**2. Legislative Framework**

Table 1 synthetizes the interwoven combination of legislative acts, considering their implementation in the Italian legislation and the main issues related to the safe use of pesticides.

**Table 1.** EU legislation, its implementation in the Italian legislation, the context, and the related main issues.


Notes: (1) In the table, only the main legislative acts are mentioned, while their further updates/amendments were not cited: e.g. concerning Directive 2006/42/EC, Directive 2009/127/EC was not mentioned. Similarly, when citing Directive 89/391/EEC we intend its consolidated text, including for example Directive 2009/39/EC. (2) Regulation (EC) no. 1907/2006 and its amendments (the so-called "REACH") have not had a strong impact on pesticide users; it was included in the table due to the new rules on safety data sheets.

From the risk assessment point of view, the combination of the mandatory requisites provided by such legislative frameworks can be translated into the following requirements, which companies have to deal with:


• The preventive and protective measure applied, as well as the re-entry safety periods for each type of application have to be defined.

The practical implementation of these provisions requires a high level of expertise and, although a specific training for operators is mandatory to use pesticides, difficulties can arise in correctly understanding and putting them into practice [41]. Moreover, if we consider the Italian context, the companies operating in the agricultural sector are mainly small-sized or family-run enterprises [42–44], where the lack of human and financial resources often represents a drawback in properly implementing occupational health and safety duties [45].

To reduce these difficulties, which are similar for small and medium sized enterprises (SMEs) operating in different contexts [46–48], a specific easy-to-use risk assessment procedure was developed, which can allow companies to comply with the abovementioned requirements.

#### **3. Background Analysis**

The risk assessment activities are based on the estimation of the probability of occurrence of a certain event and the severity of its consequences. In the case of chemical risk, such a basic rule can be translated into the evaluation of the toxicity of the chemical product (e.g., based on the acceptable operator exposure level (AOEL) or the "lethal dose" criterion) and the level of exposure of the worker, intended as intensity of exposure per exposure time [49]. The assessment of the exposure to pesticides has been discussed by numerous studies [1,27,29,30,50,51] and different approaches have been proposed, which can be roughly distinguished into: (1) biomonitoring of exposure to pesticides, i.e., the measurement of a pesticide, its metabolite(s), or biotransformation products in biological fluids such as urine or blood [52]; (2) environmental monitoring, consisting in the measure of the exposure in the working environment [53].

Nevertheless, measuring the level of exposure is a difficult and complex task, due to the particularities that characterize the agricultural activities, the differences among the operators and working environments [54]. Accordingly, alternative methods for exposure and risk assessment have been developed, which vary from the use of expert opinion [55,56] and pre-marketing models [57,58] to the use of combination of data from the literature, measurements, and expert opinion [54,59]. In particular, premarketing models were introduced due to the mandatory need for evaluating the exposure of operators as well as for residents and bystanders [60], providing calculators based on databases, such as the BROWSE model [61], EUROPOEM [57], or TOXSWA [62]. Despite their ease of access, these models suffer from several drawbacks when applied in a practical context for occupational risk assessment [63,64]. Similarly, the so-called job exposure matrices (JEMs) [65] present with limited effectiveness when considering the agricultural activities [59,66], and for effective results in the determination of intensity of exposure they should be used in combination with algorithms that take into account other parameters such as the application rate, the type of equipment, and the characteristics of the crop [67].

The latter category includes the widespread Agricultural Health Study (AHS) model [68], developed for the estimation of the exposure to more than 50 individual pesticides, using questionnaire responses and pesticide information published in the literature. Such an approach is considered the most effective means for estimating the intensity level of the exposure to pesticides [69]. Without going into detail since a large literature can be found on the AHS augmentation (e.g., in [18,27,54,70–72]), the approach can be summarized in the following equation:

$$\text{II} = (\text{MDX} + \text{APPL} + \text{REPAIR}) \times \text{PPE},\tag{1}$$

where


As noted by Mandi´c-Raj´ceviˇc [73], the weights of these factors are based on the monitoring data published in the scientific literature. Although both positive and negative examples of using the AHS algorithm exist in the literature, this system was constructed for epidemiological studies and not for the purposes of the mandatory issues of the OHS legislation. In order to provide a method for the risk assessment of activities related to the use of pesticides, Colosio et al. [54] provided an augmented algorithm based on the AHS approach, which allows a semi-quantitative estimation of the occupational exposure and risk level consequent to pesticide application. Such a study provided practical criteria for the assessment of the exposure factors, which are summarized in Table 2.



Also, modifying factors such as the type of tractor, the type of the PPE used, and the training/skill of the operator were considered. In order to evaluate the toxicity, the criterion used consisted in deriving the toxicity scores based on the risk phrases allocated to the compound of the pesticides used. The advantage of this approach is that it uses the data readily available on the label of the product and does not require any extra training in toxicological evaluation of active substances for the workers. In line with such a framework, other studies (e.g., [74,75]) also provided practical criteria for the estimation of the exposure during re-entry (especially for specific types of farming, such as cultivation in greenhouses), which are based on the density of the plants and the type of pesticide.

Even though the above mentioned studies provide effective guidelines for the assessment of operators' exposure to pesticides and the related risk levels, they were developed for the purposes of historical exposure assessment in epidemiological studies. Therefore, the models and their various extensions fail to provide a documented risk assessment needed for companies to comply with the health and safety legislation. One example is the option of non-use of proper PPE, which is a legitimate option in epidemiological studies, but is not allowed by the legislation (Directive 89/391/EEC). Additional examples include the specific training of operators as well as the maintenance/calibration of the equipment (Directive 2009/128/EC), which are also cogent requirements.

Additionally, the risk phrases and precautionary statements have been replaced by the "hazard statements" and the related codes due to the entry into force of Regulation (EC) 1272/2008, which would make the approach proposed by Colosio et al. (2012) obsolete [54].

Based on the above considerations, we tried to extend the above-mentioned approaches considering the perspective of an employer that has to draw-up a document where:

