Franciscana University - UFN, Research Group in Food and Nutritional Security (GESAN/CNPq), Santa Maria - RS, Brazil, 1614, Andradas St, Centro, Santa Maria, RS, 97010-030, Brazil.
Department of Preventive Medicine - Universidade Federal de São Paulo - UNIFESP, Brazil. 740, Botucatu St, fourth floor, Vila Clementino, São Paulo, SP, 04024-002, Brazil.
Food Res Int. 2022 Jun;156:111127. doi: 10.1016/j.foodres.2022.111127. Epub 2022 Mar 12.
Risk scores are used worldwide to predict foodborne disease (FBD) outbreaks in the food service industry. This study aims to develop and validate a new method for the calculation of the FBD risk score for the checklist used to categorize food service outlets. The proposed novel method is based on a risk score for each item using a risk matrix (consequence × probability), overcoming the limitations of the previous scoring process used during the World Cup in Brazil, which was based on a factorial analysis. The classification of consequences was based on critical points identified by experts prior to the World Cup in Brazil. Probability was defined based on the violation percentage of each item evaluated during inspections from 1536 food service outlets. Validation was performed using a secondary database of 3072 food service assessments in two inspection cycles. The risk scores of the new method were compared with those used during the World Cup. Each food service was classified based on their risk score into four categories: A, B, C, and pending. Good concordance (Lin's correlation coefficient = 0.8711 and 0.9205) was observed between the new and previous scores in the two inspection cycles, respectively. Comparison of the classifications showed substantial agreement (Kappa = 0.749, p < 0.001) to the first inspection cycle and near-perfect agreement (Kappa = 0.821; p < 0.001) to the second inspection cycle. This new method allows the inclusion and exclusion of assessment elements depending on local reality. Simpler methods can be used throughout Brazil and serve as a model for other countries' food safety assessments.
风险评分被广泛用于预测餐饮服务业中的食源性疾病（FBD）暴发。本研究旨在开发和验证一种新方法，用于计算用于分类餐饮服务场所的检查表的 FBD 风险评分。所提出的新方法基于使用风险矩阵（后果×可能性）为每个项目计算风险评分，克服了巴西世界杯期间使用的评分过程的局限性，该过程基于因子分析。后果的分类基于巴西世界杯前专家确定的关键点。根据从 1536 个餐饮服务场所评估中评估的每个项目的违规百分比来定义可能性。使用两个检查周期中 3072 个餐饮服务评估的二级数据库进行验证。新方法的风险评分与世界杯期间使用的评分进行了比较。根据风险评分，将每个餐饮服务分为四个类别：A、B、C 和待处理。在两个检查周期中，新方法和旧方法的评分之间观察到良好的一致性（Lin 的相关系数= 0.8711 和 0.9205）。两次检查的分类比较显示出高度一致（Kappa= 0.749，p<0.001），与第一次检查周期接近完美一致（Kappa= 0.821；p<0.001）。这种新方法允许根据当地实际情况包含和排除评估要素。更简单的方法可以在整个巴西使用，并为其他国家的食品安全评估提供模型。
