Interventions are actions taken during beef processing to reduce microbial contamination of carcasses. They can be GHP-based and hazard-based in their nature. GHP-based measures are pre-requisites to hazard-based measures, are qualitative in nature, and based on empirical knowledge and experience. Some examples of GHP-based control measures applied throughout slaughter and dressing process are: cleaning and disinfection of lairage-to-stunning areas, hide cleanliness assessment, bunging, rodding, hide removal methods, trimming, chilling, and sanitation of tools/equipment. On the other hand, hazard-based intervention measures are developed from scientific research to specifically control certain hazards and are able to provide demonstrable and quantifiable reduction in bacterial load. Some examples of hazard-based intervention measures are a range of different interventions for cattle hides and carcass meat mostly aimed at microbial removal, immobilisation and/or killing: animal coat clipping or microbial immobilisation, thermal treatments, chemical treatments (organic acids, chlorine compounds, oxidisers, sanitisers, etc).
EU legislation allows the use of treatments to remove surface contamination during slaughter, but only after following appropriate consideration and a risk assessment by EFSA, and approval of such treatments by the regulatory authorities. In order to investigate intervention’s performance (e.g. efficacy against chosen indicator microorganism or foodborne pathogen), the first step is to design the study using appropriate methodology. Evaluating intervention’s effects usually require consideration of several aspects: expected prevalence/levels of microorganism on the carcasses, sample size, expected reduction effect in the prevalence/numbers, randomisation and blinding, addressing possible confounders and bias, intervention application parameters (time, temperature, concentration, mode of application), sampling plan, data analysis and correct data reporting. Intervention studies are most often experimental in their nature, and as such, have study subjects assigned to a treated group or a control group before the start of the intervention. They can be performed under highly controlled conditions in the laboratory or research (pilot plant) facilities, or under field/commercial (abattoir) conditions, which are less controlled and evaluate real life intervention performance. As experimental studies, randomised controlled trials can be considered as providing the strongest research evidence and are a “gold standard” for evaluating interventions. In this case, study samples are allocated to intervention/comparison groups with a natural exposure to microorganisms and then evaluated for outcomes, or samples can be artificially inoculated and then allocated to the intervention groups (so-called “challenge trials”). Sometimes when it is not feasible to create control group, study observations (for intervention outcome) are made on the same population before and after receiving an intervention (so-called quasi experiments or before-and-after trials).
Systematic error (bias) is very important when designing the study. Bias is an effect at any stage of a study that produces results that systematically depart from the true value, e.g. when the estimate of intervention efficacy differs from the true value in the source population. Bias may lead to either an over- or under-estimation of the intervention effect. There may be five domains through which bias might be introduced into the result (1) bias arising from the randomisation process; (2) bias due to deviations from intended interventions; (3) bias due to missing outcome data; (4) bias in measurement of the outcome; (5) bias in selection of the reported result. The risk-of-bias assessment can be performed using pre-specified tool from the Cochrane Collaboration’s recommended tools for randomized and non-randomized study designs.