleaning to be validated (i.e. a product changeover cleaning or an end-of-production cleaning, open plant or closed plant (clean-in-place)).

It has always been best practice to use the most sensitive detection technique available for detecting allergens in food products (traditionally based on ELISA). In addition, food-processing surfaces should be cleaned such that allergens cannot be detected using lateral flow devices (LFDs).

2. Ensure organizational support

A team approach will help ensure that the validation process is effective; the team should be multi-departmental (production, engineering, technical, hygiene, hazard specialist) and have the backing of senior management to ensure seamless collaboration.

3. Know the hygienic design of your equipment

A review of the hygienic design of the equipment is recommended. The primary purpose of this step is to determine the areas of the equipment that are most difficult to clean; this is helpful in determining the worst-case scenario. These areas will be assessed for cleanability during the validation process. As such, it may be necessary to strike a balance between the place that is hardest to clean (but may need specialist access equipment or engineering support to dismantle) and places that are hard to clean but are practicably accessible.

4. Review any current cleaning programs and applicable regulations

If no cleaning program exists, then one is created at this stage. In practice, one often exists, meaning that sites should be recording that cleaning program, often by putting copies of CICs (cleaning instruction cards) into the validation pack. However, certain parameters of the clean are generally overlooked, such as the number of cleaning operatives and the cleaning window required. Certain parameters of the clean, such as those for chemical strengths and solution temperatures, are often formulated in terms of a range. In such cases, the validation should be undertaken in worst-case circumstances, i.e. at the lowest chemical strength or temperature in the range.

Those responsible for a site should also take into consideration the implications concerning health and safety legislation when carrying out the clean. The site should already have undertaken COSSH assessments for the chemicals they intend to use on the validation, and should consider whether any risk assessments are required for any dismantling of equipment for cleaning. Chemical disinfectants should have the relevant efficacy data, including the European standards EN1276 and EN 13697, and comply with the requirements of the Biocidal Products Regulation (EU 528/2012).

5. Determine the worst-case scenario for the soiling

There are several reasons to choose a particular food product for the validation: it may have the strongest adhering soil, the highest level of allergens or the hardest allergen to remove. The processing that the food product undergoes will also have an impact on the removal of the soiling; this could include the longest processing time, the highest temperature or the period of time the equipment sits idle before cleaning.

Whilst this determination of the worst-case soiling scenario will ensure a robust validation that stands up to scrutiny, an added benefit is that it increases efficiency by allowing food manufactures to carry out fewer validation processes. For example, if a food manufacturer has a number of allergens that are checked by the same cleaning and disinfection program, validating the program for the worst-case scenario for a single allergen (highest allergen presence, most difficult to clean soil) theoretically confers validation to cleaning programs for all allergens used.

6. Sample, sample and sample again 

Sites should then determine the type of sampling to assess whether the objective of the cleaning and disinfection program has been met. The simplest form of assessment, the visual inspection, is often the most overlooked: are debris visible at the sample sites? Other options are direct and indirect sampling: direct sampling is likely to be via surface swabbing, whilst indirect sampling is commonly used only for CIP (cleaning in place) applications and is generally taken via rinse water samples.

Finally, product sampling normally involves taking a sample of the first product off the line for testing after cleaning has been carried out. For CIP, it is considered good practice to take samples from the first, middle and last product from the line.

7. Choose the right analytical test

Now that we know how we are going to sample, we need to decide which analytical tests to use to determine whether the objective of the cleaning and disinfection program has been met. Such tests should be specific, sensitive, representative and reproducible. For allergens in the product itself, ELISA should be used wherever possible for validation.

For surfaces, ELISA is also useful after validation to assess residues, though LFDs are the desired option, as they will be the method of choice for performing ongoing cleaning verification. If a lab-based ELISA test is undertaken for surface residues, LFDs should be run in parallel to establish any correlation between the two methods.

For allergen testing, positive controls should be established to ensure that the target allergen, in food products and on process surfaces, can be detected under the conditions of food manufacturing under test. This requires the food manufacturer to send samples of the product due to be run and swabs of the surfaces before the validation clean begins. This is helpful in accounting for the variations in allergen detection that the food matrix may cause. If, for example, the allergen is known to be a product ingredient but LFDs are unable to detect it, then an LFD is not a suitable verification method going forward.

It is also good practice to determine if the cleaning or disinfectant residues present in the sample matrix have any effect on the sensitivity of the analytical detection technique.

8. Carry out the validation

Then, the actual validation process can go forward. It is generally accepted that it be repeated no fewer than 3 times. It is also good practice to carry out validation at varying times to account for different cleaning teams, seasonal variation in raw ingredients, variation in production pressures and other factors. Validations should be reviewed routinely, at least annually, or when any parameter changes, such as a product, machinery or a cleaning parameter.

9. Interpret the results

The cleaning has achieved its desired results when both the product and food contact surfaces are free from allergens. Conversely, the detection of allergens in the product as well as on food contact surfaces indicates that the cleaning has not achieved its desired results, meaning that the site must amend some aspect of the cleaning program to improve results before rerunning the validation. If a cleaning program can achieve surface cleanliness via LFD but not by ELISA (or other alternative, non-routine sensitive techniques), a risk assessment to ascertain whether the detectable allergen present on the surface is likely to be a significant risk to the subsequent batch of product should be undertaken.

Within the risk assessment, two factors are important. Firstly, cross-contamination from the food contact surface to the food involves a transfer coefficient: not all allergens present on the surface will transfer to the food. In practice, the transfer coefficient of the allergen to the foodstuff, and the area of the food contact surface touched by the portion size before it is packed will be unknown.

Secondly, the nature of product and surface testing is different. Product testing involves macerating the product sample in a large volume of diluent, whilst the swab used in surface testing is recovered into a small quantity of diluent. Yet as the same volume of diluent is tested, in effect a lower detection sensitivity is recorded for product samples. In reality, therefore, a detection of allergen present on a food surface would likely result in an allergen level in the food product of approximately 100 times less.

10. Verify with ATP or protein swabs

After validation, food manufacturers require a method of verification. Cleaning verification is intended to demonstrate that on subsequent cleaning occasions, the cleaning and disinfection program has met its objectives. It may be possible to implement a verification routine via ATP measurements rather than allergen lateral flow strips.

However, for this to be an acceptable method, ATP must be present on the surfaces when there is an absence of detectable allergen residues. If this is possible, ATP can be frequently measured (daily, for example), whilst allergen lateral flow strips could be used less frequently (weekly or monthly). The average results of the validation plus any ‘comfort factor’ should be set as target levels for ATP, which requires that ATP be used alongside other test methods during the validation.

Similarly, protein swabs can serve to provide ongoing verification: as the vast majority of allergens are proteins, an absence of protein implies an absence of allergen. However, a presence of protein does not necessarily indicate a presence of allergen. If ATP or protein swabs are to be used, those operating the site must understand that they are not measuring the allergen itself but general hygiene indicators.

Conclusion - Customizing the validation to your specific production environment

Cleaning and validation are complex processes, involving a variety of possible tools. Ultimately, a cleaning validation is a procedure that must be customized to the needs of a specific production environment. Detailed knowledge of the products and process lines and an understanding of current and previous cleaning programs and their efficacy should inform the decision about what constitutes the worst-case scenario. Only this way can a facility ensure that the validation is robust enough to meet both internal requirements and the standards of external accrediting bodies. An approach combining LFD and ELISA methods can give some insight into how successful a validation performs on surfaces and products, respectively. ATP and protein testing methods can help to verify the ongoing efficacy of a validation program.