Foodborne diseases, caused by contaminated food contact or ingestion, pose a significant global health concern. Contamination with bacteria, viruses, parasites, or chemicals can result in various illnesses, with over 200 identified, predominantly affecting the gastrointestinal tract but also leading to neurological, gynecological, and immunological issues [1]. Annually, nearly one in 10 individuals worldwide suffer from foodborne illnesses, resulting in over 420,000 deaths. While foodborne diseases affect all countries, low- and middle-income nations bear a disproportionate burden. Factors such as poverty, international trade, longer food chains, urbanisation, climate change, migration, and increased travel exacerbate these challenges, increasing the risk of contamination and the spread of infections across borders. In the European Union alone, over 5,000 foodborne outbreaks are reported annually, causing approximately 45,000 cases [2]. In 2022, foodborne outbreaks increased significantly compared to the previous year, highlighting the need for stringent hygiene standards and HACCP (Hazard analysis and critical control points) protocols throughout the food production chain to mitigate contamination risks and safeguard consumers.

According to the Food Packaging Forum [3], the most used packaging materials and food contact surfaces are:

• Ceramics;

The Codex Alimentarius specifies that the primary requirement for packaging is to avoid being a source of contamination [4]. Established in 1963 by the Food and Agriculture Organization (FAO) and the World Health Organization, the Codex sets international standards, guidelines, and codes of practice for food safety, quality, and fairness in international food trade, encompassing 99% of the global population. Food packaging is pivotal in maintaining food quality and safety by acting as a protective shield against external contaminants, prolonging product shelf life, and preventing spoilage. However, packaging itself can introduce contaminants, jeopardising food safety and integrity. Contaminated food packaging poses health risks, foodborne illnesses, allergenic reactions, and economic consequences, such as costly recalls and reputational damage for producers [5, 6]. Preventive measures include stringent quality control throughout the production process, using high-quality packaging materials, adhering to good practices and standards, and implementing rigorous hygiene and health risk analyses, notably through the Hazard Analysis and Critical Control Points (HACCP) method. While not explicitly mandated by legislation, adopting the HACCP methodology is widely recognised and demanded by the national and international market to ensure food safety and prepare for certification schemes. HACCP facilitates both contamination prevention and compliance with industry standards, aligning with market expectations and enhancing consumer confidence in food safety [7].

A crucial yet often overlooked aspect is the biological contamination level of packaging materials, influenced by factors like handling frequency and exposure to air. Different materials exhibit varying propensities for microbial proliferation and require distinct sanitation methods. Production processes vary widely among packaging materials, presenting opportunities for both contamination and decontamination Metals, glass, and plastics manufacturing involve temperatures incompatible with microbial survival. Conversely, cellulosic materials present contamination risks due to source contamination, processing in humid environments conducive to microbial growth, and lower processing temperatures that sustain resilient microbial forms. Paper and board production may involve biocidal agents. Post-production contamination risks persist across all materials due to handling, non-sterile air drafts, insect presence, and machine contact. Acceptable hygienic conditions are indicated by cell counts below 10⁴ cells/cm², while values exceeding 10⁷ cells/cm² denote unsatisfactory conditions [8].

Microbial growth on packaging materials depends on factors like nutrient presence, moisture, and surface biofilm formation. Effective decontamination, typically targeting a 5-log reduction in microbial load, employs various thermal, chemical, and physical methods.

Traditional decontamination approaches include heat treatments, in particular dry heat (>180 °C), hot water and steam (130°C–150°C), and the utilisation of chemicals like hydrogen peroxide, ethylene or propylene oxides [9–11]. A growing interest is directed to more innovative methods like high-pressure processing, high-intensity pulsed electric field treatment, pulsed light, ozone-based treatments and cold-atmospheric plasma, which holds promise for diminishing harmful microorganisms in the context of the food industry more safely and sustainably [12–14].

Plasma is the fourth state of matter, consisting of charged particles (ions and electrons) and neutral molecules. It can exploit several actions in many different applications, by tuning the different active agents that it includes [15]. These agents are the just-mentioned charged particles, chemically reactive species, such as free radicals, electromagnetic fields, radiations, and heat. For an application of interest, it is possible to control the operating conditions in order to maximise the presence of the active agents of interest. One main characteristic of plasmas (and the most relevant one to choose based on the application) is their macroscopic temperature. Plasmas can be consequently classified into:

- equilibrium (or thermal) plasmas, having the electron temperature in equilibrium with the heavy particles temperature, resulting in a macroscopic temperature higher than 10⁴ K;

Plasmas can also be classified on the basis of the pressure, which can be atmospheric or lower. Cold plasmas at atmospheric pressure will be hereinafter referred to as CAP (Cold Atmospheric Plasmas). Atmospheric pressure is often preferred in the industrial application perspective, since the arrangement is much simpler and economical. Furthermore, in-line continuous processes are economically sustainable only at atmospheric pressure, while low-pressure treatments should be carried out as batch processes.

The mechanisms of action related to the microbial inactivation are various. The active chemical components found in CAP demonstrate the ability to deactivate microorganisms on food surfaces due to their antimicrobial characteristics [14]. The primary effective components in the air plasma process include reactive oxygen species (ROS, including ozone, singlet oxygen, superoxide, peroxides, and hydroxyl radical) and reactive nitrogen species (RNS, mainly nitrogen oxides). The antimicrobial properties of these oxidative species can be attributed to lipid peroxidation within cell membranes and the oxidation of proteins and DNA within microbial cells [16]. Keener and Misra [17] have identified potential drivers for implementing this technology in the food industry, including lower energy requirements compared to current technologies, making it more environmentally friendly, reduced operational and maintenance costs due to simple systems with minimal upkeep and sanitation needs, improved chemical safety of foods through plasma inactivation and removal of pesticide and chemical residues, and its status as a green technology promoting environmental sustainability, as it only requires air and electricity to generate effective plasma.

Plasmas can be generated by a wide range of different sources, depending on the high voltage generator characteristics and the configuration of the source electrodes. The most used CAP sources are:

• Corona, in which a pointy electrode locally intensifies the electric field and allows the discharge;

Subsequently, the flow diagram is presented (Figure 1). The records were first identified through Scopus database searching, as previously illustrated. A few additional papers known by the authors were included as well. The collection of documents was then subjected to a preliminary screening on the software ASReview LAB (Zenodo, Switzerland), which assisted in the selection of eligible studies employing artificial intelligence tools. An additional amount was excluded by analysing the full-text articles for the above-mentioned reasons. A total number of 83 studies was included in the qualitative synthesis (i.e., the data extraction). Three papers did not present the data item considered to evaluate the outcome, i.e., the logarithmic reduction, and were consequently excluded from the quantitative analysis.

The studies outcomes are summarised in the table reported as Supplementary Material. For each study included in the quantitative analysis, the outcome indicator, chosen as the logarithmic reduction of the contaminant concentration, is reported. The effect estimates of the individual studies are graphically synthesised in the forest plot, with standard deviation if specified. Each different combination of factors (plasma source, type of treatment, pathogen, packaging material) is considered individually and associated with its outcome. The total number of individual cases is 263. Only the most efficient case is considered when different operating conditions were investigated, such as different treatment times, distance between plasma and the substrate, gas flux or generator electrical settings. The averages of these results for groups of similar pathogens are also calculated and reported in Figure 2.

Observing the average reduction rates for the sub-groups, it is evident that Gram-positive bacteria are usually less sensitive than Gram-negative bacteria to plasma treatment since the mean logarithmic reduction is respectively 3.90 ± 1.73 and 4.42 ± 1.76. This difference is in accordance with the literature and due to their distinctive structure of cell membrane [21]. The anti-viral effect of plasma on contaminated packaging is a field that needs further investigation, since only 3 studies were included in the quantitative analysis. The mean efficacy against fungi, equal to 4.06 ± 1.76 LogR, reflects the total outcome relative to the whole research. In fact, the total outcome of the meta-analysis is an average of 4.11 ± 1.75 LogR over 256 individual cases, which leads to the conclusion that plasma treatments at atmospheric pressure are effective against foodborne pathogens.